University of Pennsylvania ScholarlyCommons
Publicly Accessible Penn Dissertations
Spring 2011
Roles of Protein Factors in Regulation of Imprinted Gene Expression
Shu Lin University of Pennsylvaina, [email protected]
Follow this and additional works at: https://repository.upenn.edu/edissertations
Part of the Biology Commons
Recommended Citation Lin, Shu, "Roles of Protein Factors in Regulation of Imprinted Gene Expression" (2011). Publicly Accessible Penn Dissertations. 1546. https://repository.upenn.edu/edissertations/1546
This paper is posted at ScholarlyCommons. https://repository.upenn.edu/edissertations/1546 For more information, please contact [email protected]. Roles of Protein Factors in Regulation of Imprinted Gene Expression
Abstract Genomic imprinting is an important epigenetic phenomenon in which only one parental allele is expressed. Allele-specific DNA methylation often exists in imprinted control regions (ICRs) and is required for properly imprinted expression. Imprinting control in mammals involves an insulator mechanism that requires CTCF or a long ncRNA silencing mechanism. In this dissertation, I studied functions of protein factors in genomic imprinting. First, methylated DNA binding proteins are involved in maintenance of DNA methylation at imprinted loci and required for selective silencing of one specific allele. eW showed that MBD3 was localized to paternal H19 ICR. By RNA interference experiments in preimplantation embryos, we showed that MBD3 and its NuRD complex cofactor MTA2 were required for maintenance of the paternal methylation at the H19 ICR, and for silencing of the paternal H19. MTA2 is also required for Peg3 allelic expression. These results demonstrate new roles of the NuRD complex in genomic imprinting. Moreover, I showed allele-specific associations of MBD1 and Kaiso with imprinted loci, implicating functional requirements of these proteins in imprinted regulation. Second, I demonstrated colocalization of CTCF and the cohesin complex at three imprinted loci. CTCF and cohesins preferentially bind to the unmethylated allele of H19 ICR and Gtl2 DMR, and both alleles of KvDMR1 in mouse embryonic fibroblast cells (MEFs). To determine the functional importance of CTCF and cohesins, CTCF and cohesins were depleted in MEFs. Monoallelic expression of imprinted genes was maintained. However, mRNA levels for imprinted genes were typically increased; for H19 and Igf2 the increased expression was independent of the CTCF binding sites in the ICR. These experiments demonstrate an unappreciated role for CTCF and cohesins in the repression of imprinted genes in somatic cells. Finally, I have established F1 hybrid trophoblast stem (TS) cell lines that can be used as a model system to further study functions of protein factors in genomic imprinting and extraembryonic development. Overall, this dissertation has provided new and important insights into roles of protein factors in regulation of imprinted gene expression.
Degree Type Dissertation
Degree Name Doctor of Philosophy (PhD)
Graduate Group Biology
First Advisor Marisa S. Bartolomei
Keywords Genomic Imprinting, Imprinted gene, MBD3, MTA2, CTCF, Cohesin
Subject Categories Biology
This dissertation is available at ScholarlyCommons: https://repository.upenn.edu/edissertations/1546 ROLES OF PROTEIN FACTORS IN REGULATION OF IMPRINTED
GENE EXPRESSION
Shu Lin
A DISSERTATION
in
Biology
Presented to the Faculties of the University of Pennsylvania
in
Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
2011
Supervisor of Dissertation Graduate Group Chairperson
______
Marisa S. Bartolomei Paul D. Sniegowski
Professor of Cell and Developmental Biology Associate Professor of Biology
Dissertation Committee
Doris Wagner, Associate Professor of Biology
Richard M. Schultz, Charles and William L. Day Distinguished Professor of Biology
R. Scott Poethig, Professor of Biology
Gerd A. Blobel, Professor of Pediatrics
Jumin Zhou, Associate Professor of Gene Expression and Regulation, The Wistar
Institute ACKNOWLEDGMENTS
I would like to express my gratitude to all those who have helped and inspired me during my graduate study. I am deeply grateful to my advisor Dr. Marisa Bartolomei, whose help, stimulating suggestions and encouragement helped me throughout the course of my research. I appreciate her enthusiasm for science and her extensive knowledge. I also thank Dr. Kimberly Reese, who first welcomed me in the lab and gave me lots of help for the MBD3 project. I would like to thank our collaborators: Dr. Richard Schultz and his postdoc Dr. Pengpeng Ma for their help with the MTA2 project; Dr. Paul
Lieberman at the Wistar Institute and his former postdoc Dr. William Stedman for their support with the cohesin work; Dr. Anne Ferguson-Smith at University of Cambridge for sharing the information of the Gtl2 CTCF binding site.
I want to thank my thesis committee members Drs. Doris Wagner, Richard Schultz,
Scott Poethig, Gerd Blobel and Jumin Zhou. They have always provided me constructive
suggestions to my projects and have been a constant source of encouragement during my
graduate study.
I would like to express my deeply appreciation to all the Bartolomei lab members.
They have made the lab a convivial place to work and study. In particular, I would like to
thank Dr. Joanne Thorvaldsen who has offered me so much help and encouragement. I
am grateful to Dr. Winifred Mak for her collaboration on the TS cell derivation. I want to
thank Chris Krapp for his constant support. Drs. Sébastien Vigneau, Raluca Verona,
Rocio Rivera, Mellissa Mann gave me advice in many experiments. I want to thank Lara
ii Abramowitz, Dr. Nora Engel, Dr. Folami Ideraabdullah, Rob Plasschaert, Dr. Martha
Susiarjo, Jamie Weaver and all past and current members of the Bartolomei lab for their help.
I would also like to extend thanks to people who have been instrumental in completion of my dissertation work. I am grateful to Dr. Galina Filippova at Fred
Hutchinson Cancer Research Center for sharing the siRNA protocol. I thank Dr. Mike
Golding at Texas A&M Univeristy and Dr. M. Celeste Simon for sharing their expertise in virus-based RNAi.
I thank all my friends and family. My deepest gratitude goes to my mother, Yufen
Lin and my father, Jiting Lin for their love and support throughout my life; this
dissertation is simply impossible without them. I would like to give my special thanks to
Dr. Bing He, who is a wonderful husband and a great scientist. His love and support
enabled me to complete this work.
iii ABSTRACT
ROLES OF PROTEIN FACTORS IN REGULATION OF IMPRINTED GENE
EXPRESSION
Shu Lin
Marisa S. Bartolomei
Genomic imprinting is an important epigenetic phenomenon in which only one
parental allele is expressed. Allele-specific DNA methylation often exists in imprinted
control regions (ICRs) and is required for properly imprinted expression. Imprinting
control in mammals involves an insulator mechanism that requires CTCF or a long
ncRNA silencing mechanism. In this dissertation, I studied functions of protein factors in
genomic imprinting. First, methylated DNA binding proteins are involved in maintenance
of DNA methylation at imprinted loci and required for selective silencing of one specific
allele. We showed that MBD3 was localized to paternal H19 ICR. By RNA interference
experiments in preimplantation embryos, we showed that MBD3 and its NuRD complex
cofactor MTA2 were required for maintenance of the paternal methylation at the H19
ICR, and for silencing of the paternal H19. MTA2 is also required for Peg3 allelic expression. These results demonstrate new roles of the NuRD complex in genomic imprinting. Moreover, I showed allele-specific associations of MBD1 and Kaiso with imprinted loci, implicating functional requirements of these proteins in imprinted regulation. Second, I demonstrated colocalization of CTCF and the cohesin complex at three imprinted loci. CTCF and cohesins preferentially bind to the unmethylated allele of
iv H19 ICR and Gtl2 DMR, and both alleles of KvDMR1 in mouse embryonic fibroblast cells (MEFs). To determine the functional importance of CTCF and cohesins, CTCF and cohesins were depleted in MEFs. Monoallelic expression of imprinted genes was maintained. However, mRNA levels for imprinted genes were typically increased; for
H19 and Igf2 the increased expression was independent of the CTCF binding sites in the
ICR. These experiments demonstrate an unappreciated role for CTCF and cohesins in the repression of imprinted genes in somatic cells. Finally, I have established F1 hybrid trophoblast stem (TS) cell lines that can be used as a model system to further study functions of protein factors in genomic imprinting and extraembryonic development.
Overall, this dissertation has provided new and important insights into roles of protein factors in regulation of imprinted gene expression.
v Table of Contents
List of Tables ...... x
List of Figures ...... xi
Chapter 1. Introduction ...... 1
1.1. Genomic Imprinting ...... 2
1.1.1. Genomic imprinting and evolution ...... 3
1.1.2. Genomic imprinting and human diseases ...... 6
1.1.3. Imprinted regulation in mammals ...... 8
1.2. DNA methylation and reprogramming in mammals ...... 19
1.2.1. DNA methylation ...... 19
1.2.2. DNA methylation reprogramming in mammals...... 20
1.3. Methyl-CpG binding proteins ...... 24
1.3.1. The MBD family ...... 24
1.3.2. The Kaiso family ...... 30
1.4. CTCF ...... 31
1.4.1. Models for insulator function ...... 32
1.4.2. Discovery and multifunctionality of CTCF ...... 35
1.4.3. CTCF binding sites ...... 38
1.4.4. CTCF cofactors ...... 39
1.4.5. CTCF interacts with cohesins ...... 40
vi Chapter 2. Roles of methylated DNA binding proteins in genomic imprinting ...... 44
2.1. Methylated DNA binding proteins are associated with imprinted loci ...... 46
2.1.1. MBD3 binds preferentially to the paternal H19 ICR in ES cells ...... 47
2.1.2. A survey for occupancy of MBD1 and Kaiso at imprinted loci...... 51
2.2. MBD3 is required for the H19/Igf2 locus in preimplantation embryos ...... 55
2.2.1. MBD3 is required for the maintenance of DNA methylation at the H19 ICR in preimplantation embryos ...... 57
2.2.2. MBD3 is required for the imprinted expression of H19 in preimplantation embryos ...... 60
2.3. MTA2 is required for imprinted expression at the H19/Igf2 and Peg3 loci in preimplantation embryos ...... 64
2.3.1. MTA2 is required for the imprinted expression of H19 and Peg3 in preimplantation embryos ...... 67
2.3.2. MTA2 is required for the maintenance of DNA methylation at the H19 ICR in
preimplantation embryos ...... 69
Chapter 3. Roles of CTCF and the cohesin complex in genomic imprinting ...... 75
3.1. CTCF and cohesins colocalize at three imprinted loci ...... 77
3.1.1. CTCF and cohesins colocalize at the H19/Igf2 locus ...... 78
3.1.2. CTCF and cohesins colocalize at the Gtl2 DMR and KvDMR1 ...... 85
3.2. Functional analysis of CTCF and cohesins in MEFs ...... 93
3.2.1. Depletion of CTCF and cohesins in MEFs lead to increased Igf2 expression
vii from the normally expressed paternal allele ...... 96
3.2.2. Depletion of CTCF and cohesins in MEFs lead to non-allelic changes of expression levels at the Dlk1-Dio3 locus, but not at the Kcnq1/Kcnq1ot1 locus ...... 102
3.2.3. Using TS cell as a new model cell line to study roles of CTCF and cohesins in
genomic imprinting ...... 103
Chapter 4. Discussion and future directions ...... 112
4.1. The NuRD complex and genomic imprinting in preimplantation embryos ..115
4.2 Possible mechanisms for incomplete penetrance in MBD3- or
MTA2-depleted embryos ...... 116
4.3. Establishment versus maintenance roles of MBD3 and MTA2 at imprinted loci ...... 118
4.4. Various requirements for protein cofactors at different developmental stages
and at distinct imprinted loci...... 118
4.5 Additional functions of MBD3 and MTA2 in preimplantation embryos ..... 120
4.6 Genome-wide colocalization of CTCF and cohesins ...... 121
4.7. Cohesins’ dependence of the CTCF binding sites ...... 122
4.8. Other cohesin cofactors in transcriptional regulation ...... 123
4.9. Cohesins may not be involved in the insulator function at the H19 ICR ..... 124
4.10. Cooperation of CTCF and cohesins in transcriptional regulation of imprinted genes ...... 125
4.11. Possible mechanisms for the non-allelic functions of CTCF and cohesins at
viii imprinted loci ...... 127
4.12. Tissue and cell type specificity of CTCF and cohesins functions ...... 128
Chapter 5. Materials and methods ...... 130
Mouse strains ...... 131
Culture of ES cells ...... 131
Isolation and culture of F1 hybrid MEFs ...... 132
Derivation and culture of F1 hybrid TS cells ...... 132
PCR programming ...... 133
Transgenic mice genotyping ...... 133
Bisulfite mutagenesis ...... 134
Chromatin Immunoprecipitation ...... 135
Antibodies ...... 135
RNA extraction and reverse transcription (RT) ...... 136 siRNA treatments ...... 136
Lentivirus-RNAi ...... 137
Real-time PCRs ...... 139
Allele-specific PCRs ...... 140
Western blots ...... 141
Statistics ...... 141
References ...... 164
ix List of Tables
Table 1 Effects of MBD3- and MTA2-depletion at imprinted loci in preimplantation
embryos ...... 72
Table 2 Tissue culture medium ...... 142
Table 3 Primers and PCR conditions used for mouse genotyping ...... 144
Table 4 Primers and PCR conditions used for bisulfite mutagenesis ...... 146
Table 5 Primers and PCR conditions used for real-time PCR quantifications ...... 148
Table 6 Primers and enzymes used for allele-specific analysis by PCRs ...... 152
Table 7 Other PCR primers and conditions ...... 157
Table 8 Antibodies ...... 158
Table 9 shRNA sequences ...... 161
x List of Figures
Figure 1.1 The mouse H19/Igf2 imprinting locus ...... 10
Figure 1.2 CTCF binding sites are required for the insulator activity at the H19/Igf2 locus
...... 12
Figure 1.3 Ectopic binding of CTCF to the paternal ICR causes loss of imprinting at the
H19/Igf2 locus ...... 13
Figure 1.4 The mouse Igf2r imprinting locus ...... 15
Figure 1.5 The mouse Kcnq1/Kcnq1ot1 imprinting locus ...... 17
Figure 1.6 DNA methylation reprogramming during mammalian development ...... 21
Figure 1.7 The MBD and Kaiso families of methylated DNA binding proteins ...... 25
Figure 1.8 The looping model for insulator function at the H19/Igf2 locus ...... 34
Figure 1.9 The tracking model for insulator function at the H19/Igf2 locus ...... 36
Figure 1.10 The mitotic cohesin complex ...... 41
Figure 2.1 MBD3 preferentially localizes to the paternal H19 ICR ...... 49
Figure 2.2 MBD1 and Kaiso binding to methylated DMRs in ES cells ...... 52
Figure 2.3 MBD1 and Kaiso binding to methylated DMRs in MEFs ...... 54
Figure 2.4 Two independent RNAi approaches to deplete MBD3 ...... 56
Figure 2.5 Depletion of MBD3 causes loss of DNA methylation at the H19 ICR ...... 58
Figure 2.6 Depletion of MBD3 causes biallelic expression of H19 in preimplantation embryos ...... 61
Figure 2.7 Depletion of MTA2 causes biallelic expression of H19 and Peg3 in
xi preimplantation embryos ...... 68
Figure 2.8 Depletion of MTA2 causes hypomethylation at the H19 ICR...... 70
Figure 3.1 Cohesins colocalize with CTCF at the H19 ICR, in a CTCF binding
sites-dependent manner ...... 79
Figure 3.2 Cohesins colocalize with CTCF at the maternal H19 ICR ...... 81
Figure 3.3 Cohesins colocalize with CTCF at the CTCF binding site in the IVR ...... 83
Figure 3.4 The mouse Dlk1-Dio3 imprinting locus ...... 87
Figure 3.5 Cohesins colocalize with CTCF at the maternal allele of Gtl2 DMR ...... 88
Figure 3.6 Cohesins colocalize with CTCF at KvDMR1 ...... 90
Figure 3.7 Cohesins colocalize with CTCF at both alleles of KvDMR1...... 92
Figure 3.8 Successful depletion of CTCF, SMC1 and RAD21 from MEFs ...... 95
Figure 3.9 CTCF and cohesins depletion associate with elevated expression, but not loss of imprinted expression of Igf2 ...... 97
∆ Figure 3.10 CTCF depletion in H19 R/+ MEFs associate with elevated expression of Igf2
...... 99
Figure 3.11 CTCF and cohesins colocalize at a CTCF binding site 5.2 kb upstream of Igf2
DMR0 ...... 101
Figure 3.12 CTCF and cohesins depletion lead to changes in Gtl2 and Dlk1 RNA levels, but not in Kcnq1ot1 RNA levels ...... 104
Figure 3.13 Characterization of F1 TS cell lines ...... 107
Figure 3.14 RNAi in TS cells ...... 109
xii Figure 4.1 Protein factors are required for regulation of imprinted gene expression ...... 114
xiii Chapter 1. Introduction
1 1.1. Genomic Imprinting
Genomic imprinting is an epigenetic phenomenon in which certain genes, termed imprinted genes, are expressed in a parent-of-origin dependent manner. In mammals and flowering plants, genomic imprinting occurs in embryonic tissues as well as embryo-nourishing tissues: the placenta and the endosperm, respectively [reviewed in
(Ideraabdullah et al., 2008; Köhler and Weinhofer-Molisch, 2010)]. Loss of imprinting has been reported in a number of human diseases including cancers and genetic disorders
[reviewed in (Butler, 2009; Lim and Maher, 2010)]. Imprinted gene expression requires establishment of imprinting marks in the germ line and the early embryo, and subsequently, maintenance of these marks during cell division in somatic tissues. In extraembryonic tissues, imprinting marks persist without reprogramming, because the placenta and the endosperm are terminally differentiated. Differential DNA methylation and posttranslational histone modifications are the best-established epigenetic marks that are critical for the allele-specific expression of imprinted genes (Ideraabdullah et al.,
2008; Köhler and Weinhofer-Molisch, 2010). In mammals, differential sex-dependent deposition of DNA methylation marks happens during gametogenesis at a group of cis -regulatory elements (germ-line differentially methylated regions/DMRs) that are
required for proper imprinted expression. In contrast to mammals, plant imprinting marks
are created by differential removal of DNA methylation at the end of female
gametogenesis.
2 1.1.1. Genomic imprinting and evolution
Genomic imprinting is a phenomenon that only exists, as far as we currently know, in mammals and flowering plants. Imprinted genes express one copy of the gene; therefore, any deleterious mutations on the active allele become dominant. Because such a process overcomes the advantages of diploidy, the question arises as to why genomic imprinting is evolutionary beneficial. Among the different theories that try to explain this paradox, the parental conflict theory and the defense theory have received the most serious consideration.
1.1.1.1. The parental conflict theory
Evidence for genomic imprinting in mouse came from studies demonstrating that both maternal and paternal genomes are essential for successful development of the conceptus and normal phenotype of the adult (McGrath and Solter, 1984; Surani et al.,
1984; Barton et al., 1984; Cattanach and Kirk, 1985). Embryos with two female or male pronuclei were generated using nuclear transplantation. In embryos with two female genetic complements, embryonic development was good but extraembryonic development was poor; in embryos with two male genetic complements, the reverse was true. A similar phenomenon was observed in flowering plants by interploidy crosses, in which an increased dosage of paternal chromosomes promotes endosperm
(extraembryonic lineage) development, whereas an increased dosage of maternal chromosomes represses endosperm development (Birchler, 1993; Scott et al., 1998).
Based on these observations, the parental conflict theory or the kinship theory was
3 developed, which suggests that genomic imprinting serves to control the nutrient flow from the mother to the progeny, with maternally and paternally expressed genes having different roles in nutrient allocation (Haig and Westoby, 1989; Trivers and Burt, 1999).
Whereas paternally expressed imprinted genes are suggested to promote nutrient flow to the embryo, maternally expressed imprinted genes reduce nutrient flow to the embryo, preserving nutrition for the mother.
In the three groups of living mammals, monotremes, marsupials and eutherians, there is a correlation between complexities of the extraembryonic tissues and levels of genomic imprinting (Renfree et al., 2009). The egg-laying monotremes have a simple extraembryonic structure that is similar to the reptiles and birds. The monotremes lay eggs after a brief period of intrauterine growth supported by the uterine secretions transferred across the yolk sac (Renfree et al., 2009). Marsupials have a brief intrauterine life and are born tiny and altricial, but have sophisticated lactation that supports the young. The marsupial placenta has a vascularized yolk-sac, secretes hormones, and is essential for fetal development. However, it is short lived and does not support the full term of development (Freyer et al., 2003). Young eutherians are nurtured in utero for an extended period of time and most are delivered highly developed. The eutherian placenta has various adaptations of the chorionic membrane, which is vascularized by the allantois to form the chorioallantoic placenta. The eutherian placenta not only contributes to the nutrient and gas exchange between mother and fetus, but is also able to mediate immune responses (Renfree et al., 2009). Genomic imprinting is present in eutherians and marsupials, but has not yet been identified in monotremes. Marsupials generally have
4 fewer imprinted genes than eutherians [reviewed in (Renfree et al., 2009)]. Furthermore, many imprinted genes play important roles in embryonic and extraembryonic development, as well as maternal nursing behaviors (Barlow DP and Bartolomei MS,
2007). A recent report showed functional dominance of maternally methylated DMRs (as compared with the paternally methylated DMRs) at the time of chorioallantoic placenta formation (Schulz et al., 2010). By comparison of 15 DMRs in 16 eutherian species, the authors showed that paternally methylated DMRs tend to lose CpGs, whereas maternally methylated DMRs tend to gain CpGs during evolution. The authors concluded that functional dominance of maternally methylated DMRs is the consequence of the selective pressure to tightly regulate the expression of genes affecting the fetal-maternal interface once the placenta had evolved, therefore supporting the parent conflict theory.
Taken together, in favor of the parental conflict theory, evolution of genomic imprinting correlates well with viviparity and placenta.
1.1.1.2. The defense theory
In contrast, imprinting might have evolved as a by-product of a defense mechanism destined to silence invading foreign DNA in the gametes and the embryos (the defense theory) (Barlow, 1993). Recent studies in both mammals and flowering plants provide substantial evidence for the defense theory. Paternal expressed 10 (Peg10 ) is a retrotransposon-derived imprinted gene that has an essential role for the mouse placenta
(Ono et al., 2003). In the common ancestors of eutherians and marsupials, Peg10 was inserted next to the Sarcoglycan epsilon (Sgce ) gene (Suzuki et al., 2007). Suzuki et al.
5 showed that Peg10 is imprinted in marsupials and identified a DMR that covers the
Peg10 promoter. DMRs are also found in many other imprinted loci and a subset of
DMRs is required for imprinting control (see 1.1.3). Importantly, the Peg10 DMR extends to the neighboring Sgce promoter in eutherians and Sgce is imprinted in eutherians but not in marsupials. Therefore, imprinted expression of Peg10 and Sgce
correlates well with the repression of exogenous DNA sequences by DNA methylation.
Moreover, comparison of repeat elements in imprinted loci revealed significant difference
between monotremes and therians, that monotremes have fewer repeats of certain classes
than marsupials and eutherians (Pask et al., 2009). In addition, imprinted genes in plants
are located in the vicinity of transposons or repeat sequences, suggesting that the
insertion of transposon or repeat sequences was a prerequisite for imprinting evolution
(Gehring et al., 2009; Hsieh et al., 2009).
