IDENTIFICATION AND CHARACTERIZATION OF CIS -ACTING ELEMENTS IN THE REGULATION OF IMPRINTED GENE EXPRESSION By SARA RODRIGUEZ JATO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004
to the small things and the big ones to my family and friends
ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Thomas P. Yang, and everybody in his laboratory who taught me all the good science I know. Especially I would like to thank Dr. Sung Hae L. Kang and Dr. Chien Chen, Teresa Kunkel for her beautiful gel shifts, and particularly Christine Mione Kiefer, for her infinite patience and understanding. I would also like to express my gratitude to Dr. Daniel J. Driscoll for his support and for providing me with samples and cells when I needed them, Dr. Robert D. Nicholls for the good discussion and help on bioinformatics, Dr. Richard C. Scarpulla for the NRF-1 antibody and Dr. Edward Seto for the recombinant YY1. Friends the list is too long to write it but they know who they are. And finally, but really before anybody else, I thank my family for always staying so close despite the distance. iii
TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 Epigenetics....................................................................................................................1 Genomic Imprinting: The Beginnings..........................................................................2 Evolutionary Origin of Genomic Imprinting................................................................3 Genomic Imprinting in Development...........................................................................4 H19 and IGF2: A Well Characterized Imprinted Domain...........................................5 Prader-Willi Syndrome and Angelman Syndrome.......................................................6 Chromosome 15 q11-13 imprinted gene domain.........................................................8 The SNURF-SNRPN Promoter: Published Data.........................................................12 Molecular Characteristics of the Imprinting Center...................................................14 Nuclear Architecture and the Nuclear Matrix............................................................28 Ying Yang 1................................................................................................................32 Nuclear Respiratory Factor-1.....................................................................................33 The Mouse Model.......................................................................................................35 2 ANALYSIS OF CIS-REGULATORY ELEMENTS IN THE SNURF-SNRP 5â€™ REGION.....................................................................................................................39 Analysis of the PWS-SRO/SNURF-SNRPN 5â€™ Region by Transient Expression Assays....................................................................................................................40 Analysis of Cis-Regulatory Elements in the SNURF-SNRPN 5â€™ Region/PWS-IC by In Vivo Footprint Analysis.....................................................43 Examination of In Vivo Footprints by Transient Expression Assays.........................55 3 CHARATERIZATION OF NHS2: FUNCTIONAL IMPLICATION FOR THE IC........................................................................................................................59 Identification of an Activator Function Associated with NHS2.................................60 iv
Specificity of the Activator Function Associated with 2.2-NSH2.............................65 Identification of Elements Mediating the Activator Function....................................67 In Vitro binding of YY1 to the Putative YY1 Binding Site in ACS/NHS2...............71 Chromatin Immunoprecopitation (ChIP) Analysis of Factors Associated with SNURF-SNRPN 5â€™ Region/NHS1 and ACS/NHS2...............................................73 Association of RNA Polymerase II with ACS/NHS2................................................79 Analysis of Histone Modifications in SNURF-SNRPN 5â€™ Region/NHS1 and ACS/NHS2.............................................................................................................83 4 MATERIALS AND METHODS...............................................................................86 5 DISSCUSION AND FUTURE DIRECTIONS........................................................102 Analysis of SNURF-SNRPN 5â€™ Region/NHS1.........................................................102 Analysis of the SNURF-SNRPN gene 5â€™ region by transient expression assays.............................................................................................................102 Identification and functional analysis of cis-acting elements in the SNURF-SNRPN gene 5â€™ region.....................................................................103 Characterization of NHS2 and Associated Activation Function..............................106 Identification of a activator function associated with NHS2.............................107 Definition of an Activation Conserved Sequence (ACS) in the human and the mouse SNURF-SNRPN locus............................................................108 Analysis of cis-acting elements in the ACS/NHS2...........................................111 APPENDIX A VECTOR DESIGN FOR TRANSIENT EXPRESSION ASSAYS.........................117 B ANALYSIS OF 2.2-NHS2 ON SNURF-SNRPN PROMOTER ELEMENTS.........119 LIST OF REFERENCES.................................................................................................121 BIOGRAPHICAL SKETCH...........................................................................................138 v
LIST OF TABLES Table page 1-1 Summary of published data on SNURF-SNRPN promoter by transient transfection assays.................................................................................................13 4-1 Primers and conditions for LMPCR reactions.......................................................92 4-2 Site directed mutagenesis of SNURF-SNRPN promoter cis-acting elements: primer design.........................................................................................95 4-3 Primers and conditions for the amplification of immunoprecipitated DNA........101 5-1 Identification and functional analysis of cis-acting elements in the 5â€™region of the SNURF-SNRPN gene.................................................................................104 vi
LIST OF FIGURES Figure page 1-1 Molecular mechanisms leading to Prader-Willi syndrome and Angelman syndrome..................................................................................................................7 1-2 Schematic representation of the AS-PWS associated locus in human chromosome 15 q11-q13..........................................................................................9 1-3 Microdeletions associated with the PWS-AS IC...................................................15 1-4 Structure and nuclease hypersensitivity of the SNURF-SNRPN gene locus/IC..................................................................................................................27 2-1 Analysis of the SNURF-SNRPN promoter by transient transfection assays in SK-N-XX cells in the absence of a heterologous enhancer...............................42 2-2 Fundamentals of in vivo footprinting.....................................................................45 2-3 Schematic representation of the SNURF-SNRPN 5â€™ region analyzed by in vivo footprinting.................................................................................................47 2-4 DMS in vivo footprint analysis of the SNURF-SNRPN 5â€™region..........................49 2-5 DNase I in vivo footprint analysis of the SNURF-SNRPN 5â€™region......................51 2-6 Summary of in vivo footprint analysis of the SNURF-SNRPN 5â€™ region..............55 2-7 Functional analysis of in vivo footprints by transient expression assays. ............56 3-1 Identification of an activator function associated with 2.2-NHS2.........................62 3-2 Functional interaction between the activator associated with 2.2-NHS2 and in vivo footprints of the SNURF-SNRPN gene promoter.......................................64 3-3 Preferential activation of the SNURF-SNRPN promoter by the activator function associated with 2.2-NHS2.......................................................................66 3-4 Identification of cis-acting elements involved in the activator function associated with 2.2-NHS2: definition of a minimal activator sequence (MAS)....68 vii
3-5 Blast algorithm based alignment of the human 2.2-NHS2 region and the full length first intron in the mouse Snurf-Snrpn gene.................................................70 3-6 Functional analysis of a putative YY1 binding site in 2.2-NHS2 by transient expression assays...................................................................................................71 3-7 Gel mobility-shift assays of the putative YY1 binding site in the ACS-NHS2.....72 3-8 ChIP analysis of factors associated with the SNURF-SNRPN upstream promoter U1A, the SNURF-SNRPN 5â€™ region and ACS/NHS2............................78 3-9 ChIP analysis of RNA polymerase II association with SNURF-SNRPN upstream promoter U1A, SNURF-SNRPN 5â€™ region/NSH1 and ACS/NHS2.......81 3-10 ChIP analysis of the histone code associated with SNURF-SNRPN upstream promoter U1A, SNURF-SNRPN 5â€™ region/NHS1 and ACS/NHS2.......................84 5-1 Targeted deletions of the Snurf-Snrpn locus in mouse........................................110 B-1 Functional interaction between the activator associated with 2.2-NHS2 and components of the SNURF-SNRPN gene promoter.............................................120 viii
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IDENTIFICATION AND CHARACTERIZATION OF CIS-ACTING ELEMENTS IN THE REGULATION OF IMPRINTED GENE EXPRESSION By Sara Rodriguez-Jato May 2004 Chair: Thomas P. Yang Major Department: Biochemistry and Molecular Biology The Prader-Willi and Angelman syndromes (PWS-AS) associated region includes a cluster of imprinted genes that are coordinately regulated by an imprinting center (IC) spanning the 5â€™region of the SNURF-SNRPN gene. The IC has a bipartite structure where the PWS-IC is postulated to create an active domain on the paternal allele, while the AS-IC is believed to silence the maternal allele. Also, alternative upstream SNURF-SNRPN promoters have been implicated in the AS-IC function. This project focused on identifying and characterizing cis-acting elements within the IC that may mediate IC function and/or SNURF-SNRPN promoter activity. The PWS-IC contains major nuclease hypersensitive sites (NHS1 and NSH2) associated with the SNURF-SNRPN promoter region and the 1st intron (adjacent to the IC), both specific to the paternal allele. In vivo footprint analysis of the promoter region has identified multiple factor binding sites, 4 specific to the paternal allele and 1 specific to the maternal allele, some of which affected promoter function in transient expression assays. Transient expression assays also helped ix
to define other promoter elements in the SNURF-SNRPN 5â€™region and to identify a novel position-dependent and orientation-independent activator associated with NHS2. This element(s) preferentially activates the SNURF-SNRPN main and upstream promoters relative to the UBE3A and MKRN3 promoters. It appears to be a complex regulatory unit from which a subregion was mapped that sustains high levels of SNURF-SNRPN promoter activity. This subregion was also identified by sequence comparison of the intronic activator with the homologous region in mouse and contains several highly conserved sites including potential binding sites for SP1, NRF1 and YY1. In vivo and in vitro studies showed binding of YY1 to the intronic activator and its involvement in the activator function. Furthermore, recruitment of the non-processive form of RNA polymerase II to the activator was shown suggesting a possible role for the activator in the IC function by recruiting pol II and transferring it to the SNURF-SNRPN upstream promoters in the female germline and to the main promoter in the male germline/paternal allele. This transfer may be mediated by NRF-1 that was shown to bind in vivo to both the SNURF-SNRPN promoter and NHS2. Additionally YY1 may target the paternal allele to the nuclear matrix, which would account for the nuclear localization and early replication of that allele. x
CHAPTER 1 INTRODUCTION Genomic imprinting is an epigenetic process through which certain genes acquire different expression patterns depending on the parental origin of the allele. This thesis project is aimed at uncovering some of the molecular mechanisms that underlie this process. Epigenetics Epigenetic is a general term to describe all the modifications of the chromatin and DNA that form part of the transcriptional memory of a cell, i.e., the epigenetic information complements the genetic information to determine what genes are transcribed and at what level (Wu and Morris, 2001) and reviewed in (Jaenisch and Bird, 2003). Epigenetic modifications can be stably transmitted through cell mitosis and meiosis. However, epigenetic processes are also dynamic and during development and cell differentiation, epigenetic changes direct the necessary adjustments in gene transcription that are then maintained in the terminal state of the cell. Aberrant changes in cellular epigenetics can lead to phenomena like aging (Bandyopadhyay and Medrano, 2003) or cancer (reviewed in (Momparler, 2003)). Epigenetics also play a crucial role in genomic imprinting, in which paternallyand maternally-inherited alleles in the same nuclear environment and sharing essentially the same DNA sequence are present in two different transcriptional states. Each allele exhibits a parent-of-origin-specific set of epigenetic modifications or paternal/maternal 1
2 epigenotype, that result in the active expression of one allele and the silencing of the other. Genomic Imprinting: The Beginnings The discovery of genomic imprinting marked the beginning of the era of post-Mendelian genetics. Plants (Kermicle and Alleman, 1990), insects (Crouse, 1960) and mammals (Surani et al., 1984) are all subjected to this process in some way or another however, the scope of this dissertation is limited to its occurrence in eutherian mammals. Genomic imprinting was first described in 1984 by two independent groups (McGrath and Solter, 1984) and (Surani et al., 1984) by means of pronuclear transplantation studies. Their studies analyzed one-cell-stage mouse embryos with two female pronuclei (parthenogenotes), two male pronuclei (androgenotes) or a biparental pronucleous. They observed that only the embryos with both parental genomic contributions, but not parthenogenotes or androgenotes, would complete normal embryogenesis, which demonstrated the functional non-equivalence of maternal and paternal genomes in mammals. It was then hypothesized that some genes were expressed exclusively from one parental genome and that if that parental genome was not present in the embryo, those genes would not be expressed and the embryo would not develop appropriately. Mice with uniparental disomies (UPD; inheritance of both copies of a chromosome or a chromosomal region from one parent only) for a specific chromosome or subchromosomal region permitted the identification and localization of chromosomal regions subjected to genomic imprinting (Cattanach and Kirk, 1985), (Searle and Beechey, 1985), (Searle and Beechey, 1990). The first imprinted gene to be identified was a transgene in mouse, RsvIgmyc, that fortuitously showed paternal-specific
3 expression1 (Chaillet et al., 1995). Mutations in the murine insulin-like growth factor 2 Igf2, which resulted in a dwarfing phenotype only when paternally inherited, led to the discovery of the first endogenous imprinted gene (DeChiara et al., 1991). Over 80 imprinted genes have been identified to date; however, the final number is likely to be higher (visit www.geneimprint.com for updated information). Several approaches have been used to identify imprinted genes including screening of cDNA libraries from parthenogenic and androgenic embryos (Mann et al., 1995), genome-wide search for differentially methylated regions which are a landmark of imprinted genes (Strichman-Almashanu et al., 2002) (see below) and analysis of genes in a chromosomal region associated with an imprinted phenotype (Ozcelik et al., 1992). Evolutionary Origin of Genomic Imprinting Genomic imprinting entails a genetic paradox since it implies the loss of the advantage of diploidy. Having multiple alleles protects organisms against the effects of deleterious recessive mutations, which raises the question of why monoallelic expression has developed in the evolution of eutherian mammals. There are three major schools of thought to answer that question (Wilkins and Haig, 2003): The enhanced adaptability model (Beaudet and Jiang, 2002). This model is based on a) the advantage conferred by the functional haploicy in genomic imprinting; masking of one allele from natural selection in each generation allows for the accumulation of more complex mutations and reduces the pressure of selection in changing environments and b) the development of mechanisms to adjust the rates of gene expression by many 1 Maternal expression: expression of a given gene from the maternally-inherited allele Paternal expression: expression of a given gene from the paternally-inherited allele
4 imprinted genes. This allows for a rapid and reversible adaptability to fluctuating conditions and thus a selective advantage. The ovarian time bomb model (Varmuza and Mann, 1994). Ovarian teratomas originate from oocytes that spontaneously activate and initiate development. These teratomas are relatively benign but they can differentiate into invasive trophoblast disease. This model proposes that imprinting arose as a way to protect females from this process, since the genes involved in trophoblast development are silent in the oocyte. The conflict model . It was initially proposed as a conflict between maternal and paternal interests in respect to the genes that affect the resources that the offspring acquire from the mother. However, this model has developed further to take into account all the advantages and disadvantages (not merely growth and resources) that the rate of expression of a given imprinted gene represents for each individual of the same kin, parents, parents to be, siblings and so forth (Wilkins and Haig, 2003). Genomic Imprinting in Development The parent-of-origin identity of the imprinted loci is established during germline development and entails a collection of epigenetic marks or imprints that ultimately result in differential patterns of gene expression from the paternallyand maternally-inherited alleles (Tremblay et al., 1995) (Stoger et al., 1993) (Shemer et al., 1997). During fertilization, the zygote acquires a set of chromosomes from the mother (oocyte) in which imprinted loci have a maternal epigenotype, and a set of chromosomes from the father (sperm) in which imprinted loci have a paternal epigenotype. The identity or epigenotype of these imprinted alleles is retained during somatic development in a process called somatic maintenance. During gametogenesis, however, the marks that confer that identity are erased and replaced by a new set of marks so that all the alleles have a
5 maternal epigenotype in the female germline and a paternal epigenotype in the male germline. This process is referred to as the imprint switch. The nature of the imprinting mark, its somatic maintenance and switch in the germline are not completely characterized. However, evidence indicates that several epigenetic modifications may be involved. These modifications include chromatin structure and modification of core histone amino-terminal tails (Pedone et al., 1999), DNA methylation (Reik et al., 1987), DNA/factor interactions (Szabo et al., 2000), replication timing (Kitsberg et al., 1993) and nuclear localization (Kagotani et al., 2002). Imprinted genes are often grouped in clusters in the genome. This chromosomal organization suggests that imprinted genes are not regulated individually but as a chromosomal imprinted domain. Most imprinted domains include small regions that direct and coordinate the imprinting process across the entire gene cluster (reviewed in (Verona et al., 2003)). The small regions or imprinting control centers include elements that may act in cis to regulate imprinted gene expression. Well-characterized examples of imprinted gene clusters or domains are the H19/IGF2 locus and the Prader-Willi syndrome and Angelman syndrome associated region which is the main focus of this dissertation. H19 and IGF2: A Well Characterized Imprinted Domain H19 and IGF2 (insulin-like growth factor 2) are two oppositely imprinted genes in human chromosome 11 p11.5 and are linked to the Beckwith-Wiedemann syndrome (BWS[MIM 130650]) (Henry et al., 1991). Analyses of a syntenic region in mouse chromosome 7 have led to an enhancer competition model to explain the regulation of this imprinted domain. h19 and Igf2 are 90 kb apart and share a group of enhancers upstream from the h19 gene. Expression of h19 and Igf2 specifically from the
6 maternallyand paternally-inherited alleles, respectively, is dependent upon a 2.2 kb intervening imprinting control region (ICR) that has differential DNA methylation (Thorvaldsen et al., 1998). On the maternally-inherited unmethylated allele, CTCF is associated with the ICR and functions as an insulator or boundary element, blocking the effect of the h19 enhancers on the Igf2 promoter. Conversely, DNA methylation of the ICR inhibits CTCF binding to the paternally-inherited allele so that the Igf2 promoter can be activated by the h19 enhancers and expressed (Hark et al., 2000), (Szabo et al., 2000). Additionally, h19 expression on the paternally-inherited allele is blocked by DNA methylation at the promoter region. Furthermore, apart from acting as a insulator, CTCF is also involved in maintaining the unmethylated state of the maternal chromosome during embryo development (Schoenherr et al., 2003). Prader-Willi Syndrome and Angelman Syndrome Prader-Willi syndrome (PWS [MIM 176270]) and Angelman syndrome (AS [MIM 105830]) are two distinct neurobehavioral disorders with an incidence of 1 in every 10,000 to 15,000 births each. Both disorders are linked to the same imprinted domain in chromosome 15 q11 to q13, which includes several genes expressed exclusively from the paternally-inherited allele and at least two genes expressed exclusively from the maternally-inherited allele. PWS arises from defects on the paternally-inherited allele that cause the loss of expression of the first group of genes. Patients are characterized by neonatal hypotonia with failure to thrive, hypergonadism, hyperphagia and obesity, short stature and small hands and feet, craniofacial dysmorphism, mild mental retardation with learning disabilities and obsessive compulsive disorder. AS arises from defects on the maternally-inherited allele that cause the loss of expression/function of maternally
7 Figure 1-1: Molecular mechanisms leading to Prader-Willi syndrome and Angelman syndrome. Horizontal lines labeled P and M represent the paternallyand the maternally-inherited alleles of chromosome 15 respectively. Blue and red indicate paternal and maternal epigenotype respectively. Grey lines represent other chromosomal regions involved in translocations. Five mechanisms that result in PWS and AS are described, as well as the frequency of occurrence among patients (Nicholls et al., 1998). expressed genes. Patients are characterized by ataxia and tremors, lack of speech, severe mental retardation, inappropriate laughter and in most cases microcephalus, seizures and abnormal electroencephalograms (Cassidy et al., 2000). There are several genetic mechanisms leading to the loss of gene expression/ function deficiency that results in PWS and AS as depicted in Figure 1-1. The most common one, which occurs with an incidence of 70 %, is a 4 to 4.5 Mb deletion that includes the entire imprinted domain. The origin of the deletion is always paternal in PWS patients and maternal in AS patients. Uniparental disomy (UPD) constitutes 25-28% of the PWS cases (maternal-UPD) although paternal-UPD is rare amongst AS patients (2-5%). No individual gene mutations have been described in PWS which is
8 considered to be a contiguous gene syndrome, while 10-15 % of AS patients carry a mutation in the maternally expressed UBE3A gene (Kishino et al., 1997). About 5 % of PWS and AS patients have imprinting defects in which the imprinted domain on the paternallyand maternally-inherited chromosome 15 q11-q13, respectively, show the reverse epigenotype and patterns of gene expression. Half of these patients have no detectable mutation or deletion and their imprinting defect has been postulated to originate as a stochastic failure to reset the imprints during gametogenesis and/or to maintain them through embryogenesis (Buiting et al., 1998). The other half of the patients have imprinting mutations that in most cases consist of microdeletions on the 5â€™ region of the SNURF-SNRPN gene (Buiting et al., 1995). These microdeletions have been the key to define the existence of an imprinting center (IC) that regulates imprinted gene expression across the entire domain. Less frequent are balanced translocations within the imprinted domain (<1%) that can disrupt the structure of some of the genes in the cluster or can disrupt the function of the IC itself. Chromosome 15 q11-13 imprinted gene domain The PWS-AS associated region is located on chromosome 15 q11-q13. Figure 1-2 shows a map of the region that includes at least six paternally-expressed genes and two maternally-expressed genes. Four of the genes MKRN3, NDN, MAGEL2 and C150RF2 are intronless genes, which suggests a common derivation by retrotransposition. In addition, there are several paternally-expressed antisense transcripts to some of the genes (MKRN3 antisense or MKRN3-AS and UBE3A-AS) which, in this and other imprinted systems as well, have been proposed to play a role in the regulation of their respective sense genes. The paternal MKRN3 gene (formerly ZNF 127) codes for a polypeptide
9 Figure 1-2: Schematic representation of the AS-PWS associated locus in human chromosome 15 q11-q13. Blue ovals represent genes expressed from the paternally-inherited allele and pink ovals represent genes expressed from the maternally-inherited allele. The arrows indicate the direction of transcription. The position of the imprinting center or IC is designated with a yellow box. ICMKRN3 / MKRN3-A S CenTelMAGEL2SNURF-SNRPNIPWUBE3A / UBE3A-ASHuman: 15q11Human: 15q11--1313 Paternally expressed Maternally expressedNDN ATP10C snoRNA snoRNA ICMKRN3 / MKRN3-A S CenTelMAGEL2SNURF-SNRPNIPWUBE3A / UBE3A-ASHuman: 15q11Human: 15q11--1313 Paternally expressed Maternally expressedNDN ATP10C snoRNA snoRNA characterized by a RING zinc-finger and several additional C3H (Cys 3 His) zinc fingers. Its function is not known however, knockout studies of the imprinted mouse homologue, Mkrn3, showed that animals lacking this gene were viable, fertile and non-obese (Glenn et al., 1993), (Jong et al., 1999). However, neurobehavioral phenotypes are often difficult to assess in the mouse. The testis-specific gene, C150RF2, is not a classical imprinted gene, since it is expressed biallelically in the male germline. However, all the alleles in the male germline have a paternal epigenotype once the imprint switch has occurred. Therefore, if expression of C150RF2 starts after the imprint switch has occurred, the gene would be imprinted and expressed exclusively from the paternally-inherited allele (Farber et al., 2000). The paternally-expressed NDN and MAGEL2 are members of the MAGE family of genes. NDN is a neuronal growth suppressor that can repress transcription by interaction with transcription factors (Nakada et al., 1998) and is essential in neuron development (Takazaki et al., 2002). The function of MAGEL2 has not yet been determined (Lee et al., 2000). Several Ndn knockout models have been constructed in the mouse however, the observed phenotypes varied depending on the background strain
10 used. In two studies deleting Ndn resulted in a certain degree of postnatal lethality (Muscatelli et al., 2000) (Gerard et al., 1999); the surviving mice were fertile and showed some of the hypothalamic and behavioral characteristic of PWS patients (Muscatelli et al., 2000). Conversely, a third study reported no apparent phenotype (Tsai et al., 1999a). SNURF-SNRPN (S NRPN u pstream r eading f rame-s mall n uclear r ibonucleop rotein N ) is a bicistronic gene expressed mainly in the brain. SNURF-SNRPN exons 1 through 3 code for SNURF, a polypeptide of unknown function, and exons 4 through 10 code for SmN, a core spliceosome factor that binds small nuclear RNA (snRNA) and is involved in mRNA splicing (Gray et al., 1999). The 5â€™ flanking region of SNURF-SNRPN is located within the IC and is likely to play a role in the regulation of the entire imprinted domain (see below). In addition to the main transcription start site at exon 1, SNURF-SNRPN has two alternative upstream transcription start sites and several upstream non-coding exons. These exons are spliced in many alternative patterns, creating a complex network of upstream transcripts observed only in the brain. Presumably, the upstream exons originated by duplication of the SNURF-SNRPN promoter/exon 1 region and, like SNURF-SNRPN, are expressed from the paternally-inherited allele. Nevertheless, these upstream transcripts or IC transcripts are associated with the part of the IC that regulates the maternal epigenotype (see below AS-SRO). Moreover, Dittrich et al. (Dittrich et al., 1996) have speculated that the upstream promoters are silent in the male germline and only expressed in the female germline, where they might assist in establishing the transcriptionally silent state. More recently, alternative upstream transcripts have been found in the mouse Snurf-Snrpn locus (Bressler et al., 2001). Furthermore, Otha et al. (Otha et al., manuscript in preparation) have confirmed that those mouse transcripts are
11 expressed in the oocyte and silenced in spermatogenesis, and, during early embryo development, they are expressed from the maternally-inherited chromosome and silenced from the paternally-inherited chromosome. SNURF-SNRPN is also the start site for a long transcript that extends hundreds of kilobases downstream. This transcript accounts for the expression of the non-coding UBE3A-AS, IPW and several families of C/D box snoRNAs (small nucleolar RNAs) which result from splicing of the large primary transcript. C/D box snoRNAs are known to methylate the 3â€™-hydroxyl group of the ribose in ribosomal RNA and other small nuclear RNAs involved in splicing. These sno RNAs are characterized by two short conserved sequence motifs, box C and box D, located a few nucleotides away from the 5' and 3' ends respectively. Additionally they include a 10-20 nucleotides region that is complementary to the RNA target for the O-methylation (Smith and Steitz, 1997). Imbedded between the 148 SNURF-SNRPN exons identified to date are two large clusters of tandemly repeated snoRNA genes HBII-52 and HBII-85, three single-copy snoRNA genes, HBII-436/13/437, and two copies of HBII-438 snoRNA gene that are 240 kb apart (Runte et al., 2001b). Study of PWS patients that carry translocations with breakpoints within the SNURF-SNRPN locus has shown that lack of expression through the region containing the HBII-85 cluster of snoRNAs and one of the copies of HBII-438 snoRNA is responsible for most of the features observed in the phenotype associated with PWS patients. Since neither of these snoRNA has complementarity to any ribosomal RNA or small nuclear RNA, it has been proposed that they might target messenger RNAs. In addition to the paternally-expressed genes described above, there are two genes expressed from the maternally-inherited allele. UBE3A codes for the ubiquitin-protein
12 ligase UBE3A/E6-AP, and ATP10C encodes a putative aminophospholipid translocase (Herzing et al., 2001; Meguro et al., 2001a). These genes, UBE3A and ATP10C, are located in the telomeric end of the cluster and are transcribed in opposite orientation to SNURF-SNRPN. However, imprinted monoallelic expression of these genes is limited to the brain and, at least for UBE3A, likely linked to expression of UBE3A-AS. UBE3A-AS constitutes the known 3â€™ end of the paternally-expressed long transcript that originates at SNURF-SNRPN and is expressed exclusively in the brain. It is widely accepted that transcription of UBE3A-AS silences expression of the sense gene on the paternal allele, although the mechanism of that repression remains elusive (Chamberlain and Brannan, 2001). Accordingly, UBE3A is expressed from both alleles in tissues where UBE3A-AS is silent. Additionally, UBE3A includes a 5â€™ CpG island, but no differential DNA methylation between expressing and silent alleles has been found. This suggests that the silencing of this gene on the paternally-inherited allele in the brain is not mediated by DNA methylation. On the other hand, not much is known about the imprinting regulation of ATP10C, since there is no evidence for an antisense transcript to ATP10C or for a DMR. (Kashiwagi et al., 2003) The SNURF-SNRPN Promoter: Published Data The SNURF-SNRPN 5â€™ flanking sequence has been shown to act as a promoter (Huq et al., 1997). This region lacks a TATA element as well as a classic initiator, but it does contain several CG boxes representing potential SP1 binding sites, which is characteristic of TATA-less promoters. Additionally, it includes a number of direct and inverted repeats that are very frequent within imprinted genes (Neumann et al., 1995). The transcriptional activity of the SNURF-SNRPN promoter has been examined by transient expression analysis by several groups (Green Finberg et al., 2003), (Hershko et
13 al., 2001), (Huq et al., 1997). Table 2-1 summarizes the main findings from those studies. It is important to note that the promoter was examined in the presence of a exogenous (simian virus 40 (SV40)) enhancer. Table 1-1: Summary of published data on SNURF-SNRPN promoter by transient transfection assays* Researcher Position SV40 Enh Me-DNA Mut -57/51 Activity -207 to +53 Yes Yes +4 to +53 Yes Huq et al. Genome Res., 1997 -207 to +53 Yes Yes No -216 to +201 No No -216 t o +201 Yes No Hershko et al. Genes to Cells, 2001 -216 to +53 Yes Low -71 to +51/59 Yes Yes -71 to +51 Yes Yes No -71 to +12 Yes No Finberg et al. Gene, 2002 -71 to +78 Yes No *Positions are indicated respect to the SNURF-SNRPN transcription start site; SV40 Enh refers to the presence of the SV40 enhancer downstream from the CAT gene in the reporter construct; Me-DNA indicates methylation of CpG dinucleotides in the reporter construct; Mut indicates mutation of the indicated nucleotides. These findings indicate that the minimal promoter of the SNURF-SNRPN gene is contained between bases -71 and +51 and includes an element necessary for promoter activity between nucleotides -57 to -51. This site is conserved in sequence and function in the homologous promoter in the mouse gene, and data suggest that expression of the SNURF-SNRPN gene is sensitive to methylation of this site (Green Finberg et al., 2003), (Hershko et al., 2001). However, according to Huq et al., the sequence from +4 to +53, which lacks the abovementioned element, retains as much as half of the maximum activity (Huq et al., 1997). Additionally, a negative regulatory element has been identified around position +60 (Green Finberg et al., 2003). Significantly, no promoter activity was detected in the absence of the SV40 enhancer (Hershko et al., 2001).
14 Molecular Characteristics of the Imprinting Center It was previously mentioned that a fraction of PWS-AS patients with imprinting defects carry microdeletions on the 5â€™ flanking region of the SNURF-SNRPN gene. These microdeletions, which range from just a few kb to several hundreds of kb as shown in Figure 1-3, affect the entire imprinted domain by preventing the appropriate epigenotype to be established and/or maintained during gametogenesis and/or early embryo development. This led to the definition of the imprinting center or IC within the SNURF-SNRPN locus that regulates imprinted gene expression throughout the entire PWS-AS associated domain (Sutcliffe et al., 1994). As can be readily observed in Figure 1-3, the IC has a bipartite structure. Paternally-inherited deletions that result in Prader-Willi syndrome cluster closer to SNURF-SNRPN and define the PWS-IC which is involved in the establishment and/or maintenance of the paternal epigenotype in the paternally-inherited imprinted domain. However, these mutations have no effect on the epigenotype of the maternally-inherited imprinted domain. All the PWS associated microdeletions share a 4.3 kb region termed small region of overlap or PWS-SRO that is represented in Figure 1-3 by the blue vertical block. The PWS-SRO includes the SNURF-SNRPN promoter and first exon (Ohta et al., 1999) and six sequences phylogeneticaly conserved between the human and the syntenic region in mouse chromosome 7C. The maternally-inherited deletions that result in Angelman syndrome cluster further upstream, defining the AS-IC which is thought to function in the establishment the maternal epigenotype in the maternally-inherited imprinted domain.