Taken together, the parental conflict theory may explain why genomic imprinting is
important in evolution, while the defense theory may explain how imprinting is
manifested.
1.1.2. Genomic imprinting and human diseases
Imprinted genes are expressed from only one parental allele. Such imprinted expression is tightly regulated by a number of different cis -regulatory elements, as well as epigenetic marks such as DNA methylation. Therefore, mutations or changes in the epigenome that occur in imprinted loci may be deleterious. Indeed, a number of human diseases and syndromes are associated with imprinted genes. These include human
6 cancers [reviewed in (Lim and Maher, 2010)], as well as genetic disorders such as:
Beckwith-Wiedemann Syndrome (BWS), Silver-Russell Syndrome (SRS), Prader-Willi
Syndrome (PWS), Angelman Syndrome (AS), Albright hereditary osteodystrophy and uniparental disomy 14 [reviewed in (Butler, 2009)]. Typically, these imprinting-related diseases are associated with DNA methylation abnormalities at imprinted loci, which are caused by DNA duplications, inversions, insertions, micro-deletions and translocations.
For example, many BWS patients have abnormal methylation of genes in the
Chromosome (Chr) 11p15.5 region, specifically at the H19/IGF2 and
KCNQ1/KCNQ1OT1 imprinting loci. Biallelic expression of INSULIN LIKE GROWTH
FACTOR 2 (IGF2 ), KCNQ1 OVERLAPPING TRANSCRIPT 1 (KCNQ1OT1 ) and reduced
expression of CYCLIN-DEPENDENT KNASE INHIBITOR 1 (CDKN1C ) are common in
BWS patients.
In addition, increased numbers of imprinting disorders (especially BWS and AS) have also been associated with children conceived through Assisted Reproductive
Technologies (ART) (Owen and Segars, 2009). ART involves superovulation of the mother, in vitro fertilizations (IVF), intracytoplasmic sperm injections (ICSI) and in vitro culture of eggs and/or embryos. A number of ART studies in human and mouse have discovered gain or loss of DNA methylation in imprinted loci of ART mouse embryos and patients (Mann et al., 2004; Rivera et al., 2008; Owen and Segars, 2009;
Market-Velker et al., 2010a, b). Therefore, investigating the mechanisms of genomic imprinting, especially DNA methylation-dependent mechanisms, will provide important knowledge for human health.
7 1.1.3. Imprinted regulation in mammals
In mammals, approximately 100 imprinted genes have been identified and many are located within about 1 megabase (Mb) clusters that harbor DMRs [reviewed in
(Bartolomei, 2009)]. Within a cluster, imprinted gene expression is co-regulated by a cis -acting element, designated imprinting control region (ICR), which is also a DMR.
The ICR, typically a few kilobase (kb) pairs in length, acquires parental allele-specific
DNA methylation in the germline and exhibits posttranslational histone modifications.
Deletion of ICRs results in loss of imprinting of linked genes (Wutz et al., 1997;
Thorvaldsen et al., 1998; Fitzpatrick et al., 2002; Lin et al., 2003; Williamson et al., 2006;
Mancini-Dinardo et al., 2006). In addition to the ICR, many imprinted loci harbor DMRs that are not required for the imprinting control. These DMRs do not acquire differential methylation during gametogenesis and/or do not survive the genome-wide reprogramming after fertilization (Bartolomei, 2009). For example, the promoters of
Cdkn1c (Bhogal et al., 2004) and Insulin like growth factor 2 receptor (Igf2r ) (Stöger et al., 1993) obtain DNA methylation after fertilization and are likely a consequence of transcriptional silencing of the repressed allele. Additionally, the Igf2 DMR2 is methylated during spermatogenesis, then demethylated together with other paternal sequences immediately after fertilization, and is subsequently remethylated during post-implantation development (Oswald et al., 2000).
Two major mechanisms that regulate imprinted gene clusters have been characterized in mammals (Bartolomei, 2009). One employs allele-specific insulation mediated by the insulator protein CTCF. The second mechanism requires the transcription of a long
8 non-coding RNA (ncRNA), which silences genes in cis , allowing the alternate allele to be expressed. Some clusters employ aspects of both mechanisms. Recently, a new mechanism employing allele-specific alternative polyadenylation sites has also been described (Wood et al., 2008).
1.1.3.1. The insulator mechanism
The insulator mechanism of imprinted regulation has been well described at the
H19/Igf2 locus (Fig. 1.1). In this context, an insulator is a DNA element that blocks enhancer-promoter interaction when placed between the gene and enhancer element
(Chung et al., 1993). The H19/Igf2 locus is located on human Chr 11p15.5 and mouse
Chr 7F. H19 encodes a ~2.2 kb ncRNA expressed from the maternal chromosome
(Bartolomei et al., 1991), and is a primary microRNA precursor for miR-675 (Mineno et al., 2006; Cai and Cullen, 2007). The Igf2 gene encodes a fetal growth factor expressed from the paternal chromosome (DeChiara et al., 1990, 1991). The imprinted expression of the mouse locus, which is largely conserved in humans, is controlled by two elements, the ICR located 2 kb upstream of H19 but still downstream of Igf2 (Thorvaldsen et al.,
1998, 2002), and the shared enhancers located downstream of H19 (Leighton et al., 1995;
Ishihara et al., 2000; Kaffer et al., 2000, 2001). The ICR, 2 kb in length, is methylated on the paternal chromosome. This methylation is established in the gametes and maintained throughout development (Tremblay et al., 1995, 1997). When the ICR is deleted from the maternal chromosome, H19 expression is reduced and Igf2 is expressed biallelically; conversely, when the ICR is deleted from the paternal chromosome, H19 is expressed on
9
Figure 1.1 The mouse H19/Igf2 imprinting locus
Schematic showing the mouse H19/Igf2 locus. Igf2 is located about 85 kb upstream of H19 . The endodermal enhancers are shown (the mesodermal enhancers are located further downstream of
H19 ). On the maternal chromosome, the ICR recruits CTCF through four binding sites. A
CTCF-dependent insulator is formed on the ICR, blocking interaction between the Igf2 promoter and 3’ shared enhancers. On the paternal chromosome, DNA methylation on the ICR repels
CTCF binding, allowing Igf2 activation by the enhancers. In addition, paternal ICR methylation spreads to the H19 promoter and therefore silences H19 .
10 both alleles and Igf2 expression is reduced (Thorvaldsen et al., 1998, 2002). Furthermore, when an extra set of the endodermal enhancers are inserted between the ICR and Igf2 ,
Igf2 is expressed from both alleles (in endodermal tissues) while H19 expression is not affected (Webber et al., 1998). These observations led to a model that the unmethylated maternal ICR serves as an enhancer blocker to insulate Igf2 from shared enhancers,
whereas the methylated paternal ICR does not have this insulator activity but functions as
a silencer of H19 on the paternal allele. Therefore, expression of H19 and Igf2 is
dependent on the accessibility of these genes to the enhancers. This model is further
supported by the discovery that the ICR binds to CTCF (Bell and Felsenfeld, 2000; Hark
et al., 2000; Kanduri et al., 2000; Szabó et al., 2000), a protein that mediates insulator
activity at other loci including the β-globin locus (Bell et al., 1999). Importantly, CTCF
binds to the H19/Igf2 ICR in a DNA methylation-sensitive manner (Szabó et al., 2000;
Bell and Felsenfeld, 2000; Hark et al., 2000; Kanduri et al., 2000), through four
21-basepair (bp) repeats (Stadnick et al., 1999).
Consistent with the insulator model, when a mutant allele that harbors deletion or mutation of the CTCF binding sites in the ICR is inherited from the maternal chromosome, Igf2 exhibits biallelic expression and H19 expression is reduced (Fig. 1.2)
(Schoenherr et al., 2003; Pant et al., 2003, 2004; Szabo et al., 2004; Engel et al., 2006).
Moreover, in a mutant that ectopically attracts CTCF to the paternal allele, forming the
insulator on both alleles, Igf2 expression is reduced and H19 is expressed biallelically
(Fig. 1.3) (Engel et al., 2004).
To date, the insulator mechanism of imprinted regulation has only been described for
11
Figure 1.2 CTCF binding sites are required for the insulator activity at the H19/Igf2 locus
Schematic showing the maternal transmission of the ∆R (deletion of repeats) allele, in which four
CTCF binding sites are deleted from the endogenous locus. Upon maternal transmission of the mutant allele, the insulator activity is lost because CTCF cannot bind to the ICR, Igf2 exhibits biallelic expression and H19 expression is reduced. Moreover, the maternal ICR ectopically acquires DNA methylation in postimplantation tissues (Engel et al., 2006).
12
Figure 1.3 Ectopic binding of CTCF to the paternal ICR causes loss of imprinting at the
H19/Igf2 locus
Schematic showing the pat ernal transmission of the 9CG allele, in which nine CpGs are mutated in the CTCF binding sites while retaining CTCF binding. Upon paternal transmission of the mutant allele, the paternal ICR loses DNA methylation and forms an insulator due to binding of CTCF.
Igf2 expression is reduced and H19 exhibits biallelic expression (Engel et al., 2004).
13 the H19/Igf2 locus, despite the fact that CTCF also binds to many other imprinted loci,
including the Dlk1-Dio3, Kcnq1/Kcnq1ot1 , Growth factor receptor bound protein 10
(Grb10 ) and RAS protein-specific guanine nucleotide-releasing factor 1 (Rasgrf1 ) loci
(Hikichi et al., 2003; Yoon et al., 2005; Fitzpatrick et al., 2007; Kernohan et al., 2010). At
the Rasgrf1 locus, there is also evidence suggesting a CTCF-dependent insulator
mechanism (Yoon et al., 2005). However, it is unclear whether the presence of CTCF is
always associated with insulator activity at these other loci, and further evidence, such as
the presence of cis -regulatory elements is required to assess the suitability of an insulator
model at these loci.
1.1.3.2. The long ncRNA mechanism
In contrast to the seemingly limited use of the insulator mechanism at imprinted loci,
a number of imprinted loci appear to use the long ncRNA mechanism to regulate
imprinting. These loci contain at least one long ncRNA, transcription of which is required
for silencing of other imprinted genes in cis . The best characterized example is the Igf2r
imprinting locus (Fig. 1.4). This locus resides on mouse Chr 17 and contains three
maternally expressed imprinted genes, Igf2r , solute carrier family 22 member 3 (Slc22a3 )
and Slc22a2 , and one paternally expressed imprinted gene, Antisense Igf2r RNA (Airn ).
Additionally, three non-imprinted genes, including Plasminogen (Plg ), Slc22a1 and Mas1
oncogene (Mas1 ), also reside at the locus. This locus has two DMRs, DMR1 and DMR2,
which encompass the promoters of Igf2r and Airn , respectively. DMR1 is methylated on
the paternal Igf2r allele. The DNA methylation is acquired progressively during
14
Slc22a2 , , Slc22a3 required for silencing of paternal locus. The paternally expressed longAirn ncRNA is Igf2r imprintinglocus Igf2r . Igf2r and and Schematic showing the mouse Figure1.4 The mouse
15 embryogenesis and is a consequence of silencing of the paternal allele (Stöger et al.,
1993). DMR2, which is maternally methylated, is the ICR of this imprinted locus because the DNA methylation on DMR2 is obtained in oocytes but not sperm, and persists during embryonic development (Stöger et al., 1993; Brandeis et al., 1993; Lucifero et al., 2002).
In addition, deletion of DMR2 causes loss of imprinting of all three maternally expressed genes in the locus (Wutz et al., 1997; Zwart et al., 2001) . The Airn transcript, which is not spliced, initiates at the second intron of Igf2r and is antisense to Igf2r (Wutz et al., 1997) .
The Airn transcript extends more than 100 kb and its 3’ end overlaps the 5’ end of Mas1
(Lyle et al., 2000). Transcription of the full-length paternal Airn is required for repression
of the paternal allele of Igf2r , as well as Slc22a3 and Slc22a2 , which do not overlap with
the Airn gene (Sleutels et al., 2002) . Therefore, the full-length paternal Airn transcript, or
transcription through the region, is required for the silencing of imprinted genes on the
paternal allele at the Igf2r imprinted locus.
The Kcnq1/Kcnq1ot1 locus exhibits features of both the long ncRNA silencing and the CTCF-related mechanisms; the former serves as the primary mechanism. Similar to the Igf2r locus, the imprinted expression of genes in the Kcnq1/Kcnq1ot1 locus requires transcription of a full-length long ncRNA, Kcnq1ot1. The Kcnq1/Kcnq1ot1 imprinted locus is located in proximity to the H19/Igf2 locus in both human and mouse. However, the imprinted regulation in the Kcnq1/Kcnq1ot1 is independent of the H19 ICR (Caspary et al., 1998; Ainscough et al., 1998). The Kcnq1/Kcnq1ot1 locus (Fig. 1.5) contains eleven maternally-expressed genes, including Oxysterol binding protein-like 5 (Osbpl5 ),
Tumor necrosis factor receptor superfamily member 23 (Tnfrsf23 ), Nucleosome assembly
16
g g ncRNA Kcnq1ot1 is required for locus (not to scale). The paternally expressed lon on the on chromosome. paternal imprinting locus Kcnq1ot1 / Kcnq1 Kcnq1/Kcnq1ot1 Schematic showing the mouse Figure1.5 The mouse silencing other of imprinted genes bidirectionally
17 protein 1-like 4 (Nap1l4 ), Pleckstrin homology-like domain family A member2 (Phlda2 ),
Slc22a18 , Cdkn1c , Potassium voltage-gated channel subfamily Q member 1 (Kcnq1 ),
Tumor-suppressing subchromosomal transferable fragment 4 (Tssc4 ), CD81 antigen
(Cd81 ), Tetraspanin 32 (Tspan32 ) and Achaete-scute complex homolog 2 (Ascl2 ); as well as one paternally-expressed long ncRNA, Kcnq1ot1, which is an antisense transcript of
Kcnq1. The locus also contains five genes that appear to not be imprinted, including
Tnfrsf22 , Tnfrsf26 , Cysteinyl-tRNA synthetase (Cars ), Transient receptor potential cation channel subfamily M member 5 (Trpm5 ) and Tyrosine hydroxylase (Th ). Two DMRs,
KvDMR1 and the Cdkn1c DMR, are located in the Kcnq1/Kcnq1ot1 locus. As mentioned earlier, the Cdkn1c DMR becomes methylated after fertilization, which is likely a consequence of transcriptional silencing of the repressed allele (Bhogal et al., 2004) and thus is not likely an ICR . The maternally methylated KvDMR1, which covers the
Kcnq1ot1 promoter, is the ICR of the locus (Smilinich et al., 1999). On the paternal allele, where KvDMR1 is unmethylated, Kcnq1ot1 is expressed, which results in silencing the adjacent genes. Paternal deletion of the KvDMR1 causes loss of expression of Kcnq1ot1 and results in activation of genes in cis in embryonic and extra-embryonic tissues
(Fitzpatrick et al., 2002; Lewis et al., 2004). Transcription of the intact Kcnq1ot1 RNA is required for imprinted expression of the entire locus (Thakur et al., 2004;
Mancini-Dinardo et al., 2006; Shin et al., 2008). An exception to this is that in several fetal and neonatal tissues, Cdkn1c imprinting was maintained on a Kcnq1ot1 truncation allele (Shin et al., 2008). Moreover, two CTCF binding sites were identified in the
KvDMR1 (Fitzpatrick et al., 2007). These data, together with the evidence that the
KvDMR1 has enhancer-blocking and silencer activities in vitro (Fitzpatrick et al., 2007;
18 Mancini-DiNardo et al., 2003; Thakur et al., 2003), indicate that CTCF may form an insulator at the KvDMR1 as a secondary mechanism in the Kcnq1/Kcnq1ot1 imprinted locus. Further characterization of enhancer elements is required to determine the involvement of CTCF in imprinted regulation at this locus.
1.2. DNA methylation and reprogramming in mammals
1.2.1. DNA methylation
DNA methylation of cytosine within CpG dinucleotides is an important epigenetic modification that is involved in genomic imprinting. In general, DNA methylation is correlated with gene repression (Bird and Wolffe, 1999), and the proper establishment and maintenance of methylation is required for normal development (Chen and Li, 2006).
In mammals, DNA methylation occurs predominantly at the 5’ carbon of cytosines in
CpG dinucleotides. For both genome-wide methylation and imprinting associated methylation, DNA methyltransferase 3A (DNMT3A) and DNMT3B are responsible for the major de novo methylation in gametes and during early development, while DNMT1 is responsible for maintenance of methylation patterns at each cell division, by recognizing hemi-methylated CpGs and adding methyl groups to the newly synthesized daughter strand (Bestor et al., 1988; Okano et al., 1998, 1999). DNMT3-like (DNMT3L) is a member of the DNMT3 family that lacks the catalytic domain but stimulates the activity of DNMT3A and DNMT3B through direct interactions (Suetake et al., 2004).
DNMT3L recognizes and binds to the tail of histone H3 and can thereby recruit
DNMT3A to chromatin and to its target DNA sequences (Ooi et al., 2007).
19 Proper DNA methylation is essential for imprinted expression because mutations in the Dnmt genes exhibit loss of methylation of the ICRs and loss of imprinted gene expression (Li et al., 1993; Bourc'his et al., 2001; Hata et al., 2002; Kaneda et al., 2004,
2010; Bourc'his and Bestor, 2004; Weaver et al., 2010). Moreover, DNMT3L cannot bind to H3 tails when H3 lysine 4 (H3K4) is methylated, suggesting H3K4 methylation is inhibitory to DNA methylation (Ooi et al., 2007). When histone H3K4 demethylase
KDM1B (lysine-specific demethylase 1B) is depleted from mouse oocytes, a number of imprinted DMRs fail to establish maternal-specific methylation, and embryos derived from KDM1B-deficient oocytes have biallelic expression of related imprinted genes
(Ciccone et al., 2009), presumably because of disruption in binding of DNMT3L to H3 tails. These results provide a nice example of crosstalk between histone modifications and
DNA methylation in regulation of imprinted gene regulation.
1.2.2. DNA methylation reprogramming in mammals
Two waves of genome-wide reprogramming of DNA methylation occur during the mammalian life cycle (Fig. 1.6) (Reik et al., 2001). During gametogenesis, genome-wide methylation is erased and a new methylation pattern is set at sex-specific times by de novo methylation, which includes the setting of methylation imprints. Mammalian germ cells are derived from a small population of extraembryonic mesoderm cells of the epiblast, as a consequence of signaling from adjacent tissues [reviewed in (Kota and Feil,
2010)]. The founder population of germ cells, called primordial germ cells (PGCs), which give rise to either spermatozoa or oocytes, is formed around embryonic day (E) 7.25 in
20
Figure 1.6 DNA methylation reprogramming during mammalian development
(Modified from Reik et al., 2001) Schematic showing DNA methylation changes during mammalian development. The vertical line separates gametogenesis and embryogenesis. PGCs in t he mouse become demethylated early in development. Remethylation (including remethylation at the germline DMRs) begins in E16 male germ cells, and after birth in growing oocytes. DNA methylation of some DMRs is reset in a sex-specific manner in the germ li nes. Following fertilization, the paternal genome is demethylated by an active mechanism, whereas the maternal genome is demethylated by a passive mechanism that depends on DNA replication. The DNA methylation level reaches its lowest extent around the blastocyst stage. Both genomes are remethylated after implantation. The embryonic lineages (Em) display higher levels of DNA methylation than the extraembryonic lineages (Ex) (Santos et al., 2002). Throughout the reprogramming during embryogenesis, methylated germline DMRs in imprinted loci and some repeats sequences remain methylated; whereas unmethylated germline DMRs remain unmethylated.
21 the mouse (Ginsburg et al., 1990). The PGCs proliferate and migrate toward the genital ridges that they colonize (at around E10.5 in the mouse), after which the gonads form.
During and after the migration, PGCs undergo global epigenetic reprogramming including DNA demethylation and changes in histone modifications (Seki et al., 2005).
Demethylation of DNA in PGCs depends partially upon the cytidine deaminase AID
(Popp et al., 2010), but the mechanism is not fully understood. Subsequently, DNA methylation, including differential methylation at germline ICRs, is re-established during gametogenesis. Remethylation begins in prospermatogonia in E16 male germ cells, and after birth in growing oocytes. In both germlines, the acquisition of DNA methylation at imprinted DMRs involves the DNMT3A-DNMT3L complex (Bourc'his et al., 2001;
Kaneda et al., 2004; Kato et al., 2007). In addition, at the Guanine nucleotide binding protein alpha stimulating complex locus (Gnas ) imprinted locus, transcription through the
DMRs is required for remethylation in oocytes (Chotalia et al., 2009).
Mature spermatozoa and oocytes remain highly methylated until fertilization, soon
after which a second wave of demethylation takes place. The paternal genome undergoes
DNA demethylation by an active mechanism (Mayer et al., 2000; Oswald et al., 2000),
whereas DNA replication-dependent demethylation occurs on the maternal genome by a
passive mechanism (Rougier et al., 1998). The mechanism that actively demethylates the
paternal genome is still under debate and remains largely unknown (Ooi and Bestor,
2008). Genome-wide DNA methylation reaches its lowest level at the blastocyst stage
and recovers to its approximately final methylation level around gastrulation [reviewed in
(Mann and Bartolomei, 2002; Li, 2002)]. The de novo methylation was observed
22 specifically in the inner cell mass (ICM), but not in the trophoectoderm of the blastocyst, consistent with higher methylation levels in the embryonic lineages than in the extraembryonic lineages (Santos et al., 2002).
The differential methylation patterns established at germ-line DMRs during gametogenesis are resistant to the genome-wide demethylation process after fertilization
(Weaver et al., 2009). One possibility is that these DMRs can survive the massive genome-wide reprogramming, which would require specific recognition and protection of these DMRs from the demethylation. Alternatively, active methylation may also play a role. It is suggested that DNA methylation is maintained in the preimplantation embryos by a combination of DNMT1 and the oocyte-specific form of DNMT1 (DNMT1o)
(Howell et al., 2001; Hirasawa et al., 2008; Cirio et al., 2008). In addition, DNMT3A and
DNMT3B are also required for the maintenance of CpG methylation at some loci (Chen et al., 2003).
In addition to the DNMTs, a number of proteins appear involved in maintenance of
DNA methylation at imprinted DMRs. For instance, two proteins encoded by AT-rich interaction domain (Arid) family genes, ARID4A/RBBP1 and ARID4B/RBBP1L1 are involved in maintenance of the maternal methylation at the Small nuclear ribonucleoprotein N (Snrpn ) locus (Wu et al., 2006b). Moreover, the maternally deposited protein DPPA3 (Developmental pluripotency-associated 3)/STELLA is required at multiple maternally and paternally methylated loci (Nakamura et al., 2007).
Furthermore, loss of DNA methylation in DMRs has been observed in both human and mouse when a KRAB zinc finger protein, ZFP57 (Zinc finger protein 57) is mutated or
23 deleted (Mackay et al., 2008; Li et al., 2008) .
In summary, it is apparent that multiple proteins maintain methylation at different
DMRs, and it is likely that other proteins involved in histone modifications and transcriptional regulation are also involved. However, the underlying mechanism by which the imprinted ICRs are specifically recognized and subsequently protected from demethylation remains to be uncovered. A group of DNA binding proteins that specifically bind to methylated DNA may be the missing link.
1.3. Methyl-CpG binding proteins
The Methyl-CpG-Binding-Domain (MBD) protein family and the Kaiso zinc finger protein family, which both have methyl-CpG binding activity, bind preferentially to methylated DNA and to recruit repressor complexes [reviewed in (Klose and Bird, 2006)].
In addition, proteins from both families bind to imprinted loci (Fujita et al., 1999;
Prokhortchouk et al., 2001; Drewell et al., 2002; Fournier et al., 2002; Samaco et al.,
2005; Makedonski et al., 2005; Filion et al., 2006) (see below). These activities make the
MBDs exciting candidates that can specifically recognize imprinted DMRs and silence linked genes.
1.3.1. The MBD family
The mammalian MBD family includes five members (Fig. 1.7 A): MeCP2 (Methyl
CpG binding protein 2), MBD1, MBD2, MBD3 and MBD4 (Hendrich and Bird, 1998).
24
Figure 1.7 The MBD and Kaiso families of methylated DNA binding proteins
Schematic showing the MBD and Kaiso families of methylated DNA binding proteins in the mouse. The abbreviations are defined as follows: methyl-CpG binding domain (MBD), transcription repression domain (TRD), the CxxC zinc finger domain (CxxC), glycine & arginine repeats (GR), glutamic acid-rich region (E-rich) and zinc finger (ZF). (A) The MBD family shares a conserved MBD. The MBD and TRD of MBD2 overlap. The MBD of MBD3 harbors mutations
(black bar), causing the inability of MBD3 to bind to methylated CpG (Saito and Ishikawa, 2002).
(B) The Kaiso family proteins share an N-terminus POZ/BTB domain and three ZFs that bind to methylated DNA. ZBTB4 and ZBTB38 contain additional ZFs that are not homologous among the three proteins.
25 All MBD family proteins contain the MBD, which binds methylated DNA in vitro (Nan et al., 1993). Moreover, all members are associated with histone deacetylase
(HDAC)-containing repressor complexes and transcriptional repression (Klose and Bird,
2006).
MeCP2 was the first member of the MBD family identified. Heterozygous mutations in the human X-linked MECP2 gene cause a severe and progressive neurodevelopmental
disorder called Rett syndrome (RS), almost exclusively in girls (Gonzales and LaSalle,
2010). Besides the MBD, MeCP2 also contains a transcriptional repression domain (TRD)
(Nan et al., 1997). MeCP2 has both methyl-CpG binding activity and sequence-specific
binding activity (Weitzel et al., 1997; Bowen et al., 2004a), and can mediate
transcriptional repression by two independent mechanisms. The enzymatic mechanism
involves interactions between the TRD and Sin3A, a co-repressor that exists in a complex
with HDACs (Jørgensen and Bird, 2002; Bowen et al., 2004a). The HDAC-independent
structural mechanism is indicated by the ability of MeCP2 to mediate chromatin
compaction regardless of the methylation state of DNA (Jørgensen and Bird, 2002;
Georgel et al., 2003; Bowen et al., 2004a). Besides its transcriptional repression roles,
MeCP2 can participate in activation of transcription, nuclear organization and splicing
(Bogdanovi ć and Veenstra, 2009). Interestingly, MeCP2 binds to a number of imprinted loci, including the H19 ICR (Drewell et al., 2002), the methylated maternal Snrpn DMR
(Samaco et al., 2005; Makedonski et al., 2005), and the maternal allele of the Zinc finger
(CCCH type) RNA binding motif and serine/arginine rich 1 (Zrsr1 ) DMR, which is maternally methylated (Fournier et al., 2002; Georgel et al., 2003). The binding of
26 MeCP2 in the Snrpn DMR has been correlated to imprinted regulation of a neighbor imprinted gene, Ubiquitin protein ligase E3A (Ube3a ) (Makedonski et al., 2005) .
However, it is unclear if MeCP2 functions to regulate imprinted expression at these loci
because two reports show that mutations in MeCP2 cause changes in expression levels but not loss of imprinting (Balmer et al., 2002; Samaco et al., 2005).
MBD1 binds to both methyl-CpG and nonmethylated-CpG sites via its MBD and two or three cysteine-rich (CxxC) motifs, respectively (Jørgensen and Bird, 2002;
Jørgensen et al., 2004). In addition, MBD1 has a TRD, which has no sequence similarity to the TRD of MeCP2 (Ng et al., 2000). Both the CxxC motifs and the TRD confer the transcriptional repression activity of MBD1 (Cross et al., 1997; Fujita et al., 1999, 2003;
Ng et al., 2000). MBD1 recruits the Suv39h1-HP1 complex and HDACs (Fujita et al.,
2003), as well as the H3 lysine 9 (H3K9) methyltransferase SETDB1 (Sarraf and
Stancheva, 2004) to enhance DNA methylation-mediated transcriptional repression.
Mbd1 -null mice have normal life-span without obvious developmental defects, but have deficits in adult neurogenesis and hippocampal function (Zhao et al., 2003). Similar to
MeCP2, MBD1 also binds to the maternal allele of the Zrsr1 DMR (Fournier et al., 2002).
In addition, human MBD1 associates with and suppresses expression of the human
SNRPN gene in vitro (Fujita et al., 1999) . Further characterization is required to
understand MBD1’s role in imprinted loci.
MBD2 and MBD3 are closely related to each other, not only in the MBD but also in
the sequence outside the MBD (Hendrich and Bird, 1998). The shared region of MBD2
and MBD3 contains a coiled-coil domain, which might be responsible for formation of
27 hetero- and homo-dimers of MBD2 and MBD3 (Tatematsu et al., 2000). MBD2 binds methylated DNA and confers DNA methylation-mediated transcriptional silencing through its TRD, which overlaps with the MBD and is not homologous to either the
MeCP2 or MBD1 TRDs (Boeke et al., 2000). The Mbd2-null mice are fertile with subtle immune system defects and impaired maternal behavior (Hendrich et al., 2001). A number of imprinted genes, including H19 , Igf2 , Igf2r , Snrpn , and Paternally expressed
gene 3 (Peg3 ), were tested in adult tissues from the Mbd2 -null mice. None of these genes were affected by depletion of MBD2 (Hendrich et al., 2001). Therefore, MBD2 seems to be dispensable for imprinted gene regulation.
MBD3 is the only member of the MBD family proteins that lacks the capacity to
selectively recognize methylated DNA in vitro , which is due to two amino changes in the
conserved residues of the MBD (Hendrich and Bird, 1998; Saito and Ishikawa, 2002). As
stated above, it is suggested that MBD2 and MBD3 form heterodimers though the
coiled-coil domain (Tatematsu et al., 2000). Therefore, MBD3 may be recruited to
methylated DNA elements by MBD2. However, unlike the mild phenotype observed in
Mbd2 -null mice, Mbd3 -null mice are embryonic lethal, indicating that MBD3 has
essential functions during embryonic development that is MBD2 independent (Hendrich
et al., 2001). MBD3 is required for the pluripotency of mouse embryonic stem (ES) cells
(Kaji et al., 2006), and is required for development of the ICM of preimplantation
embryo to epiblast (Kaji et al., 2007), likely by repressing the expression of trophoblast
lineage genes (Zhu et al., 2009). Similar to MeCP2 and MBD1, MBD3 also binds to the
Zrsr1 DMR (Fournier et al., 2002). Cofactors of MBD3, likely MBD2, may be involved
28 in this interaction.
Both MBD2 and MBD3 are components of the nucleosome remodeling and histone deacetylation (NuRD) repressor complex (Zhang et al., 1999; Wade et al., 1999;
Brackertz et al., 2002; Le Guezennec et al., 2006). Depletion of MBD3 from mouse ES cells causes disassembly of the NuRD complex, suggesting MBD3 is a core component of the NuRD complex (Kaji et al., 2006). MBD2- and MBD3-containing NuRD complexes are suggested to be distinct protein complexes with different biochemical and functional properties (Le Guezennec et al., 2006).
Unlike other members of the MBD family, MBD4 has a DNA glycosylase domain that targets sites of cytosine deamination, where MBD4 acts as a thymine DNA glycosylase and is involved in DNA break repair (Hendrich et al., 1999). MBD4 deficient mice have an increased rate of CpG to TpG mutations (Millar et al., 2002; Wong et al.,
2002). MBD4 is also suggested to be associated with transcriptional repression in a DNA methylation- and HDAC-dependent manner (Kondo et al., 2005).
As already stated, MeCP2, MBD1 and MBD3 bind to a variety of imprinted loci
(Drewell et al., 2002; Fournier et al., 2002; Samaco et al., 2005; Makedonski et al., 2005), indicating involvement of these proteins in imprinted regulation. Further investigation is required to identify potential interactions between MBD proteins and imprinted loci, as well as consequences of such interactions.
29 1.3.2. The Kaiso family
Another family of proteins, the Kaiso family, also binds to methylated DNA. The three members of the family, Kaiso, ZBTB4 (Zinc finger and BTB domain containing 4) and ZBTB38, share three homologous zinc finger domains and an N-terminal POZ/BTB
[poxvirus and zinc finger (POZ) or Broad complex, Tramtrack, Bric-a-brac (BTB)] domain (Collins et al., 2001; Filion et al., 2006) (Fig. 1.7 B). The POZ/BTB domain is an evolutionarily conserved protein-protein interaction domain that belongs to the C 2H2 class of zinc fingers (Collins et al., 2001). In the Kaiso family, the homologous zinc finger domains are responsible for the methylated-CpG-binding activity (Prokhortchouk et al., 2001). Similar to the MBD family proteins, Kaiso family proteins act as transcription repressors in a DNA methylation-dependent manner. Kaiso interacts with
CTCF, and thereby negatively regulates CTCF insulator activity (Defossez et al., 2005).
Kaiso-deficient mice are viable and fertile, with no detectable defects in development
(Prokhortchouk et al., 2006). Interestingly, these mice show resistance to intestinal cancer when crossed with tumor-susceptible mice, suggesting a role for Kaiso in suppressing intestinal cancer (Prokhortchouk et al., 2006). The other Kaiso family members, ZBTB38 and ZBTB4, preferentially bind to the methylated paternal ICR of H19/Igf2 locus, whereas Kaiso interacts with both alleles with a bias towards the paternal allele (Filion et al., 2006). The association of MBD and Kaiso families of proteins with imprinted loci prompted our interest in testing the potential role of methyl-CpG binding proteins in genomic imprinting.
30 1.4. CTCF
The methyl-CpG binding proteins may function to establish and/or maintain parental-specific methylation at ICRs. In contrast, the insulator protein, CTCF, may function to establish and/or maintain parental-specific hypomethylation at ICRs
[reviewed in (Filippova, 2008)]. CTCF is an eleven-zinc finger protein with multiple functions (discussed later in 1.4.2). The role of CTCF in imprinted gene expression has been best characterized at the H19/Igf2 locus. CTCF binds to the unmethylated allele of
the H19 ICR and plays important roles in the imprinted regulation of the H19/Igf2 locus
(see 1.1.1.1). Mutations or deletions of CTCF binding sites on the maternal H19 ICR result in gain of DNA methylation in the ICR during development, indicating CTCF binding is required for maintenance of hypomethylation at the maternal H19 ICR
(Schoenherr et al., 2003; Pant et al., 2003, 2004; Szabo et al., 2004; Engel et al., 2006).
Indeed, CTCF binding sites are enriched in DNA methylation-free regions genome-wide
(Mukhopadhyay et al., 2004). In addition to the H19 ICR, CTCF binds to DMRs at a
number of other imprinted loci including Dlk1-Dio3, Kcnq1/Kcnq1ot1 , Grb10 and
Rasgrf1 (Hikichi et al., 2003; Yoon et al., 2005; Fitzpatrick et al., 2007; Kernohan et al.,
2010). Moreover, CTCF binding appears to be methylation-sensitive at the
Kcnq1/Kcnq1ot1 , Grb10 and Rasgrf1 loci (Hikichi et al., 2003; Yoon et al., 2005;
Fitzpatrick et al., 2007), and depletion of CTCF in mouse oocytes led to reduction in two
imprinted transcripts: Grb10 and Gene trap locus 2 (Gtl2 ) / Maternally expressed 3
(Meg3 ), which is located in the Dlk1-Dio3 locus (Wan et al., 2008). Interestingly, the
Kcnq1/Kcnq1ot1 locus (see 1.1.1.2) and possibly the Dlk1-Dio3 locus (Lin et al., 2003;
31 Tierling et al., 2006) use the ncRNA mechanism to regulate imprinted expression. It is intriguing that CTCF is present and may possibly function at imprinted loci that are regulated by different imprinting control mechanisms (insulator versus ncRNA).
However, how CTCF manifests its functions at the H19/Igf2 locus and whether it has similar roles at other imprinted loci remains to be fully elucidated.
1.4.1. Models for insulator function
The insulator function of CTCF has been well-characterized. However, the mechanism of insulation is still hotly debated (Ohlsson et al., 2010). The looping model proposes that CTCF facilitates the formation of three-dimensional structures around the insulator element, resulting in different domains of transcriptional activity. This model is supported by the following evidence: CTCF interacts with itself (Pant et al., 2004), and therefore it can mediate interactions between distal binding sites; CTCF binds to the nucleolar protein nucleophosmin (B23) and tethers a transgenic insulator element to subnuclear sites (Yusufzai et al., 2004); CTCF binding sites are enriched at boundaries between genes that undergo and escape X chromosome inactivation (Filippova et al.,
2005), as well as boundaries of histone methylation domains (Kim et al., 2007; Xie et al.,
2007; Barski et al., 2007; Xi et al., 2007); CTCF is related to nucleosome positioning
(Kanduri et al., 2002; Fu et al., 2008), and is enriched at nuclear lamina-associated domains (Guelen et al., 2008); and moreover, CTCF mediates inter- and intra-chromosomal interactions [reviewed in (Williams and Flavell, 2008)]. These observations suggest that CTCF can provide anchor points and tether remote regions in
32 the genome together, supporting the looping model. Indeed, a study of the β-globin locus suggests that functional insulator elements form loops whereas non-functional insulator elements seem to have a linear conformation, by chromatin conformation capture (3C) experiments (Tolhuis et al., 2002). At the H19/Igf2 locus, 3C experiments have revealed
association of the ICR with the Igf2 promoters and DMRs (Murrell et al., 2004; Ling et
al., 2006; Kurukuti et al., 2006), suggesting that maternal Igf2 is isolated in a loop that
has a silencing environment, apart from the active domain where H19 and the enhancers are located [Fig. 1.8 A shows interaction of the ICR with the Igf2 DMR1, which is maternally methylated and was shown to function as a mesodermal silencer element to repress Igf2 maternal expression (Constância et al., 2000)]. In addition, the ICR has also been found to be associated with the endodermal enhancers (Yoon et al., 2007; Qiu et al.,
2008) (Fig. 1.8 B), suggesting a competition mechanism: the ICR binds to the enhancers, preventing possible interactions between these enhancers and the Igf2 promoter.
Alternatively, a tracking model suggests that RNA polymerase II (Pol II) and other activating factors bind to an enhancer and track along the chromatin until they contact an insulator or a promoter that has high binding affinity with the enhancer. The chromatin loops are formed by the intervening region between the enhancer and the scanned elements. This model is supported by the observation that histone acetylations are associated with increased chromatin mobility and flexibility, and therefore impacts formation of chromatin loops and further determines the interaction between enhancers and genes (Li et al., 2006). Consistently, it has been shown that CTCF binding sites at the chicken β-globin insulator 5’HS4 stall Pol II and inhibit enhancer-induced histone
33
Figure 1.8 The looping model for insulator function at the H19/Igf2 locus
Schematic showing two alternatives for the looping model at the H19/Igf2 locus. The maternal chromosome is shown. (A) ICR associates with Igf2 DMR1, isolating Igf2 to a silencing environment (gray area). (B) ICR associates with the 3’ shared enhancers, blocking possible interaction between the Igf2 promoter and the enhancers. H19 is located in an active chromatin domain (orange area). (A) and (B) together sugg est a competition mechanism: ICR interacts with both Igf2 and the enhancers, preventing possible interactions between Igf2 and the enhancers
(Yoon et al., 2007; Qiu et al., 2008).
34 acetylation (Zhao and Dean, 2004). Moreover, despite the fact that some CTCF binding sites also have Pol II occupancy, most of these sites are not transcribed (Chernukhin et al.,
2007), which can result from tracking Pol II being stalled by CTCF-bound DNA elements.
At the H19/Igf2 locus, the tracking model is supported by 3C experiments demonstrating that the enhancers associate with Igf2 and the entire intervening region between Igf2 and
H19 on the paternal chromosome, whereas the enhancers do not associate with regions upstream of the ICR on the maternal chromosome (Engel et al., 2008) (Fig. 1.9). The tracking process of an enhancer and its associated activating proteins may result in clustering of a transcriptional factory, which may serve as a subnuclear domain to which
CTCF tethers chromatin (Cook, 2003).
1.4.2. Discovery and multifunctionality of CTCF
It remains unclear if imprinted loci other than the H19/Igf2 locus use CTCF binding sites as insulators. Instead, CTCF may play distinct functions at these loci, as it has been shown to be multifunctional. CTCF was initially identified as a CCCTC binding factor in the chicken V-myc Myelocytomatosis viral oncogene homolog (Myc ) locus where in vitro studies indicated that it suppressed Myc expression (Lobanenkov et al., 1990). After its
discovery, CTCF was found associated with the 5’ end of the chicken β-globin locus
(5’HS4) (Bell et al., 1999), which functions as an insulator that blocks enhancer-promoter
interactions (Chung et al., 1993). The CTCF binding site at 5’HS4 is necessary and
sufficient for the insulator activity (Bell et al., 1999). Further studies of the H19/Igf2 and
Myc sequences also demonstrated requirement of CTCF binding sites for insulator
35
Figure 1.9 The tracking model for insulator function at the H19/Igf2 locus
Schematic showing the tracking model. Enhancers and active factors (not shown) track along the chromosome (arrows). On the maternal chromosome, the enhancers are blocked by the
CTCF-bound ICR. On the paternal chromosome, an insulator is not formed because CTCF cannot bind to the ICR, and the enhancers are able to activate Igf2 expression.
36 activities (Gombert et al., 2003; Wan and Bartolomei, 2008).
The CTCF protein has eleven zinc fingers and is conserved from fruit fly to human
(Moon et al., 2005). Among vertebrates, the CTCF proteins from zebrafish, Xenopus and human have 98% identical amino acids within the zinc fingers (Burke et al., 2002;
Pugacheva et al., 2006), whereas the avian CTCF zinc fingers are 100% identical to the human counterpart (Filippova et al., 1996). Surprisingly, despite the strict conservation in the primary sequence of the CTCF protein, CTCF binding sites are somewhat divergent among different species and target genes, suggesting that the zinc fingers are flexible
[reviewed in (Filippova, 2008)]. Consistently, CTCF was found to be multi-functional.
Besides the insulator activity, CTCF is associated with transcriptional activation and repression of a number of target genes. CTCF binding sites are involved in transcriptional repression in the Myc locus (Klenova et al., 1993; Filippova et al., 1996), the composite thyroid response element of the lysozyme gene (Burcin et al., 1997), the microsatellite-containing silencer of the HLA-DRB1 gene (Arnold et al., 2000), Paired box gene 6 (Li et al., 2004; Wu et al., 2006a), and others. In contrast, CTCF binding sites positively regulate transcription in the amyloid β-protein precursor promoter (Vostrov and Quitschke, 1997; Yang et al., 1999), the Interleukin 1 receptor associated protein kinase 2 gene (Kuzmin et al., 2005), Retinoblastoma promoter (De La Rosa-Velázquez et
al., 2007), and others. Furthermore, clusters of CTCF binding sites are identified at the X
inactivation center, likely having a role in X chromosome inactivation (Chao et al., 2002;
Pugacheva et al., 2005; Boumil et al., 2006; Navarro et al., 2006; Xu et al., 2007;
Donohoe et al., 2009). In conclusion, these data suggest that CTCF has multiple functions
37 in transcriptional regulation.
1.4.3. CTCF binding sites
It is hypothesized that CTCF binds to different DNA sequences through different
combinations of zinc fingers (Filippova et al., 1996). Indeed, various zinc finger
truncations cause loss of CTCF binding to different binding sites in vitro [reviewed in
(Ohlsson et al., 2001)]. However, the changes in CTCF binding affinity may be due to structural changes in the protein conformation, and/or due to the interdependence of
DNA-binding properties of individual zinc fingers. In addition, these mutations can also affect interactions between CTCF and cofactors.
In order to understand the mechanisms by which CTCF manifests its multiple functions, significant efforts have been made to map CTCF binding sites. In the past few years, genome-wide mapping of CTCF binding sites by chromatin immunoprecipitation
(ChIP) technologies (ChIP-on-chip and ChIP-sequencing) has identified thousands of
CTCF binding sites (e.g., ~26,000 sites in the human genome) (Kim et al., 2007;
Chernukhin et al., 2007; Xie et al., 2007; Barski et al., 2007; Xi et al., 2007; Chen et al.,
2008; Heintzman et al., 2009; Goren et al., 2009; Cuddapah et al., 2009; Kang et al., 2009;
Kunarso et al., 2010). Computational analysis of these datasets has provided insights into the functions of CTCF. For example, CTCF is enriched at boundaries of histone modification domains, suggesting CTCF’s role in formation of chromatin domains.
Analysis by Essien et al. identified three classes of CTCF binding sites according to their sequence similarity to a common consensus sequence (Essien et al., 2009). These classes
38 of CTCF sites, termed LowOc, MedOc and HighOc for low-, medium- and high-occupancy, have distinct properties. The majority of CTCF binding sites in imprinted loci belong to the LowOc and MedOc classes, which seem to have lower binding affinity to CTCF, and tend to be more cell-type specific than the HighOc class.
Interestingly, the LowOc sites also tend to cluster (e.g., the CTCF binding sites in the
H19 ICR), suggesting another level of regulation by recruiting multiple CTCF proteins to compensate for individual low affinity binding sites. The authors hypothesized that the low affinity may be important for the allele-specific binding of CTCF at imprinted loci.
Moreover, the LowOc sites are not as conserved as the HighOc sites. Consistently, imprinted genes evolved recently in mammals (see 1.1.1), whereas some HighOc sites, such as the CTCF binding sites in the β-globin locus, are conserved between avian and
mammalian species (Farrell et al., 2002).
1.4.4. CTCF cofactors
In addition to binding at different target sites, CTCF interacts with various protein
partners. For instance, its insulator function has been correlated with binding of
nucleophosmin (B23) (Yusufzai et al., 2004), Kaiso (Defossez et al., 2005), the
SNF2-like chromodomain helicase DNA binding protein 8 (CHD8) (Ishihara et al., 2006)
and Pol II (Chernukhin et al., 2007). The repression function of CTCF may be attributed
to the interaction between CTCF and the Sin3A corepressor complex (Lutz et al., 2000).