15 PWSShort Regionof Overlap(PWS-SRO)ASShort Regionof Overlap(AS-SRO) PWSDeletion on paternalalleleASDeletion on maternalallele Involved inmaternal alleleidentit y Involved inpaternal alleleidentit y 1 234-1011-2 U1AU2U3U1B Bâ€™ // 10 kb SNURF-SNRPN PWSShort Regionof Overlap(PWS-SRO)ASShort Regionof Overlap(AS-SRO) PWSDeletion on paternalalleleASDeletion on maternalallele Involved inmaternal alleleidentit y Involved inpaternal alleleidentit y 1 234-1011-2 U1AU2U3U1B Bâ€™ // 10 kb SNURF-SNRPN PWSShort Regionof Overlap(PWS-SRO)ASShort Regionof Overlap(AS-SRO) PWSDeletion on paternalalleleASDeletion on maternalallele Involved inmaternal alleleidentit y Involved inpaternal alleleidentit y 1 234-1011-2 U1AU2U3U1B Bâ€™ // 1 234-1011-2 U1AU2U3U1B Bâ€™ // 10 kb 10 kb SNURF-SNRPN Figure 1-3: Microdeletions associated with the PWS-AS IC. The map on the top of the figure shows the SNURF-SNRPN gene locus. Black boxes represent exons and bent arrows represent transcription initiation sites. The extent of the microdeletions in PWS and AS imprinting mutation patients is shown below. Horizontal lines represent the deletions. The parental origin of the microdeletions is indicated on the right. The PWS-SRO and the AS-SRO are indicated by blue and pink vertical blocks respectively. These mutations have no effect on the epigenotype of the paternally-inherited chromosome. A 0.8 kb AS-SRO common to all the AS-associated deletions is located approximately 35 kb upstream from SNURF-SNRPN exon 1 (Buiting et al., 1999) and is represented in Figure 1-3 by the pink block. The AS-SRO includes U3, one of several upstream non coding exons in the SNURF-SNRPN locus (Buiting et al., 1999) suggesting a role of the IC transcripts in the function of the AS-IC. Most PWS and AS patients carrying imprinting mutations are familial cases in which the deletions are inherited from the paternal grandmother in PWS, or the maternal grandfather in AS. The other two alternatives in which the paternal grandfather carries a
16 PWS-IC deletion or the maternal grandmother carries a AS-IC deletion would result in a father affected by PWS or a mother affected by AS, respectively (Brannan and Bartolomei, 1999). In contrast, deletions that occur de novo are not subject to this strict inheritance pattern and can affect either allele regardless of the grandparental origin. Some PWS patients carry deletions in their paternally-inherited chromosome that include both the PWSand AS-IC (Saitoh et al., 1996). In these patients, the imprinted domain lacking the IC shows a maternal epigenotype suggesting that the AS-IC is not required for the establishment of a maternal epigenotype in the absence of the PWS-IC. However, the deletion has been inherited from the paternal grandmother in all the cases. Therefore, it is possible that the maternal epigenotype had never been erased. No AS patients carrying a similar deletion have been identified, since the mother inheriting the mutated chromosome from the grandfather would be affected by PWS. As a result, the study of PWS patients carrying de novo deletions including both PWS-IC and AS-IC in the chromosome of grandpaternal origin or AS patients carrying similar deletions in the maternally-inherited chromosome would provide a better understanding of the role of the AS-IC in the erasure and re-setting of the maternal epigenotype. However, there are no published reports of PWS and AS patients with those characteristics which would suggest that the maternal epigenotype may be appropriately re-set and may result in no phenotype. Nevertheless, no final conclusions can be drawn since the number of patients with de novo deletions is very low (Tilghman and Schoenherr, 1998). From all these observations Brannan and Bartolomei (Brannan and Bartolomei, 1999) proposed a model in which the PWS-IC is functional in the paternally-inherited chromosome and drives the active transcriptional state while in the maternally-inherited
17 allele the AS-IC negatively regulates the PWS-IC allowing for the establishment of a transcriptionally inactive state. According to this model the AS-IC is not required to establish a maternal epigenotype per se, rather, it prevents the establishment of a paternal epigenotype and is, therefore, dispensable when the PWS-IC is deleted. Alternatively, Dittrich et al. (Dittrich et al., 1996) and more recently Otta et al. (manuscript in preparation) proposed a more active participation of the AS-IC in the establishment of the maternal epigenotype in the female germline and early embryogenesis. According to this model, the IC transcripts are essential components in the process of establishing a silent or repressed state across the domain. In 2000, Bielinska et al. (Bielinska et al., 2000) reported the study of a male individual with mild PWS clinical symptoms in which an IC deletion including both PWS-SRO and AS-SRO in his paternally-inherited chromosome had occurred postzygotically. The individual was therefore mosaic for the deletion that was present in the germline and was transmitted to his son, who was affected by PWS. DNA methylation analysis revealed DNA methylation in sites that are normally methylated in the maternally-inherited imprinted domain and lack of DNA methylation in sites that are normally methylated in the paternally-inherited imprinted domain. Based on these findings, the authors proposed that the deleted paternally-inherited chromosome had acquired a maternal epigenotype. This hypothesis was confirmed in chimaeric mice derived from an ES cell line in which a paternal PWS-IC deletion had been introduced. Interestingly no changes on DNA methylation were observed in the ES cells before injection into the recipient blastocyst. The observation that the PWS-AS associated region in the paternally-inherited chromosome can acquire a maternal epigenotype upon
18 postzygotic deletion of the PWS-IC indicates that the PWS-IC is involved and required to maintain the paternal epigenotype during early embryo development. El-Maarri et al. (El-Maarri et al., 2001) studied the role of the PWS-IC in establishing the paternal epigenotype in the male germline. They analyzed the sperm of male individuals carrying a PWS-IC deletion on their maternally-inherited chromosome 15. These individuals are healthy, however, the offspring that inherited the chromosome carrying the deletion will be affected by PWS. DNA methylation analysis revealed that the correct patterns of DNA methylation were established along the imprinted domain in the male germline despite the loss of the PWS-IC. The same results were observed in the mouse. The authors conclude that the PWS-IC is not necessary for the correct reset of the DNA methylation imprint in the male germline and that the aberrant DNA methylation observed in IC deletion PWS patients is established de novo during embryogenesis. This indicates that the PWS-IC not involved in the DNA methylation switch in the male germline but is critical for the maintenance of paternally-inherited DNA methylation imprints during embryogenesis. Some of the epigenetic modifications associated with the IC have been already characterized. These epigenetic marks are the basis to understand the mechanism though which the IC regulates imprinted gene expression throughout the PWS-AS associated imprinted domain. Described next are known data on the DNA methylation and chromatin structure in the IC. DNA methylation. In mammals DNA methylation is restricted to position 5 of cytosine residues within CpG dinucleotides. 5-methyl-C accounts for approximately 1% of all DNA bases or 70-80% of all CpG dinucleotides in the genome (Ehrlich et al.,
19 1982). The bulk of the genome is, therefore, mostly methylated. The 5â€™ ends of many genes include regions of high relative densities of CpG, or CpG islands (Razin and Cedar, 1977). In contrast to the bulk of the genome, CpG islands are largely unmethylated (Tazi and Bird, 1990). However, a significant fraction of these CpG-rich regions do acquire DNA methylation during development, as is the case of CpG islands associated with genes in the inactive X or silent imprinted genes (Yen et al., 1984) (Stoger et al., 1993). DNA methylation is generally associated with transcriptional silencing (Keshet et al., 1986). For a number of loci, DNA methylation has been shown to be a late event in the process of silencing a gene suggesting that it acts as a locking mechanism for the repressed state after it has been established by other epigenetic mechanisms (Lock et al., 1987), (Gautsch and Wilson, 1983), (Tamaru and Selker, 2001). Other schools of thought support an active role of DNA methylation in the creation of a transcriptional repressive state by recruiting specific DNA methyl-binding proteins (MDBâ€™s) that in turn recruit other factors, including chromatin modification enzymes (Antequera et al., 1989) (Chandler et al., 1999) (reviewed in (Hendrich and Tweedie, 2003)). DNA methylation patterns in the genome are reset during gametogenesis (Tada et al., 1997) and early embryo development (Monk et al., 1987), (Kafri et al., 1992) where waves of global de-methylation followed by de novo methylation allow for the necessary reprogramming of gene expression patterns. In mammals, DNA methylation patterns are dependent on DNA methyltransferases. There are three confirmed DNA methyltransferases Dnmt1, 3a and 3b. Dnmt3a and 3b have partially redundant functions and are thought to recognize and â€œde novo methylateâ€ unmodified DNA resulting in hemimethylated DNA (Okano et al., 1999; Okano et al.,
20 1998). Loss of function of Dmnt3a and Dnmt3b resulted in death within 4 weeks of birth in Dnmt3a-/mice and embryonic lethality in Dnmt3b-/mice. Inactivation of both methyltransferases resulted in a more severe phenotype accompanied by partial impairment of de novo DNA methylation (Okano et al., 1999). Dnmt1 is thought to maintain the DNA methylation patterns established by Dnmt3a and 3b (Bestor, 1992) (Pradhan et al., 1999). It recognizes hemimethylated DNA and methylates the complementary CpG, thus creating a heritable fully methylated site. Consequently, knockout of Dnmt1 in mouse led to global loss of methylation and embryonic lethality (Li et al., 1992). DNA methyltransferases do not appear to have any sequence specificity. Therefore, targeting of these enzymes to specific sequences in vivo may be mediated by DNA binding factors or chromatin structure. However, how this targeting occurs remains unknown. Most imprinted loci show DNA methylation patterns that differ between expressing and silent alleles (Tremblay et al., 1995), (Sutcliffe et al., 1994), (Liu et al., 2000). This differential DNA methylation imprint, in particular at some key regulatory regions, is acquired during gametogenesis and is maintained though the waves of deand remethylation during embryogenesis and somatic development (Davis et al., 2000), (Shemer et al., 1997). For those reasons, establishment of differential DNA methylation in the germline is believed a fundamental epigenetic event in the regulation of genomic imprinting. Furthermore, differential DNA methylation is considered a hallmark of imprinted loci. The PWS-IC is associated with a CpG island at the 5â€™ region of the SNURF-SNRPN gene (Sutcliffe et al., 1994) (Shemer et al., 1997). This CpG island becomes
21 methylated on the maternally-inherited allele in the female germline and remains methylated throughout embryogenesis (Geuns et al., 2003), (Shemer et al., 1997). In contrast, this region is unmethylated in the paternally-inherited allele. Interestingly, during spermatogenesis, but not embryogenesis, this unmethylated state is independent of the AS-IC (El-Maarri et al., 2001). DNA methylation in the maternally-inherited allele can be reversed in cell culture by treatment with the DNA methyltransferase inhibitor 5-azacytidine, which reactivates expression of the SNURF-SNRPN gene. However, mouse Dnmt1-/ES cells that lack differential DNA methylation in Snurf-Snrpn 5â€™ region, retain imprinted gene expression (Xin et al., 2003). Analysis of other CpG dinucleotides in the SNURF-SNRPN locus also showed monoallelic DNA methylation. Specifically, the U1B and U1A upstream exons and transcription initiation sites showed DNA methylation on the maternally-inherited allele, while the upstream pseudoexon U1D, which is upstream of U1B and is not transcribed, showed methylation on the paternally-inherited allele (Runte et al., 2001a). On the contrary, the AS-SRO is partially methylated in both paternallyand maternally-inherited alleles to the same extent, suggesting that DNA methylation may not be part of the imprinting mechanism in this region (Perk et al., 2002). Alternatively, the AS-IC may only be functional in the female germline, where the DNA methylation patterns of the AS-IC have not been analyzed. The significance of DNA methylation in genomic imprinting and, in particular, in the PWS-AS associated domain has been confirmed in mice deficient in DNA methyltransferases. Dnmt1 knockout mice showed hypomethylation and loss of monoallelic expression of some imprinted genes including h19, Igf2 (Li et al., 1993) and
22 Snurf-Snrpn (Xin et al., 2003), although other imprinted genes such as Mash2 (Caspary et al., 1998) were unaffected. Mutation of both Dmnt3a and Dnmt3b in mouse ES cells did not affect DNA methylation patterns at the h19 5â€™ region, patterns that are established in the germline and maintained in embryogenesis. In order to study the role of Dnmt3a and 3b in the establishment of methylation imprints in the germ line, Hata et al. (Hata et al., 2002) transplanted ovaries from [Dnmt3a-/-, Dnmt3b+/-] female mice into wild type females. They obtained two embryos in which maternal DNA methylation imprints at Igf2r, Peg1, Peg3 and Snurf-Snrpn had been lost. These results suggested that Dnmt3a and/or 3b are required for the establishment of de novo DNA methylation imprints in the germline (Hata et al., 2002). Chromatin structure. In the nucleus of eukaryotes genomic DNA is tightly packed with proteins in a series of hierarchical structures to form chromatin. It was initially thought that the role of chromatin was limited to compact genomic DNA into the nucleus. However, over the last few years evidence has accumulated showing that chromatin actively participates in regulating nuclear functions like transcription (see below), DNA repair (Green and Almouzni, 2002) or centromere identity and stability of kinetochore-microtubule attachments (Sharp and Kaufman, 2003). At the most fundamental level of chromatin organization, genomic DNA is associated with histones to form nucleosomes. The nucleosome core contains two of each histones H2A, H2B, H3 and H4. H3 and H4 form homodimers that interact with each other to create a stable complex that assembles a H2A and H2B heterodimer on each side. Linker histone such as H1 or H5, bind nucleosomal DNA as it enters and exits the nucleosome. Linker histones assist to stabilize and compact chromatin into higher order
23 structures. Therefore, they are mostly associated with silent but not active chromatin in vivo. Nucleosomes are non-static structures separated by a stretch of linker DNA of variable length. The core particle includes 147 bp of DNA wrapped around the histone octamer. However, in vivo the length of DNA sequence in a nucleosome oscillates between 100 and 170 bp. Several nucleosome remodeling complexes have been identified that can alter the nucleosomal structure (Narlikar et al., 2002). These complexes may function by sliding the histone octamer onto a neighboring stretch of DNA or by modifying the interaction of histones with DNA. Activation of genes is often associated with recruitment of nucleosome remodeling complex to gene promoters and the subsequent changes in chromatin structure allow binding of transcription factors to DNA. This suggests that nucleosomes may affect the interaction of DNA and transcription factors and assist in the activation or repression of gene expression. Additionally, each of the core histones has a highly mobile, positively charged amino-terminal domain and histones H2A and H3 also have analogous regions on their C-termini. These domains or histone tails are the target of multiple posttranslational modifications including acetylation (Turner, 2000), methylation (Rea et al., 2000), phosphorylation (Clayton and Mahadevan, 2003), ubiquitination (Zhang, 2003). Recent evidence suggests that these histone modifications may result in altered chromatin structure which, in turn, may lead to downstream events such as changes in the transcriptional status of a chromosomal region or definition of centromeres. In this manner, histone modifications, either sequentially or in combination, form a â€˜histone codeâ€™ (Strahl and Allis, 2000), (Fischle et al., 2003). The histone code may be recognized by chromatin associated factors and be translated into gene activation, gene
24 silencing etc. In fact, some proteins have been shown to interact with specifically modified histones. This is the case of heterochromatin protein 1 (HP1), which interacts with histone H3 methylated at lysine 9 (H3 K9-Me) (Lachner et al., 2001). Additionally, hyperacetylation of H3 and H4 (Turner, 2000) tails and methylation of H3 at lysine 4 (H3-MeK4) (Strahl et al., 1999) are generally associated with transcriptionally active chromatin while hypoacetylation of H3 and H4 (Turner, 2000) and H3-MeK9 (Lachner et al., 2001) are generally associated with transcriptionally silent chromatin. In solutions containing low concentrations of monovalent cations and no multivalent cations (Widom, 1986), the nucleosomal array adopts an extended structure or nucleosome filament. The nucleosome fiber then compacts into a approximately 30 nm wide filament or 30 nm fiber (Felsenfeld and McGhee, 1986) that can be further compacted into higher order chromatin structures. The level of compaction and accessibility of chromatin can be assessed by sensitivity of the underlying DNA to digestion with nucleases. This type of assays has shown that transcriptionally competent chromatin is less compact or more open and accessible to transcription factors (active or euchromatic state). On the contrary transcriptionally repressed chromatin tends to be more compact or closed and less accessible to the transcriptional machinery (inactive or heterochromatic state). Additionally, further sensitivity or hypersensitivity to nuclease digestion has been found associated with regions in DNA that function in cis as regulatory regions. These nuclease hypersensitive sites (NHS) likely reflect the disruption of chromatin at the nucleosome level and the interaction of the underlying DNA and proteins in trans. Analysis of NHS has become a common technique to identify cis-acting regulatory regions in chromatin.
25 Modification of histone tails associated with the PWS-AS IC have been examined by chromatin immunoprecipitation (ChIP). Analysis of SNURF-SNRPN 5â€™ region revealed preferential association of hyperacetylated H3 and H4 (Gregory et al., 2001) (Fulmer-Smentek and Francke, 2001; Saitoh and Wada, 2000)) and H3-K4-di-Me (Xin et al., 2001) with the paternally-inherited allele. Conversely, on the maternally-inherited allele H3 and H4 were found to be hypoacetylated (Gregory et al., 2001) (Fulmer-Smentek and Francke, 2001; Saitoh and Wada, 2000)) and H3 was found to be dimethylated at K9 (Xin et al., 2001). According to Xin et al., methylation of H3 (H3-K4-diMe and H3-K9-diMe) is restricted to a small well defined region flanking SNURF-SNRPN exon 1. However, similar studies in mouse, showed H3-K9 hypermethylation (combination of diand tri-methylation) associated with a region further downstream in intron 1 in the maternally-inherited allele (Fournier et al., 2002). Methylation of H3-K9 is thought to be related to DNA methylation (Tamaru and Selker, 2001). Accordingly, it has been shown in mouse ES cells that DNA methylation patterns in the PWS-IC maternally-inherited allele are dependent on the G9a histone H3 Lys9/Lys27 methyltransferase (Xin et al., 2003). G9a -/ES cells also showed the loss of Snurf-Snrpn monoallelic expression. However, this was not a consequence of the loss of differential DNA methylation since Snurf-Snrpn imprinted expression was retained in Dnmt1 -/ES cells that lacked differential DNA methylation but showed H3-K9-diMe associated with the PWS-IC. Restoring G9a in G9a -/ES cells, re-establish H3-K9 methylation but no DNA methylation or monoallelic expression. However, G9a -/-embryos derived from heterozygote intercrosses showed normal methylation patterns in the PWS-IC in both paternallyand maternally-derived alleles.
26 Histone modifications characteristic of open chromatin (H3 and H4 hyper-acetylation and H3-K4 methylation) have been found to associate with the AS-SRO maternally-inherited allele (Perk et al., 2002). On the contrary, analysis of other sites in the SNURF-SNRPN locus showed no differences in H3 and H4 acetylation between alleles. Likewise, the 5â€™ regions of the imprinted genes NDN, IPW and MAGEL2 showed no detectable H3 and H4 acetylation on either allele, suggesting that histone modifications may no play a direct role in the transcriptional regulation of those genes (Fulmer-Smentek and Francke, 2001; Saitoh and Wada, 2000) and indicating that differential histone modifications are restricted to the IC. The chromatin structure of the PWS/AS associated region has been studied using several approaches. Watanabe et al. used a method based on sonication of nuclei followed by chromatin fractionation by centrifugation which allows separation of heterochromatin from euchromatin. The enrichment of paternallyand maternally-inherited alleles of several mouse imprinted genes in those fractions was assessed by PCR (Watanabe et al., 2000). For Mkrn3, Snurf-Snrpn and Ndn they observed enrichment of the paternally-inherited allele in the euchromatic fraction. However, this result was only observed in one of two interspecies crosses except for Mkrn3, that showed increased DNase I sensitivity in the paternally-inherited copy in both crosses. Meguro et al. analyzed the nuclease sensitivity of a 13.6 kb region located between SNURF-SNRPN and IPW in the snoRNAs region (Meguro et al., 2001b). This region showed increased general sensitivity to restriction nucleases in the transcriptionally-active paternally-inherited allele. In contrast, Perk et al. analyzed the DNase I sensitivity of the AS-SRO and flanking region and found that the maternally-inherited allele is more
27 PWS-SRO1 234-1011-2SNURFSNRPN U1AU2U3U1B Bâ€™ AS-SRO Paternal HSMaternal HSCenTel PWS-SRO1 234-1011-2SNURFSNRPN U1AU2U3U1B Bâ€™ AS-SRO Paternal HSMaternal HSCenTel Figure 1-4: Structure and nuclease hypersensitivity of the SNURF-SNRPN gene locus/IC. The SNURF-SNRPN gene locus is depicted above where black boxes represent exons and bent arrows represent transcription initiation sites. The approximate position of paternal-specific NHSs (blue arrows above the line) and maternal-specific HSs (pink arrows below the line) is indicated, as well and the location of the PWS-SRO and the AS-SRO. accessible and DNase I sensitive (Perk et al., 2002). Schweizer et al. analyzed the nuclease hypersensitivity of the entire SNURF-SNRPN locus (Schweizer et al., 1999). Their results indicated that the maternally-inherited allele of SNURF-SNRPN exhibits several weak nuclease hypersensitive sites, one of which localizes to the AS-SRO. The paternally-inherited allele shows two prominent hypersensitive sites, NHS1 located within the PWS-SRO in the promoter region of SNURF-SNRPN immediately upstream from the first exon, and NSH2 located just downstream from the PWS-SRO in the first intron of SNURF-SNRPN (Figure 1-3). NSH1 is associated with the promoter region of SNURF-SNRPN and therefore is likely to contain cis-acting elements involved in promoter function as well as cis-acting elements that may mediate the PWS-IC function; NSH2 may include SNURF-SNRPN promoter distant regulatory elements and its close proximity to PWS-SRO suggests it may also contain elements that influence the PWS-IC function. It was stated before that deletion of the AS-IC on the maternally-inherited chromosome 15 q11-q13 results in gene expression patterns and epigenetic marks characteristic of the paternally-inherited allele. Perk et al. examined the precise changes
28 that such deletions trigger on DNA methylation and chromatin structure of the maternally-inherited PWS-SRO (Perk et al., 2002). As expected, DNA methylation patterns, nuclease hypersensitivity, and association of H3-K4-Me were typical of the paternally-inherited allele. However, association of acetylated H3 and H4 and replication timing observed in the PWS-SRO in a normal maternally-inherited chromosome 15 are retained after deletion of the AS-IC. This suggests that the AS-IC plays a crucial role at setting up the maternal epigenotype in PWS-SRO, although it may not be the only determinant. Conversely, deletion of the PWS-IC on the paternally-inherited allele has no effect on the methylation or chromatin structure of the AS-SRO. This suggests that the PWS-IC plays no role at setting up the paternal epigenotype in the AS-SRO. The authors conclude that their observations support the model proposed by Brannan and Bartolomei (Brannan and Bartolomei, 1999) in which the function of the AS-IC is to repress the PWS-SRO on the maternal chromosome. Nuclear Architecture and the Nuclear Matrix Chromosomes in the interphase nucleus are organized in distinct territories (Lichter et al., 1988), (Zink et al., 1999) in which transcriptionally active and inactive chromatin appears to be highly compartmentalized with active chromatin located predominantly at or near the surface of more compact chromatin domains (Dietzel et al., 1999), (Volpi et al., 2000). In addition, early replicating, gene rich, active chromatin and late replicating, gene poor, inactive chromatin form distinct domains that are distributed throughout the interphase chromosomal territories and are stably maintained for at least several cell generations (Zink et al., 1999). These and similar findings support the idea that chromatin is highly organized around these chromosomal domains that might concentrate a number of associated functions by assembling different sets of factors (for replication,
29 transcription or repair) depending on the cell cycle stage (Berezney and Wei, 1998). In fact, a number of spatially distinct functional nuclear domains including replication, transcription and RNA processing sites (reviewed in (Strouboulis and Wolffe, 1996) have been defined by immunostaining, FISH and pulse labeling. Furthermore, the distinct 3D distribution of these sites is retained in nuclear preparations in which the chromatin and soluble component of the nucleus have been extracted. This suggests that the replication (Hozak et al., 1993) and transcriptional machineries (Jackson et al., 1993) may be associated with some non chromatin insoluble nuclear structure or nuclear matrix (reviewed in (Cook, 1999). The nuclear matrix can be defined as the residual scaffold that is left in the nucleus after the removal of the nuclear envelope, chromatin and soluble components of the nucleus by sequential extractions (Zbarskii and Debov, 1948) (Smetana et al., 1963), (Monneron and Bernhard, 1969), (Berezney and Coffey, 1974), (Berezney and Coffey, 1977). Berenzeny and Coffey first proposed that the nuclear matrix represented the basis for structure and function in a highly organized nucleus. However, no absolute structure/function link has been made and, for many, the biological reality of the nuclear matrix remains uncertain. DNA in eukaryotic nucleus is packed into nucleosomes to form nucleosome filaments, which are further compacted into a 30 nm fiber. This 30 nm fiber is folded into loop domains that range from 5 to 200 kb of DNA (Razin and Gromova, 1995). These looped structures are believed to be attached to the nuclear matrix which is thought to play an active role in organizing chromatin loops in the nucleus (reviewed in (Heng et al., 2001). In fact, the distribution of DNA sequences retained at the base of the loops in
30 nuclear matrix preparations mirrors the distribution of chromosomal territories. The sequences bound to the nuclear matrix at the base of the loop have been proposed to function as anchors or matrix attachment regions (MARs) (Cockerill and Garrard, 1986; Ivanchenko and Avramova, 1992). MARs have been often found to colocalize with hypersensitive sites and cis-acting regulatory sequences (Kieffer et al., 2002) and are often associated with regions involved in the regulation of nuclear processes such as replication and transcription (reviewed in (Cook, 1999)). In fact, many chromatin remodeling complexes and transcription factors (YY1, CTCF etc) required for transcription regulation are associated with the nuclear matrix (Hendzel et al., 1994), (Reyes et al., 1997), (Guo et al., 1995), (Dunn et al., 2003). Visualization of DNA in nuclear matrix preparations has revealed that transcriptionally active sequences are more tightly associated with the nuclear matrix at the base of the loops while transcriptionally inactive sequences often extend into the looped region (Gerdes et al., 1994). Furthermore, Cai and Kohwi-Sigematsu (Cai and Kohwi-Shigematsu, 1999) showed that activation of the T-cell-specific SATB1 factor after T-cell differentiation is associated with attachment of the gene to the nuclear matrix. Cook and others have proposed a model that recapitulates all the evidence that functionally links chromatin loops and gene regulation ((Cook, 1999), (Lemon and Tjian, 2000). In this model, chromatin loops interact with the transcription machinery which is located at the nuclear matrix. This supports a system that is highly dynamic to adapt to the specific requirement of a developmental or cell cycle state. Chromatin loops have also been proposed to play a role in transcription regulation by physically separating domains. Thus, insulators and boundary elements may lead to
31 the formation of chromatin loops that act as topological barriers that modulate promoter enhancer interactions (Cai and Shen, 2001). Additionally, the base of the looped structures may also prevent the spreading of specific chromatin modifications among chromatin domains. The PWS-IC contains a high density of MARs (Greally et al., 1999), though the association of the PWS-IC with the nuclear matrix is somewhat controversial. An initial report showed preferential association of the maternal allele with the nuclear matrix (Greally et al., 1999). More recently, Kagotani et al. examined the association of SNURF-SNRPN with the nuclear matrix using DNA-FISH combined with a tyramide signal amplification technique which is more sensitive that standard DNA-FISH (Kagotani et al., 2002). Contrary to what was reported before, they demonstrated that association with the nuclear matrix occurred at the transcriptionally active paternally-inherited allele of SNURF-SNRPN while the maternally-inherited allele was located in the nuclear halo in nuclear matrix preparations. There is in general a good correlation between active chromatin and early replication and, equally, between inactive chromatin and late replication. Additionally, it has been suggested that the transcription and replication functions in the nucleus are associated with the nuclear matrix. Accordingly, it was not surprising that for SNURF-SNRPN the transcriptionally-active nuclear matrix-associated paternally-inherited allele replicates earlier in S-phase than the maternally-inherited allele. Furthermore, this asynchronous replication is established in the germ line (Gunaratne et al., 1995), (Simon et al., 1999). These data suggest compartmentalization of SNURF-SNRPN paternallyand maternally-inherited alleles in functionally different nuclear domains after
32 fertilization. Moreover this distribution may be mediated by interaction with the nuclear matrix. Ying Yang 1 Ying Yang 1 or YY1 is a ubiquitously expressed zinc finger transcription factor that regulates expression of many cellular and viral genes, many of which participate in cell growth and differentiation (for review see (Shi et al., 1997) and (Thomas and Seto, 1999)). YY1 binds DNA in a sequence specific manner through the consensus binding sequence (C/g/a)(G/t)(G/t/a)CATN(T/a)(T/g/c) (Hyde-DeRuyscher et al., 1995). Additionally, YY1 has been found to not bind methylated DNA (Satyamoorthy et al., 1993) at promoters and enhancers of the genes it regulates. Early studies showed that YY1 is able to bind the initiator element of some gene promoters (Seto et al., 1991). Furthermore, is necessary and sufficient for accurate basal transcription of the AAV p5 promoter, which shows that YY1 can direct and activate transcription in vitro (Seto et al., 1991). Additionally, YY1 has been proposed to recruit RNA polymerase II in the presence of TFIIB in vitro (Usheva and Shenk, 1994). YY1 can also act as a repressor or an activator of transcription at least partly by interacting with chromatin remodeling enzymes such as histone deacetylases (HDACS) histone acetyltransferases and arginine-methyltransferases (HATs and Arg-HMTs) ((Thomas and Seto, 1999),(Rezai-Zadeh et al., 2003) (Yao et al., 2001). YY1 can interact with other nuclear factors binding to neighboring DNA sequences such as SP1 (Lee et al., 1993), NRF-1 (Wong-Riley et al., 2000), p300 (Lee et al., 1995) or some members of the E2F family of transcription factors (Schlisio et al., 2002). The latter interaction between YY1 and E2F is indirect and requires an additional factor, RYBP (ring-1 YY1 binding protein).