Furthermore, CTCF interacts with the Y-box DNA/RNA-binding factor YB-1, which is
also multifunctional and involved in transcription, replication and RNA processing
39 (Chernukhin et al., 2000). The transcriptional repressor protein YY1 often binds to DNA in proximity to CTCF sites (Kim, 2008; Kang et al., 2009). YY1 and POU domain class 5 transcription factor 1(POUF51)/OCT4, a transcription factor required for pluripotency, were found to be important CTCF cofactors in X-inactivation (Donohoe et al., 2007,
2009).
1.4.5. CTCF interacts with cohesins
In addition to the protein cofactors mentioned above, CTCF interacts with cohesin complex subunits (Stedman et al., 2008). The cohesin complex is highly conserved in eukaryotes, and is required for sister chromatid cohesion and chromosome segregation during cell division [reviewed in (Bose and Gerton, 2010)]. The mitotic cohesin complex consists of four subunits, including RAD21/Sister chromatid cohesion protein 1 (SCC1),
SCC3 (SA1/SA2), and two members of the structural maintenance of chromosomes
(SMC) family, SMC1/SMC1A and SMC3. The cohesin complex is proposed to form a ring-like structure and hold the sister chromatids together during mitosis (Fig. 1.10)
(Gruber et al., 2003; Ivanov and Nasmyth, 2005; Haering et al., 2008). In vertebrates, cohesins are loaded to the chromatin during late telophase (Sumara et al., 2000), with the assistance of the SCC2/SCC4 adherin complex (Ciosk et al., 2000; Watrin et al., 2006).
The bulk of cohesin complex dissociates at the prophase/metaphase transition, while a small portion of the complex persists at centromeres and dissociates at the onset of anaphase, due to the activity of separase [reviewed in (Barbero, 2009)]. As part of its mitotic function, the cohesin complex also participates in post-replication DNA break
40
Figure 1.10 The mitotic cohesin complex
Schematic showing the four sub units of the mitotic cohesin complex. SMC1 and SMC3 dimerize through the hinge domains and embrace the DNA (grey lines) by the coiled-coil arms. The N- and
C-terminus of SMC1 and SMC3 form the head domains, which contain the ATPase motifs. The head domains of SMC1 and SMC3 interact with C- and N-terminus of RAD21, respectively, to close the ring. SCC3 binds to RAD21 to stabilize the ring [reviewed in (Bose and Gerton, 2010)].
41 repair (Watrin and Peters, 2006) and heterochromatin formation (Gullerova and
Proudfoot, 2008).
Cohesin and cohesin-related genes are associated with human diseases. Mutations in the NIPBL gene (the human ortholog of Scc2 ), and subsequently, in the SMC1A and
SMC3 genes, were described in Cornelia de Lange syndrome (CdLS) patients (Krantz et al., 2004; Tonkin et al., 2004; Deardorff et al., 2007). Surprisingly, some CdLS individuals do not have severe defects in sister chromatid cohesion (Kaur et al., 2005), suggesting that cohesins are involved in gene regulation and development independent of their mitotic role. Following this discovery, a number of studies revealed that cohesins are expressed in post-mitotic cells, involved in transcriptional regulation and chromatin structure, and required in neuronal morphogenesis and embryonic development [reviewed in (Dorsett, 2007; Barbero, 2009)], likely through their ability to interact with chromosomes. Importantly, Drosophila SMC1, RAD21 and SCC3 are postulated to
prevent enhancer-promoter interactions at the cut locus (Rollins et al., 2004; Dorsett et al.,
2005), indicating that cohesins function as insulator proteins. Thus, cohesins may
function in concert with CTCF at insulator elements.
In my dissertation, I have investigated the role of protein factors in genomic
imprinting. I have studied the function of methyl-CpG binding proteins, CTCF and
cohesins at imprinted loci. In Chapter 2, I show that MBD and its NuRD complex
cofactor Metastasis tumor antigen 2 (MTA2) are required for maintenance of differential
methylation and imprinted expression at the H19/Igf2 locus. Depletion of both MBD3
42 and MTA2 from preimplantation embryos causes hypomethylation of the paternal H19
ICR, as well as biallelic expression of H19 . Furthermore, depletion of only MTA2 but not
MBD3 causes biallelic expression of the Peg3 gene. In Chapter 3, I have identified
cohesins as general cofactors of CTCF at three imprinted loci: the H19/Igf2 ,
Kcnq1/Kcnq1ot1 and Dlk1-Dio3 loci. RNA interference (RNAi) experiments targeting
CTCF and two cohesin subunits, SMC1 and RAD21, causes elevated expression of imprinted genes in mouse embryonic fibroblast cells (MEFs) . However, imprinted expression is maintained, suggesting a non-allelic role of CTCF and cohesins in genomic imprinting. Furthermore, in order to study cell-type specific functions of CTCF and cohesins. I have derived F1 hybrid trophoblast stem (TS) cell lines that can be used as a model cell line.
This dissertation provides important knowledge about the regulation of imprinted gene expression: 1) how DNA methylation marks at imprinted loci are interpreted and maintained during mammalian preimplantation development; and 2) how the insulator protein CTCF and its cofactor, the cohesin complex, are involved in non-allelic regulation of imprinted genes. This knowledge will be critical as an increasing number of epimutations at imprinted loci are reported in individuals with BWS, AS, SRS, PWS, many types of human cancers and other human diseases. A greater comprehension of the mechanisms regulating imprinted gene expression will contribute to the understanding of the etiology of these diseases. Furthermore, the incidence of these diseases is increased in babies conceived with ART such as in vitro fertilization, so understanding imprinting takes on added importance.
43 Chapter 2. Roles of methylated DNA binding proteins in genomic
imprinting
44
Germline DMRs are resistant to the genome-wide DNA demethylation during preimplantation development (see 1.2.2 and Fig. 1.6). Therefore, an important question is the identity of proteins that maintain DMR methylation. The methylated DNA binding proteins, including the MBD and Kaiso families, bind specifically to methylated DNA, and are associated with transcriptional repression (see 1.3). Thus, these proteins are candidates for the dual activities of recognizing DNA methylation at imprinted
ICRs/DMRs and silencing transcription from the imprinted locus. We proposed that the
MBD and Kaiso families of proteins may bind to germline DMRs and prevent their demethylation during preimplantation development. Moreover, as part of the exquisite regulation of imprinted gene regulation, DNA methylation at these DMRs is maintained through each cell division in somatic tissues. To assist the regulation of imprinted gene expression, the MBD and Kaiso families of proteins may bind to imprinted DMRs before and after implantation to 1) maintain the methylation memory; and 2) assist the transcriptional repression of silenced imprinted genes.
Indeed, another methylated DNA binding protein that does not belong to the MBD or the Kaiso families, NP95, is associated with the inheritance of DNA methylation after each cell cycle (Sharif et al., 2007). The study by Sharif et al. (Sharif et al., 2007) suggested that a complex containing NP95 and DNMT1 is localized to replicating foci in dividing cells and methylates the newly synthesized DNA strand. NP95 deficient mouse embryos exhibit lethality around E10.5, similar to Dnmt1 -null embryos, and have global
loss of DNA methylation (Li et al., 1992; Sharif et al., 2007). Importantly, DNA
45 methylation is lost in the H19 ICR, KvDMR1 and the ICR of the Dlk1-Dio3 locus
[Intergenic DMR (IG-DMR)], which results in misexpression of linked imprinted genes
(Sharif et al., 2007). Thus, NP95 functions as a general cofactor of DNMT1 and is important for maintenance of DNA methylation at imprinted loci, repetitive elements and heterochromatin regions (Sharif et al., 2007). In contrast, the MBD and Kaiso families of proteins, except MeCP2 and MBD3, are not essential in development (see 1.3), and therefore may function to maintain DNA methylation in a locus-specific manner.
To test the hypothesis that methylated-CpG binding proteins participate in regulation of genomic imprinting, I have used three approaches. First I have used ChIP assays of ES cells and mouse embryonic fibroblasts (MEFs) chromatin to investigate the binding of methylated DNA binding proteins at five imprinted loci: the H19/Igf2 , Dlk1-Dio3 , Snrpn,
Peg3 and Kcnq1/Kcnq1ot1 loci. Second, in collaboration with Dr. Kimberly Reese, a
former member of the Bartolomei laboratory, I have studied function of MBD3 using
RNAi to deplete MBD3 in mouse preimplantation embryos and then assay DNA
methylation and gene expression at imprinted loci. Finally, in collaboration with Drs.
Pengpeng Ma and Richard Schultz, I have studied the involvement of a MBD3 cofactor,
MTA2, in genomic imprinting by RNAi experiments.
2.1. Methylated DNA binding proteins are associated with imprinted loci
To investigate potential physical interactions between the MBD and Kaiso families
and imprinted loci, ChIP assays employing antibodies against MBD3, MBD1 and Kaiso
46 were performed. Two model cell lines, F1 hybrid mouse ES cells (Thorvaldsen et al., unpublished) and F1 hybrid MEFs were used. ES cells are derived from the ICM of the mouse blastocysts (E3.5) (Evans and Kaufman, 1981; Martin, 1981), representing the preimplantation embryonic lineage. MEFs are derived from E12.5-14.5 mouse embryos, representing embryonic lineage in a later developmental stage. The F1 hybrid cells are derived from crosses between two different mouse strains that harbor polymorphisms at imprinted loci (see Chapter 5), which allows the distinction between parental alleles. The parent-of-origin of imprinted loci can be recognized by restriction enzyme digestion of
PCR products, which result in strain-specific fragment sizes. I have studied associations of MBD3, MBD1 and Kaiso at five imprinted loci ( H19/Igf2 , Dlk1-Dio3 , Snrpn , Peg3 , and Kcnq1/Kcnq1ot1 ) in ES cells and MEFs.
2.1.1. MBD3 binds preferentially to the paternal H19 ICR in ES cells
Because Mbd3 -null embryos die right after implantation (Hendrich et al., 2001), we
chose MBD3 as the first candidate and tested occupancy of MBD3 at imprinted loci.
With help from Drs. Kimberly Reese and Raluca Verona, I performed ChIP experiments
utilizing an antibody against MBD3 and found that MBD3 was enriched at the paternal
H19 ICR but not other imprinted loci. Some data from the MBD3 ChIP assays was
published in (Reese et al., 2007).
Differential histone modifications between two parental alleles, including acetylated
histone H3 (AcH3) and dimethyl-histone H3 lysine 4 (H3K4me2), are known to be
associated with imprinted loci including the H19 ICR, IG-DMR, Snrpn DMR and
47 KvDMR1 [(Lewis et al., 2004; Carr et al., 2007; Verona et al., 2008); Verona et al. unpublished]. Therefore, an antibody against AcH3 was included as a positive control in the ChIP assays. In ES cells, the AcH3 antibody precipitated significantly more chromatin (p<0.01) than the no antibody negative control (No-Ab) at the H19 ICR and
the Snrpn DMR (Fig. 2.1 A). In contrast, an antibody against MBD3 precipitated significantly more chromatin (p<0.05) than No-Ab at the H19 ICR but not the Snrpn
DMR (p=0.24) (Fig. 2.1 A) (Reese et al., 2007). Therefore, MBD3 is associated with the
H19 ICR but not the Snrpn DMR in ES cells, suggesting a role of MBD3 at the H19 ICR during preimplantation development.
To determine if MBD3 localization at the H19 ICR is enriched on the methylated allele, and if MBD3 localizes at other imprinted loci allele-specifically, we performed allele-specific PCR analysis of DNA isolated from ChIP experiments. Five imprinted regions were analyzed, including two paternally methylated DMRs (the H19 ICR and
IG-DMR) (Tremblay et al., 1997; Lin et al., 2003) and two maternally methylated DMRs
(Snrpn DMR and KvDMR1) (Glenn et al., 1996; Smilinich et al., 1999), as well as the
H19 promoter. The AcH3 antibody was included as a positive control because it is enriched to the transcribed allele at imprinted loci in multiple cell types (Verona et al.,
2008). At the H19 ICR, AcH3 is preferentially enriched the unmethylated maternal allele:
69% of the total DNA that is ChIPed using the AcH3 antibody is from the maternal allele
(Fig. 2.1 B). In contrast to AcH3, MBD3 is associated preferentially with the methylated paternal allele (23% of ChIPed DNA is maternal) (Fig. 2.1 B). We tested four other imprinted regions, including the H19 promoter, IG-DMR, Snrpn DMR and KvDMR1
48
Figure 2.1 MBD3 preferentially localizes to the paternal H19 ICR
(A) ChIPs followed by real-time PCRs showing that MBD3 is associated with the H19 ICR but not the Snrpn DMR. *: p<0.05, **: p<0.01. p values show differences between No-Ab control and antibodies against AcH3 and MBD3. The data are normalized to AcH3. Error bars indicate standard deviation of at least three biological replicates. (B) to (F) Allele-specific ChIP PCRs showing that MBD3 is enriched allele-specifically at the paternal H19 ICR but not other imprinted loci. PCR primers specific for (B) the H19 ICR, (C) H19 promoter, (D) IG-DMR, (E) Snrpn DMR and (F) KvDMR1 were used. In: Input; B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bott om. At least three experiments were performed and similar results were obtained.
49 (Fig. 2.1 C-E). At all four regions, no enrichment of MBD3 was observed and the background levels of MBD3 binding are around 50% maternal (Fig. 2.1 C-G).
Furthermore, consistent with the maternal expression of H19 , AcH3 is enriched at the maternal H19 promoter (60%) (Fig. 2.1 C). AcH3 is also enriched on the maternal allele of IG-DMR (70%), which is paternally methylated (Fig. 2.1 D). At two maternally methylated DMRs, the Snrpn DMR and KvDMR1, AcH3 is enriched on the paternal allele (23% maternal and 4% maternal, respectively) (Fig. 2.1 E and F). Therefore, in ES cells, MBD3 is localized preferentially at the paternal H19 ICR, but not the H19 promoter, nor other imprinted loci tested. In contrast, in MEFs, preliminary data indicates that
MBD3 is not enriched at the H19 ICR (data not shown). Altogether, these observations
suggest that MBD3 play a role in the maintenance of DNA methylation at the H19/Igf2 locus during the preimplantation period but not after implantation.
MBD3, in contrast to other MBD proteins, does not bind directly to DNA in vitro because two amino acid residues are mutated in the MBD (Fig. 1.7) (Hendrich and Bird,
1998; Saito and Ishikawa, 2002). Consistently, the association of MBD3 with the H19
ICR is largely variable between different ChIP experiments (Fig. 2.1 A). MBD3 may bind to the H19 ICR indirectly through the interaction with other proteins such as MBD2 or other NuRD complex components (see 1.3.1). Furthermore, MBD3 only recognizes the
H19 ICR but not other imprinted loci tested (Fig. 2.1), suggesting different methylated
DNA binding proteins may bind to distinct imprinted loci.
50 2.1.2. A survey for occupancy of MBD1 and Kaiso at imprinted loci
In order to test if other methylated DNA binding proteins are recruited to imprinted loci, I performed ChIP assays using antibodies against MBD1 and Kaiso in both F1 hybrid ES cells and MEFs. MBD2 was not included in the analysis because the MBD2 antibody I used (see Chapter 5) was not suitable for ChIP experiments (data not shown).
Four imprinted DMRs, including two paternally methylated loci (the H19 ICR and
IG-DMR) (Tremblay et al., 1997; Lin et al., 2003) and two maternally methylated loci
(the Snrpn DMR and Peg3 DMR) (Glenn et al., 1996; Li et al., 2000) are included in this survey.
In ES cells, no significant enrichment of MBD1 or Kaiso was seen at the H19 ICR
(Fig. 2.2A). At the Snrpn DMR, the antibody against Kaiso precipitated significantly more chromatin (p<0.05) than the No-Ab control (Fig. 2.2 A). Consistent with a previous report that MBD1 associates with the human SNRPN gene in vitro (Fournier et al., 2002), the antibody against MBD1 precipitated more chromatin than No-Ab (Fig. 2.2 A).
However, it remains uncertain that if this enrichment is statistically significant because only two biological repeats were obtained. No statistically significant enrichment of
MBD1 and Kaiso was detected at the Peg3 DMR (n=3), although the average binding signals for MBD1 and Kaiso were higher than No-Ab at the Peg3 DMR (p=0.24 for
MBD1, p=0.17 for Kaiso). However, allele-specific PCRs showed maternal-specific enrichment of MBD1 and Kaiso at the Snrpn DMR and Peg3 DMR (Fig. 2.2 B), which are both methylated on the maternal allele (Glenn et al., 1996; Li et al., 2000). As shown in Fig. 2.2 B, at the Snrpn DMR, the active histone mark AcH3 is enriched on the
51
Figure 2.2 MBD1 and Kaiso binding to methylated DMRs in ES cells
ChIPs showing MBD1 and Kaiso occupancy at imprinted loci in ES cells. (A) ChIPs followed by real-time PCRs. *: p<0.05. p value shows differences between No-Ab control and antibodies against Kaiso. Error bars indicate standard deviation of two or three biological replicates. (B, C)
Allele-specific ChIP PCRs using primers specific to (B) Snrpn DMR and (C) Peg3 DMR. MBD1 and Kaiso bind preferentially to the maternal alleles. In: Input; B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom.
52 unmethylated paternal allele (24% maternal), whereas MBD1 and Kaiso are enriched on the methylated maternal allele (97% and 70%, respectively). At the Peg3 DMR, two
positive histone modifications, AcH3 and H3K4me2, are enriched on the unmethylated
paternal allele (32% and 33% maternal, respectively). In contrast, MBD1 and Kaiso are
enriched on the methylated maternal allele (68% and 63%, respectively). Taken together,
these data suggest that in ES cells, MBD1 and Kaiso bind to the methylated alleles of the
Snrpn DMR and Peg3 DMR.
To investigate roles of methylated DNA binding proteins at a different developmental stage, MEFs, which represent a postimplantation stage, were used for ChIP assays. Two paternally methylated DMRs (the H19 ICR and IG-DMR) (Tremblay et al., 1997;
Smilinich et al., 1999) were examined for occupancy of MBD1 and Kaiso in MEFs. As shown in Fig. 2.3 A, a significant enrichment was seen at the H19 ICR with the antibody against Kaiso (p<0.01). Moreover, Kaiso is enriched at the paternal allele of the H19 ICR
(34% maternal) (Fig. 2.3 B) . This observation was consistent with a previous study that
Kaiso binds to both alleles of the H19 ICR with a bias towards the paternal allele, in adult mouse brain (Filion et al., 2006). Although not statistically significant, the antibody against MBD1 precipitates more chromatin than the No-Ab at both DMRs in average
(Fig. 2.3 A). Interestingly, allele-specific analysis suggests that MBD1 is enriched at the paternal alleles of H19 ICR and IG-DMR (28% maternal and 31% maternal, respectively)
(Fig. 2.3 B and C), suggesting possible involvement of MBD1 at these two loci after implantation. AcH3 was included as a positive control and was enriched on the unmethylated alleles (Fig. 2.3 B and C).
53
Figure 2.3 MBD1 and Kaiso binding to methylated DMRs in MEFs
ChIPs showing MBD1 and Kaiso occupancy at imprinted loci in MEFs. (A) ChIPs followed by real-time PCRs. **: p<0.01. p value shows differences between No-Ab control and antibodies against Kaiso. Error bars indicate standard deviation of two or three biological replicates. (B, C)
Allele-specific ChIP PCRs using primers specific to (B) the H19 ICR and (C) IG-DMR. In: Input; B: maternal allel e; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom.
54 I hypothesized that distinct imprinted loci use different protein factors to maintain the differential DNA methylation. Consistent with this idea, the ChIP survey provides evidence suggesting that methylated-CpG binding proteins are associated with various imprinted loci. In ES cells, MBD3 is localized at the methylated paternal H19 ICR, and
Kaiso binds to the methylated maternal Peg3 DMR. In addition, at the H19 ICR, MBD3 is associated with the methylated paternal allele in ES cells, whereas Kaiso binds to the same allele in MEFs. These results suggest that imprinted loci require different methylated DNA binding proteins at distinct developmental time and there is a functional difference between these proteins. At other loci tested, these data also indicate association of MBD1 and Kaiso with the methylated DMRs. Further characterization is required to determine if MBD1 and Kaiso are functionally required at these loci.
2.2. MBD3 is required for the H19/Igf2 locus in preimplantation embryos
We have shown that MBD3 is associated with the paternal H19 ICR in ES cells but not MEFs (see 2.1.1), suggesting that MBD3 is involved in maintenance of DNA methylation at the H19 ICR during preimplantation development. Because the Mbd3 -null mice die right after implantation (Hendrich et al., 2001), we employed an RNA interference (RNAi) approach targeting MBD3 to generate hypomorphic alleles. This work was done in collaboration with Dr. Kimberly Reese and published in (Reese et al.,
2007). Two independent RNAi methods (injection and transgenic RNAi) targeting the same sequence of the Mbd3 mRNA were used to deplete MBD3 in preimplantation embryos (Fig. 2.4). In the first method, in vitro transcribed double stranded RNA (dsRNA)
55
Figure 2.4 Two independent RNAi approaches to deplete MBD3
Schematic showing the injection RNAi (A) and transgenic RNAi (B) methods targeting MBD 3 in preimplantation embryos. (A) F1 hybrid (see Chapter 5) 1-cell embryos are injected with dsMbd3 or dsGfp RNA and cultured in vitro to the blastocyst stage. (B) Transgenic females carryin g the transgene are mated with males that come from a different strain (see Chapter 5). For both methods, blastocysts are collected and analyzed for DNA methylation and imprinted gene expression.
56 [dsMbd3 or a dsRNA against Green fluorescent protein (Gfp )] was injected into one-cell embryos. Embryos were then cultured in vitro to the blastocyst stage and collected for
analysis. The second method utilized a Zona pellucida 3 (Zp3 ) promoter-driven transgene, which is expressed in growing oocytes (Fedoriw et al., 2004; Reese et al., 2007). The transgene, termed Zp3-dsMbd3 contains an Mbd3 inverted repeat and generates a dsRNA identical to the in vitro transcribed dsRNA, targeting the first 510 bp of the Mbd3 mRNA
(Fig. 2.4). Because the transgene is transcribed in growing oocytes, all embryos from a transgenic mother have maternal stores of dsMbd3, regardless of the genotypes of the embryos. In RNAi blastocysts generated by both the injection RNAi and the transgenic
RNAi, the Mbd3 RNA and protein levels were reduced, whereas Mbd2 RNA levels were not affected (Reese et al., 2007). As previously stated (see 1.3.1), MBD2 and MBD3 are the most closely related members of the MBD family. Because Mbd2 RNA levels were unaffected, the RNAi approaches used to target knock-down of MBD3 were both effective and specific.
2.2.1. MBD3 is required for the maintenance of DNA methylation at the H19 ICR in preimplantation embryos
To test if MBD3 depletion causes a loss of DNA methylation at imprinted loci, we
employed bisulfite mutagenesis and DNA sequencing to screen the DNA methylation
patterns of ICRs/DMRs in several pools of RNAi and control embryos. Because MBD3
localizes at the H19 ICR but not Snrpn DMR in ES cells (see 2.1.1), we first tested the
effects of MBD3 depletion at the H19 ICR. The 5’ portion of the H19 ICR (Fig. 2.5 A)
57
Figure 2.5 Depletion of MBD3 causes loss of DNA methylation at the H19 ICR
(A) Schematic showing the paternal H19/Igf2 locus. 17 CpGs were assayed in the 5’ portion of the
H19 ICR as indicated by the black line below the diagram. (B) Bisulfite mutagenesis of RNAi treated embryos and control embryos. Percentages of hypomethylated strands are indicated on top of each group. Each line of circles represents a single DNA strand with the number to the left of the line corresponding to the number of times this pattern was seen. Each circle represents a single CpG. If the CpG was methylated, the circle is filled. Those strands with less than half of the
CpGs methylated are considered hypomethylated. Only the paternal strands are shown.
58 was analyzed by cloning and sequencing of the bisulfite treated DNA. We could distinguish the maternal and paternal strands by using single nucleotide polymorphisms
(SNPs) between the two alleles in our F1 hybrid embryos (see Chapter 5). Fig. 2.5 B shows the paternal strands of H19 ICR, as all sequenced maternal strands were unmethylated as expected (data not shown). Strands with less than half of the CpGs methylated are considered hypomethylated. In control embryos, only a few hypomethylated strands are identified (13% and 22% for uninjected and dsGfp-injected blastocysts, respectively), but in dsMbd3-injected embryos, 42% of the paternal strands are hypomethylated. Zp3-dsMbd3 transgenic blastocysts showed similar results with 42% of the paternal strands hypomethylated compared with 16% of the strands in non-transgenic embryos. Similar analysis was also performed at the Snrpn DMR, but the normally methylated maternal alleles from RNAi embryos did not differ from controls
(data not shown), consistent with the result the absence of MBD3 binding (see 2.1.1).
These results indicate that MBD3 is required to maintain the paternal DNA methylation at the H19 ICR during the preimplantation period. In addition, they indicate that other
imprinted DMRs utilize different protein factors to maintain differential DNA
methylation during preimplantation development. Consistently, MBD1 and Kaiso are
suggested to be associated with the methylated allele of Snrpn DMR and Peg3 DMR in
ES cells, whereas no binding of MBD1 or Kaiso was observed in the H19 ICR in ES cells
(see 2.1.2 and Fig. 2.2).
59 2.2.2. MBD3 is required for the imprinted expression of H19 in preimplantation
embryos
Loss of DNA methylation of the H19 paternal ICR is associated with activation of
H19 expression from the paternal allele (Engel et al., 2004). Therefore H19 allelic
expression may be affected by MBD3 depletion, together with loss of DNA methylation
at the H19 ICR. To investigate whether MBD3 is required for imprinted expression of
H19 , we performed allele-specific reverse-transcription (RT) PCR analysis of single blastocysts (Mann et al., 2003). In blastocysts derived from injection of one-cell embryos with dsMbd3, 26% (9/35) of the embryos showed biallelic H19 expression whereas only
4% (2/26) of uninjected blastocysts and no dsGfp-injected blastocysts (0/28) exhibited biallelic H19 expression (Fig. 2.6A). The numbers of dsMbd3 embryos that exhibit biallelic H19 expression are significantly different from the numbers of control embryos
(p<0.01, Chi-square) (Fig. 2.6 A). Similarly, 39% (15/38) of blastocysts derived from
Zp3-dsMbd3 transgenic mothers exhibited biallelic H19 expression, whereas none of the non-transgenic blastocysts (0/5) showed biallelic expression (Fig. 2.6 B). Although allelic expression of H19 is closely coordinated with Igf2 (see 1.1.3.1), Igf2 is not expressed in early embryos until after implantation. Thus, we cannot use this approach to assess the role of MBD3 in the regulation of Igf2 expression. Nevertheless, these experiments demonstrate that MBD3 is required for proper imprinted expression of H19 in
preimplantation embryos.
RNAi treated embryos were also assayed for the other imprinted genes expressed at this early stage, including Snrpn , Peg3 , Gtl2 and ATPase Cu++ transporting alpha
60
Figure 2.6 Depletion of MBD3 causes biallelic expression of H19 in preimplantation embryos
(A) H19 is biallelically expressed in 26% (9/35) of dsMbd3-injected embryos. This is statistically significant different from controls [p<0.01 compared w ith uninjected (2/46) and p<0.01 compared with dsGfp -injected (0/28)]. (B) H19 is biallelically expressed in 39% (15/38) of transgenic embryos. None (0/5) of the non-transgenic controls showed biallelic expression.
61 polypeptide (Atp7a ). Atp7a is located on the X chromosome and is silenced on the paternal allele due to imprinted X-inactivation in the preimplantation embryo (Huynh and
Lee, 2003). Female embryos were tested for Atp7a and all embryos were tested for Snrpn ,
Peg3 and Gtl2 expression. No change in allelic expression was detected for any of these genes in MBD3-depleted embryos (data no shown). These results suggest that MBD3 may be only required for H19 at the blastocyst stage.
By using two independent RNAi methods (injection RNAi and transgenic RNAi), we observed the same effects on the imprinted expression and DNA methylation of H19 , demonstrating that MBD3 is required for maintenance of the DNA methylation and repression of the paternal H19 , but not at other imprinted loci tested. These results indicate that the silent alleles at imprinted loci other than the H19/Igf2 locus employ other
MBD or Kaiso families of proteins to confer silencing of the inactive allele.
However, the paternal H19 ICR is totally demethylated on some strands but not on others (Fig. 2.5). Similarly, not all RNAi embryos exhibited biallelic expression of H19
(Fig. 2.6). The variable phenotypes between embryos are probably due to incomplete depletion of the MBD3 protein from these embryos because detectable amounts of Mbd3
RNA and protein remained in the RNAi embryos (Reese et al., 2007). Furthermore, several other methylated DNA binding proteins, including MBD1 and Kaiso, are also expressed in preimplantation embryos and may contribute to the maintenance of paternal
H19 ICR methylation [(Zeng et al., 2004); Reese and Bartolomei, unpublished].
Consistently, our preliminary ChIP data suggest that MBD1 and Kaiso bind to the
62 paternal H19 ICR, in both ES cells and MEFs (see 2.1.2 and Fig. 2.2, 2.3).
We examined expression of four other imprinted genes in MBD3-depleted embryos,
but none showed activation of the silent allele. One explanation is that residual amount of
MBD3 is sufficient to maintain appropriate monoallelic expression of these genes.
Alternatively, MBD3 is not involved in the repression of these genes. In this case, other
proteins or protein complexes, such as ZFP57 (Mackay et al., 2008; Li et al., 2008),
MBD1 and Kaiso, may be required for their imprinted expression. Alternatively, other
methylated DNA binding proteins that are expressed in preimplantation embryos might
be sufficient to compensate for the reduction of MBD3 at these loci, but not at H19 . By
reducing several methylated DNA binding proteins in concert by RNAi, such
compensatory mechanisms might be uncovered.
Our analysis has been focused on the preimplantation stage because it is the period
when genome-wide reprogramming takes place. In addition, our RNAi methods only
provide transient depletion of MBD3. The MBD3 protein level is presumably recovered
later as the dsRNAs are diluted during embryonic growth after implantation, and the
dsRNAs are degraded. For the dsMbd3 RNAi experiments, despite the early phenotypes
described above, mice from Zp3-dsMbd3 transgenic mothers can be developed full term
and are healthy and fertile. This suggests that the RNAi effect is only brief and does not
affect the viability. Alternatively, all the embryos that did not have sufficient amount of
MBD3 died before birth. In embryos that survived, temporary depletion of MBD3 may
still impact later development in a subtle way. To test if MBD3 knock-down has a
long-existing effect in genomic imprinting, embryos from later developmental stages
63 need to be analyzed for DNA methylation status and imprinted expression of imprinted loci. Reese et al. recovered seven E7.5 embryos from a Zp3-dsMbd3 transgenic female.
In all embryos, expression of H19 was monoallelic (Reese et al., 2007). This suggests that either the active paternal H19 expression observed in Zp3-dsMbd3 blastocysts is repressed during/after implantation or the embryos have biallelic H19 expression did not survive. Analysis of both H19 expression and DNA methylation at the H19 ICR in additional postimplantation Zp3-dsMbd3 embryos (e.g., E6.5 embryos) was therefore of interest. I have conducted several experiments using E6.5 embryos, but that transgene was unfortunately repressed. Based on experience of others using a transgenic technique in the B6 mice (Bartolomei and Weaver, personal communications), such transgene repression is not rare because B6 mice are prone to high levels of DNA methylation
(Sapienza et al., 1989). To resolve this problem, I introduced the Zp3-dsMbd3 transgene into a different strain, DBA/2J (The Jackson laboratory), which has less DNA methylation activity (Sapienza et al., 1989; Allen et al., 1990). Similar experiments were performed in mice with a different transgene by Jamie Weaver, a graduate student in the
Bartolomei laboratory. Successful reactivation of the transgene was observed after it was crossed into DBA/2J mice for 5-6 generations (Weaver et al. unpublished). I have successfully derived the DBA/2J Zp3-dsMbd3 mice, now allowing examination of the knock-down efficiency, as well as further analysis of the long-term effects of MBD3 depletion.
2.3. MTA2 is required for imprinted expression at the H19/Igf2 and Peg3 loci in
64 preimplantation embryos
MBD3 is a component of the NuRD complex, which can remodel nucleosomes and deacetylate histones, and thereby is associated with transcriptional repressions [reviewed in (McDonel et al., 2009)]. The NuRD complex is composed of at least seven proteins in mammals and shares four components, HDAC1, HDAC2 and two histone-binding proteins (Retinoblastoma binding protein 4 (RBBP4) and RBBP), with the Sin3A histone deacetylase complex (McDonel et al., 2009). The other three proteins, Chromodomain helicase DNA binding protein (CHD)/Mi-2 (CHD3/Mi-2α or CHD4/Mi-2β), MBD
(MBD2 or MBD3), and MTA (MTA1, MTA2 or MTA3), appear to be unique to the
NuRD complex (McDonel et al., 2009).
HDAC1 and 2 are histone deacetylases that catalyze removal of lysine residues in
histone tails (Thiagalingam et al., 2003), a process associated with epigenetic silencing.
CHD3/4 is the largest subunit of the NuRD complex, and has a chromodomain that
exhibits ATP-dependent nucleosome-remodeling activity (Zhang et al., 1998).
MTA proteins (MTA1, 2, and 3) are the last characterized components of the NuRD
complex (Manavathi and Kumar, 2007). It is suggested that only one MTA protein is
present in a given NuRD complex (Yao and Yang, 2003; Bowen et al., 2004b; Fujita et al.,
2004; Manavathi and Kumar, 2007; Wang et al., 2009), presumably contributing to target
specificity. The founding member of the MTA family, MTA1, was first described in rat
metastatic breast tumors (Toh et al., 1997). Since then, MTAs have been implicated in
numerous biological processes [e.g., estrogen receptor α (ER α)-associated signaling]
65 (Manavathi and Kumar, 2007). MTA2 deficient mice exhibit multiple phenotypes, including partial embryonic and perinatal lethality, female infertility, abnormal T cell activation, and lupus-like autoimmune disease (Lu et al., 2008). Although it is likely that the three MTAs form different NuRD complexes with distinct functions (Yao and Yang,
2003; Bowen et al., 2004b; Fujita et al., 2004; Manavathi and Kumar, 2007; Wang et al.,
2009), the role of MTA proteins in the function of NuRD complexes remains elusive.
MBD3 is an integral member of the NuRD complex because depletion of MBD3 causes disassembly of the NuRD complex and degradation of some NuRD complex components in ES cells (Kaji et al., 2006). Consistently, in our MBD3-depleted blastocysts, MTA2 protein levels were reduced by 53% in the nuclei (Reese et al., 2007).
The requirement for MBD3 at the H19/Igf2 locus and the accompanying reduction in
MTA2 levels strongly implicates a role for the NuRD complex in genomic imprinting.
In this section, I will present experiments done in collaboration with Drs. Pengpeng
Ma and Richard Schultz, investigating requirements of MTA2 for genomic imprinting in preimplantation embryos. We have shown that: 1) MTA2 is required for the imprinted expression of two genes, H19 and Peg3 ; and 2) similar to MBD3, MTA2 is required for
maintenance of the paternal methylation at the H19 ICR. This work was published in (Ma et al., 2010).
66 2.3.1. MTA2 is required for the imprinted expression of H19 and Peg3 in preimplantation embryos
We demonstrated that MBD3 is required for maintenance of the paternal DNA
methylation in the H19 ICR, and is required for imprinted expression of H19 in preimplantation embryos (see 2.2). MTA2 protein levels were reduced in the nuclei of
MBD3-depleted embryos (Reese et al., 2007), indicating involvement of the NuRD complex in regulating H19 expression. To define further the function of the NuRD complex in genomic imprinting, and in particular the function of MTA2, we used an injection RNAi approach to deplete selectively MTA2 in mouse preimplantation embryos.
Similar to the injection RNAi approach used to target MBD3 (Fig. 2.4 A), 1-cell embryos were injected with in vitro transcribed dsMta2 or dsGfp RNA. The embryos were then cultured to the blastocyst stage and collected for analysis. MTA2 protein was largely depleted (less than 10% of the dsGfp-injected control) in these embryos as determined by immunocytochemistry and western blot (Ma et al., 2010).
To determine the effect of directly depleting MTA2 on imprinted genes, allele-specific RT PCR was performed, allowing examination of allelic expression of H19 in MTA2-depleted blastocysts (Fig. 2.7 A). Consistent with the phenotypes observed in
MBD3-depleted embryos (Fig. 2.6), we found that H19 expression was biallelic in 32% of the dsMta2-injected embryos (16/50) whereas only 10% (4/39) of the dsGfp-injected embryos exhibited biallelic H19 expression (p<0.05, Chi-square). Moreover, no difference was observed in the expression of Snrpn between dsMta2- and dsGfp-injected embryos [0% (0/36) biallelic and 1% (1/21) biallelic, respectively] (Fig. 2.7 B); as well as
67
Figure 2.7 Depletion of MTA2 causes biallelic expression of H19 and Peg3 in preimplantation embryos
Shown are percentages of embryos that biallelically express (A) H19 , (B) Snrpn and (C) Peg3 . *: p<0.05. p values show difference between RNAi embryos and control embryos. (A) H19 is biallelically expressed in 32% (16/50) of dsMta2-injected embryos. This is statistically significantly different from dsGfp-injected [10% (4/39)]. (B) Monoallelic expression of Snrpn is maintained in both dsGfp- [1% (1/21) biallelic] and dsMta2-injected embryos [0% (0/36) biallelic] (p>0.05). (C)
Monoallelic expression of Kcnq1ot1 is maintained in both dsGfp- [0% (0/3) biallelic] and dsMta2-injected embryos [0% (0/5) biallelic]. (D) Peg3 is biallelically expressed in 32% (12/38) of dsMta2-injected embryos. This is statistically significantly different from dsGfp-injected [7%
(2/28)].
68 in the expression of Kcnq1ot1 (3 dsGfp-injected embryos and 5 dsMta2-injected embryos
were analyzed, and all of them showed monoallelic expression of the paternal allele) (Fig.
2.7 C). In contrast, a different phenotype between the dsMbd3 embryos and dsMta2
embryos was seen at the Peg3 locus: 32% (12/38) of the MTA2-injected embryos
exhibited biallelic Peg3 expression when compared with dsGfp-injected embryos [7%
(2/28)] (Fig. 2.7 D), whereas Peg3 remained imprinted in the dsMbd3 embryos (see
2.2.2). These results suggest that MBD3 and MTA2 function in concert to regulate
imprinted expression of H19 , but not Snrpn , in preimplantation embryos, and that MTA2
functions independently of MBD3 at the Peg3 locus.
2.3.2. MTA2 is required for the maintenance of DNA methylation at the H19 ICR
in preimplantation embryos
The H19 ICR is hypomethylated in MBD3-depleted embryos (Fig. 2.5), whereas
Snrpn DMR is not affected (see 2.2.1). To investigate the effects of MTA2 depletion in
DNA methylation at imprinted loci, we performed bisulfite mutagenesis and DNA
sequencing using pools of dsMta2- and dsGfp-injected embryos (Fig. 2.8). The normally
unmethylated maternal H19 ICR, paternal Snrpn DMR and paternal Peg3 DMR were all
hypomethylated as expected (data not shown). DNA methylation profiles of the paternal
H19 ICR, maternal Snrpn DMR and maternal Peg3 DMR are shown in Fig. 2.8. We
observed a significant increase (p<0.01, Chi-square) in the number of DNA
hypomethylated strands in MTA2-depleted embryos (43%) compared with controls (12%)
(Fig. 2.8 A). These results strongly suggest that MTA2 is necessary for both the proper
69
Figure 2.8 Depletion of MTA2 causes hypomethylation at the H19 ICR
Bisulfite mutagenesis o f RNAi treated embryos and control embryos. Percentages of hypomethylated strands are indicated on top of each group. Strands with less than half of the
CpGs methylated are considered hypomethylated. 17, 16 and 24 CpGs were analyzed for (A)
H19 ICR, (B) Snrpn DMR, (C) Peg3 DMR, respectively. Only the normally methylated strands are shown. (A) The dsMta2-injected embryos have significantly more hypomethylated strands compared with the dsGfp-injected embryos (p<0.01). (B, C) The RNAi embryos are not different from the controls (p>0.05).
70 imprinted gene expression of H19 and the maintenance of H19 paternal methylation in
mouse preimplantation embryos. In contrast, no changes in DNA methylation were
detected at the Snrpn DMR (Fig. 2.8 B), consistent with the monoallelic expression of
Snrpn in both RNAi embryos and controls (Fig. 2.7 B). However, although the
dsMta2-injected embryos had more strains that were hypomethylated than the
dsGfp-injected embryos (40% versus 22%) (Fig. 2.8 C), the difference was not
statistically significant (p>0.05, Chi-square). Effects of MBD3- and MTA2-depletions at
imprinted loci are summarized in Table 1.
In the dsMta2-injected embryos analyzed, roughly equal percentages of embryos
exhibited biallelic expression of Peg3 and H19 (Fig. 2.7 A and D), suggesting that these
two imprinted loci require similar amounts of MTA2 protein and/or MTA2-containing
complexes. However, only the DNA methylation in the paternal H19 ICR was
significantly affected. These data suggest that cis elements other than the Peg3 DMR may be required for imprinted expression of Peg3 . Consistently, the Peg3 DMR has not been demonstrated to function as an ICR by deficiency studies. Szeto et al. (Szeto et al., 2004) analyzed three different lines of transgenic mice that carry a bacterial artificial chromosome (BAC) covering 120 kb around the Peg3 locus. Although all BACs included the Peg3 DMR sequence, only one out of three lines of transgenic mice exhibited partial imprinted expression of Peg3 and differential methylation in the Peg3 DMR. These results suggest that the Peg3 DMR alone is not sufficient to control the imprinted expression of Peg3 and our data indicate that MTA2 functions through an unknown element, in addition to the Peg3 DMR, to repress the maternal Peg3 .
71 Table 1 Effects of MBD3- and MTA2-depletion at imprinted loci in preimplantation embryos
Imprinted locus (DNA methylation and/or gene MBD3-depleted MTA2-depleted expression) embryos embryos
Paternal H19 ICR methylation Reduced Reduced H19/Igf2 H19 expression Biallelic Biallelic
Maternal Snrpn DMR No effect No effect methylation Snrpn
Snrpn expression Monoallelic Monoallelic
Maternal Peg3 DMR N/A No effect methylation Peg3
Peg3 expression Monoallelic Biallelic
Kcnq1/Kcnq1ot1 Kcnq1ot1 expression Monoallelic Monoallelic
Dlk1-Dio3 Gtl2 expression Monoallelic N/A
X-Chromosome Atp7a Monoallelic N/A
N/A: data not available. The phenotypes that are different from control embryos are underlined.
72 During mammalian preimplantation development, genome-wide DNA demethylation occurs at most methylated sequences except at some repeats and germline imprinted
DMRs (see 1.2.2 and Fig. 1.6). The paternally methylated H19 ICR must survive the active DNA demethylation on the paternal genome and then stay methylated in subsequent stages during passive demethylation (Tremblay et al., 1995, 1997), whereas the maternally methylated Peg3 DMR must survive the maternal-specific passive demethylation (Li et al., 2000). How these processes occur has remained largely unknown. We proposed that methylated DNA binding proteins are involved in these processes, and that maternally- and paternally-methylated DMRs require protection from different proteins. We have shown that MBD3, as well as its NuRD complex cofactor
MTA2, are intimately involved at the maintenance of DNA methylation at the H19 ICR during preimplantation development.
We found that depletion of MTA2, but not MBD3, led to biallelic expression of Peg3 , suggesting an independent function of MTA2 that is not related to MBD3. Because
MTA1 is translocated from the nucleus to the cytoplasm in MTA2-depleted embryos as shown by western blot and immunocytochemistry experiments performed by Dr.
Pengpeng Ma (Ma et al., 2010), MTA1-containing NuRD complexes may be the actual functioning complexes at the Peg3 locus, whereas MTA2- and MBD3-cotaining NuRD complexes control the H19 locus.
In conclusion, data presented in Chapter 2 have provided the first link between
MBD3- and MTA2-containing NuRD complex and genomic imprinting. It remains
73 elusive how loss of MBD3 and MTA2 is mechanistically linked to biallelic expression of
H19 and Peg3 , and linked to loss of DNA methylation at the H19 ICR. Although MBD3 localizes at the H19 ICR, no binding of MTA2 at the H19 ICR and Peg3 DMR was observed in ChIP experiments using multiple MTA2 antibodies (data not shown). This could be because the MTA2 antibodies were not suitable for ChIP experiments.
Nevertheless, our results uncover essential roles for MBD3, MTA2 and the NuRD complex in maintaining differential DNA methylation and controlling allelic expression at imprinted loci.
74 Chapter 3. Roles of CTCF and the cohesin complex in genomic
imprinting
75
CTCF is a multi-functional protein that plays important roles in chromosomal organization (see 1.4). CTCF binds to a number of imprinted loci, including H19/Igf2¸
Dlk1/Dio3, Kcnq1/Kcnq1ot1 , Grb10 and Rasgrf1 (see 1.1.3.1 and 1.4). As previously mentioned, CTCF binding sites are required for the insulator activity of the H19 ICR.
Deletion of these binding sites from the maternal chromosome leads to biallelic expression of Igf2 and reduced H19 expression (Schoenherr et al., 2003; Pant et al., 2003,
2004; Szabo et al., 2004; Engel et al., 2006) . In contrast, CTCF may play other functions
such as transcriptional activation or repression at other imprinted loci.
It remains largely unknown how CTCF can manifest multiple functions in the
nucleus. One hypothesis is that CTCF recruits disparate protein cofactors to distinct
binding sites and therefore functions differently. Indeed, a number of CTCF cofactors are
associated with various functions of CTCF (see 1.4.4). The cohesin complex plays
important roles in chromosome biology and transcriptional regulation (Barbero, 2009),
and was recently found to interact with CTCF (Stedman et al., 2008). Lieberman and
colleagues described colocalization of CTCF and cohesins at the major latency control
region of the Kaposi’s sarcoma-associated herpesvirus (KSHV). Deletion of the CTCF
binding site from the viral genome disrupted cohesins binding and caused derepression of
two lytic transcripts (Stedman et al., 2008). The authors additionally showed
colocalization of CTCF and cohesins at the cellular Myc locus. These data prompted our interest to investigate possible involvement of cohesins at imprinted loci.