33 YY1 is known to associate with the nuclear matrix (Guo et al., 1995) (Stein et al., 1991). Guo et al. determined that a significant fraction of YY1 is localized in the nuclear matrix which suggested that YY1 may mediate gene-matrix interactions. This is consistent with a functional role of the nuclear matrix in the regulation of expression of YY1 target genes. The PEG 3 imprinted locus has been recently shown to include several arrays of tandemly repeated YY1 binding sites (Kim et al., 2003). In addition, in vitro experiments showed that the YY1 binding sites in the PEG3 locus function as chromatin insulators. These data provided evidence for YY1 association with imprinted domains and suggested that YY1 may play a role in the regulation of imprinted gene expression. YY1 is expressed very early in the embryo. Knockout studies in mouse resulted in peri-implantation lethality, which suggested that YY1 is an essential component in the regulation of development (Donohoe et al., 1999). Interestingly, YY1 is localized to the cytoplasm and excluded from the nucleous of Xenopus oocytes where it is bound by ribonucleoproteins particles that render it inactive (Ficzycz et al., 2001). The same phenomenon has been observed in mammalian oocytes (Donohoe et al., 1999). Nuclear Respiratory Factor-1 Nuclear respiratory factor-1 or NRF-1 is a nuclear transcription factor initially identified because of its involvement in the regulation of nuclear-encoded mitochondrial genes (Evans and Scarpulla, 1989) (Evans and Scarpulla, 1990). The mitochondrial genome encodes some of the factors involved in oxidative phosphorylation. However, the majority of the genes involved in that process are expressed from the cellular nuclear genome. NRF-1 is found in a number of promoters and enhancers of genes of the respiratory chain (Chau et al., 1992), (Evans and Scarpulla, 1990) as well as other genes
34 that participate in the expression and assembly of mitochondrial function, such as components of its replication (Clayton, 1992) and transcriptional machinery (Virbasius and Scarpulla, 1994), enzymes of the heme biosynthetic pathway (Braidotti et al., 1993), and factors involved in mitochondrial protein transport and assembly (Hernandez et al., 1999). Furthermore, NRF-1 expression is increased upon stimulation of mitochondrial respiratory function either by exercise or electrical impulses(Xia et al., 1997). Consequently NRF-1 has been proposed to play an integrative role in the interaction between the nucleus and the mitochondria in response to the availability of metabolic resources (reviewed in (Scarpulla, 2002)). In addition, NRF-1 has also been shown to regulate genes encoding rate-limiting enzymes in several biosynthetic pathways including purine biosynthesis (Chen et al., 1997) which is consistent with the theory that NRF-1 may mediate the communication between cellular respiration and other metabolic pathways. It is therefore possible that NRF-1 plays a part in the conflict model (see Evolutionary Origin of Genomic Imprinting) between the maternal and the paternal genomes related to the use of the maternal resources by the developing embryo. Correspondingly, NRF-1 is known to be expressed in the early developing embryo and targeted deletion results in small size and increased lethality in heterozygous animals and in peri-implantation lethality in homozygous animals (Huo and Scarpulla, 2001). NRF-1 binds unmethylated DNA as a homodimer through the consensus sequence YGCGCAYGCGCU, where Y is a pyrimidine nucleotide and U is a purine nucleotide (Evans and Scarpulla, 1989) (Evans and Scarpulla, 1990). NRF-1 is generally associated with gene promoter regions where it directly or indirectly interacts with other transcription factors such as SP1 (Chen et al., 1997), YY1 (Wong-Riley et al., 2000) and
35 P300/CBP. Additionally binding of NRF-1 to DNA can be stimulated by acetylation (Izumi et al., 2003) and phosphorilation (Gugneja and Scarpulla, 1997). The Mouse Model There is a striking conservation on the organization of the genes and patterns of imprinted gene expression in the PWS-AS associated domain between human and mouse. The PWS-AS associated region in human chromosome 15, and its syntenic region in mouse chromosome 7, also share many patterns of allele specific epigenetic modifications such as patterns of differential DNA methylation, of differential histone tail modifications or replication timing among others. However, there seems to be some differences in structure of the IC between both species. A number of knockout studies have been performed in the mouse in order to identify the elements and mechanisms that control the IC function in mouse. Yang et al. (Yang et al., 1998) deleted 43 kb of DNA sequence flanking Snurf-Snrpn 5â€™ end (23 kb upstream and 20 kb downstream from the transcription start site). Upon paternal transmission, this deletion resulted in loss of expression of paternally-expressed genes and gain of expression of maternally-expressed genes in the paternally-inherited chromosome. The change in expression pattern on the paternally-inherited copy of the PWS-AS associated domain, was accompanied by a switch in the DNA methylation patterns from paternal to maternal. This suggested that the PWS-IC is contained within that 43 kb sequence. On the other hand, maternal transmission of the deletion had no effect on gene expression patterns or epigenotype of the maternally-inherited allele, suggesting the AS-IS was not included in the 43 kb sequence. Additionally, deletion of a smaller 4.8 kb fragment around Snurf-Snrpn exon 1 (from ~-2500 to ~+2300) resulted, when paternally-transmitted, in 50% lethality and in partial imprinting defect on the paternally-inherited allele, which was characterized by
36 partial gain of a maternal epigenotype (Bressler et al., 2001). As observed with the 43 kb deletion, maternal transmission had no effect on the PWS-AS associated domain. These results suggested that the 4.8 kb sequence contains some but not all the elements that are necessary for the function of the PWS-IC. The location of the AS-IC, or element equivalent to the AS-SRO, has not been defined in mouse yet. Deletion of a 100 kb sequence extending from 11 to 111 kb upstream of Snurf-Snrpn, had no effect on imprinted gene expression of the associated imprinted domain (Chen et al., 2002). This suggested that the structure of the human and mouse AS-IC may differ at least in the position/distance respect to SNURF-SNRPN. Nonetheless, an imprinting defect of maternal origin is observed when a 1.5 Mb sequence from 111 to 1611 upstream from Snurf-Snrpn is deleted (Chen et al., 2002). Other deletions in the PWS-AS associated domain have been generated but showed no effect on the imprinted gene expression/imprinting status of this region. A deletion of 900 bp of the Snurf-Snrpn gene promoter region (from ~-600 to ~+300) (Bressler et al., 2001) had no effect on the IC function or phenotype. Therefore, PWS-IC elements within the previously described 4.3 kb deleted region ought to be located outside of the promoter region of Snurf-Snrpn. A second deletion of the Snurf-Snrpn coding region had no effect on IC function and showed no phenotype either (Yang et al., 1998). Elimination of the sequence between Snurf-Snrpn exon 2 and Ube3A (Tsai et al., 1999b) resulted in a PWS phenotype probably due to the elimination of the snoRNAs. However, no imprinting defect was observed and the remaining paternally-expressed genes were appropriately expressed.
37 The function of the IC has also been studied with the assistance of transgenes. Blades et al. (Blaydes et al., 1999) generated several mouse transgenic lines carrying a 76 kb human transgene that included the SNURF-SNRPN gene and 45 kb of SNURF-SNRPN 5â€™flanking sequence containing the PWS-SRO and the AS-SRO. The SNURF-SNRPN gene was expressed in all the transgenic lines regardless of parental origin. Furthermore, DNA methylation of the transgene did not affect active transcription of the SNURF-SNRPN gene. It is possible that the transgene did not include all the elements necessary for silencing the maternally-inherited allele. Alternatively, it is also possible that the human IC was not recognized by murine trans-acting factors or that the IC mechanism itself differs between both species. Similarly, moue transgenic lines were generated carrying transgenes that included the entire 22-kb murine Snrpn gene and approximately 33 and 30 kb of 5â€™ and 3â€™ flanking sequence, respectively. One single-copy transgenic line failed to acquire imprinted expression, which compared to what was observed for the human transgene. On the other hand a second two-copy transgenic line was imprinted and expressed exclusively after paternal transmission (Blaydes et al., 1999). However, due to the diversity of the results depending on copy number and the small amount of transgenic lines, no definite conclusions about the human and mouse IC function can be drawn. Smaller transgenes in mouse encompassing 1.2 to 5 kb of Snurf-Snrpn 5â€™ flanking region showed lack of imprinted expression (Shemer et al., 2000). In addition, the same group generated a minitransgene that included approximately 1kb of sequence spanning the human AS-SRO and approximately 200 bp of Snurf-Snrpn 5â€™ flanking region spanning the promoter and exon 1 (Shemer et al., 2000). They found that those
38 sequences were sufficient to establish differential DNA methylation, monoallelic expression and asynchronous replication. However, it was not clear if this was a copy number dependent effect. This thesis project focused on untangling the mechanisms through which the IC regulates imprinted gene expression in the PWS-AS associated region by identifying and characterizing cis-acting regulatory elements within two paternal-specific nuclease hypersensitive sites that are associated with the IC.
CHAPTER 2 ANALYSIS OF CIS-REGULATORY ELEMENTS IN THE SNURF-SNRP 5â€™ REGION Hypersensitivity to nuclease digestion is often associated with cis-regulatory regions in chromatin (Gross and Garrard, 1988). Hence, in order to identify candidate regulatory regions in the IC, Schweizer et al. (Schweizer et al., 1999) examined the nuclease hypersensitivity of the SNURF-SNRPN locus. Their analysis of the paternally-inherited allele revealed two prominent nuclease hypersensitive sites, NHS1 and NHS2. NHS1 lies within the PWS-SRO in the promoter region of the SNURF-SNRPN gene and will be discussed in this chapter. The SNURF-SNRPN 5â€™ region has been implicated in IC function as a nucleation site for the establishment of an active domain on the paternally-inherited allele (Brannan and Bartolomei, 1999; Dittrich et al., 1996). Furthermore, binding of transcription factors and activation of transcription have been proposed as a possible mechanism for initiating the imprint switch and the establishment of an active domain on the paternally-inherited allele (Nicholls et al., 1998). Therefore, the transcriptional activity of the PWS-SRO and in particular NHS1 in SNURF-SNRPN 5â€™region was analyzed by transient expression assays. Additionally, the SNURF-SNRPN 5â€™ region includes a cluster of six phylogenetic footprints1, which are conserved between human and mouse (Glenn et al., 1996). These sequences are the only conserved regions between both species across the entire human 1 Phylogenetic footprint is defined as at least 6 contiguous bases of absolute conservation in orthologous genes of evolutionary distant species 39
40 PWS-SRO and the homologous region in mouse. It is, therefore, possible that these phylogenetic footprints represent cis-acting elements involved in the PWS-IC function. Therefore, analysis of the SNURF-SNRPN 5â€™ region/NHS1 was extended to include the six phylogenetic footprints. Analysis of the PWS-SRO/SNURF-SNRPN 5â€™ Region by Transient Expression Assays Cis-acting elements within the PWS-SRO and 5â€™ region of the SNURF-SNRPN gene were analyzed by transient expression assays. The SNURF-SNRPN 5â€™ flanking sequence is hypersensitive to nuclease digestion on the paternally-inherited transcriptionally active allele (Schweizer et al., 1999) and has been shown to act as a promoter by transient expression assays (Green Finberg et al., 2003), (Hershko et al., 2001), (Huq et al., 1997). These types of experiments are based on the use of a reporter plasmid containing the fragment of DNA with suspected promoter activity in front of a reporter gene. The reporter construct is introduced into cells by transient transfection and, after a set period of time, the activity of the reporter gene product is analyzed as a measurement of expression and, by inference, of transcription. It is important to note that in the reports mentioned above, the SNURF-SNRPN promoter activity was analyzed in the context of the heterologous (SV40) enhancer. Furthermore, one of the studies reported lack of promoter activity in the absence of such enhancer (Hershko et al., 2001). Therefore, in order to minimize the alteration of the endogenous promoter, it was necessary to establish a system in which the transcriptional activity of the SNURF-SNRPN gene 5â€™ region could be measured in the absence of a heterologous enhancer. A collection of reporter constructs was generated carrying varying size fragments of the PWS-SRO/SNURF-SNRPN 5â€™ flanking region in front of the firefly luciferase gene
41 (pGL3, Promega). Equal molar amounts of each construct were individually transfected into human SK-N-SH neuroblastoma cells and human HT1080 fibrosarcoma cells and were harvested 24h later. The activity of the firefly luciferase present in the cell lysate was assayed in a luminometer (Sirius Luminometer) as the amount of light emitted after addition of the appropriate substrate (see Materials and Methods). A constant amount of an internal control construct containing the Renilla Luciferase gene (pRL-TK, Promega) was cotransfected with the experimental construct in order to correct for the efficiency of the transfection. The activity of the Renilla luciferase was measured in a luminometer in the same sample as the firefly luciferase and used to normalize the experimental values. All reactions were performed in duplicate a number of times (indicated in Figure 2-1). Figure 2-1 shows the results for transient expression assays in SK-N-SH with reporter constructs that included different fragments of SNURF-SNRPN 5â€™ region (indicated on the left of the figure). Similar results were obtained for HT1080 cells. The largest fragment examined (a) included most of the PWS-SRO sequence. Constructs (a) -4000 to +80, (b) -676 to +160 and (c) -676 to +80, showed very small differences in reporter activity, which suggested that the regions between -4000 and -676 and between +80 and +160 do not include elements involved in SNURF-SNRPN promoter function as detected in transient expression assays. Further reduction of the 3â€™ end (construct (d) -676 to +53) induced a 1.5-fold increase in the luciferase activity. Even tough this difference was not statistically significant, these luciferase data implied that the region in exon 1 between +53 and +80 can act as a repressor of transcription. This interpretation agrees with a previous report that identified a similar function associated with that same region in the SNURF-SNRPN promoter by transient expression assays in the context of
42 012345678 Luciferase-676+80 Luciferase-207+53 Luciferase-676+53 Luciferase-366+53 Luciferase-207+80 Luciferase-676+165Relative Luciferase Activity(a)(b)(c)(d)(d)(e)(f) 6 5 7 4 8 3 2 1 0 Luciferase-3998+80// 012345678 Luciferase-676+80 Luciferase-207+53 Luciferase-676+53 Luciferase-366+53 Luciferase-207+80 Luciferase-676+165Relative Luciferase Activity(a)(b)(c)(d)(d)(e)(f) 6 5 7 4 8 3 2 1 0 Luciferase-3998+80 // Figure 2-1: Analysis of the SNURF-SNRPN promoter by transient transfection assays in SK-N-XX cells in the absence of a heterologous enhancer. Increasingly truncated portions of the SNURF-SNRPN 5â€™ region (depicted on the left) were cloned into a luciferase reporter construct. These constructs were assayed in transient expression assays in human SK-N-SH neuroblastoma cells. Relative luciferase activities are shown respect to the largest construct (a) which was arbitrarily assigned the value 1. Error bars represent the standard deviation of the means. the SV40 enhancer (Green Finberg et al., 2003). However, the increase in reporter gene activity that Green Finberg et al. observed upon deletion of this element from the SNURF-SNRPN promoter was substantially larger than the fold increase observed in Figure 2-1. This effect may be due to the artificial presence of the SV40 enhancer in the reporter construct. Shortening of the 5â€™ end in construct (e) -366 to +53 did not have a significant effect on promoter activity (compare constructs (d) and (e)). On the other hand, further shortening in construct (g) -207 to +53 resulted in two to three-fold increase of the luciferase activity with respect to constructs (d) and (e) indicating the existence of a second negative regulatory element between positions -366 and -207. This second and novel regulatory element was confirmed in construct (f) (-207 to +80) in which the first
43 negative regulatory element is included and the second negative regulatory element is absent. This construct showed a 2.5-fold increase in luciferase activity with respect to construct (c). This also indicated that the upstream and downstream repressors can act independently of each other. Since these elements modulate the SNURF-SNRPN gene transcription, it is likely that they are functional in the paternally-inherited active allele rather than in the maternally-inherited silent allele, where silencing is likely mediated by DNA methylation and modifications associated with repressed chromatin. Alternatively, it is also possible that the gradual increase in luciferase activity observed in promoter fragments progressively truncated at the 5â€™ and 3â€™ ends may be due to the effect of bringing a hypothetical cis-acting element in the vector itself in increasing proximity to the promoter. In summary, the experiment described above revealed promoter function to be associated with the SNURF-SNRPN 5â€™ region in the absence of an exogenous (SV40) enhancer, and identified two negative regulatory elements in that region. Analysis of Cis-Regulatory Elements in the SNURF-SNRPN 5â€™ Region/PWS-IC by In Vivo Footprint Analysis In vivo footprinting was used to perform high resolution identification of cis-acting elements in the 5â€™ flanking region of SNURF-SNRPN. Transient expression assays described in the previous section showed that sequences in the PWS-SRO upstream from -366 and downstream from +80 with respect to the SNURF-SNRPN transcription start site are not likely to include cis-acting regulatory elements involved in the promoter activity of the SNURF-SNRPN gene. Therefore, the analysis was limited to a region in 5â€™ SNURF-SNRPN that included NHS1 and the six phylogenically conserved sequences. Allele-specific in vivo DNA-protein interactions in that region were examined by
44 dimethylsulfate (DMS) and DNase I ligation-mediated PCR (LMPCR) in vivo footprint analysis. The maternally-inherited allele was analyzed in lymphoblasts derived from PWS patients with UPD or with a deletion of the imprinted domain. The paternally-inherited allele was analyzed in lymphoblasts derived from AS patients with UPD or the above mentioned deletion. These cells were kindly provided by Dr.Daniel J. Driscoll. In vivo footprint analysis is based on the use of a chemical or enzymatic agent upon intact cells to modify/cleave genomic DNA, which is protected from such modification/ cleavage at the sites where there are factors bound. To identify DNA-protein interactions in vivo the pattern of modification of genomic DNA treated within cells is compared to the pattern of modification of purified naked genomic DNA (purified genomic DNA striped of all protein factors). Figure 2-2 shows a schematic representation of the fundamentals of in vivo footprinting. Ligation-mediated PCR (LMPCR) was used to selectively amplify the genomic region interest after modification and cleavage. Different DNA modifying/cleavage agents can be used to identify sites of factor binding and to obtain information about the interaction with DNA. In this particular study dimethylsulfate (DMS) and DNase I have been used. DMS is a small hydrophobic chemical probe that methylates guanine residues at position N7 through the major groove and, at a lower frequency, adenine residues at position N3 through the minor groove. Subsequent incubation of DMS-modified DNA with a mild base results in preferential cleavage of the DNA at the modified guanine residues. Consequently, DMS footprinting reveals intimate contacts between guanines in the DNA and factors interacting with them, although other contacts through nucleotides other than guanines can not be detected. In
45 NakedDNA In vivoPrimerDMSNaked DNA PiperidineCleavage G G G G G G DMSIn vivo G G G G G G Ligation-MediatedPCR (LMPCR)NakedDNA In vivoPrimerDMSNaked DNA PiperidineCleavage G G G G G G DMSIn vivo G G G G G G Ligation-MediatedPCR (LMPCR) Figure 2-2: Fundamentals of in vivo footprinting. The panel on the right describes in vivo footprinting based on modification of genomic DNA with DMS, which diffuses into the cell passively and methylates guanine residues. Intimate contact of DNA binding factors with guanine nucleotides prevents that modification while nucleotides in proximity to hydrophobic pockets in the protein are modified at higher frequency. The panel on the left illustrates treatment with DMS of purified naked genomic DNA stripped of proteins and therefore modified at all the guanine residues. In both cases treatment with a mild base cleaves the DNA at the modified guanine residues generating a collection of nested fragments that end at each of the modified nucleotides. The fragments ending at protected nucleotides will be underrepresented in the in vivo DMS-treated sample compared to the DMS-treated naked DNA. The fragments ending at nucleotides modified at higher frequency will be overrepresented. Following LMPCR amplification, this pattern can be observed on a sequencing gel (at bottom of the cartoon) as bands missing or of reduced intensity (protected nucleotides) and band of enhanced intensity (nucleotides with enhanced reactivity to DMS) which are indicative of DNA-protein interactions. addition, DMS footprints can also be detected as increased modification of guanine residues with respect to naked DNA. It is believed that hydrophobic pockets within
46 proteins bound to the DNA accumulate DMS and, therefore, increase the rate of DMS modification of neighboring guanine residues. DNase I is a enzymatic probe that binds and cleaves DNA through the minor groove. DNA-protein interaction sites are protected from DNase I cleavage and are flanked by nucleotides of enhanced reactivity. In addition to identifying sites of DNA/protein interaction, DNase I footprinting can also identify rotational positioning of nucleosomes. Rotational phasing of nucleosomes has been postulated as a potential mechanism to regulate the accessibility of transcription factors to its binding sites (Pina et al., 1990). In rotationally phased nucleosomes the orientation of the double helix for a given DNA position with respect to the histone core is the same in all the cells in a population. As a result, on every turn of the helix DNase I preferentially accesses and cleaves the same site. This results in a favored periodicity of digestion every 10 nucleotides that can be visualized as a ladder of bands with an approximate spacing of 10 bases in a DNA sequencing gel. The in vivo footprint analysis of the SNURF-SNRPN 5â€™region was designed to include the SNURF-SNRPN minimal promoter as defined by Green Finberg et al. (Green Finberg et al., 2003), NHS1, and the six sequences phylogenetically conserved between the human and the mouse genes, that are located within 400 bp upstream from SNURF-SNRPN transcription start site. Figure 2-3 illustrates the position of these sequences as well as possible transcription factor binding sites. Also included are the relative position of the primers designed for LMPCR and the extent of the sequence analyzed with each primer set.
47 +150-490 AUG Exon1 NRF-1 E2F CTCF SP1 CTCF NRF-1 E2F CTCF SP1 SP1Not to scalePhylogeneticallyConserved Regions TY542TY652TY540TY641 TY551TY648TY528 TY760 SNURF-SNRPN +150-490 AUG Exon1 NRF-1 E2F CTCF SP1 CTCF NRF-1 E2F CTCF SP1 SP1Not to scalePhylogeneticallyConserved Regions TY542TY652TY540TY641 TY551TY648TY528 TY760 SNURF-SNRPN Figure 2-3: Schematic representation of the SNURF-SNRPN 5â€™ region analyzed by in vivo footprinting. The black box and bent arrow indicate the positions of the first exon and the transcription initiation site respectively. Also indicated are potential transcription factors binding sites (colored boxes) and sequences conserved between the human and the mouse loci (red brackets). The horizontal arrows specify the position of the overlapping primer sets used in the LMPCR and the extent of the sequence analyzed with each one. Primer sets above the line assayed the upper strand and primer sets below the line assayed the lower strand. Each region was analyzed in multiple gels from at least two separate experiments. Footprints were confirmed in two different cell lines for each allele: lymphoblasts derived from AS patients with UPD or deletion were used for analysis of the paternally-inherited allele, while lymphoblasts derived from PWS patients with UPD or deletion were used for the analysis of the maternally-inherited allele. Figure 2-4 shows representative results for the DMS in vivo footprint analysis. Shown are sequencing gels where the pattern of bands indicates the position of the guanine residues in the sequence. The intensity of the bands reflects the reactivity of the guanine residue to DMS treatment. Paternallyand maternally-inherited alleles were analyzed separately in the indicated cell lines derived from AS and PWS patients and in
48 vivo-treated genomic DNA was compared to purified naked genomic DNA in order to identify factors bound in vivo. Differences in band intensity between the in vivo treated DNA an the naked DNA had to fulfill the following criteria to be considered a footprint: 1the pattern had to be reproducible in all or most of the sequencing gels for each experiment; 2the pattern had to be specific to one allele and not present in both alleles; 3the pattern of increased or decreased intensity of the bands had to be in the context of other bands with constant intensity. Six DNA-protein interaction sites were identified exclusively on the paternally-inherited allele (designated P1-6) and one on the maternally-inherited allele (designated M6). The close circles in Figure 2-4 denote bands of decreased in intensity with respect to the naked DNA, and represent the protection of that particular guanine residue from DMS modification due to intimate DNA-protein contacts. The open circles denote bands of increased intensity with respect to the naked DNA, which reflect enhanced reactivity to DMS modification of that particular guanine residue. Open and close circles, therefore, indicate sequence-specific DNA-protein in vivo interactions or footprints. Figure 2-4A shows footprint P1, which was identified with primer set TY 527on the lower strand of the paternally-inherited allele at positions +56 and +58. P1 was detected by comparing the banding pattern from in vivo treated cells with that from naked DNA. This analysis revealed that the band corresponding to the guanine residue +56 on the paternally-inherited allele was protected, while the band corresponding to guanine residue +58 showed enhanced reactivity in the in vivo-treated sample with respect to naked DNA. This pattern was not observed on the maternally-inherited allele. Additionally, the bands above and below the bands corresponding to guanines +56 and
49 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 +58+56 P1A) -85M6 -86P6 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 -34 P4-27 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 B)-56-58P5 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 P2 -5P3 -13 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 E)D)C)Sec.Sec.Sec.Sec.Sec.+70 +46 -80 -91 -19 +6 -44 -38 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 06090120060901200609012006090120 +58+56 P1A) -85M6 -86P6 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 -85M6 -86P6 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 06090120060901200609012006090120 -34 P4-27 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 0609006090 B)-56-58P5 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 -56-58P5 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 Mat. allele(PWS-Del)Pat. allele(AS-Del) 0609006090 0609006090 P2 -5 -5P3 -13 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 0609012006090120 Maternal allele(PWS-UPD)Paternal allele(AS-UPD) 06090120060901200609012006090120 E)D)C)Sec.Sec.Sec.Sec.Sec.+70 +46 -80 -91 -19 +6 -44 -3 8 Figure 2-4: DMS in vivo footprint analysis of the SNURF-SNRPN 5â€™region. Maternal alleles were analyzed in lymphoblasts derived from PWS UPD and deletion (Del) patients. Paternal alleles were analyzed in lymphoblasts derived from AS UPD and deletion patients. Cells were treated with DMS for the indicated periods of time. Time 0 seconds indicates naked genomic DNA treated with DMS. Shown above are representative sequencing gels in which the parental origin of the allele is indicated. Similar results were obtained on the UPD and Del cell-lines for each paternal and maternal allele. Therefore only the results for one of either UPD or Del cell line are shown for each allele. In vivo footprinted sites on the paternal allele (P1-6) and on the maternal allele (M6) are indicated with close and open circles that represent protection or enhanced reactivity respectively of guanine residues in vivo. The numbers on the left and the right of each figure indicate position with respect to the transcription initiation site which is indicated by a bent arrow. The bracket indicates the position of a phyogeneticly conserved sequence. Identification of A) P1 on the lower strand with primer set TY 527; B) M6/P6 on the lower strand with primer set TY527; C) P2 on the upper strand analyzed with primer set TY 639; D) P3 on the upper strand with primer set TY 639; E) P4 on the upper strand with primer set TY 639; F) P5 on the upper strand with primer set TY 639.
50 +58, showed comparable intensities in the in vivo-treated samples compared to naked DNA for both paternallyand maternally-inherited alleles. A sequence around -85 and -86 showed footprints in both paternallyand maternally-inherited alleles with primer set TY 527on the lower strand as shown in Figure 2-4B. Footprint M6 was identified on the maternally-inherited allele at position -85 as a band of enhanced reactivity in the in vivo-treated DNA with respect to naked DNA. Intensity of the bands above and below did not show those differences. On the other hand, analysis of the paternally-inherited allele showed enhanced reactivity at position -86. This indicated the existence of a paternal-specific footprint (P6) in the paternally-inherited allele at approximately the same position that the maternal-specific footprint M6 occupied in the maternally-inherited allele. This footprinted site M6/P6 is likely to play different roles in the paternallyand maternally-inherited alleles, possibly by interacting with two different factors or with one factor in two a different conformations. On the transcriptionally inactive, heavily DNA methylated, maternally-inherited allele, M6 may participate in the silencing of the SNURF-SNRPN gene. Conversely, on the transcriptionally active paternally-inherited allele, P6 may play a role in SNURF-SNRPN promoter function. Figure 2-4B-E show footprints P2, P3, P4 and P5 respectively, which were detected on the upper strand with primer set TY 639. Footprints P2, P3 and P4 were identified as bands of increased intensity at positions -5 (P2), -13 (P3) and -34 (P4); footprint P5 at -56 and -58, which was also detected with primer set TY 539, was observed as the protection of both guanine residues with respect to naked DNA. Analysis of the SNURF-SNRPN 5â€™
51 [DNase I] -139-125-115-106-98-86-77-65-56-47-154-165 Maternal (PWS-UPD) Paternal (AS-UPD) DNACellsCellsDNA DNACellsCellsDNA Maternal (PWS-UPD) Paternal (AS-UPD)[DNase I] -55-63-75-85-97-103-111-117-131-145 TY526Lower StandTY539Upper Stand * *[DNase I] -139-125-115-106-98-86-77-65-56-47-154-165 Maternal (PWS-UPD) Paternal (AS-UPD) DNACellsCellsDNA Maternal (PWS-UPD) Paternal (AS-UPD) Maternal (PWS-UPD) Paternal (AS-UPD) DNACellsCellsDNA DNACellsCellsDNA DNACellsCellsDNA DNACellsCellsDNA DNACellsCellsDNA Maternal (PWS-UPD) Paternal (AS-UPD)[DNase I] -55-63-75-85-97-103-111-117-131-145 TY526Lower StandTY539Upper Stand * * Figure 2-5: DNase I in vivo footprint analysis of the SNURF-SNRPN 5â€™region. Maternal alleles were analyzed in lymphoblasts derived from PWS UPD and deletion (Del) patients. Paternal alleles were analyzed in lymphoblasts derived from AS UPD and deletion patients. Permeabilized cells were treated with increasing concentrations of DNase I. Shown above are representative sequencing gels in which the parental origin of the allele and the primer set used in the LMPCR amplification reaction are indicated. DNA refers to purified naked genomic DNA treated with DNase I, and cells refers cells treated with DNase I in vivo. Brackets indicate the position of DNase I footprints and asterisks indicate the position of previously identified DMS footprint M6. Numbers to the right of each gel denote positions with respect to the SNURF-SNRPN transcription initiation site. An approximately 10 bp ladder on the maternal allele from roughly nucleotide -140 to nucleotide -60 is indicated by the horizontal arrows to the left of the gel for primer set TY 526. flanking region with the remaining four primer sets identified no additional in vivo footprints but did detect a GA polymorphism at position -125. The maternal-specific site (M6) was confirmed on the upper and lower strand by DNase I in vivo footprinting (Figure 2-5). The sequence between -91 and -98 was
52 protected from DNase I cleavage as indicated by decreased intensity of that region which is flanked by two sites of enhanced reactivity. Site M6 was detected by DMS and DNase I in vivo footprinting, however, the position of the footprints does not perfectly coincide. This is not surprising since DMS and DNase I do not necessarily interact with DNA in the same manner. While DMS modifies guanines through the mayor groove, DNase I binds the double helix though the minor groove. Analysis of the paternally-inherited allele by DNase I footprinting yielded inconclusive results. This may be the result of the low level of expression of SNURF-SNRPN in lymphoblasts which may lead to low levels of transcription factor occupancy at the promoter of SNURF-SNRPN on the paternally-inherited allele. Further analysis of the pattern of DNase I cleavage revealed a 10 bp ladder on the maternally-inherited allele that extends at least 80 nucleotides or 8 turns of the helix from approximately -60 to -140 on the lower strand. The naked DNA and paternally-inherited allele also display the same 10 bp periodicity to some extent; however, the pattern is more regular and noticeable in the maternally-inherited allele. This DNase I cleavage pattern might be an indication of the rotationally phasing of a nucleosome over the SNURF-SNRPN promoter region. This phasing might participate in the silencing of the maternally-inherited allele of SNURF-SNRPN. On the other hand, on the paternally-inherited allele, which is nuclease hypersensitive and includes several in vivo footprints, a weaker 10 bp ladder may reflect the positioning of a non-canonical nucleosome that permits binding of transcription factors. The DNase I cleavage pattern is not apparent on the opposite strand as might be expected. However, similar patterns have been observed in vivo for PGK1 (Pfeifer and Riggs, 1991) and HPRT (Chen and Yang, 2001).
53 Footprint P5 is the only one to concur with one of the six sequences phylogeneticly conserved in the human and the mouse SNURF-SNRPN 5â€™ region. Additionally, Green Finberg et al. and Hershko et al. had already determined by transient expression assays the existence of a functional element associated with a sequence that includes footprint P5 in the human (Green Finberg et al., 2003) and the mouse (Hershko et al., 2001) SNURF-SNRPN promoter. However, this is the first evidence in vivo that a functional element is present at the endogenous locus. Analysis of transcription factor database (TRANSFAC) revealed a potential NRF-1 binding site associated with footprint P5. NRF-1 is a transcription factor that activates transcription of many genes involved in oxidative phosphorylation and metabolic pathways. Footprint P2 shows some sequence similarity to an initiator element (gcagagt, where the nucleotides in bold correspond to the initiator sequence). Initiator elements share a very loose consensus sequence: Py Py A+1 N T/A Py Py, where Py represents a pyrimidine (Javahery et al., 1994). Transcription factor database analysis identified P2 as a putative binding site for E2F, which is known to both activate and repress transcription. Footprint P1 lies within a previously reported negative regulatory element within SNURF-SNRPN exon 1 (Green Finberg et al., 2003). The investigators used gel retardation assays to show the formation of several DNA-protein complexes upon incubation of a 45 bp oligonucleotide with nuclear extract. Mutation of positions +56 and +60 abolished that interaction; however, the mutations were never tested functionally in the transient expression assay that had originally identified the repressor. Additionally, site P1 is located at the translation start site and transcription factor database analysis showed a weak similarity to an AP1 binding site.