In Chapter 3, I tested the hypothesis that CTCF and cohesins function together for
76 regulation of imprinted gene expression, by conducting the following experiments: 1)
Using ChIP assays, I studied colocalization of CTCF and two cohesin subunits, SMC1 and RAD21, at five imprinted regions: the H19 ICR, conserved sequence at the
intervening region (IVR) between H19 and Igf2 , a CTCF binding site upstream of Igf2 ,
Gtl2 DMR and KvDMR1. 2) I further analyzed the functional requirement of CTCF and cohesins at these imprinted loci by small interfering RNA (siRNA) treatments to knock-down CTCF, SMCI and RAD21 in F1 hybrid primary MEFs. MEF was chosen as a model cell line because it exhibits monoallelic expression of imprinted genes that are linked to CTCF binding sites. Furthermore, another model cell line, TS cells were derived to study cell-type specific functions of CTCF and cohesins.
3.1. CTCF and cohesins colocalize at three imprinted loci
To investigate potential physical interactions between cohesins and imprinted loci,
ChIP assays employing antibodies against two cohesins subunits, SMC1 and RAD21, as
well as CTCF were performed. Three imprinted loci, the H19/Igf2 , Dlk1-Dio3 and
Kcnq1/Kcnq1ot1 loci, were included in the study. The H19/Igf2 locus employs an
insulator mechanism to regulate imprinted gene expression in a CTCF binding
sites-dependent manner (see 1.1.3.1 and 1.4.1), whereas both the Dlk1-Dio3 (see 3.1.2)
and Kcnq1/Kcnq1ot1 loci (see 1.1.3.2) are suggested to use the long ncRNA mechanism.
There are CTCF binding sites located in these loci that employs different mechanisms to
regulate imprinting (Szabó et al., 2000; Bell and Felsenfeld, 2000; Hark et al., 2000;
Kanduri et al., 2000; Fitzpatrick et al., 2007; Kernohan et al., 2010), suggesting that
77 CTCF may manifest multiple functions in genomic imprinting.
3.1.1. CTCF and cohesins colocalize at the H19/Igf2 locus
The four CTCF binding sites located in the H19 ICR have been well characterized
(see 1.1.3.1). In addition to the H19 ICR, there are a number of other CTCF binding sites at the IVR (Chen et al., 2008) and the Igf2 gene (Du et al., 2003; Kurukuti et al., 2006;
Yoon et al., 2007; Chen et al., 2008; Li et al., 2008). CTCF’s role at the IVR remains largely unknown. Some CTCF binding sites around the Igf2 gene recruit CTCF in a maternal-specific manner, and were hypothesized to be required for maternal repression of Igf2 expression (Kurukuti et al., 2006; Li et al., 2008) (see 3.2.1.2). I have examined colocalization of CTCF and cohesins in the H19 ICR and the IVR (see below).
3.1.1.1. CTCF and cohesins colocalize at the maternal H19 ICR
To investigate cohesins’ presence at imprinted loci, ChIP experiments using antibodies against CTCF and two cohesin subunits, SMC1 and RAD21, were carried out in MEFs (Fig. 3.1 B and Fig. 3.2 B). Consistent with the colocalization of CTCF and cohesins in the KSHV genome (Stedman et al., 2008), both SMC1 and RAD21 colocalize with CTCF at the H19 ICR (Fig. 3.1 B), but not at three other H19 regions analyzed, including a region upstream of the ICR (DMD_UP), H19 promoter (Prom) and Exon 5
(Ex5) (Fig. 3.1 A, B) (Verona et al., 2008). In addition, PCRs of ChIPed DNA using primers specific for the c-Myc CTCF binding site was also included as a positive control
(Fig. 3.1 A). These data demonstrate that cohesins bind specifically to the H19 ICR
78
Figure 3.1 Cohesins colocalize with CTCF at the H19 ICR, in a CTCF binding sites-dependent manner
(A) Schematic showing the H19/Igf2 locus. The white bars indicate four CTCF bi nding sites that
∆ are deleted in the H19 R/+ mutant. The regions analyzed in (B, C) are indicated by black bars below the diagram. DMD_UP: upstream of ICR; Rep1&2: CTCF binding sites 1&2 in the ICR;
Prom: H19 promoter; Ex5: H19 Exon 5. (B, C) ChIPs followed by real-time PCRs. *: p<0.05; **: p<0.01. p value shows differences between No-Ab control and antibodies against CTCF, SMCI and RAD21. Error bars indicate standard deviation of at least three biological replicates. (B)
CTCF and cohesins colocalize at the H19 ICR but not other regions (DMD_UP, Prom and Ex5) in the H19/Igf2 locus. The Myc promoter was included as a positive control. (C) Shown are binding
∆ of CTCF and cohesins in wild-type (WT) MEFs and neonatal liver from the H19 R/+ mutant at
∆ Rep1&2 region. Cohesins binding is lost in the H19 R/+ mutant.
79 which has four CTCF binding sites, but do not bind to other regions assayed at the H19
locus.
The binding of cohesins at the KSHV latency control region is dependent on the
CTCF binding sites (Stedman et al., 2008). In order to test if the same dependence is also
true at the H19 ICR, I performed ChIP assays targeting CTCF and cohesins in the
∆ ∆ H19 R/+ mutant, which has maternal transmission of the H19 R mutant allele, an allele that lacks the four CTCF binding sites at the H19 ICR (Fig. 1.2) (Engel et al., 2006). As expected, CTCF no longer binds to the H19 ICR in the mutant, and the same is true for cohesins (Fig. 3.1 C). These data demonstrate that the maternal CTCF binding sites are required for cohesins binding at the H19 ICR. In addition, because cohesin binding is
∆ completely abolished in H19 R/+ mutant, cohesins likely only bind to the wild-type maternal allele. Indeed, allele-specific PCRs of ChIPed DNA from F1 hybrid MEFs suggest that both cohesin subunits preferentially bind to the maternal allele of H19 ICR
(Fig. 3.2 B). In contrast, at the Ex5 region where cohesins are not enriched (Fig. 3.1 B), the background levels of ChIP signals are equal between maternal and paternal alleles, similar to signals from the No-Ab control (Fig. 3.2 D). Similarly, in neonatal liver from
∆ the H19 R/+ mutant, ~50% of background signals at the H19 ICR are from either parental allele (Fig. 3.2 C). In conclusion, cohesins colocalize with CTCF at the maternal H19
ICR, in a CTCF binding site-dependent manner. These results were published together with the KSHV data in (Stedman et al., 2008).
80
Figure 3.2 Cohesins colocalize with CTCF at the maternal H19 ICR
(A) Schematic showing the H19/Igf2 locus. The white bars indic ate the four CTCF binding sites
∆ that are deleted in the H19 R/+ mutant. The regions analyzed in (B) to (D) are indicated by black bars below the diagram. Rep3: CTCF binding sites 3 in the ICR; Ex5: H19 Exon 5. (B-D)
Allele-specific ChIP PCRs. (B) CTCF and cohesins are enriched on the maternal allele at Rep3 in
∆ wild -type (WT) cells. (C) The allelic binding of CTCF and cohesins is lost in the H19 R/+ mutant.
(D) No allelic binding of CTCF and cohesins is seen at Ex5 in WT. In: Input; B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom.
81 3.1.1.2. CTCF and cohesins colocalize at the intervening region between H19 and
Igf2
A DNase I hypersensitivity site (HSS) intergenic to H19 and Igf2 (~40 kb from either gene) that colocalizes to a conserved sequence between human and mouse has previously been described (Fig. 3.3 A) (Koide et al., 1994; Ainscough et al., 2000). Other studies have suggested that this HSS region contains an Igf2 muscle specific repressor and/or brain specific enhancer (Ainscough et al., 2000; Jones et al., 2001). Interestingly, a CTCF binding site was reported to be located in the conserved sequence within the HSS (Chen et al., 2008). Recently, using a mouse imprinting region tiling array (MIRTA) of embryonic tissues, Huang and Barlow have identified an abundant transcript adjacent to the intergenic HSS and nearer to Igf2 (Ru Huang and Denise Barlow, Research Center for
Molecular Medicine of the Austrian Academy of Sciences, unpublished) . This transcript is called the intervening region (IVR) transcript. The IVR transcript is ~8 kb long and its transcription start site is 1.5 kb distal to the CTCF binding site within the HSS (Fig. 3.3
A). Allele-specific expression analysis of the IVR transcript in wild-type embryonic tissues demonstrated it is preferentially expressed from the paternal allele (75% to 85% bias) (Thorvaldsen and Bartolomei, unpublished). When this analysis was performed in embryonic tissues inheriting deletion of the H19 maternal ICR (Thorvaldsen et al., 1998), preferential expression of the paternal IVR transcript is no longer detected (Thorvaldsen and Bartolomei, unpublished). These data demonstrate that imprinted expression of the
IVR transcript is dependent upon the H19 ICR. I set out to determine if CTCF and cohesins bound to the conserved sequence adjacent to the IVR transcription start site,
82
Figure 3.3 Cohesins colocalize with CTCF at the CTCF binding site in the IVR
(A) Schematic showing the IVR region which is ~40 kb from H19 and Igf2. The direction of the IVR transcript is shown with the arrow. The green boxes represent two conserved regions, one of which overlaps with a HSS and contains a CTCF binding site (Koide et al., 19 94; Ainscough et al.,
2000; Chen et al. 2008). The regions analyzed in (B, C) are indicated by black bars below the diagram. Ctrl1 & 2: control 1 & 2; IVR CBS: IVR CTCF binding site. (B) ChIPs followed by real-time PCRs show that CTCF and cohesins bind to the IVR CBS but not Ctrl1 & 2. The H19
ICR is included as a positive control. *: p<0.05; **: p<0.01. p value shows differences between
No-Ab contro l and antibodies against CTCF, SMCI and RAD21. Error bars indicate standard deviation of at least three biological replicates. (C) Allele-specific ChIP PCRs show that CTCF and cohesins bind to both alleles of IVR CBS in MEFs. In: Input; B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom. At least three experiments were performed and similar results were obtained.
83 where a CTCF binding site had been identified (Fig. 3.3 A) (Chen et al., 2008).
To investigate potential interactions of cohesins at the IVR locus, I performed ChIP experiments in F1 hybrid MEFs using antibodies targeting CTCF, SMCI and RAD21.
Real-time PCRs of ChIPed DNA showed that CTCF and cohesins colocalize at the IVR
CTCF binding site (CBS), but not in two control regions, Ctrl1 and Ctrl2 (Fig. 3.3 A, B).
Ctrl1 is located ~2.1 kb proximal to the CBS, whereas Ctrl2 resides at the 5’ end of the
IVR transcript Exon 2 (Fig. 3.3 A) (Thorvaldsen and Bartolomei, unpublished). Because the IVR transcript is partially imprinted and expresses from mostly the paternal allele
(Thorvaldsen and Bartolomei unpublished), I tested if CTCF and cohesins binding was enriched at one particular allele by allele-specific PCRs. As shown in Fig. 3.4 C, CTCF and cohesins bind to both alleles of the IVR CBS. Two histone modifications, AcH3 and
H3K4me2 are associated equally with both alleles. The binding of CTCF, SMCI and
RAD21 are slightly biased to the paternal allele (48%, 40% and 43% maternal, respectively), whereas the input DNA is 59% maternal. The same bias was seen in three independent experiments but was not statistically different from the input (data not shown). Consistent with our data that CTCF and cohesins bind to both alleles at the IVR
CBS in MEFs, an approximately 1 kb region around the IVR CBS is hypomethylated on both parental alleles in embryos (Koide et al., 1994).
These data show that CTCF and cohesins bind to both alleles of the IVR CBS, indicating non-allelic function of CTCF and cohesins at this region. This may be because the IVR transcript is not detectable in MEFs (Thorvaldsen and Bartolomei, unpublished),
CTCF and cohesins may not have allele-specific functions at the IVR in MEFs. Therefore,
84 the biallelic binding of CTCF and cohesins may be a consequence of the silencing of the
IVR transcript. Given that a consistent small bias of CTCF and cohesins binding towards the paternal allele, they may bind to this site at a paternal-specific manner in tissues that express the IVR transcript. ChIP experiments of such tissues will provide information to distinguish allelic versus non-allelic functions of CTCF and cohesins at this region.
3.1.2. CTCF and cohesins colocalize at the Gtl2 DMR and KvDMR1
As previously mentioned, a number of imprinted loci other than the H19/Igf2 locus contain CTCF binding sites. To examine if cohesins are general cofactors of CTCF at imprinted loci, I expanded the investigation to two more CTCF-associated imprinted loci: the Dlk1-Dio3 and Kcnq1/Kcnq1ot1 loci (Fitzpatrick et al., 2007; Kernohan et al., 2010).
I showed that CTCF and cohesins colocalize at these two loci, with a maternal bias at the
Gtl2 DMR and biallelic binding at KvDMR1.
3.1.2.1. CTCF and cohesins colocalize to the unmethylated allele of Gtl2 DMR
The Dlk1-Dio3 imprinting cluster was initially identified in uniparental disomy
(UPD) of distal mouse Chr 12 deficiencies (Takada et al., 2000), and UPDs of the
orthologous human Chr 14q32 (Kagami et al., 2005). The locus contains three paternally
expressed protein-coding genes: Delta-like 1 homolog (Dlk1 ), Retrotransposon-like 1
(Rtl1 ) and Deiodinase iodothyronine type III (Dio3 ) and a number of
maternally-expressed ncRNAs, including Gtl2 , an antisense transcript of Rtl1 (Rtl1as ), a
number of genes encoding small nucleolar RNAs (snoRNAs), Micro RNA containing
85 gene (Mirg ) and numerous microRNAs (Fig. 3.4). Gtl2 encodes a long ncRNA with multiple spliced forms of unknown function (Schuster-Gossler et al., 1998; Paulsen et al.,
2001). It was proposed that these ncRNAs can make up a large polycistronic transcription unit that represses the protein-coding genes on the maternal chromosome, suggesting a ncRNA-dependent mechanism of imprinting control. In human and mouse, the paternally methylated germline DMR, Intergenic DMR (IG-DMR), functions as the ICR of the cluster (Lin et al., 2003, 2007; Kagami et al., 2008). Deletions or mutations of IG-DMR cause loss of imprinting across the entire locus. There is a second postfertilization-derived paternally methylated Gtl2 DMR, spanning the promoter, exon 1 and part of the intron 1 of Gtl2 (Takada et al., 2000) (Fig. 3.4). Maternal deletion of the
Gtl2 DMR in human results in regional silencing of imprinted genes (Kagami et al.,
2010). The Gtl2 DMR contains a CTCF binding site [(Paulsen et al., 2001; Kernohan et
al., 2010); Takada and Ferguson-Smith, unpublished], raising the possibility that cohesins
bind to the region.
To test if cohesins colocalize with CTCF at the Gtl2 DMR, ChIP experiments were
performed in MEFs. At the Gtl2 DMR, the antibodies against CTCF and cohesins
precipitated significantly more chromatin (p<0.01) than a No-Ab control. However, no
enrichment of CTCF and cohesins was seen at the IG-DMR (Fig. 3.5 A), consistent with
the absence of CTCF binding sites (Paulsen et al., 2001). CTCF binds to the Gtl2 DMR at
a similar level compared with the H19 ICR, whereas less SMC1 and RAD21 bound to the
Gtl2 DMR as bound to the H19 ICR (Fig. 3.5 A), indicating that the affinity of cohesins
differ between the Gtl2 DMR and the H19 ICR. Therefore, cohesins may interact with
86
DMR is also Glt2 The The paternally methylated Gtl2. Mirg. ssed microRNA genes spread out an over 200 DMR is located at the 5’ end of Gtl2 is located more than 500 kb downstream of Dio3 ilization DMR, the (not shown). locus (not to scale). number A of maternally expre Mirg to imprintinglocus Dlk1-Dio3 Gtl2 Dlk1-Dio3 methylatedon the paternal chromosome. IG-DMR is the ICR of the locus. A second, post-fert Schematic showing the mouse kb region from downstream of Figure3.4 The mouse
87
Figure 3.5 Cohesins colocalize with CTCF at the maternal allele of Gtl2 DMR
(A) ChIPs followed by real-time PCRs show that CTCF and cohesins bind to the Gtl2 DMR but not
IG-DMR. The H19 ICR is included as a positive control. *: p<0.05; **: p<0.01. p value show s differences between No-Ab control and antibodies against CTCF, SMCI and RAD21. Error bars indicate standard deviation of at least three biological replicates. (B) Allele-specific ChIP PCRs show that CTCF and cohesins preferentially bind to the maternal allele of Gtl2 DMR in MEFs. In:
Input; B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom. At least three experiments were performed and similar results were obtained.
88 CTCF differently and exhibit disparate functions at these two loci.
To determine if the binding of CTCF and cohesins is exclusive to one parental allele, we performed allele-specific PCR analysis of DNA isolated from ChIP experiments.
Because the CTCF binding site resides near the Gtl2 promoter, and histone modifications that are associated with active transcription were reported to be enriched at maternal allele at this region (Carr et al., 2007), we included AcH3 and H3K4me2 as controls.
AcH3 and H3K4me2 are preferentially associated with the transcribed, maternal allele
(Fig. 3.5 B), as do CTCF and the two cohesin complex subunits (Fig. 3.5 B). Binding of
CTCF on the unmethylated maternal allele is consistent with results of experiments showing a methylation-sensitive binding pattern (Filippova, 2008). In conclusion, CTCF and cohesins colocalize in an allele-specific manner at the Gtl2 DMR, similar to the H19
ICR. This result suggests that CTCF, as well as cohesins, are involved in the imprinted regulation of the Dlk1-Dio3 locus.
3.1.2.2. CTCF and cohesins colocalize to both alleles of the KvDMR1
At the Kcnq1/Kcnq1ot1 locus (see 1.1.3.2 and Fig. 1.5), allelic silencing of linked imprinted genes is regulated by expression of the long ncRNA Kcnq1ot1. This locus also harbors two CTCF binding sites (Fitzpatrick et al., 2007). These two CTCF binding sites,
CTCF BS1 and 2, reside in one of the two CpG islands within KvDMR1 (Fig. 3.6 A).
To determine if cohesins colocalize with CTCF at ICR of the Kcnq1/Kncq1ot1 locus,
I analyzed chromatin from F1 hybrid MEFs (see Chapter 5) (Verona et al., 2008).
Following ChIP, real-time PCRs showed that CTCF, SMC1 and RAD21 associated with
89
Figure 3.6 Cohesins colocalize with CTCF at KvDMR1
(A) Schematic showing the KvDMR1 (Black box), which is enlarged at the bottom. The striped boxes represent two CpG islands in the KvDMR1. The regions an alyzed in (B) are indicated below the diagram. (B) ChIPs followed by real-time PCRs show that CTCF and cohesins bind to both CTCF binding sites (BS) 1&2. *: p<0.05; **: p<0.01. p values show differences between
No-Ab controls and antibodies against CTCF, S MC1 and RAD21. The data are normalized to the
No-Ab controls of each PCR. Error bars indicate standard deviation of at least three biological replicates.
90 both CTCF BS1 and 2, but not the control region (Ctrl) (Fig. 3.6 B). Allele-specific analysis revealed that the allelic bias of CTCF binding at KvDMR1 in MEFs was very weak, in contrast to AcH3 and H3K4me2, which were strongly biased to the paternal allele. Compared with the input DNA, the CTCF antibody precipitated 5-20% KvDMR1
DNA from of the paternal allele than the maternal allele. Similarly, the two cohesin complex subunits showed a small bias (~5%) towards the paternal allele (Fig. 3.7 B, C).
Furthermore, the No-Ab controls showed a background signal that were biased towards the maternal allele (Fig. 3.7 B, C), suggesting the small maternal enrichments observed in
CTCF and cohesins ChIPed DNA may be a consequence of a PCR bias. My results contrast those of Fitzpatrick et al., who showed that CTCF binding at the KvDMR1 was exclusively on the paternal allele in C57BL/6 X SD7 F1 hybrid MEFs (Fitzpatrick et al.,
2007). Because I used different mouse strains to generate MEFs [C57BL/6 X
B6(CAST7P12X), see Chapter 5], the contrasting results could reflect mouse strain differences.
In summary, I have shown that CTCF and cohesins colocalize to four imprinted
regions, including two ICRs ( H19 ICR and KvDMR1), a non-ICR DMR ( Gtl2 DMR),
and the IVR region between H19 and Igf2. The allele-specific analysis revealed three patterns of CTCF and cohesins binding at DMRs: 1) they bind allele-specifically to the
H19 ICR; 2) they bind biallelically to the KvDMR1 (ICR); 3) they bind
allele-specifically to the Gtl2 DMR (which is not an ICR). These results suggest that
CTCF and cohesins play distinct functions at different imprinted loci. Interestingly,
91
Figure 3.7 Cohesins colocalize with CTCF at both alleles of KvDMR1
(A) Schematic showing the KvDMR1 (Black box), which is enlarged at the bottom. The striped boxes represent two CpG islands in the KvDMR1. The regions analyzed in (B, C) are indicated below the diagram. (B, C) Allele-specific ChIP PCRs showing biallelic binding of CTCF and cohesins in (B) CTCF BS1 and (C) CTCF BS-2. AcH3 and H3K4me2 are included as controls. B: maternal allele; C: paternal allele. The percentage of protein binding on the B allele over total binding (B+C) is indicated at the bottom. At least three experiments were performed and similar results were obtained.
92 CTCF and cohesins always behave the same at these sites. When they are bound to one specific allele, they always bind to the unmethylated allele. At the IVR, cell types or tissues that express the IVR transcript need to be included in future analysis. If CTCF and cohesins still bind to both alleles, they may play non-allelic functions such as transcription activation or repression at this region. In contrast, if CTCF and cohesins bind to one parental allele specifically, they could play important roles in the imprinted expression of this transcript. In conclusion, the data presented in 3.1 suggest that cohesins are general cofactors of CTCF, and may play important functions in concert with CTCF at imprinted loci.
3.2. Functional analysis of CTCF and cohesins in MEFs
The results described above demonstrated that CTCF and cohesins colocalize at four imprinted regions. To assess the functional significance of CTCF and cohesins binding in regulating expression of imprinted genes, I performed RNAi experiments to knock-down levels of CTCF and/or cohesins. F1 hybrid MEFs (see Chapter 5) were treated with siRNAs targeting Ctcf , Smc1 and Rad21 . An siRNA that does not target any known transcript in mouse was included as a control (Ctrl siRNA). I first determined conditions to achieve the best knock-down efficiency and then demonstrated that CTCF and cohesins were specifically depleted.