54 Site P3 lies within one of three possible CTCF binding sites located in the SNURF-SNRPN 5â€™ region. The sequence shared by sites M6 and P6 lies within a second potential CTCF binding site that partially overlaps with a potential NRF-1 binding site. CTCF has been implicated in the regulation of the H19/IGF2 imprinted domain (Hark et al., 2000), where it binds the maternal, in that case, unmethylated allele and acts as an insulator. However, as will be shown later, in vivo chromatin immunoprecipitation analysis of the SNURF-SNRPN promoter region showed undetectable levels of CTCF at both parental alleles. No significant match was identified in the transcription factor database for the site P4. Figure 2-6 summarizes the footprinting data on the SNURF-SNRPN 5â€™ region as well as the analysis of transcription factors in the same region. SP1 SP1 SP1CTCFCTCFCTCF NRF-1NRF-1 E2FE2F ATGTACPaternal footprintMaternal footprintPhylogenetic boxes (human/mouse) SP1 SP1 SP1CTCFCTCFCTCF NRF-1NRF-1 E2FE2F ATGTAC ATGTACPaternal footprintMaternal footprintPhylogenetic boxes (human/mouse)
55 Figure 2-6: Summary of in vivo footprint analysis of the SNURF-SNRPN 5â€™ region. The nucleotide sequence corresponds to the SNURF-SNRPN 5â€™ region. The transcription start site is indicated by the bent arrow and the translation initiation site is indicated in red. The position of the DMS and DNase I in vivo footprints as well as the position of potential transcription factor binding sites and phylogenetically conserved sequences are shown. Examination of In Vivo Footprints by Transient Expression Assays The function of the cis-acting elements identified by in vivo footprinting was analyzed by transient expression assays in human SK-N-SH neuroblastoma cells. The reporter construct used in these assays included approximately 700bp of SNURF-SNRPN 5â€™region cloned into pGL3 (Promega) upstream from the luciferase gene. Mutations at the footprinted sites P1, P2, P4, P5 and M6/P6 were introduced individually by site directed mutagenesis (QuikChange XL Site-Directed Mutagenesis Kit, Stratagene) into the cloned SNURF-SNRPN promoter. Mutations in each of the factor binding sites were designed to replace at least the DMS-footprinted guanine residues that directly interact with DNA binding proteins and to incorporate a novel restriction site. This restriction site allowed for the screening of the mutated clones by restriction digestion that was then verified by direct sequencing. Additionally the mutant sequence was analyzed in a transcription factor database (TRANSFAC) in order to eliminate the introduction of new transcription factor binding sites in the SNURF-SNRPN promoter region. The mutant sequence for each of the in vivo footprinted sites is indicated next: P1: AATTCA; P2: ATGCAT; P4: GATATC; P5: AGTACT; M6/P6: CATATG AND M6/P6â€™: AATATT. Nucleotides in bold correspond to the mutant sequence. Site M6/P6 was mutated to two different sequences for reasons that will be explained later. The effect of the mutations on the levels of luciferase activity was analyzed. As shown in Figure 2-7, mutations of the paternal specific sites P2 (potential E2F binding
56 0.00.20.40.60.81.01.2Wild typeMutant P1Mutant P2Mutant P4Mutant P5Mutant M6/P6Relative Luciferase Activity Luciferase-676+80 1.2 0.6 0.4 0.8 1.0 0.2 0 0.00.20.40.60.81.01.2Wild typeMutant P1Mutant P2Mutant P4Mutant P5Mutant M6/P6Relative Luciferase Activity Luciferase-676+80 Luciferase-676+80 1.2 0.6 0.4 0.8 1.0 0.2 0 Figure 2-7: Functional analysis of in vivo footprints by transient expression assays. Luciferase reporter constructs including the SNURF-SNRPN promoter (depicted on the top of the figure) were mutated at each of the DMS in vivo footprints, P1, P2, P4, P5 and M6/P6. Reporter constructs were transiently transfected into in human SK-N-SH neuroblasoma cells. Relative luciferase activities are shown compared to an empty construct containing no promoter, and to a construct including the wild type unmutated promoter, which was arbitrarily assigned the value 1. Error bars represent standard deviations. site/initiator) and P5 (potential NRF-1 binding site) significantly reduced the expression of the reporter gene decreasing the luciferase activity to almost background levels (compare the mutated constructs with the empty construct). This result clearly indicates the involvement of these elements in transcription from the SNURF-SNRPN promoter. Similarly, Green Finberg et al., (Green Finberg et al., 2003) mutated and DNA-methylated a site that corresponds to P5 in the SNURF-SNRPN promoter in a reporter construct activated by an exogenous (SV40) enhancer. They found that both mutation and DNA methylation of this site abolished expression from the reporter gene. However, the analysis of the P5 mutation shown above is the first to establish a functional correlation between SNURF-SNRPN transcription and site P5 in the absence of the SV40
57 enhancer. Conversely, Huq et al. (Huq et al., 1997) analyzed a fragment of the SNURF-SNRPN 5â€™region from position +4 to +53 by transient expression assays in the presence of the SV40 enhancer. Despite the deletion of sites P2 and P5, which were shown in figure 2-7 to be essential for promoter function, the authors only observed a two-fold reduction in reporter activity compared to a construct containing the SNURF-SNRPN minimal promoter which included sites P2 and P5 (-207 to +53). These contradictory results may be explained by the artificial effect of the robust SV40 enhancer on the SNURF-SNRPN promoter. Mutation of the paternal-specific footprint P1 did not alter the expression levels of the reporter gene in these transient expression assays, however, a potential role for P1 in repressing the function of a downstream activator will be described in the next chapter. Mutation of the footprinted site P4 had no effect in SNURF-SNRPN promoter function. This suggests that the site P4 may play a role in chromatin structure or IC function rather than in SNURF-SNRPN promoter activity. Footprint P3 which resembles a CTCF binding site was not examined functionally in transient expression assays. Mutation of the M6/P6 footprinted site reduced the expression levels of the reporter gene by half. This decrease in luciferase activity was significantly less severe than the effect observed upon mutation of sites P2 or P5. To rule out the possibility that the mutation did not completely abolish binding of the associated factors, a second mutation was created at this site with the same outcome. This suggests that the sequence associated with sites M6/P6 may play a role in SNURF-SNRPN promoter activity. It is likely that such role in promoter function is associated with P6, which resides in the
58 transcriptionally active paternally-inherited allele, rather than with M6, which resides in the transcriptionally silent maternally-inherited allele.
CHAPTER 3 CHARATERIZATION OF NHS2: FUNCTIONAL IMPLICATION FOR THE IC Nuclease hypersensitivity assays are a standard method for identifying distant control regions including enhancers (Picard and Schaffner, 1984), locus control regions or LCRs (Talbot et al., 1989), silencers (Sawada et al., 1994), insulators (Udvardy et al., 1985) and matrix attachment regions (MARs) (Levy-Wilson and Fortier, 1989). The SNURF-SNRPN gene locus includes of two NHSâ€™s on the paternally-inherited allele that are absent from the maternally-inherited allele (Schweizer et al., 1999). NHS1 was examined in the previous chapter. NHS2 is located just downstream from the PWS-SRO within the first intron of the SNURF-SNRPN gene. Association of transcription factors with the SNURF-SNRPN promoter and transcription of the SNURF-SNRPN gene may contribute to the mechanism through which the PWS-IC regulates imprinted gene expression throughout the PWS-AS associated domain during gametogenesis and/or early embryo development (Nicholls et al., 1998). Therefore, cis-acting elements involved in regulating transcription of the SNURF-SNRPN gene could potentially play a role in the PWS-IC function. Considering the proximity of NSH2 to the SNURF-SNRPN promoter and to the PWS-SRO, it is possible that NHS2 acts as a modulator of the transcription of the SNURF-SNRPN gene and, by extension, of PWS-IC function. Consequently, we analyzed the effect of NSH2 on SNURF-SNRPN transcription. 59
60 Identification of an Activator Function Associated with NHS2 NHS2 is located roughly 1.5 to 1.7 kb downstream from the SNURF-SNRPN transcription start site (Schweizer et al., 1999). In the studies that are described below, a large 2.2 kb DNA fragment from +624 to +2888 with respect to the SNURF-SNRPN transcription initiation site was analyzed. This fragment, termed 2.2-NHS2, included NHS2 and flanking sequences. The flanking sequences were included in order to ensure that all the elements within the potential control region identified by nuclease hypersensitivity were examined. 2.2-NHS2 lies outside the PWS-SRO, which was defined by microdeletions in PWS patients. The PWS-SRO has been proposed to be an essential component of the PWS-IC function. However, this does not necessarily exclude that other regions outside the PWS-SRO, such as NHS2, may also play a role in the PWS-IC function. There are several reports of allele specific DNA methylation and histone modification patterns in the sequence included in 2.2-NHS2, which is suggestive of function. Most of the sequence in 2.2-NHS2 is located outside from the SNURF-SNRPN 5â€™ CpG island. However, analysis with methyl-sensitive restriction enzymes revealed maternal-specific DNA methylation in a group of 4 CpG dinucleotides at the 5â€™ region of 2.2-NHS2, approximately 1 kb downstream from the SNURF-SNRPN transcription initiation site (Glenn et al., 1996). On the other hand, a single Sma I site located at the 3â€™ end of 2.2-NHS2, approximately 2.9 kb downstream of SNURF-SNRPN transcription initiation site, showed paternal-specific DNA methylation (Schweizer et al., 1999). Histone methylation at H3-K4 in the paternally-inherited allele and H3-K9 in the maternally-inherited allele, have been found to be limited to the SNURF-SNRPN promoter region, although only 1 kb of sequence on each side of exon 1 has been analyzed (Xin et al., 2001). On the other hand, histone H3 and H4 hyperacetylation was
61 found associated with the paternally-inherited allele of a region included in 5â€™ end of 2.2-NHS2 (Fulmer-Smentek and Francke, 2001) (Saitoh and Wada, 2000). The role of 2.2-NHS2 in SNURF-SNRPN transcription was analyzed in transient expression assays using the system described in the previous chapter. 2.2-NHS2 was introduced into a reporter construct in which the firefly luciferase gene was driven by either one of two fragments of the SNURF-SNRPN promoter (from -760 to +80 or -4000 to +80). These constructs were transiently transfected into human neuroblastoma SK-N-SH cells. Transfected cells were lysed 24 h later and the activity of the luciferase gene product determined. Figure 3-1A shows relative luciferase activities for reporter constructs including 2.2-NHS2 at different positions and orientations (constructs (b)-(e)), with respect to a construct including approximately 700 bp of SNURF-SNRPN promoter but no 2.2-NHS2 (construct (a)). Reporter constructs (b) in which 2.2-NSH2 was cloned downstream from the luciferase gene in the forward orientation showed an 8 fold induction in the relative reporter activity compared to construct (a). This suggested that 2.2-NSH2 can act as an activator of transcription. However, reversing the orientation of 2.2-NHS2 drastically reduced the level of activation, suggesting that the activation function associated with 2.2-NHS2 may dependent on the orientation respect to the promoter. Alternatively, since the elements responsible for the activation function in 2.2-NHS2 may be close to one end of the fragment, it is also possible to interpret the difference between constructs (b) and (c) as a distance effect. Next, 2.2-NHS2 was introduced upstream from the SNURF-SNRPN promoter in the forward (construct (d)) and the reverse (construct (e)) orientations. This resulted in an approximately 6-fold and
62 Luc -3998+80// Luc -3998+80// A) 0123456789101112 11 6 5 7 4 9 8 3 12 10 2 1 0(a)(b)(c)(d)(e) Luc-676+80 Luc-676+80 Luc-676+80 Luc-676+80 Luc -676+80B) 01234567891011121314 12 6 4 8 14 10 2 0Relative Luciferase Activity(a)(b)(c)(d)(e)2.2NHS2 2.2NHS22.2NHS22.2NHS22.2NHS2 2.2NHS22.2NHS22.2NHS2 Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// A) 0123456789101112 11 6 5 7 4 9 8 3 12 10 2 1 0(a)(b)(c)(d)(e) Luc-676+80 Luc-676+80 Luc-676+80 Luc-676+80 Luc -676+80B) 01234567891011121314 12 6 4 8 14 10 2 0Relative Luciferase Activity(a)(b)(c)(d)(e)2.2NHS2 2.2NHS22.2NHS22.2NHS22.2NHS2 2.2NHS22.2NHS22.2NHS2 Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Luc -3998+80// Figure 3-1: Identification of an activator function associated with 2.2-NHS2. luciferase reporter constructs including 2.2 kb of intronic sequence encompassing HS2 (grey arrow) at different positions and orientations with respect to the SNURF-SNRPN promoter (black line), were assayed by transient expression assays in human neuroblastoma cells. Relative luciferase activities are shown in each case compared to a construct containing only the promoter, which was arbitrarily assigned the value 1. Error bars represent standard deviations. a) Analysis of reporter constructs including approximately 750 bp of the SNURF-SNRPN 5â€™ region. b) Analysis of reporter constructs including approximately 4 kb of upstream sequence encompassing most of the PWS-SRO was included in the reporter construct. 8-fold induction of the reporter activity, respectively, which suggested that the activator function associated with 2.2-NHS2 may be independent of the position with respect to the promoter. The effect of reversing the orientation of 2.2-NHS2 in constructs (d) and (e), where 2.2-NHS2 is located approximately 500 bp upstream from the promoter, was less dramatic than in constructs (b) and (c), where 2.2-NHS2 is located approximately 2 kb
63 downstream from the promoter. This finding was more consistent with a distance dependence effect rather than with an orientation dependence effect on the activator function. This conflict was further investigated in Figure 3-1B by using reporter constructs in which the promoter consisted of a 4 kb fragment of SNURF-SNRPN 5â€™ sequence. Constructs (b) and (c), included 2.2-NHS2 downstream from the luciferase gene in the forward and reverse orientations, respectively. These constructs behaved in a similar manner to constructs (b) and (c) in Figure 3-1A. However, constructs (d) and (e) in which 2.2-NHS2 was positioned upstream approximately 4.5 kb from the SNURF-SNRPN promoter showed a significant decrease in reporter activity with respect to constructs (d) and (e) in Figure 3-1A. This effect is independent of the orientation of 2.2-NHS2, which is consistent with a distance dependence rather with an orientation dependence of the activation function associated with 2.2-NHS2. The existence of a functional interaction between the activator associated with 2.2-NHS2 and specific cis-acting elements in the SNURF-SNRPN promoter was examined next by transient expression assays. The previous chapter describes the identification of several cis-acting elements in the 5â€™ flanking region of the SNURF-SNRPN gene by in vivo footprint analysis (sites P1 through M6/P6, see Figure 2-4). Some of these footprinted sites (P2, P4, M6/P6) were shown to participate in SNURF-SNRPN promoter function. However, sites P1 and P5 did not seem to play a role in transcription in transient expression assays (see Figure 2-6). The role of these footprints in the context of the activation function within 2.2-NHS2 was analyzed by transient expression assays in SK-N-SH cells. For that purpose, 2.2-NHS2 was cloned into reporter constructs that included approximately 700 bp of SNURF-SNRPN 5â€™ region
64 02468101214Wild type promoterMutation P1Mutation P2Mutation P4Mutation P5Mutation M6/P6No activator Luciferase-676+80HS2 12 6 4 8 14 10 2 0Relative Luciferase Activity 02468101214Wild type promoterMutation P1Mutation P2Mutation P4Mutation P5Mutation M6/P6No activator Luciferase-676+80HS2 12 6 4 8 14 10 2 0 12 6 4 8 14 10 2 0Relative Luciferase Activity Figure 3-2: Functional interaction between the activator associated with 2.2-NHS2 and in vivo footprints of the SNURF-SNRPN gene promoter. This analysis was performed by transient expression assays. Luciferase reporter constructs included approximately 700 of SNURF-SNRPN promoter (black line) and 2.2-NHS2 (grey line). The SNURF-SNRPN promoter was individually mutated at each of the indicated DMS in vivo footprints. Relative luciferase activities are shown compared to a construct containing a wild type SNURF-SNRPN promoter and lacking 2.2-NHS2 (No Activator construct) and a construct including a wild type SNURF-SNRPN promoter and 2.2-NHS2. Error bars represent standard deviation of the means. (construct (e) in Figure 3-2)in which sites P1, P2, P4, P5 and M6/P6 had been individually mutated (see Figure 2-6). Figure 3-2 shows the relative luciferase activity for these constructs. Mutation of site P1, which had no effect in the absence of 2.2-NHS2, resulted in an increase in reporter activity compared to the wild type promoter sequence, when 2.2-NHS2 was present. This suggested that this element may be involved in repressing the SNURF-SNRPN promoter response to 2.2-NHS2 activation function. This is consistent with a published report in which Green Finberg et al. (Green Finberg et al., 2003) analyzed the SNURF-SNRPN promoter in the presence of the SV40
65 enhancer by transient expression assays. They identified a repressor function associated with a small region in the SNURF-SNRPN first exon that contained the footprinted site P1. It becomes clear now that the repressor identified by Green Finberg et al. was regulating the activation of the SNURF-SNRPN by the SV40 enhancer. On the other hand, mutation of site P2 (potential E2F binding site), site P4 and site M6/P6 (potential NRF-1/CTCF binding site) had a relative effect comparable to that observed in the absence of 2.2-NHS2 (compare Figure 3-2 with Figure 2-6). However, the reduction in transcription that is observed upon mutation of site P5 in the absence of 2.2-NHS2, is partially relieved in the presence of 2.2-NHS2. This suggests that that the function associated with site P5, which is a potential NRF-1/CTCF binding site, may be partially contributed to by elements within 2.2-NHS2. Site P3 was not tested with the activator. Specificity of the Activator Function Associated with 2.2-NSH2 The ability of 2.2-NHS2 to activate transcription from promoters of other genes in the cluster was examined next by transient expression assays in SK-N-SH cells. Reporter constructs included the promoter region of the indicated genes upstream from the luciferase gene and 2.2-NHS2 downstream from the luciferase gene. The promoters analyzed in these constructs included: 1-the human SNURF-SNRPN gene promoter; 2-the SNURF-SNRPN upstream promoters U1A and U1B; 3-the mouse Snurf-Snrpn gene promoter; 4-the Snurf-Snrpn upstream promoters U1-A and U1-C; 5-the MKRN3 gene promoter; 6-the UBE3A gene promoter (Figure 3-3). Figure 3-3 shows that the activation of expression of the human SNURF-SNRPN upstream promoters U1A and U1B induced by 2.2-NHS2 was almost as strong as the
66 Human upstreampromotersMouse upstreampromoters Human SNURF-SNRPNU1BU1AMouse Snurf-snrpnU1-CU1-AMKRN3UBE3A1.92.03.63.220.127.116.11.1Relative Luciferase Activity 22 24 26 12 10 14 8 18 16 6 28 20 4 2 0 Promoter Promoter+HS2 Luciferase 2.2-NHS2 PromoterHuman upstreampromotersMouse upstreampromoters Human SNURF-SNRPNU1BU1AMouse Snurf-snrpnU1-CU1-AMKRN3UBE3A1.92.03.63.18.104.22.168.1Relative Luciferase Activity 22 24 26 12 10 14 8 18 16 6 28 20 4 2 0 Promoter Promoter+HS2 Luciferase 2.2-NHS2 Promoter Luciferase 2.2-NHS2 Promoter Figure 3-3: Preferential activation of the SNURF-SNRPN promoter by the activator function associated with 2.2-NHS2. The effect of the intronic activator on the promoter of various genes in the PWS-AS associated domain was examined by transient expression assays in human SK-H-NH neuroblastoma cells. Black bars represent the relative luciferase activity for reporter constructs containing the indicated promoters and lacking 2.2-NHS2. The construct containing the human SNURF-SNRPN promoter without 2.2-NHS2 was arbitrarily assigned the value 1. Grey bars represent the relative luciferase activity for reporter constructs containing the promoter regions of the indicated genes and 2.2-NHS2. The fold increase induced by the activator on each promoter is indicated on the right margin of the chart. Error bars represent standard deviations. activation of expression of the human SNURF-SNRPN main promoter induced by 2.2-NHS2. Interestingly, the upstream promoters U1A and U1B seemed to be as active, if not more so, than the SNURF-SNRPN promoter. This was observed despite the lack of expression of the upstream transcripts in most tissues and the low abundance of the SNURF-SNRPN upstream transcripts in the tissues where they are expressed. The Snurf-Snrpn promoter in the mouse was activated by 2.2-NHS2 as well, suggesting some degree of conservation between the human and mouse promoters in response to the
67 activator function. Furthermore, the upstream Snurf-Snrpn promoters in mouse U1-Aand U1-C also showed activation by 2.2-NHS2, although the levels of expression fromthese promoters were considerably lower (10 fold for U1-C and 6 fold for U1-A) than that of the Snurf-Snrpn main promoter in the mouse gene. These lower levels of expression from the upstream promoters in mouse are consistent with the endogensituation, which contrasts to what was observed for the SNURF-SNRPN main and upstream promoters in the human. This may be a consequence of the intrinsic weaof the human SNURF-SNRPN promoter, which requires the activation function associatedwith 2.2-NHS2. MKRN3 and ous kness UBE3A are located in the PWS-AS imprinted domain and are expre the r ction The 2lanking sequenon ssed from the paternallyand maternally-inherited alleles, respectively. BothUBE3A and the MKRN3 promoters showed in a higher level of basal activity that the SNURF-SNRPN promoter in the transient expression system and it is interesting to note that the UBE3A promoter was able to support transcription in the reverse orientation to the same level as in the forward orientation. MKRN3 and UBE3A promoters were only weakly activated by 2.2-NHS2, suggesting that the activator function has a preference fothe SNURF-SNRPN main and upstream promoters in the human and the mouse genes. Interestingly, 2.2-NHS2 which functions as an activator of some paternally-expressed genes did not repress UBE3A, which is a maternally-expressed gene. Identification of Elements Mediating the Activator Fun .2 kb region in 2.2-NHS2 included NHS2 and a large portion of f ce. In order to narrow down the position of the elements involved in the activatifunction, systematic deletions of the 2.2-NHS2 sequence were analyzed by transient
68 Figure 3-4: Identification of cis-acting elements involved in the activator function associated with 2.2-NHS2: definition of a minimal activator sequence (MAS). Luciferase reporter constructs including ~700 bp of the SNURF-SNRPN promoter region and the indicated subfragments of intronic sequence (depicted on the top of the figure) were analyzed in transient expression assays in human SK-N-SH neuroblastoma cells. A restriction map of the 2.2-NHS2 region is shown on the top left corner and the size and position of each subfragment are described below the restriction map. Enzymes: E-EcoR I, N-Not I, B-Bgl II, T-Taq I, S-Sty I, H-Hpa I, Sâ€™-Sma I. An R next to the construct N-T indicates that the subfragment was introduced in the reverse orientation. Numbers refer to the position respect to SNURF-SNRPN transcription initiation site. Relative luciferase activities are shown compared to a reporter construct containing the SNURF-SNRPN promoter and lacking 2.2-NHS2. The stars mark a region that is common in all of the intronic fragments that show a significant activation function, termed minimal activator sequence or MAS. Error bars indicate standard deviations. expression assays. These experiments defined a minimal activator sequence (MAS), a 80 bp region flanked by Taq I and Bgl II restriction sites that retained a significant portion of the activation function. R No Activator Luciferase 2.2-NHS2 Subfragments E+624N+881B+1179T+1270B+1347S +1499// H+2093// S+2880 E-NE-TE-SE-BN-SB-BT-SE-HB-SS-SH-SE-S 01234567891011 SNURF-SNRPNPromoter Relative Luciferase Activity 11 6 5 7 4 9 8 3 10 2 1 0 R No Activator Luciferase 2.2-NHS2 Subfragments E+624N+881B+1179T+1270B+1347S +1499// H+2093// S+2880 E-NE-TE-SE-BN-SB-BT-SE-HB-SS-SH-SE-S 01234567891011 SNURF-SNRPNPromoter Relative Luciferase Activity 11 6 5 7 4 9 8 3 10 2 1 0
69 Reporter constructs used in the transient expression assays included approximately 700 bp of SNURF-SNRPN 5â€™ region upstream from the luciferase gene and subfragment of 2.2-NHS2 downstream from the luciferase gene. As shown in Figure 3-4, the truncated constructs E-H, E-Sâ€™, N-Sâ€™, T-Sâ€™ and B-B retained 50% or more of the activation potential associated with the full length 2.2-NHS2. All these fragments share a approximately 80 bp common region that is bracketed between the Taq I and Bgl II restriction sites and is marked with a star in the figure. This region was termed minimal activator sequence (MAS) and it colocalizes with the approximate position of NHS2 in the first intron of the SNURF-SNRPN gene. For other fragments there were none or significantly less activation function remaining and, therefore, were not investigated further. In order to identify conserved sequences between the 2.2-NHS2 region in human and the homwith the enThis analysis id approximaSignificanthe approximafactor databive binding site MAS recognized in the transcription factor search. ologous region in mouse, a sequence comparison of the entire 2.2-NHS2 tire sequence for first intron of the Snurf-Snrpn gene was carried out. entified a single region that showed significant conservation (Figure 3.4). Thistely 80 bp conserved region was termed activator conserved sequence (ACS). ly, the ACS extensively overlaps with the MAS and is also coincident with tte location of NHS2 in SNURF-SNRPN first intron. Analysis of transcription ase (TRANSFAC) sites within the conserved ACS/NHS2 indicated putates for SP1 and YY1 within the portion of the ACS that overlaps with th and for NRF-1 immediately downstream of the Bgl II restriction site. In addition, there is another highly conserved 9 bp element adjacent to the YY1 putative binding site (see Figure 3-6) which was no
70 Figure 3-5: Blast algor ithm based alignment of the human 2.2-NHS2 region and the full length first intron in the mouse Snurf-Snrpn gene. The only region of significant sequence conservation between human (H) and mouse (M) sequences is shown above together w ith the partially overlapping minimal activator sequence (MAS) flanked by Taq I and Bgl II restriction sites. The position of potential transcri ption factor binding sites identified by analysis of transcription factor database (T RANSFAC) is indicated. YY1 is a ubiquitous factor that has been de scribed to participate in gene activation and gene repression both by interacting with promoter (including with initiators elements) and with distal elements. It has al so been shown to act as an insulator and a nuclear matrix attachment factor. The ro le of the putative YY1 binding site in the activation function associated with 2.2-NHS2 was examined by introducing a mutation in that sequence by site directed mutagenesis in a luciferase reporte r construct including approximately 700 bp of SNURF-SNRPN 5â€™ region and 2.2-NHS2 (Figure 3-1A, construct (b) ). A 4 fold reduction in the expressi on levels was observed in the mutated construct, which suggested that the putati ve YY1 binding site in the ACS/NHS2 is a major component of the activator function. H 1269 tcgaaataacctg | | || | | M1798 ttggtattatcaa H 1282 ggggtc tgtgttgc--gtca-ctgccattgtgca -gctcct-t |||||| |||| || |||| | ||||||||| | | | || | M 1811 ggggtcgtgtcgcatgtcaaa c-gccattgtg-a ggat-ctgt H 1336 c aagatggccgc cgctgcagcg--gcttagatct gcgcaagcgc |||||||||||| | ||||| | ||| || |||||| |||| M 1867 a aagatggccgc tggtgcagaggtgct--gaagt gcgcaggcgcYY1NRF-1 SP1 ACSTaqI BglIIH 1269 tcgaaataacctg | | || | | M1798 ttggtattatcaa H 1282 ggggtc tgtgttgc--gtca-ctgccattgtgca -gctcct-t |||||| |||| || |||| | ||||||||| | | | || | M 1811 ggggtcgtgtcgcatgtcaaa c-gccattgtg-a ggat-ctgt H 1336 c aagatggccgc cgctgcagcg--gcttagatct gcgcaagcgc |||||||||||| | ||||| | ||| || |||||| |||| M 1867 a aagatggccgc tggtgcagaggtgct--gaagt gcgcaggcgcYY1NRF-1 SP1 ACSTaqI BglII
71 Figure 3-6: Functional analysis of a putative YY1 binding site in 2.2-NHS2 by transient expression assays. Luciferase reporter constructs included ~700 bp of SNURF-SNRPN promoter region (black line) and 2.2-NHS2 (grey arrow). Constructs were transfected into human SK-N-SH neuroblastoma cells. The putative YY1 binding site in 2.2-NHS2 was mutated (indicated by the 0123456789 Luciferase Luciferase-676+80HS2-676+80 Luciferase-676+80HS2*Relative Luciferase Activit y 0123456789 Luciferase Luciferase-676+80HS2-676+80 Luciferase-676+80HS2*Relative Luciferase Activit y 6 5 7 4 9 8 3 2 1 0 6 5 7 4 9 8 3 2 1 0 asterisk) and the relative luciferase activity compared 1st to a construct S2. In VThe ain the ACS/NHS2 was ading he e n was 2.2-dition, a non-specific competitor (HK-ATPase intronic oligonucleotide) did not compete (lane 5). including the SNURF-SNRPN promoter and lacking 2.2-NHS2 and 2nd to a construct including the SNURF-SNRPN promoter and wild type 2.2-NHError bars represent standard deviations. itro binding of YY1 to the Putative YY1 Binding Site in ACS/NHS2 bility of YY1 to interact with the putative YY1 binding site nalyzed next. Crude nuclear extract from human SK-N-SH neuroblastoma cells was incubated with a radioactively labeled probe containing the putative YY1 binsequence in electrophoretic mobility shift assays (EMSA). As shown in Figure 3-4, one prominent and several weak shifted bands were detected (lane 1), which indicates tformation of DNA-protein complexes. These bands were readily competed with a 30 fold excess of cold probe (lane 2) or with an unrelated oligonucleotide that included thYY1 consensus binding sequence (aagatggccgcc; lane 4). However, no competitioobserved with an oligonucleotide similar in sequence to the probe except for a mutation at the putative YY1 binding sequence that also disrupts the activation function of the NHS2 as shown in figure 3-6 (lane 3). In ad
72 Figure 3-7YY1 binding site in the ACS-NHS2. site ract with cold probe (lane 2), probe mutated at the YY1 binding site (lane 3), ed oligonucleotide (lane 5). Lanes 6-9, supershift analysis with antibodies preincubation, lane 8-simultaneous incubation, lane 9-postincubation of the 10 and 11, incubation of the probe with purified YY1 protein. YY1-b (Santacruz, sc1703), were used in supershift assays to confirm the identity of the preincubation with the nuclear extract before addition of the probe resulted in the pre-incubation (lane 7), simultaneous incubation (lane 8) and post-incubation (lane 9) of f all the bands except for : Gel mobility-shift assays of the putative C-2 C-3C-4C-5 C-2 C-3C-4C-5 C-2 C-3C-4C-5 CompetirorYY1 AbHis-YY1------++++---bbba-----+++---------1110987654321Super-shiftsC-1 CompetirorYY1 AbHis-YY1------++++---bbba-----+++---------1110987654321Super-shiftsC-1 CompetirorYY1 AbHis-YY1------++++---bbba-----+++---------1110987654321CompetirorYY1 AbHis-YY1------++++---bbba-----+++---------1110987654321------++++---bbba-----+++---------1110987654321Super-shiftsC-1 Lane 1, a radiolabeled oligonucleotide probe including the YY1 binding in the ACS-NHS2 and flanking sequence was incubated with nuclear ext from human SK-N-SH neuroblastoma cells. Lanes 2-5, competition reactions oligonucleotide containing the consensus YY1 binding site (lane 4), unrelatspecific for YY1 DNA binding domain (lane 6) and full length YY1 (lane 7-Ab and the nuclear extract with respect to addition of the DNA probe). LanesTwo different antibodies against YY1, anti-YY1-a (Santa Cruz, sc281) and anti-factor bound to the probe. Anti-YY1-a recognizes the DNA binding domain of YY1 andelimination the bands associated with complexes 1-4 and the intensification of the band associated with complex 5 (lane 6). Anti-YY1-b recognizes the full length protein and the antibody with the nuclear extract resulted in a supershift o
73 the band associated with complex 5. Furthermore, purified recombinant YY1 (kindly provided by Dr. Edward Seto) yielded the same gel shift pattern than the crude nuclear extract upon incubation with the radiolabeled probe (lane 10 includes lower amounts of recombinant YY1; lane 11 included higher amounts of recombinant YY1). The band associated with complex 5 that was not competed with antiYY1-a or supershifted with anti-YY1-b, was present after incubation of the probe with purified YY1. This indicated that complex 5 represented, nonetheless, a DNA-YY1 complex. However, the reason why this complex was not affected by the antibodies is unclear. Altogether, the EMSA analysis showed that YY1 does bind to the presumptive YY1 binding site in the CS/NHS2 in vitro. BindACS/NHS2technique aDNA. In aposttrn p s A Chromatin Immunoprecopitation (ChIP) Analysis of Factors Associated with SNURF-SNRPN 5â€™ Region/NHS1 and ACS/NHS2 ing of YY1, NRF-1 and CTCF to SNURF-SNRPN 5â€™ region/NHS1 and in vivo was analyzed by chromatin immunoprecipitation (ChIP) assays. This llows the in vivo detection of factors interacting with specific regions in the ddition, ChIP assays are sensitive enough to distinguish between specific anslational modifications associated with those factors. These assays are based othe use of a crosslinking agent to â€œfreezeâ€ all the protein-protein-DNA interactions occurring within a cell at the time of the crosslinking. This is followed by isolation of the crosslinked chromatin and shearing by sonication to an average size between 200 b(mononucleosomes) and 1000 bp depending on the application. The sheared chromatin ithen immunoprecipitated with an antibody raised against a specific factor. After reversing the crosslinking, the presence of the genomic region of interest in the immunoprecipitate is tested by PCR, dot blot or other methods for DNA detection.
74 Alternatively, instead of sonicating the chromatin, cells can be treated with micrococalnuclease (MNase). The in vivo MNase digestion of DNA is followed by crosslinking. In order to distinguish between SNURF-SNRPN paternallyand maternally-inherited alleles, lymphoblast derived from AS and PWS patients, containing either the paternal or the maternal allele, respectively, were used in the ChIP assays. The immunoprecipitated DNA was analyzed by PCR. The PCR products were size-fractionated by polyacrylamide gel electrophoresis (PAGE), stained with SyBr green, avisualized and quantitated in a fluorescence scanner. At least two independent immunoprecipitation (IP) reactions were performed with each antibody and each reaction was analyzed nd IP with at least two independent PCR reactions. no antibody was added, was performed in each experiment. The precipitated fraction of the DNA in this sample was used to control for non-specific precipitation of the DNA region of interest. The non-precipitated fraction of the DNA in that sample was considered as an input DNA control. The input DNA control was used to normalize across experiments the amount of DNA immunoprecipitated with the various antibodies in the different cell lines. For some genomic regions, the PCR reaction was more efficient on the DNA immunoprecipitated from the PWS-derived cell line than on the DNA immunoprecipitated from the AS-derived cell line. Although PCR reactions were performed within the linear range (see Material and Methods), these differences in amplification efficiency may result in a shift from the linearity in the amplification reaction of the immunoprecipitated DNA from the PWS cell line. Therefore, for each experiment and each region, the amount of input DNA control from ASand PWSSeveral controls were included in the ChIP analysis. A mock IP reaction in which
75 derived cell lines used in the PCR reaction, was corrected in order to yield the same amount of amplified product. That same correction was then applied to the antibodyspecifted with ificant e st is reds of base pairs away ol arates r ed nner ic immunoprecipitated DNA in each experiment and for each region. Shearing the chromatin by sonication resulted in DNA fragments with an average size of 500 to 1000 bp. Therefore, some of the DNA fragments immunoprecipitaan antibody for a given factor binding to a specific DNA sequence included signportions of neighboring DNA that was not associated with that factor. In those cases, PCR-amplification of the immunoprecipitated DNA with primers specific for thneighboring sequence would result in a PCR product even though the factor of interenot associated with that region and the actual binding sequence is hund . This effect will be referred to as â€œamplification of neighboring regionsâ€. Since theregions amplified in the SNURF-SNRPN 5â€™ region is only 1.3 kb upstream from ACS/NSH2, controls were generated to account for the amplification of neighboring regions, in which the amplification of one of these two regions may occur after immunoprecipitation with antibodies for factors that bind the other region. The contrfor SNURF-SNRPN 5â€™ region consisted of a region (1.3 kb upstream or 1.3 kb us) locatedupstream from SNURF-SNRPN 5â€™ region, and at the same distance (1.3 kb) that septhe regions amplified in SNURF-SNRPN 5â€™ region and the ACS/NHS2. The control foACS/NHS2 consisted of a region (1.3 kb downstream or 1.3 kb ds) located 1.3 kb downstream from ACS/NHS2. Figure 3-8 shows the results for the allele specific ChIP analysis of the indicatregions using antibodies against YY1, NRF-1 and CTCF. The top panel shows representative gels. The intensity of the bands was determined in a fluorescence sca
76 and the values for each immunoprecipitated DNA sample were normalized with respto the input DNA controls. The average and standard deviation of the normalized valufor each antibody and region were calculated and plotted in the bottom panel ect es of Figure 3-8. Chd a , was me d band , n lymphoblasts, does not include a clear YY1 binding site. e IP analysis showed in vivo clear association of YY1 with the ACS/NHS2. This association was only detected in the AS-derived cell-line indicating that the association of YY1 with the ACS/NHS2 occurs specifically in the paternally-inherited chromosome annot on the maternally-inherited allele. As a positive control for IP with YY1 antibody,region 5â€™ of the glucocorticoid receptor gene, which is known to interact with YY1analyzed. Association of this region with YY1 was detected in both ASand PWS-derived cell lines (data not shown). A weak band representing a possible interaction of YY1 with the SNURF-SNRPN 5â€™ region was observed in AS-derived cells. However, another band of similar intensity was detected in 1.3-ds, which is located at the sadistance from ACS/NHS2 than 5â€™ SNURF-SNRPN and controls for the effect of amplification of neighboring regions (see above). Therefore, the abovementionecorresponding to the SNURF-SNRPN 5â€™ region was considered unlikely to represent an in vivo interaction of YY1 with this region. In addition, a very weak band representing a possible interaction of YY1 with the SNURF-SNRPN upstream promoter U1A was detected in the paternallyinherited allele. However, this band was also detected for the maternally-inherited allele in the PWS-derived cell line. The U1A promoter regionwhich is not active in either allele i Therefore, it is likely that the observed PCR products at the U1A promoter were th
77 Maternal allele(PWS)Input No Ab YY1 NRF-1 Input No Ab CTCF Paternal allele(AS)Input No Ab YY1 NRF-1 Input No Ab CTCF 1.3 kb downstream ASC/NHS2 UIA 1.3 kb upstream SNURF-SNRPN ACS/NHS2 SNURF-SNRPN 5â€™region DM1 Maternal allele(PWS)Input No Ab YY1 NRF-1 Input No Ab CTCF Paternal allele(AS)Input No Ab YY1 NRF-1 Input No Ab CTCF Input No Ab CTCF 1.3 kb downstream ASC/NHS2 UIA UIA 1.3 kb upstream SNURF-SNRPN 1.3 kb upstream SNURF-SNRPN ACS/NHS2 SNURF-SNRPN 5â€™region ACS/NHS2 SNURF-SNRPN 5â€™region DM1 DM1 AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 5' SNRPN ACSNHS2 1.3 kb us 1.3 kb ds U1A AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 5' SNRPN ACSNHS2 1.3 kb us 1.3 kb ds U1A 1 0AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In 5' SNRPN ACSNHS2 1.3 kb us 1.3 kb ds U1A1 0AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In 5' SNRPN ACSNHS2 1.3 kb us 1.3 kb ds U1AAS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In AS Input AS NRF-1 AS CTCF AS YY1 AS No Ab PWS CTCF PWS NRF-1 PWS YY1 PWS No Ab PWS In 5' SNRPN ACSNHS2 1.3 kb us 1.3 kb ds U1A
78 Figure 3-8: ChIP analysis of factors associated with the SNURF-SNRPN upstream promoter U1A, the SNURF-SNRPN 5â€™ region and ACS/NHS2. Paternal alleles were analyzed in lymphoblasts derived from AS patients maternal alleles were analyzed in lymphoblasts from PWS patients. Chromatin was immunoprecipitated with antibodies against YY1 (Santacruz, sc1703), NRF-1 (provided by Dr. Richard C. Scarpulla (Northwestern Medical School, Chicago, Il)) and CTCF (Upstate Biotechnology). Purified DNA was analyzed by PCR using primers for the indicated regions. Input DNA was diluted 1:20 and used to normalize the amount of DNA used in the amplification reaction for each cell type. On the top are shown representative gels. The PCR products were quantitated in a fluorescence scanner and the values norma lized to the input DNA. The chart at the center represents the average values across several experiments. The lower values are expanded in the chart at the bottom. result of non-specific immunoprecipitation. Alternatively, a barely perceptible association of YY1 with both paternallyand maternally-U1A alleles may be possible. ChIP analysis revealed the association of NRF-1 with the ACS/NHS2 and the SNURF-SNRPN 5â€™ region/NHS1. As was previously shown for YY1, this association was only observed in the AS-derived cell line, indicating that NRF-1 interaction with ACS/NHS2 and SNURF-SNRPN 5â€™ region/NHS1 is specific to the paternally-inherited allele and does not occur in the maternally-inherited allele. As a positive control for IP with NRF-1 antibody, a region in the DM1 (myotonic dystrophy 1) locus in chromosome 19, which was suspected to interact with NRF-1, was analyzed (Filippova et al., 2001). Association of this region with NRF-1 was detected in both ASand PWS-derived cell lines. After amplification of the NRF-1 immunoprecipitated DNA, the normalized intensity of the PCR-amplified ACS/NHS2 sequence was higher than the normalized intensity for the PCR-amplified SNURF-SNRPN 5â€™region/NHS1. This likely reflects a stronger association of NFR-1 with the ACS/NHS2 compared to SNURF-SNRPN 5â€™ region/NHS1. The SNURF-SNRPN upstream promoter U1A showed a very weak and
79 biallelic band for NRF-1, which was likely to be the result of non-specific cipitation. SNURF-SNRPN gene 5â€™ region includes several potential CTCF binding efore the association of CTCF with that region was examined by chromatin IP. ion with CTCF was detected in the SNURF-SNRPN 5â€™ region or any of the ns analyzed. As a positive control for IP with CTCF antibody, the same e DM1 (myotonic dystrophy 1) locus, in which CTCF is known to assoc immunopreThesites. TherNo interactother regioregion in thiate with to sequences 176 bp apart, was analyzed (Filippova et al., 2001). Association of this region with CTCF was detected in both ASand PWS-derived cell lines. In summary, ChIP analysis showed the association of YY1 with the ACS/NHS2 and the association of NRF-1 with the ACS/NHS2 and at lower levels with SNURF-SNRPN 5â€™ region/NHS1. Association of CTCF with these regions was not detected. Association of RNA Polymerase II with ACS/NHS2 YY1 has been reported to recruit RNA Pol II and TFIIB to activate transcription (Usheva and Shenk, 1994). Therefore, it is possible that the activator function associated with ACS/NHS2 is based in the recruitment of RNA Pol II by YY1, followed by transfer of the polymerase to the SNURF-SNRPN promoter. Consequently, the association of RNA Pol II with the ACS/NHS2 was analyzed by ChIP assays. Two antibodies were used in the immunoprecipitation reaction: a general antibody that recognizes RNA Pol II (Santa Cruz, sc-899) and a second more specific antibody that recognizes the unphosphorylated non processive form of RNA Pol II (Covance, MMS126R). Transcription is preceded by the formation of a preinitiation complex at gene promoters that includes general transcription factors and RNA Pol II in a unphosphorylated form. Phosphorylation of the C-terminal domain of RNA Pol II is required to disrupt the
80 interactions of the polymerase with other transcription factors at elongationtranscription. Therefore, unphosphorylated RNA Pol II is generally considered to be the of non-pro II s CR II and, in particug Regions 1.3 kb upd the ACS/N ternally-allycessive form associated with transcription initiation while phosphorylated Polis generally considered to be the processive form associated with transcription elongation. Figure 3-9 shows the results of the analysis of RNA Pol II. The top panel showrepresentative gels and the bottom panel shows the average relative intensity of the Pproducts with respect to the input DNA control in each experiment. PCR amplification of DNA immunoprecipitated with antibodies for either RNA Pol II or the unphosphorylated form of RNA Pol II revealed the association of RNA Pol lar the unphosphorylated form, with both the ACS/NHS2 and SNURF-SNRPN 5â€™ region/NHS1. This association was only detected on AS-derived cell lines, indicatinthat it occurs specifically with the transcriptionally active paternally-inherited allele. stream and downstream of SNURF-SNRPN 5â€™ region an HS2, respectively, were also analyzed. However, no PCR-amplification of these regions was observed, eliminating the possibility that detection of RNA Pol II in ACS/NHS2 was the result of amplification of neighboring regions effect. As a control a transcribed region in the SNURF-SNRPN second intron was analyzed. No initiation oftranscription takes place in this region and, therefore, should not be associated with unphosphorylated RNA Pol II. As is shown in Figure 3-9, such association was not observed. Furthermore, the levels of RNA Pol II detected in the transcribed painherited allele are comparable to the levels detected in the not transcribe materninherited allele, indicating that the signal observed is likely due to unspecific
81 Figure 3-9: ChIP analysis of RNA polymerase II association with SNURF-SNRPN upstream promoter U1A, SNURF-SNRPN 5â€™ region/NSH1 and ACS/NHS2. Paternal alleles were analyzed in ly mphoblasts derived from AS patients, maternal alleles were analyzed in lymphoblasts from PWS patients. Chromatin was immunoprecipitated with antibodies against RNA Pol II -Pol II(Santa Cruz, sc-899 ) and the unphosphor ylated form of RNA Pol II -Pol II unP(Covance, MMS126R). Purified DNA was analyzed by PCR using primers for the indicated regions. In put DNA was diluted 1:20 and used to normalize the amount of DNA used in th e amplification reaction for each cell type. On the top are shown represen tative gels. The PCR products were quantitated in a fluorescence scanner a nd the values normaliz ed to the input DNA. The bottom chart represents the average values across several experiments. 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Intron2 ACS/NHS2 5' SNRPN 1.3 kb us 1.3 kb ds U1A 5' SNRPN ACS/NHS2 1.3 kb us 1.3 kb ds U1A Intron2AS Input AS No Ab PWS No Ab PWS In AS PolII AS PolUnP PWS PolII PWS PolUnP1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Intron2 ACS/NHS2 5' SNRPN 1.3 kb us 1.3 kb ds U1A Intron2 ACS/NHS2 5' SNRPN 1.3 kb us 1.3 kb ds U1A 5' SNRPN ACS/NHS2 1.3 kb us 1.3 kb ds U1A Intron2AS Input AS No Ab PWS No Ab PWS In AS PolII AS PolUnP PWS PolII PWS PolUnP AS Input AS No Ab PWS No Ab PWS In AS PolII AS PolUnP PWS PolII PWS PolUnP Pat. allele(AS)Mat. allele(PWS) UIA Input No Ab PolII-un P PolII Input No Ab PolII-un P PolII ACS/NHS2 Intron 2 SNURF-SNRPN 5â€™region 1.3 kb upstream SNURF-SNRPN 1.3 kb downstream ASC/NHS2Pat. allele(AS)Mat. allele(PWS) UIA Input No Ab PolII-un P PolII Input No Ab PolII-un P PolII ACS/NHS2 Intron 2 SNURF-SNRPN 5â€™region 1.3 kb upstream SNURF-SNRPN 1.3 kb downstream ASC/NHS2 UIA Input No Ab PolII-un P PolII Input No Ab PolII-un P PolII ACS/NHS2 ACS/NHS2 Intron 2 SNURF-SNRPN 5â€™region SNURF-SNRPN 5â€™region 1.3 kb upstream SNURF-SNRPN 1.3 kb downstream ASC/NHS2
82 immunoprecipitation of the region of interest. Therefore, association of RNA Pol II with the second intron was not detected, which is consistent with the low level of transcription SNURF-SNRPN gene in the cell line used for the ChIP analysis. Similarly, RNA Pol II was not detected on the upstream promoter U1A, consistent with the fact that the transcripts are brain-specific and not expressed in the lymphoblast cell type ployed in the immunoprecipitation. The average values indicated in the chart in Figure 3-9 suggested that the association of the unphosphorylated form of pol II is stronger at the ACS/NHS2 that at SNURF-SNRPN promoter. This suggested that ACS/NHS2, probably with the assistance of YY1, may recruit the unphosphorylated form of RNA pol II. NRF-1 and likely SP1 may be involved in stabilizing the association of the polymerase with this region. Bending of the DNA probably facilitates the transfer of RNA Pol II to the oter that contains two additional NRF-1 binding site, several putative SP1 binding sites and a putative E2F binding site. SNURF-SNRPN upstream promoters also contain potential NRF-1 binding sites, therefore, activation of those promoters by ACS/NHS2 y use the same mechanism. EST database analysis identified a transcript that starts in the vicinity of of the upstreamemthe prommaACS/NHS2 Hence, it is bound by several Neverthelerather than and splices once within SNURF-SNRPN intron 1 before splicing into exon 2. possible that, as a result of the recruitment of RNA Pol II to a regiontranscription factors, the ACS/NHS2 may act as a weak promoter. ss, this would be a consequence or secondary effect of the activator function part of the activation mechanism. Additionally is also possible that the
83 transcript in the EST database is an incomplete version of a larger transcript that starts in one of the SNURF-SNRPN known promoters. Analysis of Histone Modifications in SNURF-SNRPN 5â€™ Region/NHS1 and ACS/NHS2 Histone tail modifications are among the mechanisms that establish and maintain distinct chromatin states which are key regulators of gene expression. Complex combinations of histone modifications specify a wide variety of chromatin states and associated genomic function. Therefore, patterns of allele-specific histone modifications associated with ACS/NHS2 and the SNURF-SNRPN 5â€™ region were characterized. Figure 3-10 shows the results for the ChIP analysis of histone H3 and H4 with specific modifications. The top panel shows representative gels and the bottom panel shows the average relative intensity of the PCR products with respect to the input DNA control in each experiment. The acetylated state of histone H4 was analyzed using antibodies that recognized acetylation of H4 at all lysines in the histone tail (K5, K8, K12 and K16) and antibodies specific for acetylated H4 lysine 5 and H4 lysine 8. Figure 3-10 shows that acetylated histone H4 was enriched in the paternally-inherited allele of SNURF-SNRPN 5â€™ region/NSH1 and the ACS/NHS2. On the other hand, hypoacetylated H4 was associated with both regions in the maternally-inherited chromosome. Enrichment for acetylated H3 lysine 9 and methylated of H3 lysine 4 was also observed in the paternally-inherited allele compared to the maternally-inherited allele of SNURF-SNRPN 5â€™ region/HS1 and the ACS/NHS2. Interestingly the level of histone modifications in SNURF-SNRPN 5â€™ region and ACS/NHS2 is similar except for the level of H3-K4 methylation, which is higher in ACS/NHS2. This is consistent with the higher levels of
84 Figure 3-10: ChIP analysis of th e histone code associated with SNURF-SNRPN upstream promoter U1A, SNURF-SNRPN 5â€™ region/NHS1 and ACS/NHS2. Paternal alleles were analyzed in lymphoblasts derived from AS patients and maternal alleles were analyzed in lymphoblasts from PWS patients. Chromatin was immunoprecipitated with antibodies agains t histones modified at specific residues: acetylated H4-K5, acetylated H4-K8, acetylated H4 and acetylated H3-K9 (Upstate Biotechnologies), methyl ated H3 K4 (Abcam). Purified DNA was analyzed by PCR using primers for the indicated regions. Input DNA was diluted 1:20 and used to norm alize the amount of DNA used in the amplification reaction for each cell type. On the top are shown representative gels. The PCR products were quantitated in a fluorescence scanner and the values normalized to the input DNA. Th e bottom chart represents the average values across several experiments. PWS No Ab PWS In AS Input AS No Ab AS H4 Ac AS H4K5 Ac AS H3K4 Me AS H3K9 Ac AS H4K8 Ac PWS H4 Ac PWS H4K5 Ac PWS H3K4 Me PWS H3K9 Ac PWS H4K8 Ac 5' SNRPN ACS/NHS2 1.3 kb up 1.3 kb ds U1A9 8 7 6 5 4 3 2 1 0 PWS No Ab PWS In AS Input AS No Ab AS H4 Ac AS H4K5 Ac AS H3K4 Me AS H3K9 Ac AS H4K8 Ac PWS H4 Ac PWS H4K5 Ac PWS H3K4 Me PWS H3K9 Ac PWS H4K8 Ac PWS No Ab PWS In AS Input AS No Ab AS H4 Ac AS H4K5 Ac AS H3K4 Me AS H3K9 Ac AS H4K8 Ac PWS H4 Ac PWS H4K5 Ac PWS H3K4 Me PWS H3K9 Ac PWS H4K8 Ac 5' SNRPN ACS/NHS2 1.3 kb up 1.3 kb ds U1A9 8 7 6 5 4 3 2 1 0 Paternal allele(AS)H3-K4-Me H4-Ac Input No Ab Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac H3-K4-Me H4-Ac Input No Ab Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac Maternal allele(PWS) SNURF-SNRPN promoter HS2 U1A 1.3 kb uptream SNURF-SNRPN 1.3 kb downstream ASC/NHS2 Paternal allele(AS)H3-K4-Me H4-Ac Input No Ab H3-K4-Me H4-Ac Input No Ab Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac H3-K4-Me H4-Ac Input No Ab H3-K4-Me H4-Ac Input No Ab Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac Input No Ab H4-K5-Ac H3-K9-Ac H4-K8-Ac Maternal allele(PWS) SNURF-SNRPN promoter HS2 HS2 U1A U1A 1.3 kb uptream SNURF-SNRPN 1.3 kb downstream ASC/NHS2 1.3 kb uptream SNURF-SNRPN 1.3 kb downstream ASC/NHS2
85 NRF-1 and RNA Pol II associated with that same region compared to SNURF-SNRPN 5â€™ region/NHS1. Histones associated with regions 1.3 kb upstream and 1.3 kb downstream from SNURF-SNRPN 5â€™ region/HS1 and ACS/NHS2, respectively, are largely unmodified on both alleles. The differential histone modifications described earlier are mostly absent in the upstream promoter U1A. U1A is not active in the cell type examined and that was shown in the previous section not to associate with NRF-1 or RNA Pol II. In summary, the paternally-inherited alleles of SNURF-SNRPN 5â€™ region/NHS1 and ACS/NHS2 were characterized by histone modifications representative of active chromatin, which is consistent with the nuclease hypersensitivity, factor binding and active transcription associated with the SNURF-SNRPN locus in the paternally-inherited chromosome 15. On the other hand, the maternally-inherited alleles of SNURF-SNRPN 5â€™ region/NHS1 and ACS/NHS2 were characterized by histone modifications representative of silent chromatin.
CHAPTER 4 MATERIALS AND METHODS Cell culture: EBV transformed lymphoblasts derived from PWS and AS pwere kindly provided by Dr. DJ Driscoll and were grown in suspension in RPMI medsupplemented w atients ium ith 10% FBS and 1% penicillin-streptomycin. Human neuroblastoma SK-N rived form ated in the figures were cloned into pGL3-basic ector (Promega) using commercial enzymes: in SNURF/SNRPN (Pvu II), -676 Bgl II), -455 (Rsa I), -207 (Xba I), +80 (Pvu II). UBE3A -455 (Ava I); +75 (Ava I). MKRN3 -890 (Sap I) 45 (Taq I). Snurf-Snrpn -567 (Nde I); +56 (Pvu II). Fragments for U1A, U1B, U1-A and U1-C were generated by PCR (Continues in Appendix A) Transient transfection and reporter assay: SK-N-SH and HT1080 cells were transfected with SuperFect (Qiagen) or FuGene 6 (Promega) according to the manufacturer specifications. Transfections performed with SuperFect were optimized to 4 l of reagent and a total amount of 825 ng of DNA per well on 24 well-plates. Transfections performed with FuGene were optimized to 1 l of reagent and a total amount of 165 ng of DNA per well on 24 well plates. In order to maintain the -SH and human fibrosarcoma HT1080 cell-lines were grown in monolayer in Earl minimal essential medium (EMEM) supplemented with 10% FBS and 1% penicillin-streptomycin. All the cells were maintained in culture at 37 C in 5% CO2. Vector design: SNURF-SNRPN fragments were derived from plasmid pHTRM1.8;MKRN3 fragments were derived from the plasmid pDN34 (Jong et al., 1999); UBE3A fragments were derived plasmid pUBE3A-3.3. Snurf-Snrpn fragments were deplasmid pGN73 The fragments indic v ( 86
87 transfection conditions consistent, approximately 1x10-13 moles of the experimental constructs were supplemented with pBluescript S/K II plasmid in order to keep the total amount of DNA constant. A construct containing the Renilla luciferase gene driven by the herpes simplex virus thyr, pRL-TK (Promega), was used that requirion . Those averaYang, midine kinase (HSV-TK) promote in a 1:11 mass ratio respect to the total DNA as an internal control for transfection efficiency. 24 hours post-transfection, cells were lysed in passive lysis buffer (Promega) and the Firefly and Renilla luciferase activities sequentially and individually measured with the Dual-Luciferase Reporter Assay System (Promega). The photinus pyralis and the Renilla reniformis luciferase gene products have diverged in their substrate specificity, the first using the beetle luciferin in a chemoluminescent reaction es ATP, Mg2+ and O2, while the second uses O2 and coelenterateluciferin (coelenterazine). This system allows for the examination of both gene products individually in the same cell lysate by addition of the first substrate, quenching of the chemoluminescent reaction and addition of the second substrate. The lucifarase reactwas quantified in a Sirius Luminometer V2.2, that measures relative light units. Each experiment was performed in duplicate and repeated at least three times. The values obtained from the firefly lucifarase were normalized to the internal control, the Renilla luciferase, and the average value of each construct for each experiment calculated ges were then used to calculate the standard deviation of the means across independent experiments. In vivo and in vitro DMS treatment of cells and DNA for in vivo footprinting: DMS treatment was performed as described by Hornstra and Yang (Hornstra and 1992). Briefly, approximately 2 X 107 suspension cells were collected, washed with
88 phosphate buffer saline (PBS) and incubated at room temperature in 1ml of PBS withDMS for 60, 90 and 120 s. Cells were immediately washed with 50 ml of ice cold PBS three times and lysed overnight at room temperature in lysis buffer (50mM Tris-HCl pH8.0, 150 mM NaCl, 25 mM EDTA, 0.5% SDS and 300 g/ml proteinase K). The lysate was sequentially extracted with equal volumes of phenol, phenol:chloroform (1:1) and chloroform, then treated with RNase (RNase cocktail by Amersham), phenol:chloroform (1:1) and chloroform extracted a second time. The genomic DNA was then ethanol precipitated, resuspended in 180 l of double distilled water (ddH2O) and then cleaved atthe modified guanine residues in 10% piperidine at 95 C for 30 minutes followincubation on ice. Piperidine was eliminated by sequential addition of 1 ml of doudistilled water followed by drying in a vacuum concentrator in the absence of heprocedure was repeated one more time. DNA was resuspended in TE (pH 7.5) (1Tris-HCl (pH 7.5) and 1 mM EDTA (pH 8)) to a concentration of 1.5 mg/ml. For the treatment of naked genomic DNA with DMS, approximately 100 g of DNA in 10 l of ddH2O was incubated with 0.5 % DMS at room temperature for 45 seTo reduce the viscosity of the DNA, the sample was vortexed for 2 sec after addition the DMS. The reaction was immediately stopped by addition of 200 l of DMS stop buffer (1.5 M NaAc pH 5.0 and 1 M -mercaptoethanol) and the DNA precipitated with 750 l of 100% ethanol followed by a 20 minutes incubation in a dry ice/ethanolThe precipitated DNA was th 1% ed by ble at. This 0 mM c. of mixture. en washed in 75% ethanol and cleaved with piperidine as befor and e. In vivo and in vitro DNase I treatment of cells and DNA for in vivo footprinting: This proceeding was performed as described by Chen and Yang (Chen
89 Yang, 2001). Briefly, approximately 2 X 107 lymphoblast cells were collected, washed once with PBS, once with solution A (150mM sucrose, 80mM KCl, 35mM HEPES (pH7.4), 5mM K2HPO4, 5mM MgCl2, 0.5mM CaCl2), once with solution B (150mM sucrose, 80mM KCl, 35mM HEPES (pH 7.4), 5mM K2HPO4, 5mM MgCl2, 2mM CaCl2) and incubated for 2 minutes at 37 C in solution B with 0.2% Nonident P40 (NP40) and10-60 units of DNase I (Worthington). The reaction was stopped by addition of lysis buffer and cells were lysed overnight at room temperature and genomic DNA purifiephenol/chloroform extraction as described above. For the treatment of naked genomic DNA with DNase I, three tubes were preparwith approximately 50 g of DNA resuspended in 100 l of ddH2O plus 200 l of solution B. Three different concentration points were generated by addition of DNase I to final concentrations of 0.05, 0.1 and 0.2 units/l, respectively, and the mixtures incubated at room temperature for exactly 2 minutes. The reaction was stopped d by ed by additind y F-SNRPN gene promoter as described below. ed as on of EDTA and SDS to a final concentration of 20 mM and 0.2% respectively athe DNA was purified by phenol/chloroform extraction followed by ethanol precipitation. The DNA was resuspended in TE (pH 7.5) to a final concentration of 1.5 mg/ml. Preparation of the DNA sequencing ladder: G, G+A, C+T and C sequencingladders were generated from purified human genomic DNA that was chemically modifiedand cleaved according to the method developed by Maxam-Gilbert (Meth enzymol 65 499-560,1980). Ligation-mediated PCR (LMPCR) was then used to specifically amplifthe SNUR Ligation-Mediated PCR (LMPCR): LMPCR was essentially performdescribed by Chen and Yang (Chen and Yang, 2001) with minor modifications and
90 consists of 3 steps: 1extension with a gene specific primer (extension primer);2ligatioof a double stranded asymmetric oligonucleotide linker (LP15â€™-gaattcagatc-3â€™ and LP25â€™-gcggtgacccgggagatctgaattc-3â€™); 3PCR amplification with a second gene specifprimer (PCR primer) and a linker specific primer. All three steps were performed in a thermocycler with heated lid to prevent evaporation/condensation. Approximately 3 g of DMS/piperidine-treated DNA or 5 g of DNase I-treated DNA were denatured at 95 C for 10 min. and annealed to 0.6 pmols of extension primer at the temperatures indicated in Table 4-1 for 30 min. in a 15 l of a so n ic lution containing 10 mM Tris-HCl (pH 82 C 0 mM DTT, by standard phenol/chloroform extrac .9) and 40 mM NaCl. The subsequent primer extension was carried out by addition of 15 l of a solution containing 10 mM Tris-HCl (pH 8.9), 40 mM NaCl, 0.5 mM each dNTPs and 2 units of the high fidelity Vent polymerase (New England Biolabs) and incubation at 53 C for 1 min., 55 C for 1 min., 57 C for 1 min., 60 C for 1 min., 6for 1 min., 66 C for 1 min., 68 C for 3 min. and 76 C for 3 min. 20 l of a dilutionbuffer (10 mM Tris-HCl (pH 7.5), 18 mM MgCl2, 50 mM dithiothreitol (DTT) and 0.0125% bovine serum albumin (BSA)) were added followed by 100 pmols of asymmetric double stranded linker in a 25 l solution containing 10 mM MgCl2, 2 3 mM ATP, 0.005% BSA and 4.5 units of T4 ligase (Ambion). The double stranded linker was generated by combining LP1 and LP2 to a final concentration of 20 M each in 250 mM Tris-HCl (pH 7.7) followed by denaturation at 95 C, cool down toroom temperature at a rate of 1 C/min and overnight incubation at 4 C. The ligation proceeded overnight at 17 C and DNA was then purified tions and ethanol precipitation in the presence of 10 g of tRNA.