To achieve the best knock-down efficiency, I performed sequential siRNA treatments
by treating the cells every 48 h. Maximum depletion (~75% to 90%) of targeted proteins
93 occurred around 72 h and was maintained through 144 h, with two to three treatments of siRNAs (Fig. 3.8 A shows examples of CTCF and SMC1 knock-down at various time points). Consequently, I conducted most of the experiments with two or three sequential siRNA treatments and harvested cells at 96h and 144h. Because no obvious growth or proliferation defects were observed in the siRNA treated cells, any changes in these cells were likely due to the altered function of CTCF and cohesin complex subunits in transcriptional regulation, rather than cell division.
The RNAi approach using two or three sequential treatments of one siRNA against each transcript successfully reduced the amount of the protein encoded by the targeted mRNA (Fig. 3.8 B). Around 75%~90% depletion of target proteins are achieved (Fig. 3.8
B). When compared with the amount of a control protein, αTUBULIN, the residual amounts of CTCF, SMC1 and RAD21 are 17%, 25% and 10%, respectively (Fig. 3.8 B).
Among the three siRNAs that target CTCF and cohesins, the Ctcf siRNA and Rad21 siRNA usually had stronger knock-down efficiency than the Smc1 siRNA (data not shown). In addition, because there are thousands of CTCF binding sites in the genome
(Kim et al., 2007; Chernukhin et al., 2007; Xie et al., 2007; Barski et al., 2007; Xi et al.,
2007; Chen et al., 2008; Heintzman et al., 2009; Goren et al., 2009; Cuddapah et al., 2009;
Kang et al., 2009; Kunarso et al., 2010), and cohesins are associated with many regions in the genome [reviewed in (Barbero, 2009)], it is possible that depletion of CTCF and cohesins would affect housekeeping genes. Therefore, I examined the relative RNA levels of target RNAs compared with RNA levels of three housekeeping genes [ β-actin (Actb ),
Glyceraldehyde-3-phosphate dehydrogenase (Gapdh ) and Ribosomal protein large P0
94
Figure 3.8 Successful depletion of CTCF, SMC1 and RAD21 from MEFs
(A) Western blots showing knock-down effects achieved by Ctcf siRNA (Ctcf-si) and Smc1-si treatments for various time lengths. (B) Western blots showing 75% to 98% depletion of CTCF and cohesin subunits by siRNA tre atments after 96 h of treatments. For both (A) and (B), anti-αTUBULIN is used as a control. The residual amounts of target proteins are indicated below the gels. (C) RT PCRs showing relative Ctcf , Smc1 and Rad21 mRNA levels compared with three housekeeping genes, Actb , Gapdh and Rplp0 in target siRNA treated cells. No difference of relative target mRNA levels are seen, indicating that these housekeeping genes are not affected by RNAi. Target mRNA levels are normalized to the housekeeping genes. Relative val ue of the control siRNA treatment is set to 1.
95 (Rplp0 )] in the siRNA-treated cells (Fig. 3.8 C) and found no effect on the relative
transcript abundance of target RNAs when compared with different housekeeping genes.
In general, 65%-85% reduction of target RNAs was achieved compared with the control
siRNA treated cells. Thus, for all subsequent experiments, only Rplp0 was used as the
control.
3.2.1. Depletion of CTCF and cohesins in MEFs lead to increased Igf2 expression from the normally expressed paternal allele
I first characterized the effect of CTCF and cohesins depletion at the H19 /Igf2 locus.
I hypothesized that by depleting CTCF and/or cohesin complex subunits, insulator activity at the H19 ICR would be reduced, resulting in increased Igf2 transcription due to
expression from the normally repressed maternal allele, as well as decreased H19
expression due to competition of H19 and Igf2 for the shared enhancers on the maternal
chromosome. Of note, because the IVR transcript is not expressed in MEFs (Thorvaldsen
and Bartolomei, unpublished), it is not included in the following experiments.
3.2.1.1. Depletion of CTCF and cohesins lead to increased expression of Igf2
To test the hypothesis mentioned above, I treated MEFs with individual siRNAs
against CTCF and cohesins, as well as with different combinations of siRNAs (Fig. 3.9
A). As expected, Ctcf siRNA-treated cells had elevated Igf2 expression (p<0.01),
approximately three-fold relative to control siRNA-treated cells. Rad21 siRNA treatments
also led to increased Igf2 expression (p<0.05), whereas Smc1 siRNA had no significant
96
Figure 3.9 CTCF and cohesins depletion associate with elevated expression, but not loss of imprinted expression of Igf2
(A) Real-time PCRs showing H19 and Igf2 relative RNA levels in siRNA-treated wild-type (WT)
MEFs. CTCF- and RAD21-depleted WT MEFs have elevated expression of Igf2. Ctrl: control; C:
Ctcf; S: Smc1; R: Rad21 siRNAs. *: p<0.05, **: p<0.01. p values show differences between CTCF or cohesins siRNA treatments and the Ctrl siRNA treatments. H19 and Igf2 RNA levels are normalized to Rplp0 . Relative value of the control siRNA treatment is set to 1. Error bars indicate standard deviation. (B) Allele-specific analysis of Igf2 expression in WT MEFs. CTCF- and cohesins-depleted MEFs have monoallelic expression of Igf2 B: maternal allele; C: paternal allele.
Samples were collected at 144 h after three or two sequential treatments of siRNAs. Only the second lane (Ctcf siRNA) has a low level of maternal Igf2 expression detected (arrow head).
97 effect (p>0.05), probably due to insufficient depletion of SMC1 proteins (Fig. 3.8 B).
Surprisingly, H19 levels were mostly unaffected by siRNA treatments and, in Ctcf siRNA treatments, elevated (Fig. 3.9 A). Next, I examined allelic expression of Igf2 and H19.
Unexpectedly, most cells maintained imprinted Igf2 and H19 expression (Fig. 3.9 B).
When biallelic Igf2 expression was observed, it was always in the Ctcf siRNA treated
cells. This is consistent with the insulator function of CTCF at the H19 ICR. However, such maternal expression of Igf2 upon CTCF depletion was very low (Fig. 3.9 B, arrow head). Because Igf2 levels were increased about three-fold compared with controls, I concluded that the enrichment was not from the normally silent maternal allele but rather from the normally active paternal allele. The low maternal expression of Igf2 could be a result of activation on both alleles. The residual amount of CTCF is likely sufficient to maintain imprinting. CTCF and Rad21 may function as repressors of Igf2.
∆ 3.2.1.2. Depletion of CTCF in H19 R/+ MEFs lead to increased expression of Igf2
To test whether the above phenotype (expression changes of Igf2 and H19 ) depends
on the CTCF binding sites in the H19 ICR, I conducted siRNA treatments in MEFs that
∆ have maternal transmission of the H19 R mutation (Engel et al., 2006). In these MEFs,
the insulator function at the maternal H19 ICR is disrupted due to deletion of the four
CTCF binding sites at the endogenous locus (see 1.1.3.1 and Fig. 1.2), leading to biallelic
Igf2 expression (Fig. 3.10 B, Ctrl siRNA-treated sample) (Engel et al., 2006). If the phenotype in the wild-type cells (Fig. 3.9 A, B) depended on the CTCF binding sites in
∆ the H19 ICR, I would expect to see no changes in H19 and Igf2 expression in the H19 R/+
98
∆ Figure 3.10 CTCF depletion in H19 R/+ MEFs associate with elevated expression of Igf2
∆ (A) Real-time PCRs showing H19 and Igf2 relative RNA levels in siRNA-treated H19 R/+ MEFs.
∆ CTCF-depleted H19 R/+ MEFs have elevated expression of Igf2. Ctrl: control; C: Ctcf; S: Smc1; R:
Rad21 siRNAs. *: p<0.05. p value shows differences between CTCF siRNA tr eatments and the
Ctrl siRNA treatments. H19 and Igf2 RNA levels are normalized to Rplp0 . Relative value of the control siRNA treatment is set to 1. Error bars indicate standard deviation. (B) Allele-specific
∆ analysis of Igf2 expression in H19 R/+ MEFs. B: maternal allele; C: paternal allele. Samples were collected at 96 h after two sequential treatments of siRNAs. (B) CTCF- and cohesins-depleted
∆ H19 R/+ MEFs have similar amounts of two parental alleles of Igf2 , similar to the Ctrl siRNA treated
MEFs.
99 cells. Contrary to this expectation, depleting CTCF resulted in increased Igf2 expression
∆ in CTCF knock-down H19 R/+ MEFs (Fig. 3.10 A). Furthermore, allele-specific analysis
showed that Igf2 was expressed equally from both parental alleles in both control siRNA- and target siRNA-treated MEFs. Thus, the elevated expression of Igf2 was from both
∆ alleles in the H19 R/+ MEFs. I concluded that CTCF and cohesins repress Igf2 expression, independent of the CTCF binding sites in the H19 ICR.
One explanation for these phenotypes is that CTCF may function as a transcriptional repressor of Igf2 , independent of the insulator function at the H19 ICR. Such repressor function may be more sensitive to the levels of CTCF protein than the insulator function at the H19 ICR, and was affected by our siRNA experiments, whereas occupancy of
CTCF at the H19 ICR was not affected. Igf2 has four alternative promoters, P0, 1, 2 and
3, which are suggested to exhibit tissue specificity (Engström et al., 1998). There are three DMRs around the Igf2 gene, DMR 1 and 2, as well as DMR0, which may be placenta-specific (Sasaki et al., 1992; Feil et al., 1994; Moore et al., 1997) (Fig. 3.11 A).
Other than the CTCF binding sites in the H19 ICR, ChIP assays in various human and mouse tissues have revealed multiple CTCF binding sites at the Igf2 locus (Du et al.,
2003; Kurukuti et al., 2006; Yoon et al., 2007; Chen et al., 2008; Li et al., 2008) and
CTCF appears preferentially bound to the maternal allele of Igf2 DMR1 and the P2 and
P3 promoters (Kurukuti et al., 2006; Li et al., 2008). These observations suggest that
CTCF is involved in the maternal-specific repression of Igf2. Therefore, the non-allelic repressive effect I observed at Igf2 is not likely dependent on these CTCF binding sites.
Nonetheless, CTCF binding at the Igf2 DMR0 is on both alleles (Li et al., 2008), whereas
100
Figure 3.11 CTCF and cohesins colocalize at a CTCF binding site 5.2 kb upstream of Igf2
DMR0
(A) Schematic showing the H19/Igf2 locus. The transcriptional and methylation status of the maternal and paternal chromosomes are represented on the top and bottom of the DNA strand, respectively. Three Igf2 DMRs are shown. IUCBS1 & 2 are located 1.7 and 5.2 kb upstream of
DMR0, respectively, and are indicated by the arrows below t he diagram. (B) ChIPs followed by real-time PCRs show that CTCF and cohesins bind to the IUCBS2 and CTCF also binds to
IUCBS1. The H19 ICR is included as a positive control. *: p<0.05; **: p<0.01. p value shows differences between No-Ab control and antibodies against CTCF, SMCI and RAD21. Error bars indicate standard deviation of at least three biological replicates.
101 other sites (1.7 kb and 5.2kb upstream of the Igf2 DMR0) have not been tested allelically
(Yoon et al., 2007; Chen et al., 2008). It is possible that CTCF binds to both parental
alleles at these sites and functions as a transcriptional repressor. In support of this idea,
SMC1 binds to two of these CTCF binding sites (1.7 kb and 5.2 kb upstream of the Igf2
DMR0) in MEFs (GEO accession: GSM560355) (Kagey et al., 2010). In addition, CTCF and SMC3 colocalize at the upstream of human Igf2 DMR0 on both parental alleles
(Nativio et al., 2009).
To test if CTCF and cohesins bind to the two CTCF binding sites upstream of Igf2
DMR0, I designed primers specific for the -1.7 kb and -5.2 kb regions, termed IUCBS1
and 2 for Igf2 upstream CTCF binding site 1 and 2, respectively. Real-time PCR following ChIP assays showed that CTCF and cohesins colocalize at the IUCBS2 site
(Fig. 3.11 B). Moreover, CTCF also binds to IUCBS1 (p<0.05), and preliminary data showed that the two cohesin subunits precipitated more chromatin than the No-Ab control (Figure 3.11 B), suggesting cohesins colocalize with CTCF at this site. However, no polymorphisms are available at these sites in our F1 hybrid cells, precluding identification what allele is aoosicated with CTCF and cohesins. Further characterization of CTCF and cohesins binding at Igf2 will help to clarify how they repress its transcription.
3.2.2. Depletion of CTCF and cohesins in MEFs lead to non-allelic changes of expression levels at the Dlk1-Dio3 locus, but not at the Kcnq1/Kcnq1ot1 locus
CTCF and cohesins colocalize to two additional imprinted loci, the Dlk1-Dio3 and
102 the Kcnq1/Kcnq1ot1 loci (see 3.1.2 and Fig. 3.5, 3.6, 3.7). To investigate effects of CTCF and cohesins depletion at these loci, I examined five imprinted genes in the Dlk1-Dio3 and the Kcnq1/Kcnq1ot1 loci in the CTCF knock-down and cohesins knock-down MEFs.
Imprinted expression of Dlk1, Gtl2, Kcnq1, Kcnq1ot1 and Cdkn1c did not seem to be affected by siRNA treatments (data not shown). Although the expression levels of Dlk1 and Gtl2 were not affected in most siRNA treatments, when a difference was observed, it was always an increase (Fig. 3.12 A). No significant difference in Kcnq1ot1 expression level was observed (Fig. 3.12 B). My data suggest that CTCF and cohesins function as transcriptional repressors at the Dlk1-Dio3 locus but not the Kcnq1/Kcnq1ot1 locus, which is similar to Igf2 in CTCF- and cohesins-depleted MEFs (Fig. 3.9). Importantly, the repressor roles of these proteins do not affect allele-specific expression of imprinted genes analyzed.
3.2.3. Using TS cell as a new model cell line to study roles of CTCF and cohesins in genomic imprinting
The aforementioned results have uncovered unappreciated roles of CTCF and cohesins in non-allelic regulation of imprinted genes. However, we were not able to show disruption of the insulator function at the H19/Igf2 locus by CTCF siRNA treatments in
MEFs. This suggests that the H19 ICR may require low levels of CTCF protein to maintain imprinting in MEFs. Although MEFs exhibit monoallelic expression of imprinted genes from the three loci that contain CTCF binding sites, they do not express
H19 and Igf2 at high levels (see below). I hypothesized that imprinting regulation in a
103
Figure 3.12 CTCF and cohesins depletion lead to changes in Gtl2 and Dlk1 RNA levels, but not in Kcnq1ot1 RNA levels
(A) Gtl2 and Dlk1 relative RNA levels in siRNA treated MEFs. *: p<0.05, **: p<0.01. p values re flect the differences between CTCF or cohesin siRNA treatments and the control siRNA treatment. (B) Kcnq1ot relative RNA levels in siRNA treated MEFs. No significant change is seen between CTCF or cohesins siRNA treatments and the control siRNA treatment. For both panels, the levels of imprinted gene RNA are normalized to Rplp0 . Relative value of the control siRNA treatment is set to 1. Error bars indicate standard deviation.
104 cell type that expresses high levels of H19 and Igf2 may be more sensitive to protein
levels of CTCF and cohesins, compared with MEFs. Likewise, if cohesins participate in
the insulator function at the H19 ICR, depletion of cohesins in such a cell type will lead
to biallelic expression of Igf2 and reduction in H19 expression. In addition, to show that
the binding profiles of CTCF and cohesins differ from that of MEFs, ChIP assays may
also be performed in this cell type. Because H19 is only expressed in the trophoectoderm
layer in mouse blastocysts (Poirier et al., 1991), and Igf2 is expressed later at high levels in extraembryonic tissues (Lee et al., 1990), I decided to use TS cells as another model cell line to study functions of CTCF and cohesins. Mouse TS cells are derived from the trophoectoderm layer of E3.5 mouse blastocysts and thus represent the extraembryonic lineage (Tanaka et al., 1998). I did not choose ES cells, which have been well-established in our lab as a model cell line, for two reasons: 1) ES cells are derived from the ICM
(Evans and Kaufman, 1981; Martin, 1981), which does not express H19 (Poirier et al.,
1991). 2) Our F1 hybrid ES cell lines exhibit biallelic expression of H19 (Thorvaldsen
and Lin, unpublished), probably because only low levels of H19 are expressed and basal
levels of expression may come from both alleles. Alternatively, the conditions for ES cell
culture may lead to biallelic expression of H19 .
3.2.3.1 Derivation of F1 hybrid TS cells
In collaboration with Dr. Winifred Mak, I derived six F1 hybrid TS cell lines from
crosses between 129/S1 females and B6(CAST7) or B6(CAST7P12X) males (see
Chapter 5). We used the 129/S1 strain rather than the C57BL/6 (B6) strain because
105 deriving TS cells from B6 mice has been shown to be inefficient (Tanaka, 2006; Himeno
et al., 2008). In fact, we were not able to establish TS cells from B6 females after
multiple attempts (Mak and Lin, unpublished). Three TS cell lines from crosses between
a 129/S1 mother and a B6(CAST7P12X) father, designated P12X-f, P12X-a and P12X-c,
were tested for expression of TS marker genes, including Estrogen related receptor beta
(Esrrb ), Eomesodermin homolog (Eomes ), Caudal type homeobox 2 (Cdx2 ) and
Fibroblast growth factor receptor 2 (Fgfr2 ) (Tanaka et al., 1998). All three lines, as well as a positive control TS cell line, A2 (Mak et al., 2002), expressed these marker genes
(Fig. 3.13 A), indicating that the derivations were successful.
I tested expression levels of H19 and Igf2 in these F1 hybrid TS cells. Fig. 3.13 B shows the relative levels of H19 and Igf2 RNAs in MEFs, ES cells and two lines of F1
hybrid TS cells (P12X-f and P12X-a). TS cells express both genes at a much higher level
(~6 to 7 fold) compared with the levels in MEFs and ES cells. In addition, ChIP assays
using antibodies against SMC1 and RAD21 showed enrichment of both cohesin subunits
at the H19 ICR, Gtl2 DMR and KvDMR1 CTCF BS1 (Fig. 3.13 C). Therefore, cohesins likely play a role at these imprinted loci in TS cells. CTCF is not included in these experiments because a new batch of CTCF antibody did not work for ChIP experiments.
I was able to recover one F1 hybrid TS cell line (P12X-c) that expresses monoallelic
H19 , Snrpn , Peg3 , Kcnq1ot1 and Cdkn1c (data not shown). However, all TS cell lines tested exhibited biallelic expression of Igf2 (data not shown), probably because Igf2 is either not expressed at the blastocyst stage (Lee et al., 1990) or is expressed at very low levels biallelically (Latham et al., 1994; Thorvaldsen et al., 2006). The imprinted
106
Figure 3.13 Characterization of F1 TS cell lines
(A) Three lines of F1 hybrid TS cells, P12X-f, P12X-a and P12X-c express TS cell marker genes.
An established TS cell line, A2 (Mak et al., 2002), is included as a positive control. (B) H19 and
Igf2 relative RNA levels in MEF, ES and TS cells. Shown are average expression levels from two
TS cell lines (P12X-f and P12X-a). The levels of imprinted gene RNA are normalized to Rplp0 .
Relative values of the expression levels in MEFs are set to 1. Error bars indicate standard deviation. (C) ChIPs followed by real-time PCRs using primers specific for the H19 ICR, Gt l2
DMR and KvDMR1 CTCF BS1 in F1 hybrid TS cells. SMC1 and RAD21 may bind to these regions in TS cells. Error bars show standard deviation of two biological replicates.
107 expression is not fully established until a later developmental stage (after E6.5)
(Thorvaldsen et al., 2006). Despite the biallelic expression of Igf2 , the TS cell lines can
still be useful because they express a number of other imprinted genes (e.g., Kcnq1ot1
and Cdkn1c ) and can be differentiated to multiple extraembryonic lineages (Tanaka et al.,
1998) and thus are extremely useful in studying placenta-specific imprinted genes.
Particular loci of interest include the Kcnq1/Kcnq1ot1 locus that contains a number of placenta-specific imprinted genes in mouse (e.g., Osbpl5 , Nap1l4 , Tssc4 and Ascl2 )
(Tanaka et al., 1999; Paulsen et al., 2000; Engemann et al., 2000), as well as two CTCF binding sites that also recruit cohesins in MEFs (see 3.1.2). Therefore, TS cells can be a useful model cell line for studying the functional requirements of CTCF and cohesins at the Kcnq1/Kcnq1ot1 locus.
3.2.3.1 Depletion of CTCF and cohesins in TS cells requires lentiviruses-based
RNAi
To study the roles of CTCF and cohesins in TS cells, I first conducted siRNA treatments targeting CTCF, SMC1 and RAD21 in TS cells using lipofectamine
(lipofectamine 2000, Invitrogen, see Chapter 5) that was also used to introduce siRNAs into MEFs. However, very limited knock-down of target proteins, if any, was achieved by these siRNA treatments in TS cells (Fig. 3.14 A). Furthermore, TS cells differentiated upon lipofectamine treatment and it was very difficult to recover enough cells for analysis.
Differentiated TS cells usually have multiple lineages, including tetraploid giant cells
(Tanaka et al., 1998). Because these cell types may exhibit different patterns of imprinted
108
Figure 3.14 RNAi in TS cells
(A) Western blot showing that no CTCF and RAD21 knock-down was achieved by siRNA treatments of TS cells. TS cells are treated with siRNAs (si) for 54 h and harvested for analysis.
(B) Western blot showing that RAD21 is reduced in lentiviruses-infected TS cells, compared with non affected (No-V) cells. Two lines of lentiviruses carrying shRNA sequences targeting RAD21,
Rad21-9 and Rad21-12 (see Chapter 5 ), are used. Another lentivirus line carrying a scrambled shRNA (see Chapter 5) is included as a negative control. TS cells are infected with lent iviruses and drug-selected for about two weeks. For both (A) and (B), anti-TUBULIN is used as a control.
109 gene expression, we would like to study functions of CTCF and cohesins in a homogeneous cell type. Therefore, another way to deliver siRNA or shRNAs into TS cells that does not lead to differentiation is required.
Lentivirus-based RNAi technologies have been shown to be effective in primary cell lines, including ES cells. I had conducted lentivirus RNAi experiments using commercially available lentiviruses targeting MBD3 in ES cells (Lin, Reese and
Bartolomei, unpublished). I was able to deplete MBD3 protein levels in ES cells by
~75% (Lin, unpublished). However, because wild-type ES cells exhibit biallelic H19 expression (Thorvaldsen and Lin, unpublished), we didn’t use this system to study
MBD3’s functions. To investigate the roles of CTCF and cohesins at genomic imprinting in TS cells, I applied a similar lentivirus approach to knock down these proteins in TS cells. I cloned siRNA sequences that were used in siRNA treatments of MEFs into a lentivirus vector (see Chapter 5) and infected TS cells with lentiviruses to knock-down
CTCF and cohesins. Preliminary data suggest that the lentiviruses targeting RAD21, were able to infect TS cells and knock-down target proteins (Fig. 3.14 B). Further experiments will be conducted to test the expression of TS marker genes to be assured that virus-infected TS cells have not differentiated, and to confirm the knock-down of CTCF and SMC1. Subsequently, expression of imprinted genes will be examined in CTCF- and cohesins-depleted TS cells.