91 DNase I-treated DNA conserves 3â€™ hydroxyl groups that can be extended and ligated increasing the background of the reaction. To avoid undesired extension products the extension primer was biotinylated at the 5â€™ end. This allowed for the capture of primer extension products with streptavidin-coated magnetic beads (Promega) followinthe ligation. This was accomplished as described by Tormanen et al. (Tormanen et al., 1992). Briefly, beads were washed twice with washing/binding buffer (10 mM Tris-HCl(pH 7.7), 1 mM EDTA and 2 M NaCl) and resuspended at a concentration of 1 gbeads/l. The biotinilated extension products in the ligation reaction were captured wit75 l of the prewashed streptavidin-coated beads at room temperature for 15 min. The beads were collected on a magnetic stand and washed once with 75 l of washing/binding buffer and resuspended in 37.5 l of 0.15 N NaOH. The non-biotinilated template strand was eluted from the beads by incubation at 37 C for 10-15 min. The beads were separated in a magnetic stand and the supernatant containing the product of interest was transferred to a new tube and neutralized by addition of 37.5 0.15 N HCl and 7.5 l of 1 M Tris-HCl (pH 7.7). The DNA was then ethanol precipitated in the presence of 10 g of tRNA. The purified ligation products were amplified by Taq polymerase in the presence of 10 mM Tris-HCl (pH 8.9), 40 mM3 mM MgCl2, 0.25 mM each dNTPs and 20 pmols of gene specific primer and 20 pmols of linker primer. Cycling conditions for the PCR were as follows: a single denaturation step of 3 minutes at 95 C followed by 20-25 cycles of (20 s at 95 C, 1 min at the primer-specific annealing temperature indicated in Table 4-2 and 90 s + 2 s/cycle at 76C) and a final extension step of 15 min at 76C. Extension/amplificati g h l of NaCl, on though some particularly C/G rich regions required the use of 7-deaza-dGTP. Instead of 0.25 mM
92 dGTPPrimer set Sequence Annealing temp. , those reaction contained 0.175 mM dGTP and 0.075 mM 7-deaza-dGTP. Indicated in Table 4-2 are LMPCR primer set sequences, annealing temperature (extension and PCR) and requirement for 7-deaza-dGTP: Table 4-1: Primers and conditions for LMPCR reactions. TY 528* 5â€™-ATCCCAGGTTGCTTATGGT-3â€™ 45C 5â€™-GGCCCCCTCTCATTGCAACAGTG-3â€™ 63C TY 539* 5â€™-CTTGAGAGAAGCCACCGG-3â€™ 45C 5â€™-CTGACCTTGCCCGCTCCATCG-3â€™ 63C TY 541 5â€™-CAACCTGGGATCAATGGACA-3â€™ 47C 5â€™-TGCACACACCACTGGCCAAACAATC-3â€™ 64C TY 640 5â€™-AGAACGGCACAACAGCAAG-3â€™ 47C 5â€™-GGGCCCAAATTCCGTTTATTCAGTACTCC-3â€™ 65C TY 647 5â€™-TGTGGTTATGGCGCATTT-3â€™ 47C 5â€™-CTTTTTTGTACCGCACATAGGAAGACCTGA-3â€™ 65C TY 651 5â€™-TGTTTGCCGCAGTGCAG-3â€™ 47C 5â€™CCCCACAGCACTGTTGCAATGAGAGG-3â€™ 65C TY 653 5â€™-TTGAATAGCCAGTGTTTTTGAGT-3â€™ 47C 5â€™-GGGGGAGATAATTTTCACAATTTACCCCCTC-3â€™65C TY 760 5â€™-GCAAACAAGCACGCCTG-3â€™ 47C 5â€™-CATGCTCAGGCGGGGATGTGTG-3â€™ 65C The asterisk indicates primer used in reactions that were supplemented with 7-deaza-extension; the second primer was used in the PCR. Electrophoresis, transfer and hybridization of sequencing gels: The Pamplified products from the LMPCR reaction were phenol/chloroform extracted, ethprecipitated and resuspended in 20 l of ddH2O. Initially, 3l were size-fractionated in a 5% polyacrylamide denaturing gel (Long Ranger Gel Solution from BioWhitaker Molecular Applications) sequencing gel in 1xTBE. DNA was electrotransfered to a nylon membrane (Hybond N+, Amersham) for 30 min. and dried for 1h at 80 C. Thmembrane was pre-hybridize for 20 min. at 65 C in hybridization buffer (250 mM Na2PO4, 3.5 % SDS, 1% BSA). The blot was hybridized to a gene specific P radiolabeled probe that was synthesized from a plasmid template using the Prime a Pro dGTP. The first primer/annealing temperature in each set was used in the initial CR-anol e 32be kit (Ambion) with certain modifications. A plasmid containing the region to be analyzed
93 was digested with an appropriate enzyme flanking that region and purified by phenol/chloroform extraction. Approximately 1.5 g of digested plasmid and 0.3gene specific PCR primer used for the LM-PCR were denat g of ured in 9 l of ddH2O at 95 C n was supplemented with 5 l of 5X DecaP, 10l of -32P dCTP (3,000Ci/mmol) and 1 unit ofBRL) and incubated at 37 C for 1 h. For Leae replacedw in the ptenCl (), 500 mM KCl, 1.5 mM32,0/mmol), 20 pmols of PCplasmid was denatured at 95 C for 3 minain at 76 C). The probe wd immediately purified by d film ype 57, Polaroid) and excised. The probe containing polyacrylamide gel piece was crushed and soaked in 4 ml of hybridization buffer and directly added to the pre-hybridized membrane. The hybridization proceeded overnight at 65 C and the membrane washed in (20 mM Na2PO4, 0.5% SDS). The labeled DNA was detected by autoradiography. For LMPCR reactions with primer set TY 539 and TY 541 the opposite strand primer was used in the probe synthesis. Site directed mutagenesis: Sequence-specific mutations of factor binding sites in the SNURF-SNRPN gene promoter region were performed using the QuikChange XL Site-Directed Mutagenesis Kit (Stratagene) following the manufacturerâ€™s instructions. The mutations were design to substitute the nucleotides identified by in vivo footprinting. for 10 min. followed by snap freeze in dry ice. This solutio rime Buffer minus dCTP (Ambion) Klenow DNA polymerase (GibcoMPCR r ctions with primer set TY 640, Taq polymeras kleno rimer ex sion of 1 g of digested plasmid in 200 mM Tris-H pH 8.4 MgCl2, 0.2 mM dNTP-dCTP, 10l of -P dCTP (3 00 Ci R primer and 2.5 units of Taq polymerase. The followed by 5 cycles of (20s at 95 C, 1 min at 65 C nd 2 m as then denatured at 95 C for 10 min. an denaturing PAGE in 1xTBE, from where is was detected by exposure to a Polaroi (T
94 The new sequences were compared to a transcription factor database (TRANSFrule out the creation of transcription factors binding sites. Additionally new sequences included a new restriction site in order to facilitate the screening of the mutated clones. Complementary primers including the desired mutations (see table 4-2) were extended with the high fidelity PfuTurbo DNA polymerase in 1X reaction buffer (10 mM KCl, mM (NH4)2SO4, 20 mM Tris-HCl (pH 8.8), 2 mM MgSO4 0.1% Triton X-100 and 0.1 mg/ml nuclease-free BSA) supplemented with 50 ng of template DNA and 6% QuikSolution (Stratagene). The reactions were conducted as follows 1 min denaturatioat 95C, 18 cycles of 195C for 50 sec., 260C for 50 sec. and 368C for 1 minuteof template plasmid length (5.5 and 7.5 kb respectively for the two plasmids described AC) to 10 n /kb below) followed by a final extension at 68C for 7 min. Two types of template DNA were used. The template plasmid for mutations of P1, P3, P4, P5 and P/M6 was a luciferase reporter construct including the SNURF-SNRPN promoter (see Figure 2-1 construct (c)). The template plasmid for mutation of the YY1 binding site was a luciferase reporter construct including the SNURF-SNRPN promoter and a 2.2 kb nuclear hypersensitive intronic fragment (see Figure 3-1 construct (b)). Following this reaction the dam-methylated template DNA was selectively digested with Dpn I endonuclease (target sequence: 5-Gm6ATC-3) that is specific for methylated and hemimethylated DNA and leaves intact the mutation containing newly synthesized DNA. The mutated nicked plasmid was then transformed into XL10-Gold ultracompetent cells and clones positive for the mutation were identified by restriction digest and verified by sequencing.
95 Table 4-2: Site directed mutagenesis of SNURF-SNRPN promoter cis-acting elements: Factor binding site Wild type sequencePrimer sequence P2 (839-40) CTGGCG 5â€™-cgctgctgcagcgagtatgcatcagagtggagcggcc-3â€™ P5 (835-6) GGCGCG 5â€™gccgcagaggcaggctagtactcatgctcaggcgggg-3â€™ M6/P6â€™(883-4) CGCC 5â€™cactgcggcaaacaagcaatattgcgcggccgcagaggc-3â€™ P1 though M5 represent the factor binding sites identified by in vivo footprint analysis inprimer sequence indicate the new sequence. The bolded nucleotides in the wild footprint analysis. Preparation of nuclear extracts electrophoresis mobility shift assay: Crude nuclear extracts were prepared from human neuroblastoma SK-N-XX cells and human lymphoblasts derived from AS and PWS patients. Approximately 5-10 X 107 cells wecollected (lymphoblasts) or trypsinized and collected (neuroblastoma cells). Cells wresuspended in PBS, pelleted by centrifuagation at 1,500 rpm for 3 min. at 4 C and resuspended by gentle trituration in 1 ml of lysis buffer (20 mM HEPES (pH 7.6),NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 20% glycerol, 0.1% Triton-X, 1mM DTT and protease inhibitors, PMSF, leupeptin, apoprotein, pepstatin, chymostatin). Lysis proceeded for 3 min. on ice and nuclei were collected by centrifugation at 2,000 rpm fo5 min. at 4 C and resuspended in nuclear extraction buffer (20 mM HEPES (pH 7.6), 20% glycerol, 400 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 1 mM DTT and proteaseinhibitors). Nuclei were rocked for one hour at 4 C and then collected by centrifugation at 10,000 rpm at 4 C. The supernatant, that contained the nuclear extract, was dialyzed the SNURF-SNRPN promoter (see Figure 2-4 and text). The bolded nucleotides in the type sequence indicate the sites of direct protein/DNA interaction as determined by the re ere 10 mM r for at least 6 h. at 4 C in dialysis buffer (20 mM HEPES (pH 7.6), 20% glycerol, 100 mM primer design P1 (804-5) CGCG 5â€™-ggagcggtcagtgaattcatggagcgggcaag-3â€™ P4 (837-8) TGTGTG 5â€™-catgctcaggcggggagatatccgaagcctgccgc-3â€™ M6/P6 (853-4) GCCT 5â€™gcggcaaacaagcacatatgcgcggccgcagaggc-3â€™ YY1 (885-6) GCCGCC 5â€™gcagctccttcaagatgtacgtagctgcagcggcttag-3â€™
96 NaCl, 10mM MgCl2, 0.2 mM EDTA, 1 mM DTT and protease inhibitors). Nuclear extracts were then otein cn in the ntometry.in sampled of tt 205 nm (Ais mobility shift assays (EMSA): The probes were synthesized using l run in e centrifugation through a syringe plugged with glass wool. The probe was finally purified aliquoted and stored at -80 C. The pr oncentratio uclear extract preparation was measured by spectropho The prote s were diluted 50 to 200 fold in 0.01% Brij an the absorbance he mix a 205) measured in a spectrophotometer. The protein concentration in mg/ml was estimated with the following equation: A205/(31*b), where b is the length of the path in cm and equals 1. Electrophores cartridge purified oligonucleotides. Approximately 100 ng of a 30-35 bp oligonucleotides (YY1(875-6): cagctccttcaagatggccgccgctgca) was radioactively labeledwith T4 Kinase (Gibco BRL or New England Bioloabs) in the presence of -32P ATP according to the manufacturerâ€™s instructions. The kinase was inactivated at 95 C for 5 min., followed by addition of 400 ng of a complementary oligonucleotide and ddH2O to 100 l. The primer mix was denatured at 95 C for 5 min. and annealed by incubation at room temperature for at least 2 h. and at 4 C overnight. To eliminate self-annealing products, the radiolabeled probes were further purified in a 8% polyacrylamide geTBE. The position of the probes was determined by exposure to a Polaroid film (Type 57, Polaroid) and the pieces of gel containing the probes excised, placed in a microfugtube and crushed with a toothpick. Approximately 3 volumes of extraction buffer (50 mM NaCl, 50 mM Tris-HCl (pH 7.5), 1 mM EDTA) was added to the crushed gel piece that was then incubated at -80 C for 30 min. or until frozen. DNA was extracted overnight at 37 C in a spinning rotor. Acrylamide bits were removed by slow
97 with the Gel Extraction Kit from Qiagen and eluted in 100 l of TE. Counts per min(CPM) per l were measured in a scintillation coun ute ter. om of the antibodies (described in the tely re last t ia) g buffer (5 For the binding reactions, 10,000 cpm of probe were incubated for 20 min at rotemperature with 7-10 g of nuclear extract or 1-2 l of purified protein (purified YY1 was kindly provided by Dr. Edward Seto). The conditions for YY1 binding were asfollow, 50 mM NaCl, 1 mM MgCl2, 4% glycerol, 10 mM Tris-HCl (pH 7.5), 0.5 mM EDTA and 0.5 mM DTT. For supershift assays, 1.2 g xt) were added to the reaction 10 min. before addition of the probe, simultaneouswith the probe, or 10 min. after the addition of the probe. Reaction mixtures weanalyzed by 5% non-denaturing polyacrylamide gel (0.06% APS and 0.2% TEMED) run in 0.5X TBE buffer and the probes visualized by autoradiography. Chromatin immunoprecipitation (ChIP): This technique was performed as described by Kang et al. (Kang et al., 2002). Ten million cells were used in each immunoprecipitation (IP) reaction. The experimental conditions indicated below are adjusted for 10 IP reactions and therefore start off with 1X108 cells. Human lymphobcells grown in suspension were collected, resuspended in 25 mls of cell culture media and crosslinked in 1% formaldehyde (27 l of 37% formaldehyde/ml of media) for 10 min. aroom temperature with gentle rocking. The crosslinking reaction was stopped by addition of glycine to a final concentration of 0.125 M (62.5 l 2 M glycine/ml of medand gentle rocking for 5 min. at room temperature followed by two washes with 25 mls of ice cold PBS containing protease inhibitors (protease inhibitor cocktail, Roche).Formaldehyde-treated cells were resuspended in 1 ml of ice cold nuclei swellinmM PIPES (pH 8.0), 85 mM KCl, 0.5% NP-40, 1.1 mg/ml sodium butyrate and protease
98 inhibitors), transferred to a microfuge tube and incubated on ice for 10 min. Nuclei werecollected by centrifugation for 5 min. at 5,000 rpm and 4 C in a microcentrifuge, resusp ify IP dilution buffer (0.01% were ended in 1ml of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8), 1.1 mg/ml sodium butyrate and protease inhibitors) and incubated on ice for 10 min.Lysed nuclei were sonicated on ice in a 4 ml plastic culture tube. The sonicated chromatin was cleared by centrifugation for 10 min at 13000 rpm and 4 C in a microcentrifuge. Thirty microliters of the supernatant were set aside an incubated at 65 C for at least 2h, and then phenol/chloroform extracted, treated with RNase and fractionated by electrophoreses in a 1.6 % agarose gel in TBE buffer in order to verthat the average size of the fragments was between 0.5 and 1 kb. The rest of the supernatant was transferred to a 15 ml conical tube with 9 ml of Ch SDS, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM Tris-HCl (pH 8) 167 mM NaCl, 1.1 mg/ml sodium butyrate and protease inhibitors). The immune complexes were captured with Protein A Sepharose beads. A 50%slurry was prepared by rocking 0.2g dry beads in 40 ml ddH2O for 2 h followed by two washes with ddH2O. Beads were collected each time by centrifugation at 4000 rpm in a tabletop centrifuge. Following the washes beads were resuspended in approximately 1 ml of TE (pH 8) containing 0.05% sodium azide. In all the subsequent steps beadshandled with wide bore tips. Non specific DNA/beads interactions were removed by addition of 500 l of 50% slurry of Protein A Sepharose beads to the sonicated chromatin and incubation in a spinning wheel with for 2h at 4 C. Beads were pelleted by centrifugation for 10 min. at 4000 rpm and 4 C in a tabletop centrifuge and the supernatant divided into 1 ml aliquots
99 to which the antibodies were added as follows: anti-YY1 (Santa Cruz Biotechnology sc-1703; 5 g), anti-NRF-1 (provided by Dr. Richard C. Scarpulla, (Northwestern MeSchool, Chicago, Il); 4 l), anti-CTCF (Upstate Biotechnology 06-917; 5 g and 20 g), anti-RNA polymerase II (Santa Cruz Biotechnologies sc-899; 1g), anti-RNA pol II (unphosphorylated form) (), anti-H3 dimethyl-K4 (Abcam ab7766-50; 2 g), antiacetyl-K5 (Upstate 06-759; 5 l of rabbit antiserum) anti-H4 acetyl-K8 (UpstateBiotechnology 06-760; purified rabbit IgG 5l), anti-H3 acetyl-K9 (Upstate Biotechnology 06-942; 5g), -acetyl-H4 (Upstate Biotechnology 06-866; 1 l oantiserum). A control sample using no antibody was included for each cell-line in eachexperiment. Sonicated chromatin was then incubated with each of the antibodies overnight a4 C with gentle rocking. At the same time sepharose beads in TE-Na azide were collected by a quick pulse of centrifugation in a microcentrifuge and blocked overnight as dical -H4 f rabbit t 50% s e heel for 5 min/wash sequentially with 1 ml of each of the follow Xlurry in TE (pH 8) with 0.05% Na azide and 3% BSA. The next day 60 l of the blocked Protein A Sepharose beads were added to each reaction and incubated for 2 h at4 C on a rotating wheel to capture immune complexes. Beads were collected by a quick pulse of centrifugation and supernatants discarded except for 500 l of supernatant of thcontrol sample with no antibody that was saved to use as input control. Beads were then washed at 4 C on the rotating w ing: low salt wash solution (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 200 mM Tris-HCl (pH 8) and 150 mM NaCl ), high salt wash solution (0.1% SDS, 1% Triton100, 2 mM EDTA, 200 mM Tris-HCl (pH 8) and 500 mM NaCl), LiCl wash solution (0.25 M LiCl, 1% NP-40, 1% Na desoxycholate, 1 mM EDTA and 100 mM Tris-HCl
100 (pH 8)), and twice with TE (pH 8). The immune complexes were eluted from the beads with 275 l of freshly made elution buffer (1% SDS and 0.1 M NaHCO3) under vigorousshaking in a dry heating block/shaker (950 rpm) at 65 C for 15 min. The elution was repeated once with 250 l of elution buffer. The collected elutants were pulled togethsupplemented with NaCl to a final concentration of 200 mM and, together with the inpsample collected earlier, were incubated at 65 C for 4 h to reverse the crosslinkinfollowed by overnight incubation at 4 C. Samples were then treated with 40RNase A for 30 min. at 37 C and proteinase K in the additional presence of 10 M EDTA and 40 M Tris-HCl (pH 6.5) for 1 h at 37 C. The DNA was then purified usinga PCR purifica er, ut g g/ml of tion kit by Qiagen and eluted in100 l 50% TE (pH 7.5). NA NA The DNA isolated after the immunoprecipitation together with a 1:20 dilution of input DNA control was then analyzed by PCR. Aproximately 2.5 l of template Dwas amplified in a 25 l reaction with 0.625 units of Hot Start Taq (Qiagen), 0.5% Q solution, 1X Hot Start Buffer with 1.5 mM MgCl2 and 0.5 M each forward and reverseprimers. The conditions for the PCR amplification were as follow: denaturation and activation of the enzyme for 15 min, followed by 25-29 cycles of (95 C for 30 s, 54 or 60 C for 30 s and 72 C for 1 min) and a final extension of 5 min at 72 C. PCR reactions with each primer set were performed using two different cycle numbers. For each experiment, the input DNA control from the PWS patient-derived cell line was sequentially diluted to 1:2, 1:4 and 1:8 and PCR-amplified with some of the primer sets used in the ChIP analysis. This ensured the linearity of the amplification reactions. Additionally, several primer sets amplified more efficiently the immunoprecipitated Dfrom the PWS patient-derived cell line. This was probably due to differences in
101 chromatin structure between paternallyand maternallyinherited alleles of the SNURF-SNRPN gene. Comparison of input DNA control from the AS patient-derived cell line to the serial dilutions from input DNA control from the PWS patient-derived cell line aftePCR-amplification allowed normalization of the template DNA of the several cell-lines. Table 4-3 shows the sequence of the primers and conditions employed to amplify the different genomic regions examined. Table 4-3: Primers and conditions for the amplification of immunoprecipitated DNA. r Genomic region Sequence Annealing temp HS1/ SNURF-SNRPN promoter5â€™-atcccaggttgcttatggt-3â€™ 54 C 5â€™-agaacggcacaacagcaag-3â€™ HS2 (855-6) 5â€™-tcttttgactgataggccaggtg-3â€™ 60 C 5â€™-ggcttcctctccttaccactgc-3â€™ 1.3 kb us (907-8) 5â€™-aatcaaaagagcctgagtccaaag-3â€™ 54 C 5â€™-tcacataaaataacatgggaagttgac-3â€™ 1.3 kb5â€™-gaatagtgaaatcaagataaggcaagg-3â€™ ds (937-8) 5â€™-cttctggtcatttggctatttgc-3â€™ 54 C Intron 2 (857-58) 5â€™-gaaaagttgagctggatgtggtg-3â€™ 60 C 5â€™-accctcaccataatttaaaacttgc-3â€™ U1A (869-870) 5â€™-tttctctctagggccttctcttagc-3â€™ 54 C 5â€™-tcaaacagctccagagggaagac-3â€™ DM1 (904-5) 5â€™-ccagttcacaaccgctccgag-3â€™ 60 C 5â€™-ttcccggctacaaggacccttc-3â€™ GR (859-60) 5â€™-ccccctgctctgacatctt-3â€™ 60 C 5â€™-cttttccgaggtggcgagtatc-3â€™ experiments calculated. The amplified products were size-fractionated in a 5% non-denaturing PAGE gel in TBE buffer, and visualized after SyBr Green staining in a fluorescence scanner that allowed for quantification of the bands. The intensity of the experimental bands was normalized to the input DNA and average value and standard deviation across
CHAPTER 5 DISSCUSION AND FUTURE DIRECTIONS The long term goal of this research is to understand the function of the PWS-IC.Specific questions were asked in this par ticular based on the hypothesis that the function his ton of cis-acS-IC. The PWS-ICh and/otain an n, and/or preven conpaternally-inherited chromosome 1ion of the PWS-IC may be mediated by assembly of tegion of the SNURF-SNRPN gene, leading to ahe paternally-inherited PWS-AS imprinted domain (Dittri (Nicholls et al., 1998). The paternally-inherited Ph two nuclease hypersensitive sites, NSH1 and NHe maternally-inherited alleleys URF-SNRPN gene and two negative regulatory elements in the SNURF-SNRPN 5â€™ region were defined by transient expression assays in the absence of an exogenous enhancer. Published data on the transcriptional activity of the SNURF-SNRPN 5â€™ region relies on the presence of an exogenous enhancer (SV40). However, the of the PWS-IC is mediated, at least in part, by cis-acting regulatory elements. Hence, t thesis project focused on the iden ification and characterizati ting el ements within the PW has been postulated to establis r main active conformatio t the establishment of an inactive formation in the 5 q11-q13. Furthermore, the funct ranscription factors at the promoter r ctivation of gene expression in t ch et al., 1996; Perk et al., 2002) WS-IC region is associated wit S2, which are absent from th . Therefore, NHS1 and NHS2 were analyzed for their potential role in transcription of the SNURF-SNRPN gene and by extension in the PWS-IC function. Analysis of SNURF-SNRPN 5â€™ Region/NHS1 Analysis of the SNURF-SNRPN gene 5â€™ region by transient expression assaThe promoter of the SN 102
103 presence of such a robust exogenous element can introduce artifacts that affect the promoter function. Therefore, the 5â€™ region of SNURF-SNRPN was analyzed in the absence of an exogenous enhancer. No cis-acting elements involved in transcriptional activation of the SNUutside of the region betwe ssion PN DNA-protein interactions in the 5â€™ region of SNURF-SNRPN were analyzed by in vivo footprinting. The sequence examined included NHS1, the six sequences phylogenetically conserved between human and mouse, and all the promoter elements identified by transient expression assays. Six footprints were identified in the paternally-nd P6) an RF-SNRPN gene were found in the PWS-SRO o en positions -336 to +80, which immediately flank SNURF-SNRPN transcription start site (see Figure 2-1). This region includes four out of six phylogenetic footprints, conserved in the human PWS-SRO and the homologous region in mouse, that are candidate cis-regulatory elements for SNURF-SNRPN transcription and PWS-IC function. Additionally, a novel negative regulatory element was identified upstream ofthe minimal promoter. A second negative regulatory element was detected within exon 1. This second element had already been described by Green Finberg et al. in the presence of the SV40 enhancer, which artificially intensified the degree of the repre(Green Finberg et al., 2003). Identification and functional analysis of cis-acting elements in the SNURF-SNRgene 5â€™ region inherited allele (P1-P6), three of which played a role in promoter function (P2, P5 a d one of which (P1) repressed the activation of the SNURF-SNRPN promoter by a endogenous activator located within SNURF-SNRPN first intron. P6 was in close proximity to the only footprint identified on the maternally-inherited allele (M6). These and other findings are summarized in Table 5-1.
104 Table 5-1: Identification and functional analysis of cis-acting elements in the 5â€™regrepression of an activator function located in intron 1. See Figure 2-4, FigFootprint Potential binding sitePosition Effect on transcription P2 E2F -5 Yes* P4 -34 No P6 NRF-1/CTCF -86 Yes* ion of the SNURF-SNRPN gene. *Involved in promoter function; **Involved in the ure 2-7 and Figure 3-2 P1 AP1 +56, +58Yes** P3 NRF-1 -13 Not tested P5 NRF-1 -56, -58 Yes* M6 NRF-1/CTCF -85 Yes* The promoter element P5 is a potential NRF-1 binding site and lies within one of the sequences phylogenetically conserved between human and mouse. This is consistent with a recent report that identified by transient expression assays a sequence in the human and mouse SNURF-SNRPN promoter that overlaps with P5 and that is involved in promoter function (Green Finberg et al., 2003; Hershko et al., 2001). P2, which is also essential for promoter function, is a potential E2F binding site and resembles an initiator element. The functional characterization of P2 and P5 differs from previous transient s reported by Huq et al. They analyzed, in the context of the SV40 enhanthe activity e mited to repressing the effect of expression assay cer, a region of the SNURF-SNRPN promoter that lacked sites P2 and P5. However, the transcriptional activity of this region was only two-fold less that of a larger region that included the minimal promoter (Huq et al., 1997). The divergencbetween the results obtained for P2 and P5 and the results shown by Huq et al. can be explained by the presence of the exogenous SV40 enhancer in the latter study. P1 was found to repress the activation of the SNURF-SNRPN promoter by an endogenous activator located within SNURF-SNRPN intron 1. In fact P1 is included within a sequence previously identified as a repressor by transient expression assays (Green Finberg et al., 2003). However, the role of P1 is li
105 activators on the SNURF-SNRPN promoter. In vivo footprint P3, which is a potential ing site, was not examined functionally. CTCF bindFoooFpression assays and was not found associated to any known transcription factor binding site. Neverthelthis element may play a role inlishinpecific endogenous chromatin conformatiter regiot cannecreated in the episomal rm employed in the expression studies. Alternatively, this site may be involved in the function of the PWS-IC, i.e., the establishment and/or maintenance of an active state of the paternally-inherited PWS-AS associated region. Nevertheless, the Snurf-Snrpn promoter and associated elements in the mouse are not necessary for PWS-IC function (Bressler et al., 2001). However, a similar finding has not been confirmed for the human SNURF-SNRPN promoter which, in contrast to the mouse, is included in the small region that constitutes the PWS-SRO. P6 and M6 are two distinct footprints that localize to approximately the same DNA sequence in the SNURF-SNRPN 5â€™ region. Transient expression assays indicated that the sequence associated with P6 and M6 is involved in promoter function. Since the SNURF-SNRPN gene is expressed from the paternally-inherited allele and silent in the maternally-inherited allele, it is likely that the abovementioned promoter function is associated with P6 in the paternally-inherited allele and not with M6 in the maternally-inherited allele. On the other hand, M6 may participate in the silencing of SNURF-SNRPN in the maternally-inherited allele. P6 and M6 are associated with potential CTCF and NRF-1 binding sites. Therefore, the DNA methyl-sensitive NRF-1 may bind P6 in the unmethylated paternally-inherited allele. On the other hand, CTCF has been shown tprint P4 ha d no role in transcripti n of SNUR -SNRPN in transient ex ess, estab g a s on at the promo n tha ot be r fo
106 in vitro to bind CTCF binding sites in the promoter of the Snurf-Snrpn gene in mouse irrespective of DNA methylation. This suggests that CTCF may associate with M6 in the methyl at may be an indication of a non-canonical nucleosome. between the human and mouse homologous regions. Furthermore, YY1 was shown to lated maternally-inherited allele. Alternatively, both alleles may be bound by the same factor in two different conformations (posttranslational modifications etc) generating two distinct footprints. A rotationally phased nucleosome is associated with the promoter region of the SNURF-SNRPN gene on the maternal allele (see Figure 2-5). This reveals an additional level of regulation on the maternally-inherited allele that would explain the lack of hypersensitivity of this region on the maternal allele and would be likely to prevent transcription factors from binding and activating the promoter. In addition, the maternaspecific cis-acting element identified by in vivo footprinting may participate in the formation of this structure. The paternally-inherited allele shows weaker phasing of a nucleosome in the same region th Taken together, the data shown for the SNURF-SNRPN 5â€™ flanking region represent a comprehensive characterization of cis-acting elements in that region. Several elements were shown to participate in transcription of the SNURF-SNRPN paternally-inheritedallele and others were proposed to play a role in chromatin structure or IC function. Anelement unique to the SNURF-SNRPN maternally-inherited allele was proposed to participate in silencing of that allele. Characterization of NHS2 and Associated Activation Function Examination of NHS2 revealed a novel activation function with a preference for the human and the mouse SNURF-SNRPN main and upstream alternative promoters. This element includes potential binding sites for YY1, NRF-1 and SP1 that are conserved
107 play a role in the activation function. YY1, NRF-1 and the unphosphorylated non processive form of RNA Polymerase II were shown to associate with NHS2 in the paternally-inherited allele in vivo. In addition, histones associated with NSH2 showed modifications characteristic of active chromatin in the paternally-inherited allele,characteristic of silent chromatin in t and he maternally-inherited allele. Ident (see ould enhanytes ilent domain. In a similar manner, chromosome 11 containsnd IGF2 share ification of a activator function associated with NHS2 The activation function associated with NHS2 is independent of orientation and position but dependent on the distance with respect to the SNURF-SNRPN promoterFigure 3-1). However, these characteristics were determined in transient expression assays. Thus, it is possible that at the endogenous locus, where other elements may be present, the activator does not depend on the distance to the target promoter. This wallow NHS2 to activate the SNURF-SNRPN upstream promoters which have been implicated in the paternal to maternal switch in the female germline. Therefore, an cer competition model can be proposed to explain the regulation of the imprint switch in the PWS-AS region. In this model NHS2 would activate the SNURF-SNRPN main promoter in the paternal allele/male germline and it would activate the upstream promoters in the female germline. Activation of the main promoter would facilitate the establishment of an open domain while activation of the upstream promoters in oocwould assist in the establishment of a s an imprinted domain that includes the H19 and IGF2 genes. H19 a a group of enhancers that activate H19 on the unmethylated maternally-inherited allele and IGF2 on the methylated paternally-inherited allele (Thorvaldsen et al., 1998).CTCF binding to an intervening sequence in the maternally-inherited allele impedes interaction of the enhancers with the IGF2 promoter. On the other hand, in the
108 methylated paternally-inherited allele CTCF binding is blocked and the enhancers cactivate IGF2 expression (Hark et al., 2000), (Szabo et al., 2000). It is possible thatsimilar situation occurs at the PWS/AS-IC, where the intronic activator may stimulate transcription exclusively from the SNURF-SNRPN upstream promoters in the female germline and mostly from the SNURF-SNRPN main promoter in the an a male germline and, ites have been identi to ns and mline SNURF-SNRPN locus Despite low level of sequence conservation in the SNURF-SNRPN locus between the human and the mouse homologous regions outside from the coding region, both human and mouse SNURF-SNRPN promoters were activated by NHS2 in transient later in development the paternal allele. Putative CTCF binding s fied at the main and upstream promoters of the SNURF-SNRPN gene. However, the studies described here showed no detectable in vivo association of CTCF with the SNURF-SNRPN promoter or NHS2 in lymphoblasts. It is possible that CTCF bindingthese sites is limited to the female germline before the imprints are re-set. In that case CTCF would participate in the establishing of a repressive state in the maternally-inherited chromosome q11-q13 that then may be propagated by histone modificatioDNA methylation. Consistent with this idea, CTCF is not expressed in the male gerat the time of the imprint switch, when the male germline restricted factor BORIS is expressed (Klenova et al., 2002; Loukinov et al., 2002). BORIS shares the same DNA binding domain and DNA target sequences with CTCF and has been implicated in genomic imprinting by having a negative dominant effect on CTCF. In the male germine, BORIS would then prevent CTCF binding to the IC and the initiation of a repressive domain in the PWS-IC. Definition of an Activation Conserved Sequence (ACS) in the human and the mouse
109 expression assays in human cells (see Figure 3-4). Therefore it was likely that the mouhomologous region included an element similar to NHS2. Sequence comparison of thehuman 2.2-NHS2 and the entire mouse Snurf-Snrpn intron 1 showed an 80 bp fragmenttermed ACS (Activating Conserved Sequence) that is conserved in both species (see Figure 3-5). Furthermore, this region partially overlaps with the minimal activator sequence (MAS) (see Figure 3-3) and corresponds to the approximate position ofin intron 1. Knockout studies in the mouse have shown that deletion of a 4.8 kb fragmearound Snurf-Snrpn exon 1, which includes the ACS, resulted in 50% lethality after paternal transmission and in a partial imprinting defect that involved gain of a maternal epigenotype in the paternally-inherited imprinted domain. A second 900 bp deletion (from -600 to +300) on this region, that included the Snurf-Snrpn promoter and firsbut did not include the ACS had no effect on the imprinting status of the region (Bressler et al., 2001). These two deletion studies further emphasized the possible involvement of the ACS in the regulation of imprinted gene expression on the paternal allele in the PAS associated domain. In contrast to the partial imprinting defect observed in the 4.8 kbdeletion mice, a complete imprinting defect was observed when a larger fragment oSnurf-Snrpn sequence was deleted (23 kb upstream and 20 kb downstream of the transcription start site) (Yang et al., 1998). This suggests that there are IC elements within this 43 kb region that were no se , NHS2 nt t exon WSf t included in the 4.8 kb deletion. Other knockout Snurf-Snrpn exon 2 to Ube3A (Tsai et al., 1999b), the deletion of the Snurf-Snrpn coding studies in which the ACS was left intact, including the deletion of the sequence between
110 Figure 5-1: Targeted deletions of the Snurf-Snrpn locus in mouse. Depicted on top is the PWS-AS homologous region in mouse chromosome 7 C. Represented on the center is a map of the Snurf-Snrpn locus in mouse. Black boxes symbolize coding exons while grey boxes symboli ze upstream non-coding exons. The bent arrow indicated the tr anscription initiation site and the red block indicates the approximate position of the ACS. Horizontal arrows correspond to the regions removed by targeted deletion. Green arrows represent deletions that had no effect on imprinted gene expr ession/IC function while red arrows represent deletions that did have an effect on imprinted gene expression/IC function. region (Yang et al., 1998) and the deletion of 110 kb of Snurf-Snrpn upstream sequence (from -11 kb to -121 kb) (Chen et al., 2002), resulted in no imprinting defects. Additionally, deletion of 1.5 Mb of Snurf-Snrpn upstream sequence (-121 kb to 1621 kb) affected imprinted gene expression in the ma ternally-inherited domai n, indicating that the AS-IC but not the PWS-IC is included in that region. However, the mechanisms of imprinting se em to have somewhat diverged between the mouse and the human. First, a 76 kb human transgene incl uding the ASand PWSSRO and the SNURF-SNRPN coding sequence was expressed from both paternal alleles regardless of copy number (B laydes et al., 1999). A si milar transgene including the mouse sequence showed imprinted expression when integrated in two copies although Mouse 7C Mouse 7C Tel Cen1 234-10 Cen Tel Mkrn3 Magel 2 S nur f S nrpn Ipw U be 3a as Ndn SnoR N A S noR N A Mkrn3 as Ube 3a< < < < < < Frat3 Approx -23 to +20 Approx -2.8 to +2 Approx -0.6 to +0.3 ACS ACS IC Mouse 7C Mouse 7C Tel Cen1 234-10 Cen Tel Mkrn3 Magel 2 S nur f S nrpn Ipw U be 3a as Ndn SnoR N A S noR N A Mkrn3 as Ube 3a< < < < < < < < < < < < Frat3 Approx -23 to +20 Approx -2.8 to +2 Approx -0.6 to +0.3 ACS ACS IC
111 not when integrated as a single copy (Blaydes et al., 1999). Second, a transgene including the human AS-SRO and the mouse Snurf-Snrpn promoter showed imprinted expression and methylation patters while a similar transgene in which the human promoter replaced the mouse did not (Shemer et al., 2000), (Hershko et al., 2001). This indicates that the AS-IC function seems to be conserved in human and mouse while the PWS-IC function does not seem to be conserved. That would explain why deletion of the mouse Snurf-Snrpn promoter has no effect on imprinted gene expression in the PWS-AS associated domain after paternal transmission while the human SNURF-SNRPN promoter elements eAnalysis oAnalNHS2 revetranscription factor that can act as a transcription repressor, perhaps by interaction with histone and chromatin modifying complexes, and as a transcription activator possibly by recruiting components of the general transcriptional machinery such as RNA polymerase II or histone and chromatin remodeling complexes (reviewed in (Thomas and Seto, 1999)). Additionally, recent studies showed that YY1 can act as an insulator element. That is the case for the PEG3 imprinted locus where YY1 binds in tandemly repeated copies and in vitro studies suggests can act as an insulator. NRF-1 is a transcription factor associated in general with genes involved in oxidative phosphorylation and metabolic processes. Indeed, NRF-1 has been proposed as a key element in the response to changes of the metabolic resource levels in the cell. Therefore, the role of NRF-1 in the SNURF-SNRPN locus may go beyond activating a which is located within the PWS-SRO is still a major candidate to include cis-acting ssential for the proper imprinting of the paternally-inherited allele. f cis-acting elements in the ACS/NHS2 ysis using the transcription factor database (TRANSFAC) of the ACS in aled possible binding sites for YY1, NRF-1 and SP1. YY1 is a versatile
112 gene which is not to a usual NRF-1 target. Instead, it is possible that NRF-1 is inin the parental conflict in genomic imprinting. According to this model, maternal and paternal interests conflict in the advantages that expression of imprinted genes havthe members of the same kin. In addition to the binding site in ACS/NHS2, there are twoNRF-1 binding sites in the SNURF-SNRPN main as well as upstream promoters in the human and the mouse genes, which may have functional implications. The putative YY1 binding site is present in both the ACS and the MAS within NHS2. Mutation of this site in transient expression analysis showed a dramatic downregulation of the activation, which indicated the involvement of the putative YYbinding sequence in the activator function. Furthermore, YY1 binding to the ACS/NHS volved e for 1 2 ChIP analysis, respectively. In additir as r ere , therefore, non-processive form HS2 II, was demonstrated in vitro and in vivo by EMSA and on, NRF-1 was also found associated in vivo with the ACS/NHS2 and, at lowelevels, with SNURF-SNRPN 5â€™ region/NHS1. However no significant association wdetected with the non-hypersensitive non-transcribed SNURF-SNRPN upstream promoteU1A. Since YY1 is known to recruit RNA Pol II to some gene promoters, the levels of RNA Pol II present at the ACS/NHS2 and SNURF-SNRPN 5â€™ region/NHS1 in vivo wexamined by ChIP analysis. Unphosphorylated RNA Pol II is generally associated withinitiation of transcription while phosphorylated RNA Pol II is generally associated withtranscription elongation. RNA pol II in its unphosphorylated was found to be associated in vivo with the paternally-inherited allele of 2.2-Nat higher levels than with the promoter. This suggests that YY1, or a complex that includes YY1 (and NRF-1), may facilitate recruitment to the ACS/NHS2 of RNA pol
113 that may become more stably associated by interaction with NRF-I and possibly SP1. The polymerase may then be transferred to the promoter through mechanisms that are discussed below. One of the footprints identified on the paternally-inherited allele of theSNURF-SNRPN 5â€™ region/NHS1 is a potential E2F binding site/initiator element. It is known that E2F and YY1 can interact upon mediation of RYBP (RING-1 and YY1 binding protein). Additionally, NRF-1 binding sites are common to all three ACS/NHS2, SNUR as it is e 1. -sible that the same mechanism of recruitment and transfer of RNA Pol II is used to act F-SNRPN 5â€™ region/NHS1 and the SNURF-SNRPN upstream promoters as wellto the Snurf-Snrpn main and upstream promoters in the mouse gene. Therefore, possible to speculate that the transfer of RNA Pol II from ACS/NHS2 to the SNURF-SNRPN promoters may occur though bending of the intervening DNA that allows the interaction between YY1 and E2F and the transfer from a single NRF-1 and SP1 in thACS/NHS2 to two NRF-1 and multiple SP1 factors in SNURF-SNRPN 5â€™ region/NHSThis is consistent with the ChIP data that showed association of NRF-1 with the SNURFSNRPN 5â€™ region/NHS1 but not with the upstream promoter U1A which is not expressed. It is pos ivate the SNURF-SNRPN upstream promoters in the female germline and the SNURF-SNRPN main promoter in the male germline. YY1 binding to DNA is methylation sensitive therefore it would not be expected to bind the maternal allele in somatic cells. Additionally, the cellular localization of YY1 is mostly cytoplasmatic in oocytes in mammals. However it is possible that YY1 binds the maternal allele and activates the upstream promoter prior to the establishment of the methylation imprint on the female germline early in gametogenesis.
114 Recruitment of RNA pol II to distant control regions has been proposed for other gene locus. The LCR in the -globin locus has been proposed to recruit RNA pol II (Johnson et al., 2001) and then loop to track and interact with the -globin gene promoter 50 kb away (Carter et al., 2002). A similar looping/tracking mechanism has been suggested for the HNF-4 upstream regulatory regions (Hatzis and Talianidis, 2002)YY1 and NRF-1 are known to interact with histone modifying enzymes (Izumi et al., 2003),(Rezai-Zadeh et al., 2003), (Izumi et al., 2003) that may participate in the initiation/maintenance of silent chromatin in the maternally-inherited imprinted domaand/or the formation/maintenance of active chromatin in the paternally-inherited imprinted domain. Histone modifications at the ACS/NHS2 site were examined on eaallele individually and compared to the SNURF-SNRPN 5â€™ region/NHS1. As expected from previously published data (Fulmer-Smentek and Francke, 2001; Saitoh and Wada, 2000) both the SNURF-SNRPN 5â€™ region and the ACS/NHS2 showed preferential histoneH3 and 4 hyperacetylation associated with the paternally-inherited allele and H3/H4 hypoacetylation associated with the maternally-inherited allele. In addition, H3-K4 methylation of paternal origin has been reported in the SNURF-SNRPN 5â€™ region/NHwith very distinct boundaries just a few hundreds of base pairs upstream and dowfrom the SNURF-SNRPN transcription initiation site (Xin et al., 2001). However, ChIPanalysis detected association of H3-K4 methylation with the paternally-inherited allele oACS/NHS2, which is located approximately 1.5 kb downstream form the SNURF-SNRPN transcription initiation site. Furthermore, the level of H3-K4 methylation was higher in the ACS/NHS2 than in SNURF-SNRP . in ch S1 nstream f N 5â€™ region/NHS1, which is consistent with the higher levels of NRF-1 and RNA Pol II in ACS/NHS2 from the two regions.
115 Sinceuit strong In e o MRe to the intervening sequence lacks this H3-K4 methylation mark (according to Xin et al.) it is possible that both ACS/NHS2 and SNURF-SNRPN 5â€™ region/NHS1 sites recrH3-K4 methyltransferases individually or that these enzymes are also recruited to the ACS/NHS2 and then transferred to SNURF-SNRPN 5â€™ region/NHS1. YY1 is known to associate with the nuclear matrix (Guo et al., 1995). Evidence ly suggests that the nuclear matrix is key at organizing nuclear functions such as transcription (Jackson et al., 1993), (Cook, 1999) (see Introduction for more details).general transcriptionally active sequences are more tightly associated with the nuclear matrix while transcriptionally inactive sequences often extend into the nuclear halo (Gerdes et al., 1994). Accordingly, the paternal and transcriptionally active allele of the SNURF-SNRPN gene has been shown to be associated with the nuclear matrix, while thmaternal allele shows up in the nuclear halo of nuclear matrix preparations. Differential association to the nuclear matrix has been reported for other imprinted genes such as IGF2 where several differentially methylated regions (DMR) are in close proximity tMARs that associate with the nuclear matrix on the paternal (expressed) allele in a Ddependent manner (Weber et al., 2003). Therefore, it is possible that differential association of imprinted loci with the nuclear matrix is a mechanism common to imprinting control region or imprinting centers that leads to differential gene expression.In the PWS-AS associated domain, YY1 may target the IC to the nuclear matrix in the male germline but not in the female germline since YY1 binding to DNA is sensitivDNA methylation and since YY1 is mostly excluded from the nucleus in oocytes. Continuous association of the paternally-inherited allele with the nuclear matrix during
116 somatic development may be responsible for maintaining the active domain and protect ifrom the reprogramming events that occur during embryo development. Finally, the process of replication is tightly associated with the nuclear matrix (Hozak et al., 1993). Transcription and replication processes are intimat t ely connected with g ene expression. In general, actively transcribed genes replicate earlier in S phasethan silent genes, which replicate late. As might be expected, the AS-PWS associated imprinted domain replicates asynchronously, with the transcriptionally active paternally-inherited copy replicating earlier in S-phase than the silent maternally-inherited copy(Gunaratne et al., 1995). This pattern of asynchronous replication is already established in the gametes (Simon et al., 1999). This finding reinforces the importance ofthe nuclear matrix in the organization of replicational and transcriptional imprints during germline development.
APPENDIX A VECTOR DESIGN FOR TRANSIENT EXPRESSION ASSAYS SNURF-SNRPN fragments were derived from plasmid pHTRM1.8; MKRN3 fragments were derived from the plasmid pDN34 (Jong et al., 1999); UBE3A fragmentswere derived plasmid pUBE3A-3.3. Snurf-Snrpn fragments were derived form plasmid pGN73 The fragments indicated in the figures were cloned into pGL3-basic vect(Promega) using commercial enzymes: in SNURF/SNRPN â€“3999 (Pvu II), -676 (B455 (Rsa I), -207 (Xba I), +80 (Pvu II). UBE3A -455 (Ava I); +75 (Ava I). MKRN3 -890 (Sap I) 45 (Taq I). Snurf-Snrpn -567 (Nde I); +56 (Pvu II). Fragments for U1A, U1B, U1-A and U1-C were gen or gl II), -erated by PCR. In order to clone the +80 position in frame with the luciferase gene a linker including a Pvu II and a Nco I site (724-5: 5â€™-agctcagctgcgatgga-3â€™) was designed and cloned into PGL3/HindIII to generate PGL3-L. The fragment -3999 to +80 was cloned into pGL3L/Pvu II (pL-PS). This construct was digested with Bgl II and re-ligated to obtain the fragment -676 to +80 (pL-PBS). The fragmnt -207 to +80 was excised from pL-PBS with Xba I and Hind III and introduced into pGL3/Nhe I+Hind III (p207). The fragment to +53 was generated by PCR on a plasmtt3â€™ and 737: 5â€™-accgctcctcagacagatgc-3â€™ annealing temp 50 C) and cloned into PCR 2.1 using the TA original cloning kit (Invitrogen) from which it was introduced into PGL3/Nhe I+Hind III after digestion with Xba I+Hind III (pL-B). The fragment -336 to +53 was obtained by combining pL-PBS/Rsa I+Apa I with pL-B/Sma I+Apa I. MKRN3 promoter was cloned from the plasmid pDN34 with EcoR I and Sap I (fill-in) and into the Sma I site in pGL3. UBE3A e id template (647: 5â€™-tgtggttatggcgcat 117
118 promoter was digested with Ava I (-455 and +70) filled-in with klenow polymerase and ligated into the Sma I site in pGL3. U1A promoter region was PCR amplified from human genomic DNA (TY 869 5â€™-tttctctctagggccttctcttagc-3â€™ TY 870 5â€™-tcaaacagctc excised with E ouse n-ly into pL-PS cagagggaagac-3â€™ 54 C) cloned into pCRII vector from which it was coR I, filled in with klenow and introduced into pGL3/Sma I. U1B promoterregion was PCR amplified from human genomic DNA (TY 871 5â€™-acaagaattttccatccacagcc-3â€™ TY 872 5â€™-ccccgcaaaccacacagaa-3â€™ 54 C) cloned into pCRII vector from which it was excised with EcoR I andinserted into pGL3/Sma I. The mSnurf-Snrpn promoter was excised from the pGN73 plasmid with Nde I and Pvu II and inserted into pGL3/Sma I. 2.2-NHS2 EcoR I-Sma I fragment from SNURF-SNRPN intron 1 was subcloned into pBluescript/EcoR I in both orientations (pB-En-F and pB-ER). HS2 was cloned downstream of the luciferase gene the forward and reverse orientations as pB-En-R and pB-En-F/BamH I+Sal I fragments respectiveand pL-PBS/BamH I+Sal I; HS2 was cloned upstream of the luciferase gene in the forward and reverse orientations as pB-En-R and pB-En-F/Xba I+Xho I fragments into pL-PS and pL-PBS/Nhe I+Xho I.
APPENDIX B ANALYSIS OF 2.2-NHS2 ON SNURF-SNRPN PROMOTER ELEMENTS The effect of 2.2-NHS2 on increasingly truncated fragments of SNURF-SNRPpromoter region was analyzed in transient expression assays. ReporFigure B-1) that included the indicated fragments of SNURF-SNRPN 5â€™ region upstream of the luciferase gene and reporter constructs that constructs that additionally include 2.2-NHS2 downstream of the reporter gene were transiently transfected into SK-N-SH cells.Figure B-1 shows the resulting relative luciferase activity for each construct. Progrestruncation of the SNURF-SNRPN promoter in constructs without 2.2-NHS2, produced a gradual increase in the luciferase activity (see also Figure 2-1). It is interesting tothat in this experiment, in which the transfection reagent FuGene 6 was used instead of Superfect, the promoter deletions showed a less dramatic effect (compare Figure B-1with Figure 2-1). The constructs including the same truncated fragments of SNURF-SNRPN promoter and 2.2-NHS2 sho N ter constructs (see sive note wed fairly uniform levels of luciferase activity. Therefore, the 2.2-NHS2 seems to compensate for the various negative regulatory elements identified in the SNURF-SNRPN promoter in transient expression assays. Alternatively, the relative fold induction caused by introduction of 2.2-NHS2 into reporter constructs containing fragments of SNURF-SNRPN promoter can be taken into consideration. These fold inductions, that are shown in Figure B-1, drop significantly after deletion of the region between -454 and -207. This suggests that the abovementioned region may be partially responsible for the interaction between the SNURF-SNRPN promoter and the activation function within 2.2-NHS2. 119
120 included different fragments of the SNURF-SNRPN promoter without (black compared to a construct containing all the regulatory elements of the SNURFFigure B-1: Functional interaction between the activator associated with 2.2-NHS2 and components of the SNURF-SNRPN gene promoter. This analysis was performed by transient expression assays. Luciferase reporter constructs bars) or with (grey bars) 2.2-NHS2. Relative luciferase activities are shown SNRPN promoter and lacking 2.2-NHS2. The fold induction observed upon introduction of 2.2-NHS2 is indicated on the right. Error bars represent standard deviation of the means. HS2-207 to +808.6X3.7XRelative Luciferase Activity 024681012141618 + HS2 -207 to +53-675 to +53-454 to +53-675 to +807.1X10.4X4.3X
LIST OF REFERENCES Antequera, F, Macleod, D, and Bird, A P (1989). Specific protection of methylated CpGs in mammalian nuclei. Cell 58, 509-517. Bandyopadhyay, D, and Medrano, E E (2003). The emerging role of epigenetics in cellular and organismal aging. Exp Gerontol 38, 1299-1307. 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, 1389-1397. Berezney, R, and Coffey, D S (1974). Identification of a nuclear protein matrix. Biochem Biophys Res Commun 60, 1410-1417. Berezney, R, and Coffey, D S (1977). Nuclear matrix. Isolation and characterizatioework structure from rat liver nuclei. J Cell Biol 73, 616-637. R, and Wei, X (1998). The new paradigm: integrating genomic function and ar architecture. J Cell Biochem Suppl 30-31, 238-242. (1992). Activation of mammalian DNA methyltransferase by cleavage oinding regulatory domain. Embo J n of a framBerezney, nucleBestor, T Hf a Zn b 11, 2611-2617. Bielinska, B, Blaydes, S M, Buiting, K, Yang, T, Krajewska-Walasek, M, Horsthemke, B, and Brannan, C I (2000). De novo deletions of SNRPN exon 1 in early human nd mouse embryos result in a paternal to maternal imprint switch. Nat Genet 25, 74-78. Blaydes, S M, Elmore, M, Yang, T, and Brannan, C I (1999). Analysis of murine Snrpn nd human SNRPN gene imprinting in transgenic mice. Mamm Genome 10, 549-555. Braidotti, G, Borthwick, I A, and May, B K (1993). Identification of regulatory sequences in the gene for 5-aminolevulinate synthase from rat. J Biol Chem 268, 1109-1117. Brannan, C I, and Bartolomei, M S (1999). Mechanisms of genomic imprinting. Curr Opin Genet Dev 9, 164-170. 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, 232-240. a a 121
122 Buiting, K, Dittrich, B, Gross, S, Lich, C, Farber, C, Buchholz, T, Smith, E, Reis, A, Burger, J, Nothen, M M, et al. (1998). Sporadic imprinting defects in Prader-Willi syndrome and Angelman syndrome: implications for imprint-switch models, genetic counseling, and prenatal diagnosis. Am J Hum Genet 63, 170-180. Buiting, K, Lich, C, Cottrell, S, Barnicoat, A, and Horsthemke, B (1999). A 5-kb imprinting center deletion syndrome reduces the shortest region of deletion overlap to 880 bp. Hum Genet 105, 665-666. Buitinwartz, S, Nicholls, R D, and Horsthemke, B (1995). Inherited microdeletions in the Angelman and Prader-Willi syndromes Cai, H N, and Shen, P (2001). Effects of cis arrangement of chromatin insulators on Cai, S, and Kohwi-Shigematsu, T (1999). Intranuclear relocalization of matrix binding e 7 gene cluster. Cassiilli and Angelman CattaChaillet, J R, Bader, D S, and Leder, P (1995). Regulation of genomic imprinting by Chamn, S J, and Brannan, C I (2001). The Prader-Willi syndrome imprinting center activates the paternally expressed murine Ube3a antisense transcript but represses Chandler, S P, Guschin, D, Landsberger, N, and Wolffe, A P (1999). The methyl-CpG binding transcriptional repressor MeCP2 stably associates with nucleosomal DNA Chau, C M, Evans, M J, and Scarpulla, R C (1992). Nuclear respiratory factor 1 -7006. in a family with Angelman g, K, Saitoh, S, Gross, S, Dittrich, B, Sch define an imprinting centre on human chromosome 15. Nat Genet 9, 395-400. enhancer-blocking activity. Science 291, 493-495. sites during T cell activation detected by amplified fluorescence in situ hybridization. Methods 19, 394-402. Carter, D, Chakalova, L, Osborne, C S, Dai, Y F, and Fraser, P (2002). Long-range chromatin regulatory interactions in vivo. Nat Genet 32, 623-626. Caspary, T, Cleary, M A, Baker, C C, Guan, X J, and Tilghman, S M (1998). Multiple mechanisms regulate imprinting of the mouse distal chromosom Mol Cell Biol 18, 3466-3474. dy, S B, Dykens, E, and Williams, C A (2000). Prader-W syndromes: sister imprinted disorders. Am J Med Genet 97, 136-146. nach, B M, and Kirk, M (1985). Differential activity of maternally and paternally derived chromosome regions in mice. Nature 315, 496-498. gametic and embryonic processes. Genes Dev 9, 1177-1187. berlai paternal Ube3a. Genomics 73, 316-322. Biochemistry 38, 7008-7018. activation sites in genes encoding the gamma-subunit of ATP synthase, eukaryotic initiation factor 2 alpha, and tyrosine aminotransferase. Specific interaction of purified NRF-1 with multiple target genes. J Biol Chem 267, 6999
123 Chen, C, and Yang, T P (2001). Nucleosomes are translationally positioned on the actallele and rotationally positioned on the inactive allele of the HPRT promoter. Mol Cell Biol 21, 7682-7695. ive Chen, K-S, Tsai, T-F, and L, B A (2002). An insertion/duplication mutation 11 kb Chen, S, Nagy, P L, and Zalkin, H (1997). Role of NRF-1 in bidirectional transcription of Clayton, A L, and Mahadevan, L C (2003). MAP kinase-mediated phosphoacetylation of Clayton, D A (1992). Transcription and replication of animal mitochondrial DNAs. Int Cockhromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing Cook, P R (1999). The organization of replication and transcription. Science 284, 1790-Crouse, H V (1960). The nature of the influence of x-translocations on sex of progeny in Sciara coprophila. Chromosoma 11, 146-166. Davis9 methylation imprint is erased and re-established differentially on the parental DeChiara, T M, Robertson, E J, and Efstratiadis, A (1991). Parental imprinting of the Dietzel, S, Schiebel, K, Little, G, Edelmann, P, Rappold, G A, Eils, R, Cremer, C, and Dittrich, B, Buiting, K, Korn, B, Rickard, S, Buxton, J, Saitoh, S, Nicholls, R D, Poustka, geted upstream of snurf-snrpn produces a mouse model of an angelman syndromeimprinting mutation. Paper presented at: ASHG Meeting. the human GPAT-AIRC purine biosynthesis locus. Nucleic Acids Res 25, 1809-1816. histone H3 and inducible gene regulation. FEBS Lett 546, 51-58. Rev Cytol 141, 217-232. erill, P N, and Garrard, W T (1986). C topoisomerase II sites. Cell 44, 273-282. 1795. , T L, Yang, G J, McCarrey, J R, and Bartolomei, M S (2000). The H1 alleles during male germ cell development. Hum Mol Genet 9, 2885-2894. mouse insulin-like growth factor II gene. Cell 64, 849-859. Cremer, T (1999). The 3D positioning of ANT2 and ANT3 genes within female X chromosome territories correlates with gene activity. Exp Cell Res 252, 363-375. A, Winterpacht, A, Zabel, B, and Horsthemke, B (1996). Imprint switching on human chromosome 15 may involve alternative transcripts of the SNRPN gene. Nat Genet 14, 163-170. Donohoe, M E, Zhang, X, McGinnis, L, Biggers, J, Li, E, and Shi, Y (1999). Tardisruption of mouse Yin Yang 1 transcription factor results in peri-implantation lethality. Mol Cell Biol 19, 7237-7244.
124 Dunn, K L, Zhao, H, and Davie, J R (2003). The insulator binding protein CTCF associates with the nuclear matrix. Exp Cell Res 288, 218-223. Ehrlich, M, Gama-Sosa, M A, Huang, L H, Midgett, R M, Kuo, K C, McCune, R A, and A El-Maarri, O, Buiting, K, Peery, E G, Kroisel, P M, Balaban, B, Wagner, K, Urman, B, , 4. m NRF-1, ATF, and intron Sp1 recognition sequences. J Biol Chem 264, 14361-14368. Evans, M J, and Scarpulla, R C (1990). NRF-1: a trans-activator of nuclear-encoded respiratory genes in animal cells. Genes Dev 4, 1023-1034. Farbeation of a r-Willi syndrome region on chromosome 15. Genomics 65, 174-183. Felsenfeld, G, and McGhee, J D (1986). Structure of the 30 nm chromatin fiber. Cell 44, 375-377. FiczyM, and Ovsenek, N (2001). Expression, activity, and subcellular localization of the Yin Yang 1 transcription Filipp, T R, Lobanenkov, V V, and Tapscott, S J (2001). CTCF-binding sites flank CTG/CAG 8, in Fournier, C, Goto, Y, Ballestar, E, Delaval, K, Hever, A M, Esteller, M, and Feil, R Fulmncke, U (2001). Association of acetylated histones with paternally expressed genes in the Prader--Willi deletion region. Hum Mol Genet Gehrke, C (1982). Amount and distribution of 5-methylcytosine in human DNfrom different types of tissues of cells. Nucleic Acids Res 10, 2709-2721. 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 27341-34 Evans, M J, and Scarpulla, R C (1989). Interaction of nuclear factors with multiple sites in the somatic cytochrome c promoter. Characterization of upstrea r, C, Gross, S, Neesen, J, Buiting, K, and Horsthemke, B (2000). Identifictestis-specific gene (C15orf2) in the Prade cz, A, Eskiw, C, Meyer, D, Marley, K E, Hurt, factor in Xenopus oocytes and embryos. J Biol Chem 276, 22819-22825.ova, G N, Thienes, C P, Penn, B H, Cho, D H, Hu, Y J, Moore, J M, Klesert repeats and form a methylation-sensitive insulator at the DM1 locus. Nat Genet 2335-343. Fischle, W, Wang, Y, and Allis, C D (2003). Histone and chromatin cross-talk. Curr OpCell Biol 15, 172-183. (2002). Allele-specific histone lysine methylation marks regulatory regions at imprinted mouse genes. Embo J 21, 6560-6570. er-Smentek, S B, and Fra 10, 645-652.
125 Gautsch, J W, and Wilson, M C (1983). Delayed de novo methylation in teratocarsuggests additional tissue-specific mechanisms for controlling g cinoma ene expression. Nature 301, 32-37. Gerare itu Geuns, E, De Rycke, M, Van Steirteghem, A, and Liebaers, I (2003). Methylation Glenn, C C, Nicholls, R D, Robinson, W P, Saitoh, S, Niikawa, N, Schinzel, A, atients. Hum Mol Genet 2, 1377-1382. GlennA methylation, and imprinted expression of the human SNRPN gene. Am J Hum Genet 58, 335-346. Gray, T A, Saitoh, S, and Nicholls, R D (1999). An imprinted, mammalian bicistronic transcript encodes two independent proteins. Proc Natl Acad Sci U S A 96, 5616-Greal99). Conserved characteristics of heterochromatin-forming DNA at the 15q11-q13 Green dynamics. EMBO Rep 3, 28-33. romoter and cis elements within it. Gene 304, 201-206. B 3, af1-rs1. Mol Cell Biol 21, 5426-5436. Gross 57, 159-197. d, M, Hernandez, L, Wevrick, R, and Stewart, C L (1999). Disruption of the mousnecdin gene results in early post-natal lethality. Nat Genet 23, 199-202. Gerdes, M G, Carter, K C, Moen, P T, Jr., and Lawrence, J B (1994). Dynamic changes in the higher-level chromatin organization of specific sequences revealed by in shybridization to nuclear halos. J Cell Biol 126, 289-304. imprints of the imprint control region of the SNRPN-gene in human gametes and preimplantation embryos. Hum Mol Genet 12, 2873-2879. Horsthemke, B, and Driscoll, D J (1993). Modification of 15q11-q13 DNA methylation imprints in unique Angelman and Prader-Willi p , C C, Saitoh, S, Jong, M T, Filbrandt, M M, Surti, U, Driscoll, D J, and Nicholls, R D (1996). Gene structure, DN 5621. ly, J M, Gray, T A, Gabriel, J M, Song, L, Zemel, S, and Nicholls, R D (19 imprinting center. Proc Natl Acad Sci U S A 96, 14430-14435. , C M, and Almouzni, G (2002). When repair meets chromatin. First in series on chromatin Green Finberg, Y, Kantor, B, Hershko, A Y, and Razin, A (2003). Characterization of the human Snrpn minimal p Gregory, R I, Randall, T E, Johnson, C A, Khosla, S, Hatada, I, O'Neill, L P, Turner,M, and Feil, R (2001). DNA methylation is linked to deacetylation of histone Hbut not H4, on the imprinted genes Snrpn and U2 , D S, and Garrard, W T (1988). Nuclease hypersensitive sites in chromatin. AnnuRev Biochem
126 Gugneja, S, and Scarpulla, R C (1997). Serine phosphorylation within a concise amino-terminal domain in nuclear respiratory factor 1 enhances DNA binding. J Biol Chem 272, 18732-1 8739. nted human Prader-Willi syndrome region. Genes Dev 9, 808-820. Guo, ein NMP-1 is the transcription factor YY1. Proc Natl Acad Sci U S A 92, 10526-10530. Hark,cking activity at the H19/Igf2 locus. Nature 405, 486-489. Hata,family 83-1993. Hendrich, B, and Tweedie, S (2003). The methyl-CpG binding domain and the evolving Hendzel, M J, Sun, J M, Chen, H Y, Rattner, J B, and Davie, J R (1994). Histone acetyltransferase is associated with the nuclear matrix. J Biol Chem 269, 22894-Heng1). Re-defining the chromatin loop domain. Cytogenet Cell Genet 93, 155-161. Henryd, C, Schwartz, C, Utermann, G, and Junien, C (1991). Uniparental paternal disomy in a genetic cancer-predisposing Hernandez, J M, Giner, P, and Hernandez-Yago, J (1999). Gene structure of the human 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 Herzing, L B, Kim, S J, Cook, E H, Jr., and Ledbetter, D H (2001). The human aminophospholipid-transporting ATPase gene ATP10C maps adjacent to UBE3A and exhibits similar imprinted expression. Am J Hum Genet 68, 1501-1505. Gunaratne, P H, Nakao, M, Ledbetter, D H, Sutcliffe, J S, and Chinault, A C (1995). Tissue-specific and allele-specific replication timing control in the impri B, Odgren, P R, van Wijnen, A J, Last, T J, Nickerson, J, Penman, S, Lian, J B, Stein, J L, and Stein, G S (1995). The nuclear matrix prot A T, Schoenherr, C J, Katz, D J, Ingram, R S, Levorse, J M, and Tilghman, S M (2000). CTCF mediates methylation-sensitive enhancer-blo K, Okano, M, Lei, H, and Li, E (2002). Dnmt3L cooperates with the Dnmt3 of de novo DNA methyltransferases to establish maternal imprints in mice. Development 129, 19 Hatzis, P, and Talianidis, I (2002). Dynamics of enhancer-promoter communication during differentiation-induced gene activation. Mol Cell 10, 1467-1477. role of DNA methylation in animals. Trends Genet 19, 269-277. 22901. , H H, Krawetz, S A, Lu, W, Bremer, S, Liu, G, and Ye, C J (200 , I, Bonaiti-Pellie, C, Chehensse, V, Beldjor syndrome. Nature 351, 665-667. mitochondrial outer membrane receptor Tom20 and evolutionary study of its family of processed pseudogenes. Gene 239, 283-291. 6, 967-975.
127 Hornstra, I K, and Yang, T P (1992). Multiple in vivo footprints are specific to the activeallele of the X-linked human hypoxanthine phosphoribosyltransferase gene 5' region: implications for X chromosome inactivation. Mol Cell Biol 12, 5345-5354. tion Huo, L, and Scarpulla, R C (2001). Mitochondrial DNA instability and peri-implantation ce. Huq, A H, Sutcliffe, J S, Nakao, M, Shen, Y, Gibbs, R A, and Beaudet, A L (1997). Hyde-DeRuyscher, R P, Jennings, E, and Shenk, T (1995). DNA binding sites for the Ivanc92). Interaction of MAR-sequences with nuclear matrix proteins. J Cell Biochem 50, 190-200. Izumi K (2003). p300/CBP-associated factor (P/CAF) interacts with nuclear respiratory factor-1 to f focal Jaenisch, R, and Bird, A (2003). Epigenetic regulation of gene expression: how the ol antisense transcript in the Prader-Willi syndrome critical region. Hum Mol Genet 8, 783-793. Hozak, P, Hassan, A B, Jackson, D A, and Cook, P R (1993). Visualization of replicafactories attached to nucleoskeleton. Cell 73, 361-373. lethality associated with targeted disruption of nuclear respiratory factor 1 in miMol Cell Biol 21, 644-654. Sequencing and functional analysis of the SNRPN promoter: in vitro methylationabolishes promoter activity. Genome Res 7, 642-648. transcriptional activator/repressor YY1. Nucleic Acids Res 23, 4457-4465. henko, M, and Avramova, Z (19 , H, Ohta, R, Nagatani, G, Ise, T, Nakayama, Y, Nomoto, M, and Kohno, regulate the UDP-N-acetyl-alpha-d-galactosamine: polypeptide N-acetylgalactosaminyltransferase-3 gene. Biochem J 373, 713-722. Jackson, D A, Hassan, A B, Errington, R J, and Cook, P R (1993). Visualization osites of transcription within human nuclei. Embo J 12, 1059-1065. genome integrates intrinsic and environmental signals. Nat Genet 33 Suppl, 245-254. Javahery, R, Khachi, A, Lo, K, Zenzie-Gregory, B, and Smale, S T (1994). DNA sequence requirements for transcriptional initiator activity in mammalian cells. MCell Biol 14, 116-127. Johnson, K D, Christensen, H M, Zhao, B, and Bresnick, E H (2001). Distinct mechanisms control RNA polymerase II recruitment to a tissue-specific locus control region and a downstream promoter. Mol Cell 8, 465-471. Jong, M T, Gray, T A, Ji, Y, Glenn, C C, Saitoh, S, Driscoll, D J, and Nicholls, R D (1999). A novel imprinted gene, encoding a RING zinc-finger protein, and overlapping
128 Kafri, T, Ariel, M, Brandeis, M, Shemer, R, Urven, L, McCarrey, J, Cedar, H, and Razin, A (1992). Developmental pattern of gene-specific DNA methylation in the moembryo and germ line. Genes Dev 6, 705-714. use imprinted genes with the nuclear matrix. Exp Cell Res 274, 189-196. Kangel method to analyze protein-DNA interactions in vivo. Nucleic Acids Res 30, e44. Kashi of the mouse Atp10c in hippocampus and olfactory bulb. J Hum Genet 48, 194-198. Kerm the angiosperm life cycle. Dev Suppl, 9-14. KesheDNA methylation affects the formation of active chromatin. Cell 44, 535-543. Kieffn in the human es and matrix attachment regions which bind SATB1 and GATA-3. J Immunol 168, 3915-3922. Kim, sitive binding of transcription factor YY1 to an insulator sequence within the paternally expressed Kishino, T, Lalande, M, and Wagstaff, J (1997). UBE3A/E6-AP mutations cause Angelman syndrome. Nat Genet 15, 70-73. Kitsb-specific replication timing of imprinted gene regions. Nature 364, 459-463. Klenol s of normal biology and cancer. Semin Cancer Biol 12, 399-414. Lachnion of Lee, J S, Galvin, K M, and Shi, Y (1993). Evidence for physical interaction between the zinc-finger transcription factors YY1 and Sp1. Proc Natl Acad Sci U S A 90, 6145-6149. Kagotani, K, Nabeshima, H, Kohda, A, Nakao, M, Taguchi, H, and Okumura, K (2002). Visualization of transcription-dependent association of , S H, Vieira, K, and Bungert, J (2002). Combining chromatin immunoprecipitation and DNA footprinting: a nov wagi, A, Meguro, M, Hoshiya, H, Haruta, M, Ishino, F, Shibahara, T, and Oshimura, M (2003). Predominant maternal expression icle, J L, and Alleman, M (1990). Gametic imprinting in maize in relation to t, I, Lieman-Hurwitz, J, and Cedar, H (1986). er, L J, Greally, J M, Landres, I, Nag, S, Nakajima, Y, Kohwi-Shigematsu, T, and Kavathas, P B (2002). Identification of a candidate regulatory regioCD8 gene complex by colocalization of DNase I hypersensitive sit J, Kollhoff, A, Bergmann, A, and Stubbs, L (2003). Methylation-sen imprinted gene, Peg3. Hum Mol Genet 12, 233-245. erg, D, Selig, S, Brandeis, M, Simon, I, Keshet, I, Driscoll, D J, Nicholls, R D, and Cedar, H (1993). Allele va, E M, Morse, H C, 3rd, Ohlsson, R, and Lobanenkov, V V (2002). The noveBORIS + CTCF gene family is uniquely involved in the epigenetic er, M, O'Carroll, D, Rea, S, Mechtler, K, and Jenuwein, T (2001). Methylathistone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410, 116-120.
129 Lee, J S, See, R H, Galvin, K M, Wang, J, and Shi, Y (1995). Functional interactions between YY1 and adenovirus E1A Nucleic Acids Res 23, 925-931. Lee, S, Kozlov, S, Hernandez, L, Chamberlain, S J, Brannan, C I, Stewart, C L, and Levy-Wilson, B, and Fortier, C (1989). The limits of the DNase I-sensitive domain of the , lity. Cell 69, 915-926. e tion. Cell 48, 39-46. -sociated with epigenetic reprogramming events, shares the same 11-zinc-finger domain with CTCF, the insulator protein involved in reading Mannon of genes showing altered expression in preimplantation and early postimplantation parthenogenetic embryos. McGrath, J, and Solter, D (1984). Completion of mouse embryogenesis requires both the Wevrick, R (2000). Expression and imprinting of MAGEL2 suggest a role in Prader-willi syndrome and the homologous murine imprinting phenotype. Hum Mol Genet 9, 1813-1819. Lemon, B, and Tjian, R (2000). Orchestrated response: a symphony of transcription factors for gene control. Genes Dev 14, 2551-2569. human apolipoprotein B gene coincide with the locations of chromosomal anchorage loops and define the 5' and 3' boundaries of the gene. J Biol Chem 26421196-21204. Li, E, Beard, C, and Jaenisch, R (1993). Role for DNA methylation in genomic imprinting. Nature 366, 362-365. Li, E, Bestor, T H, and Jaenisch, R (1992). Targeted mutation of the DNA methyltransferase gene results in embryonic letha Lichter, P, Cremer, T, Borden, J, Manuelidis, L, and Ward, D C (1988). Delineation of individual human chromosomes in metaphase and interphase cells by in situ suppression hybridization using recombinant DNA libraries. Hum Genet 80, 224-234. Liu, J, Yu, S, Litman, D, Chen, W, and Weinstein, L S (2000). Identification of a methylation imprint mark within the mouse Gnas locus. Mol Cell Biol 20, 5808-5817. Lock, L F, Takagi, N, and Martin, G R (1987). Methylation of the Hprt gene on thinactive X occurs after chromosome inactiva Loukinov, D I, Pugacheva, E, Vatolin, S, Pack, S D, Moon, H, Chernukhin, I, Mannan, P,Larsson, E, Kanduri, C, Vostrov, A A, et al. (2002). BORIS, a novel male germline-specific protein as imprinting marks in the soma. Proc Natl Acad Sci U S A 99, 6806-6811. , M, Latham, K E, and Varmuza, S (1995). Identificati Dev Genet 17, 223-232. maternal and paternal genomes. Cell 37, 179-183.
130 Meguro, M, Kashiwagi, A, Mitsuya, K, Nakao, M, Kondo, I, Saitoh, S, and Oshimura(2001a). A novel maternally expressed gene, ATP10C, encodes a pu , M tative aminophospholipid translocase associated with Angelman syndrome. Nat Genet 28, Megu, d Oshimura, M (2001b). Large-scale evaluation of imprinting status in the Prader-Willi syndrome region: an imprinted direct repeat Momparler, R L (2003). Cancer epigenetics. Oncogene 22, 6479-6483. Monk DNA pment. Development 99, 371-382. hase J Ultrastruct Res 27, 266-288. P, and alamic and behavioral alterations reminiscent of the human Prader-Willi syndrome. Hum Nakaosomal gene for necdin, a neuronal growth suppressor, in the Prader-Willi syndrome deletion region. Gene 213, 65-72. Narlikat e chromatin structure and transcription. Cell 108, 475-487. s. Ohta,ng-mutation mechanisms in Prader-Willi syndrome. Am J Hum Genet 64, 397-413. Okan for de novo methylation and mammalian development. Cell 99, 247-257. Okand for de novo and maintenance methylation of viral DNA in embryonic stem cells. Nucleic Acids Res 26, 2536-2540. 19-20. ro, M, Mitsuya, K, Nomura, N, Kohda, M, Kashiwagi, A, Nishigaki, R, YoshiokaH, Nakao, M, Oishi, M, an cluster resembling small nucleolar RNA genes. Hum Mol Genet 10, 383-394. , M, Boubelik, M, and Lehnert, S (1987). Temporal and regional changes inmethylation in the embryonic, extraembryonic and germ cell lineages during mouseembryo develo Monneron, A, and Bernhard, W (1969). Fine structural organization of the interpnucleus in some mammalian cells. Muscatelli, F, Abrous, D N, Massacrier, A, Boccaccio, I, Le Moal, M, Cau,Cremer, H (2000). Disruption of the mouse Necdin gene results in hypoth Mol Genet 9, 3101-3110. da, Y, Taniura, H, Uetsuki, T, Inazawa, J, and Yoshikawa, K (1998). The human chrom ar, G J, Fan, H Y, and Kingston, R E (2002). Cooperation between complexes thregulat Neumann, B, Kubicka, P, and Barlow, D P (1995). Characteristics of imprinted geneNat Genet 9, 12-13. Nicholls, R D, Saitoh, S, and Horsthemke, B (1998). Imprinting in Prader-Willi and Angelman syndromes. Trends Genet 14, 194-200. T, Gray, T A, Rogan, P K, Buiting, K, Gabriel, J M, Saitoh, S, Muralidhar, B, Bilienska, B, Krajewska-Walasek, M, Driscoll, D J, et al. (1999). Imprinti o, M, Bell, D W, Haber, D A, and Li, E (1999). DNA methyltransferases Dnmt3a and Dnmt3b are essential o, M, Xie, S, and Li, E (1998). Dnmt2 is not require
131 Ozcelik, T, Leff, S, Robinson, W, Donlon, T, Lalande, M, Sanjines, E, Schinzel, A, and Francke, U (1992). Small nuclear ribonucleoprotein polypeptide N (SNRPNexpressed gene in the Prader-Willi syndrome critical region. Nat Genet 2, 265-269. ), an Pedone, P V, Pikaart, M J, Cerrato, F, Vernucci, M, Ungaro, P, Bruni, C B, and Riccio, A Perk,imprinting mechanism of the Prader-Willi/Angelman regional control center. Embo Pradhan, S, Bacolla, A, Wells, R D, and Roberts, R J (1999). Recombinant human DNA Razin, A, and Cedar, H (1977). Distribution of 5-methylcytosine in chromatin. Proc Natl Razin, S V, and Gromova, II (1995). The channels model of nuclear matrix structure. Rea, S, Eisenhaber, F, O'Carroll, D, Strahl, B D, Sun, Z W, Schmid, M, Opravil, S, hyltransferases. Nature 406, 593-599. Reik,328, 248-251. Reyesed in active chromatin and are associated with the nuclear matrix. J Cell Biol 137, 263-274. (1999). Role of histone acetylation and DNA methylation in the maintenance of theimprinted expression of the H19 and Igf2 genes. FEBS Lett 458, 45-50. J, Makedonski, K, Lande, L, Cedar, H, Razin, A, and Shemer, R (2002). The J 21, 5807-5814. Pfeifer, G P, and Riggs, A D (1991). Chromatin differences between active and inactive X chromosomes revealed by genomic footprinting of permeabilized cells using DNase I and ligation-mediated PCR Genes Dev 5, 1102-1113. Picard, D, and Schaffner, W (1984). A lymphocyte-specific enhancer in the mouse immunoglobulin kappa gene. Nature 307, 80-82. Pina, B, Truss, M, Ohlenbusch, H, Postma, J, and Beato, M (1990). DNA rotationalpositioning in a regulatory nucleosome is determined by base sequence. An algorithm to model the preferred superhelix. Nucleic Acids Res 18, 6981-6987. (cytosine-5) methyltransferase. I Expression, purification, and comparison of de novo and maintenance methylation. J Biol Chem 274, 33002-33010. Acad Sci U S A 74, 2725-2728. Bioessays 17, 443-450. Mechtler, K, Ponting, C P, Allis, C D, and Jenuwein, T (2000). Regulation of chromatin structure by site-specific histone H3 met W, Collick, A, Norris, M L, Barton, S C, and Surani, M A (1987). Genomic imprinting determines methylation of parental alleles in transgenic mice. Nature , J C, Muchardt, C, and Yaniv, M (1997). Components of the human SWI/SNF complex are enrich
132 Rezai-Zadeh, N, Zhang, X, Namour, F, Fejer, G, Wen, Y D, Yao, Y L, Gyory, I, Wright,K, and Seto, E (2003). Targeted recruitment of a histone H4-specific methyltransferase by the transcription factor YY1. Genes Dev 17, 1019-1029. ormal biparental chromosomes 15. Eur J Hum Genet 9, 519-526. RunteNURF-SNRPN transcript serves as a host for multiple small nucleolar RNA species and as an antisense RNA for UBE3A Hum Mol Genet 10, Saitohnn, J, Konig, R, Malcolm, S, Horsthemke, B, and Nicholls, R D (1996). Minimal definition of the 3, 7811-7815. um ements. Mol Cell Biol 13, 6621-6628. SawaN, and Littman, D R (1994). A lineage-specific transcriptional silencer regulates CD4 gene expression during T lymphocyte Scarpulla, R C (2002). Transcriptional activators and coactivators in the nuclear control SchlisFs, y RYBP, provides a mechanism for specificity of E2F function. Embo J 21, 5775-5786. Schoel the Igf2/H19 locus. Nat Genet 33, 66-69. ies N transcription unit. Hum Mol Genet 8, 555-566. Searle, A G, and Beechey, C V (1985). Noncomplementation phenomena and their bearing on nondisjunctional effects. Basic Life Sci 36, 363-376. Runte, M, Farber, C, Lich, C, Zeschnigk, M, Buchholz, T, Smith, A, Van Maldergem, L, Burger, J, Muscatelli, F, Gillessen-Kaesbach, G, et al. (2001a). Comprehensive methylation analysis in typical and atypical PWS and AS patients with n , M, Huttenhofer, A, Gross, S, Kiefmann, M, Horsthemke, B, and Buiting, K (2001b). The IC-S 2687-2700. , S, Buiting, K, Rogan, P K, Buxton, J L, Driscoll, D J, Arnema imprinting center and fixation of chromosome 15q11-q13 epigenotype by imprinting mutations. Proc Natl Acad Sci U S A 9 Saitoh, S, and Wada, T (2000). Parent-of-origin specific histone acetylation and reactivation of a key imprinted gene locus in Prader-Willi syndrome. Am J HGenet 66, 1958-1962. Satyamoorthy, K, Park, K, Atchison, M L, and Howe, C C (1993). The intracisternal A-particle upstream element interacts with transcription factor YY1 to activate transcription: pleiotropic effects of YY1 on distinct DNA promoter el da, S, Scarborough, J D, Killeen, development. Cell 77, 917-929. of mitochondrial function in mammalian cells. Gene 286, 81-89. io, S, Halperin, T, Vidal, M, and Nevins, J R (2002). Interaction of YY1 with E2mediated b nherr, C J, Levorse, J M, and Tilghman, S M (2003). CTCF maintains differentiamethylation at Schweizer, J, Zynger, D, and Francke, U (1999). In vivo nuclease hypersensitivity studreveal multiple sites of parental origin-dependent differential chromatin conformation in the 150 kb SNRP
133 Searle, A G, and Beechey, C V (1990). Genome imprinting phenomena on mouse chromosome 7. Genet Res 56, 237-244. Seto, E, Shi, Y, and Shenk, T (1991). YY1 is an initiator sequence-binding protein that Sharp52. ouse atl Shemer, R, Hershko, A Y, Perk, J, Mostoslavsky, R, Tsuberi, B, Cedar, H, Buiting, K, e Shi, Y, Lee, J S, and Galvin, K M (1997). Everything you have ever wanted to know about Yin Yang 1. Biochim Biophys Acta 1332, F49-66. Simoon of imprinted genes is established in the gametes and maintained during development. Nature 401, 929-932. Smetak. Smith, C M, and Steitz, J A (1997). Sno storm in the nucleolus: new roles for myriad Stein,Wijnen, A J (1991). Regulation of transcription-factor activity during growth and as carrying the imprinting signal. Cell 73, 61-71. uclei in directs and activates transcription in vitro. Nature 354, 241-245. , J A, and Kaufman, P D (2003). Chromatin proteins are determinants of centromere function. Curr Top Microbiol Immunol 274, 23Shemer, R, Birger, Y, Riggs, A D, and Razin, A (1997). Structure of the imprinted mSnrpn gene and establishment of its parental-specific methylation pattern. Proc NAcad Sci U S A 94, 10267-10272. and Razin, A (2000). The imprinting box of the Prader-Willi/Angelman syndromdomain. Nat Genet 26, 440-443. n, I, Tenzen, T, Reubinoff, B E, Hillman, D, McCarrey, J R, and Cedar, H (1999). Asynchronous replicati na, K, Steele, W J, and Busch, H (1963). A nuclear ribonucleoprotein networExpt Cell Res 31, 198-201. small RNPs. Cell 89, 669-672. G S, Lian, J B, Dworetzky, S I, Owen, T A, Bortell, R, Bidwell, J P, and van differentiation: involvement of the nuclear matrix in concentration and localizationof promoter binding proteins. J Cell Biochem 47, 300-305. Stoger, R, Kubicka, P, Liu, C G, Kafri, T, Razin, A, Cedar, H, and Barlow, D P (1993). Maternal-specific methylation of the imprinted mouse Igf2r locus identifies the expressed locus Strahl, B D, and Allis, C D (2000). The language of covalent histone modifications. Nature 403, 41-45. Strahl, B D, Ohba, R, Cook, R G, and Allis, C D (1999). Methylation of histone H3 at lysine 4 is highly conserved and correlates with transcriptionally active nTetrahymena. Proc Natl Acad Sci U S A 96, 14967-14972.
134 Strichman-Almashanu, L Z, Lee, R S, Onyango, P O, Perlman, E, Flam, F, Frieman, M B, and Feinberg, A P (2002). A genomewide screen for normally methylated human CpG islands that can identify novel imprinted genes. Genome Res 12, 543-Strouboulis, J, and Wolffe, A P (1996). Functional compartmentalization of the nucleus. J Surani, M A, Barton, S C, and Norris, M L (1984). Development of reconstituted mouse Sutcliffe, J S, Nakao, M, Christian, S, Orstavik, K H, Tommerup, N, Ledbetter, D H, and Szabo, P, Tang, S H, Rentsendorj, A, Pfeifer, G P, and Mann, J R (2000). Maternal-ive J f primary dorsal root ganglion neurons. Exp Cell Res 277, 220-232. Talbot, D, Collis, P, Antoniou, M, Vidal, M, Grosveld, F, and Greaves, D R (1989). A dominant control region from the human beta-globin locus conferring integration Tama-Thomas, M J, and Seto, E (1999). Unlocking the mechanisms of transcription factor Thorvaldsen, J L, Duran, K L, and Bartolomei, M S (1998). Deletion of the H19 d Tilghman, S M, and Schoenherr, C J (1998). Reply to Schumacher et al. Nature Genetics volume 19, 325. 554. Cell Sci 109 ( Pt 8), 1991-2000. eggs suggests imprinting of the genome during gametogenesis. Nature 308, 548-550. Beaudet, A L (1994). Deletions of a differentially methylated CpG island at the SNRPN gene define a putative imprinting control region. Nat Genet 8, 52-58. specific footprints at putative CTCF sites in the H19 imprinting control region gevidence for insulator function. Curr Biol 10, 607-610. Tada, M, Tada, T, Lefebvre, L, Barton, S C, and Surani, M A (1997). Embryonic germ cells induce epigenetic reprogramming of somatic nucleus in hybrid cells. Embo16, 6510-6520. Takazaki, R, Nishimura, I, and Yoshikawa, K (2002). Necdin is required for terminal differentiation and survival o site-independent gene expression. Nature 338, 352-355. ru, H, and Selker, E U (2001). A histone H3 methyltransferase controls DNA methylation in Neurospora crassa. Nature 414, 277-283. Tazi, J, and Bird, A (1990). Alternative chromatin structure at CpG islands. Cell 60, 909920. YY1: are chromatin modifying enzymes the key? Gene 236, 197-208. differentially methylated domain results in loss of imprinted expression of H19 anIgf2. Genes Dev 12, 3693-3702.
135 Tormanen, V T, Swiderski, P M, Kaplan, B E, Pfeifer, G P, and Riggs, A D (1992). Extension product capture improves genomic sequencing and DNase I footprinby ligation-mediated PCR Nucleic Acids Res 20, 5487-5488. ting Tremblay, K D, Saam, J R, Ingram, R S, Tilghman, S M, and Bartolomei, M S (1995). A illi syndrome. Hum Mol Genet 8, 1357-1364. TurneUdvardy, A, Maine, E, and Schedl, P (1985). The 87A7 chromomere. Identification of Usheva, A, and Shenk, T (1994). TATA-binding protein-independent initiation: YY1, Varmuza, S, and Mann, M (1994). Genomic imprinting--defusing the ovarian time bomb. Veronmprinting: intricacies of epigenetic regulation in clusters. Annu Rev Cell Dev Biol 19, 237-259. Virbaan mitochondrial transcription factor A gene by nuclear respiratory factors: a potential regulatory atl Acad Sci U S A 91, 1309-1313. , romatin organization of the major histocompatibility complex and other regions of human 113 ( Watanabe, T, Yoshimura, A, Mishima, Y, Endo, Y, Shiroishi, T, Koide, T, Sasaki, H, d regions between expressed and non-expressed alleles. Hum Mol Genet 9, 3029-3035. paternal-specific methylation imprint marks the alleles of the mouse H19 gene. Nat Genet 9, 407-413. Tsai, T F, Armstrong, D, and Beaudet, A L (1999a). Necdin-deficient mice do not show lethality or the obesity and infertility of Prader-Willi syndrome. Nat Genet 22, 1516. Tsai, T F, Jiang, Y H, Bressler, J, Armstrong, D, and Beaudet, A L (1999b). Paternal deletion from Snrpn to Ube3a in the mouse causes hypotonia, growth retardationand partial lethality and provides evidence for a gene contributing to Prader-W r, B M (2000). Histone acetylation and an epigenetic code. Bioessays 22, 836-845. novel chromatin structures flanking the heat shock locus that may define the boundaries of higher order domains. J Mol Biol 185, 341-358. TFIIB, and RNA polymerase II direct basal transcription on supercoiled templateDNA Cell 76, 1115-1121. Trends Genet 10, 118-123. a, R I, Mann, M R, and Bartolomei, M S (2003). Genomic i sius, J V, and Scarpulla, R C (1994). Activation of the hum link between nuclear and mitochondrial gene expression in organelle biogenesis. Proc N Volpi, E V, Chevret, E, Jones, T, Vatcheva, R, Williamson, J, Beck, S, Campbell, R DGoldsworthy, M, Powis, S H, Ragoussis, J, et al. (2000). Large-scale ch chromosome 6 and its response to interferon in interphase nuclei. J Cell SciPt 9), 1565-1576. Asakura, H, and Kominami, R (2000). Differential chromatin packaging of genomic imprinte
136 Weber, M, Hagege, H, Murrell, A, Brunel, C, Reik, W, Cathala, G, and Forne, T (200Genomic imprinting controls matrix attachment regions in the Igf2 gene. Mol Cell Biol 23, 8953-8959. 3). Wilkins, J F, and Haig, D (2003). What good is genomic imprinting: the function of Wong-Riley, M, Guo, A, Bachman, N J, and Lomax, M I (2000). Human COX6A1 gene: promoter analysis, cDNA isolation and expression in the monkey brain. Gene 247, Wu, Cneonatal cardiomyocytes results in the sequential activation of nuclear genes Xin, Z, Allis, C D, and Wagstaff, J (2001). Parent-specific complementary patterns of Xin, Z, Tachibana, M, Guggiari, M, Heard, E, Shinkai, Y, and Wagstaff, J (2003). Role em 278, 14996-15000. syndrome imprinting-centre mutations. Nat Genet 19, 25-31. Yao, Yen, P H, Patel, P, Chinault, A C, Mohandas, T, and Shapiro, L J (1984). Differential Zbars5-798. Widom, J (1986). Physicochemical studies of the folding of the 100 A nucleosome filament into the 300 A filament. Cation dependence. J Mol Biol 190, 411-424. parent-specific gene expression. Nat Rev Genet 4, 359-368. 63-75. -t., and Morris, J R (2001). Genes, Genetics, and Epigenetics: A Correspondence. Science 293, 1103-1105. Xia, Y, Buja, L M, Scarpulla, R C, and McMillin, J B (1997). Electrical stimulation of governing mitochondrial proliferation and differentiation. Proc Natl Acad Sci U SA 94, 11399-11404. histone H3 lysine 9 and H3 lysine 4 methylation at the Prader-Willi syndrome imprinting center. Am J Hum Genet 69, 1389-1394. of histone methyltransferase G9a in CpG methylation of the Prader-Willi syndrome imprinting center. J Biol Ch 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 Y L, Yang, W M, and Seto, E (2001). Regulation of transcription factor YY1 by acetylation and deacetylation. Mol Cell Biol 21, 5979-5991. methylation of hypoxanthine phosphoribosyltransferase genes on active and inactive human X chromosomes. Proc Natl Acad Sci U S A 81, 1759-1763. kii, I B, and Debov, S S (1948). On proteins of the cell nucleus. Doklady Akad Nauk SSSR 63, 79 Zhang, Y (2003). Transcriptional regulation by histone ubiquitination and deubiquitination. Genes Dev 17, 2733-2740.
137 Zink, D, Bornfleth, H, Visser, A, Cremer, C, and Cremer, T (1999). Organization of earlyand late replicating DNA in human chromosome territories. Exp Cell Res 247, 176-188.
BIOGRAPHICAL SKETCH I was born in Spain 29 years ago where I completed all my basic education in the city of Vigo and my college undergraduate education at the Universidad de Santiago the Compostela (USC) in the city of Santiago de Compostela. The liberal education that I received at home and at school early on made me the person I am; the time I spent in Santiago made grateful for it. After a brief internship at Procter & Gamble Research and Development, European division, I returned to the USC where I entered the graduate program for the academic year 1997-98. A year later I moved to Gainesville where I entered graduate school in January 1999 and graduated with a Ph.D. degree in medical sciences in February 2004. 138