In summary, these data presented in Chapter 3 have shown colocalization of CTCF and cohesins at three imprinted loci, H19/Igf2 , Dlk1-Dio3 and Kcnq1/Kcnq1ot1 . CTCF
110 and cohesins preferentially bind to the H19 ICR and Gtl2 DMR on the unmethylated maternal allele. Functional analysis of CTCF and cohesins has been performed in MEFs by siRNA treatment, and monoallelic expression of imprinted genes at these three loci was maintained. The mRNA levels for these genes were typically increased; for H19 and
Igf2 the increased expression was independent of the CTCF binding sites in the H19 ICR.
Depletion of CTCF and RAD21 by siRNA treatments increased the levels of Igf2 and
Gtl2 expression (Fig. 3.9 and 3.12). However, depletion of SMC1 did not show such effect. Given that no evidence is available to indicate independent function of RAD21 from the cohesin complex, this result is probably due to the relatively lower efficiency of
SMC1 knock-down (Fig. 3.8 B).
In conclusion, my results have provided a systematic view of the colocalization of
CTCF and cohesins at imprinted gene loci. Distinct binding profiles of CTCF and cohesins at different imprinted DMRs indicate locus-specific functions. Furthermore, by performing allele-specific analysis, we have uncovered new roles for CTCF and cohesins in the transcriptional regulation of the active allele of imprinted genes. In addition, I have derived F1 hybrid TS cell lines that can be used as a model cell line for imprinted gene regulation, as well as extraembryonic development.
111 Chapter 4. Discussion and future directions
112
In my dissertation, I have investigated roles of protein factors in genomic imprinting.
I have established that NuRD complex components MBD3 and MTA2 are required at the
H19/Igf2 imprinted locus during preimplantation development (Fig. 4.1). MTA2, but not
MBD3, is also required at the Peg3 locus. I have also shown association of two more methylated DNA binding proteins, MBD1 and Kaiso, with various imprinted loci in ES cells and MEFs. Taken together, these methylated DNA binding proteins and their cofactors may participate in the regulation of genomic imprinting with various degrees of locus- and tissue-specificity, and they may either function alone or in concert with each other.
The second major focus of my thesis was to investigate more systematically the colocalization and function of CTCF and the cohesin complex at imprinted loci. I showed that CTCF and cohesins were bound at three imprinted loci. Depletion of CTCF and cohesins in MEFs led to changes in the total expression levels of several imprinted genes, but monoallelic expression was maintained. Therefore, my thesis work has uncovered a non-allelic role of CTCF and cohesins in the regulation of imprinted loci. Whereas numerous studies have shown the importance of allele-specific regulation of imprinted genes, my work suggests the non-allelic control of expression levels could be as well critical for imprinted gene functions.
113
Figure 4.1 Protein factors are required for regulation of imprinted gene expression
Schematic showing the H19/Igf2 locus. On the maternal allele, the ICR recruits CTCF and forms an insulator that blocks maternal Igf2 expression, whereas H19 is active. Cohesins are found at the ICR, but may not be required for the insulator function (shown by the question mark). On the paternal allele, the ICR is methylated, preventing insulator activity. The methylation spreads to the
H19 promoter, therefore H19 is silenced. Igf2 is activated on the paternal allele. MBD3 and MTA2 are required for maintaining the methylation at the paternal ICR, as well as repression of the paternal H19 . CTCF and cohesins bind to a region upstream of Igf2 , and may repress Igf2 expression on both alleles.
114 4.1. The NuRD complex and genomic imprinting in preimplantation embryos
The lethality of Mbd3 -null embryos (Hendrich et al., 2001) suggests that the NuRD complex has important functions in preimplantation embryos. However, due to the low amount of material available for biochemical experiments, there is limited information on the composition and functions of the NuRD complex in mouse preimplantation embryos.
By performing RNAi experiments targeting individual NuRD complex components, we found that MBD3 and MTA2 are required at imprinted loci, providing the first link between the NuRD complex and genomic imprinting, as well as a mechanism for maintenance of differential methylation at ICRs.
Furthermore, although the H19/Igf2 locus was affected when either MBD3 or MTA2 was depleted, the Peg3 locus was only affected in the MTA2-depleted embryos. These results suggested that MTA2-containing NuRD complexes did not totally overlap with the
MBD3-containing NuRD complexes. Because the imprinted expression of Peg3 was not affected in Mbd2 -null mice (Hendrich et al., 2001), MBD2 and MBD3 may have redundant functions, in contrast to the indispensable role of MTA2 for the Peg3 locus.
Further functional analysis of a double mutant that lacks both MBD2 and MBD3 is required to verify the involvement of MBD2 at the Peg3 locus.
It is worth noting that the nuclear levels of MTA1 were reduced in MTA2-depleted embryos (Ma et al., 2010), suggesting that MTA1-containing NuRD complexes may be the actual functioning complexes at the Peg3 locus. Examination of MTA1 levels in
MBD3-depleted embryos could provide evidence for this idea. However, because MTA1
115 proteins are largely reduced in Mbd3 -null ES cells (Kaji et al., 2006), it is likely that
MTA1 levels may also be reduced by depletion of MBD3 in blastocysts. If so, neither
MTA1 nor MBD3 is required for imprinted expression of Peg3.
4.2 Possible mechanisms for incomplete penetrance in MBD3- or MTA2-depleted embryos
The DNA methylation and allele-specific expression of imprinted genes vary among the MBD3 or MTA2 RNAi embryos tested (see 2.2, 2.3 and Fig. 2.5, 2.6, 2.7). Because the Zp3-dsMbd3 transgene is transcribed in growing oocytes (Fedoriw et al., 2004; Reese et al., 2007), the embryos from transgenic females likely receive similar amounts of the dsRNA from the maternal deposit. Both the dsRNA injection and transgenic RNAi exhibit similar variations in phenotype severity (see 2.2 and Fig. 2.5, 2.6), suggesting that the phenotypic variations among RNAi embryos is not due to technical differences in dsRNA injections. As briefly discussed in 2.2.2, one explanation is that knock-down of
MBD3 and MTA2 has inherent limitations of not being able to ablate the targeted protein completely. Because the Mbd3 -null mice are lethal, and the Mta2 -null females are infertile (Hendrich et al., 2001; Lu et al., 2008), complete depletion of MBD3 and MTA2 from preimplantation embryos is essentially technically impossible. However, it is possible to further knock-down MBD3 in preimplantation embryos, by crossing our
Zp3-dsMbd3 mice with the heterozygous mice carrying the Mbd3-null allele. Functions of MBD3 at other imprinted loci will be observed if such embryos exhibit phenotypes at other loci besides the H19/Igf2 .
116 In addition, another explanation for such incomplete penentrance is that the time-window of the RNAi treatment is not long enough. Earlier depletion of MBD3 and
MTA2 may be required to completely ablate the functions of these proteins at imprinted loci. The transgenic RNAi utilizes the Zp3 promoter that is expressed in growing oocytes
(Fedoriw et al., 2004; Reese et al., 2007), providing a longer time-window than the dsRNA injection at the 1-cell stage. The latter does not introduce the dsRNAs into embryos until after fertilization. Supporting the idea that longer exposure to dsRNAs is required for higher penetrance, more MBD3-depleted embryos generated by the transgenic RNAi than by dsRNA injection had biallelic expression of H19 (40% and 26%, respectively).
Alternatively, other protein factors could compensate for the loss of MBD3 and
MTA2 in preimplantation embryos. For example, mutations and deletions of Zfp57 are associated with hypomethylation of the Peg3 DMR in both human and mouse (Mackay et al., 2008; Li et al., 2008). For the H19 ICR, it is unknown whether factors other than
PGC7/STELLA, DNMT1 and its cofactors are required for maintaining its paternal methylation. My ChIP data suggest that MBD1 and Kaiso do not bind to the H19 ICR in
ES cells (Fig. 2.2). It is worth testing binding of MBD2 at the H19 ICR because
MBD2-containing and MBD3-containing NuRD complexes may have redundant functions. Moreover, possible redundancy between MBD2 and MBD3 can be examined by crossing our Zp3-dsMbd3 mice with Mbd2 -null mice.
117 4.3. Establishment versus maintenance roles of MBD3 and MTA2 at imprinted loci
In the dsMbd3- and dsMta2-injected embryos, the dsRNAs were introduced after
fertilization, when the maternal DNA methylation at imprinted loci has already been
established. Because the DNA methylation is lost at the paternal H19 ICR in these
embryos, MBD3 and MTA2 are required for the maintenance of differential DNA
methylation. However, because the Zp3-Mbd3 transgene is expressed in growing oocytes,
where the maternal DNA methylation imprints are established (Reik et al., 2001), it is
unclear whether the phenotypes observed in the Zp3-Mbd3 transgenic embryos were due
to MBD3 and MTA2’s functions in the maintenance or both the maintenance and
establishment of imprinting marks. Analysis of DNA methylation patterns in the female
germ-line of the Mta2 -null mice, which exhibit female infertility (Lu et al., 2008), as well as analysis of mice that have MBD3 conditionally deleted from the female germ-line, may provide information to distinguish between these two possibilities.
4.4. Various requirements for protein cofactors at different developmental stages and at distinct imprinted loci
An intriguing result from my ChIP survey of MBD proteins and Kaiso is that while
MBD3 binds to the paternal H19 ICR in ES cells but not in MEFs (see 2.1.1 and Fig. 2.1),
Kaiso binds to the same locus in MEFs but not ES cells (Fig. 2.3). In addition, Kaiso, as
well as the other two Kaiso family members, ZBTB4 and ZBTB38, bind to the H19 ICR
in a paternal allele-biased (or paternal-specific) manner in adult mouse brain (Filion et al.,
118 2006). Taken together, these results strongly suggest that imprinted loci utilize distinct proteins or protein complexes at different developmental stages: whereas MBD3 is required for the maintenance of DNA methylation in the H19 ICR and repression of the
H19 paternal allele in preimplantation embryos, Kaiso family proteins may be required
for repression of the H19 paternal allele after implantation, when protection from
genome-wide DNA demethylation is not required.
In addition, multiple proteins of MBD and Kaiso families may play redundant roles
for a given imprinted locus, at a particular developmental stage. My ChIP survey suggests
that MBD1 and Kaiso colocalize to the maternal Snrpn DMR and Peg3 DMR in ES cells,
as well as the paternal H19 ICR and IG-DMR in MEFs (Fig. 2.2 and 2.3). The three
aforementioned Kaiso family proteins all bind to the H19 ICR in adult mouse brain
(Filion et al., 2006). Moreover, MBD1, MBD3 and MeCP2 bind to the methylated allele
of the Zrsr1 DMR in adult mouse liver (Fournier et al., 2002). These observations
support the idea that MBD and Kaiso families of proteins have compensatory functions at
imprinted loci. Consistently, MBD1 and Kaiso are dispensable from embryonic
development (Zhao et al., 2003; Prokhortchouk et al., 2006). The possible redundancy
between these proteins can be examined by generating mice carrying two or more
deletion alleles for these proteins.
In contrast to the MBD and Kaiso families of proteins, which seem to have locus-
and stage-specific functions, two other factors, NP95 and ZFP57, are responsible for
maintenance of DNA methylation at multiple imprinted loci (Sharif et al., 2007; Mackay
et al., 2008; Li et al., 2008). NP95 functions as a cofactor of DNMT1 during mitosis,
119 where NP95 was suggested to facilitate the localization of DNMT1 to the newly
synthesized DNA strand (Sharif et al., 2007). ZFP57 is a KRAB zinc finger protein.
Interestingly, ZFP57 and NuRD complex components, including CHD3, CHD4 and
HDAC1, can interact with a co-repressor protein, Tripartite motif-containing 28 (TRIM28)
(Schultz et al., 2001), suggesting a functional correlation between ZFP57 and the NuRD
complex. Indeed, the Peg3 locus was affected in both the dsMta2-injected embryos (see
2.3, 2.7 and Table 1) and the ZFP57 mutants (Mackay et al., 2008; Li et al., 2008). In contrast, the H19 ICR was not affected in ZFP57 mutants (Mackay et al., 2008; Li et al.,
2008). Therefore, it is unlikely that ZFP57 functions in concert with the NuRD complex at the H19/Igf2 locus. The functional correlation between ZFP57 and the NuRD complex at other imprinted loci remains to be determined.
4.5 Additional functions of MBD3 and MTA2 in preimplantation embryos
In addition to the phenotypes described in Chapter 2, MBD3-depleted embryos developed slower and were smaller than the control embryos (Reese et al., 2007).
Reduced cell number was observed in our MBD3 RNAi embryos at the blastocyst stage
(Reese et al., 2007), as well as in Mbd3 -null E5.5 embryos (Kaji et al., 2007). The numbers of cells that undergo apoptosis in dsMbd3 embryos were not higher than in control embryos (Reese et al., 2007). Although MBD3 was suggested to repress the trophoblast lineage (Zhu et al., 2009), dsMbd3 RNAi blastocysts developed both the ICM and trophoectoderm layers and expressed marker genes for both lineages (Reese et al.,
2007). These data suggest that the reduced cell number in MBD3-depleted embryos is not
120 due to loss of lineages, but may be the consequence of mitotic defects or developmental
arrest. Indeed, there was a significant increase in numbers of metaphase chromosome
pairs in dsMbd3 RNAi embryos, suggesting a mitotic defect (Reese et al., 2007).
However, although the dsMta2-injected embryos exhibited a similar phenotype of having
increased metaphase chromosome pairs, they had normal numbers of cells and had
similar sizes compared with the dsGfp-injected embryos (Ma et al., 2010). Genome-wide
ChIP assays using antibodies against MBD3 and MTA2, as well as microarray analysis of
gene expression changes in dsMbd3 and dsMta2 RNAi embryos may be able to uncover
the underlying mechanism for these phenotypes through the indetification of new targets
for MBD3 and MTA2.
4.6 Genome-wide colocalization of CTCF and cohesins
I demonstrated that cohesins colocalize with CTCF at five imprinted regions
including the H19 ICR, IVR CBS, Igf2 IUCBS2, Gtl2 DMR, KvDMR1 (see Chapter 3 and Fig. 4.1). At the H19 ICR, such colocalization depends on the CTCF binding sites on the maternal allele (Fig. 3.1 and 3.2). The results presented in this dissertation, together with recent discoveries that cohesins colocalize with CTCF genome-wide in human and mouse cell lines (Parelho et al., 2008; Wendt et al., 2008; Rubio et al., 2008), indicate that the cohesin complex is a general cofactor of CTCF at imprinted loci, as well as at other
CTCF targets.
Thousands of CTCF and cohesins binding sites were recovered by ChIP-on-chip
121 experiments using tiling arrays, and 56 to 77% of CTCF binding sites are also cohesins
sites (Parelho et al., 2008; Wendt et al., 2008). Conversely, 65 to 89% of cohesins binding
sites are CTCF sites (Parelho et al., 2008; Wendt et al., 2008). Cohesins may be recruited
to these shared binding sites by CTCF, which has the zinc fingers that directly bind to
DNA. However, it remains unclear if cohesins binding on these sites requires the CTCF
protein, or vice versa. RNAi experiments targeting CTCF by siRNA or shRNA treatments
showed reduced association of cohesins at a subset of CTCF binding sites: 12 sites from
one study (Parelho et al., 2008) and 90 out of 167 sites from another study(Wendt et al.,
2008). Conversely, 12 other binding sites didn’t have changes in CTCF affinity when
RAD21 was depleted (Parelho et al., 2008). However, due to the limited number of sites
tested in this study (Parelho et al., 2008), it remains uncertain if CTCF requires cohesins
at other binding sites.
4.7. Cohesins’ dependence of the CTCF binding sites
I showed that cohesins bind to the maternal H19 ICR, and such binding was lost
when the CTCF binding sites were deleted from the maternal chromosome (Fig. 3.1 and
3.2). There are two possible explanations for this observation. First, cohesins binding to
the maternal H19 ICR may require the CTCF binding sites in the H19 ICR. In support of this idea, the consensus sequences of CTCF and cohesins binding sites are highly similar
(Wendt et al., 2008). In addition, two other independent studies also revealed a requirement for CTCF binding sites for cohesins binding at the KSHV latency control region and human myotonic dystrophy gene DM1 (Stedman et al., 2008; Rubio et al.,
122 2008).
∆ Second, cohesins may be sensitive to DNA methylation, because the H19 R/+ mutant
became methylated at the maternal H19 ICR in postimplantation tissues (Fig. 1.2) (Engel
et al., 2006). Consistent with this idea, cohesins are enriched on the unmethylated
maternal allele of the Gtl2 DMR (Fig. 3.5 B). Moreover, a subset of CTCF and cohesins
binding sites exhibited cell-type specificity, and these sites are hypermethylated in cell
types that are not bound by CTCF and cohesins (Parelho et al., 2008). Biochemical
experiments examining in vitro binding of cohesins to unmethylated and methylated
DNA may provide direct evidence showing cohesins’ sensitivity to DNA methylation.
4.8. Other cohesin cofactors in transcriptional regulation
Genome-wide localization of CTCF and cohesins strongly suggest functional correlation between these factors. Consistent with this idea, I observed similar changes in imprinted gene RNA levels between CTCF-depleted and cohesins-depleted MEFs (Fig.
3.9, 3.10 and 3.12). However, we cannot rule out the possibility that CTCF and cohesins function independently at imprinted loci. In addition to CTCF, other partners of cohesins in gene regulation have been described, including the Polycomb protein complex PRC2
(Schaaf et al., 2009), the Mediator complex (Kagey et al., 2010) and estrogen receptor α
(ER α) (Schmidt et al., 2010). The Mediator- and ER α-associated cohesins binding sites are largely depleted of CTCF-associated cohesins binding sites (Schmidt et al., 2010;
Kagey et al., 2010), suggesting that cohesins function through distinct cofactors.
123 4.9. Cohesins may not be involved in the insulator function at the H19 ICR
CTCF’s role at the H19 ICR as an insulator protein regulating genomic imprinting is
well-established (see 1.1.3.1 and 1.4.1). Surprisingly, I did not observe loss of insulation
at the H19 ICR in MEFs upon depletion of CTCF by RNAi (Fig. 3.9 B). This finding
could be attributed to the relatively short period (96 h to 144 h) of CTCF depletion, or to
our siRNA experiments not reducing CTCF to levels below the threshold required for
maintaining insulation at the H19 ICR in MEFs. Alternatively, MEFs with a stronger
knock-down effect may not survive and therefore were not represented in the expression
analysis. Another possibility is that CTCF may only be required for the establishment,
rather than the maintenance of Igf2 imprinting. Our previous work, however, argues
against this idea. Conditional deletion of the entire H19 ICR from mouse neonatal liver
led to loss of imprinting, suggesting the ICR is required to maintain Igf2 repression on the
maternal allele (Thorvaldsen et al., 2006). Nevertheless, the presence of CTCF prior to
the RNAi treatment could be sufficient for establishing a chromatin state that is able to
maintain imprinted gene expression in the context of MEFs.
Similar experiments involving depletion of CTCF and cohesin complex subunits by
siRNA or shRNA treatments have been reported by other groups. Wendt et al. showed
reduction of H19 RNA levels and increased Igf2 RNA levels upon depletion of both
CTCF and RAD21 in HeLa cells (Wendt et al., 2008). However, neither statistical analysis of the mRNA changes nor allele-specific expression analysis was performed in that study. Nativio et al. reported no changes in HB2 cells, a human breast epithelial cell line(Nativio et al., 2009). The disparity between these studies and our study could
124 originate from differences in cell types and species (mouse vs. human). A recent study
using MEFs (the same cell type I used) by Yao et al. showed reduced H19 and increased
Igf2 expression upon CTCF depletion by shRNA (Yao et al., 2010). However, the H19
reduction was not statistically significant and the authors did not show the allelic
expression profile of Igf2 , even though they used a F1 hybrid MEF line from the
Bartolomei lab. Therefore, it is possible that the phenotypes they described are consistent with our results and showing a non-allelic role of CTCF at H19. A better understanding of the role of CTCF and cohesins in insulator function at the H19 ICR relies on further analysis in other cell types such as TS cells (Fig. 4.1).
4.10. Cooperation of CTCF and cohesins in transcriptional regulation of imprinted genes
Although CTCF and cohesins may have independent functions at some genomic loci,
I observed colocalization of CTCF and cohesins in all the imprinted regions tested (the
H19 ICR, IVR CBS, Igf2 IUCBS2, Gtl2 DMR and KvDMR1) (see Chapter 3). Moreover, the similar effects seen upon siRNA treatments targeting CTCF and cohesins in MEFs suggest that CTCF and cohesins cooperate in regulating expression of imprinted genes.
Despite the disparate observations by siRNA or shRNA treatments targeting CTCF and cohesins in our study and others (Wendt et al., 2008; Nativio et al., 2009; Yao et al.,
2010), one phenotype is strikingly similar, namely elevated Igf2 RNA levels after CTCF and cohesins depletion, which is independent of the CTCF binding sites at the H19 ICR
125 (Fig. 3.10). The colocalization of CTCF and cohesins at the Igf2 IUCBS2 (Fig. 3.10) may explain this phenotype. As briefly discussed in 3.2.1, besides the insulator mechanism,
Igf2 expression may be repressed, non-allelically, by recruitment of CTCF and cohesins
to the IUCBS2 (Fig. 4.1). I propose that the H19 ICR has higher occupancy of CTCF and
cohesins through the four CTCF binding sites (within 2 kb), whereas the IUCBS2 and
possibly IUCBS1 (Fig. 3.11) (3.5 kb apart) may have lower occupancy of CTCF and
cohesins. By depleting CTCF and/or cohesin subunits, a significant reduction of CTCF
and cohesins binding may be observed at IUCBS 1&2 but not at the H19 ICR. Further
experiments assaying binding of CTCF and cohesins at the H19 ICR and IUCBS2 after
siRNA treatments are currently ongoing.
Although the role of CTCF at the H19 ICR is well established, its functions at the
Gtl2 DMR and KvDMR1 remain to be elucidated. The colocalization of CTCF and
cohesins at these DMRs suggests that the proteins function together. Although we did not
observe loss of imprinting in any of the genes tested following CTCF and cohesins
depletion in MEFs, mRNA levels of imprinted genes were increased or unaltered (Fig.
3.12), suggesting a role in the repression at these loci. Elucidation of the function at these
loci is complicated by the fact that CTCF and cohesins bind to thousands of genomic
sites that depletion of these proteins may affect many different genes. The mechanism(s)
by which CTCF and cohesins impact gene expression at imprinted loci other than the
H19/Igf2 locus is (are) unclear. Interestingly, depletion of both CTCF and cohesins
similarly affects the genes in the H19/Igf2 and Dlk1/Gtl2 imprinted clusters (Fig. 3.10
and 3.12), suggesting CTCF and cohesins have similar roles within the same cluster, or
126 even across different imprinted clusters (see below).
4.11. Possible mechanisms for the non-allelic functions of CTCF and cohesins at
imprinted loci
The aforementioned results suggest the following possible mechanisms: first, CTCF
and cohesins function in concert to regulate a transcriptional activator or repressor.
Removal of CTCF and cohesins would increase the activity of an activator (or diminish
activity of a repressor), which would lead to overexpression of imprinted genes in the
same cluster. Second, given that CTCF mediates inter- and intra-chromosomal
interactions [for review, see (Williams and Flavell, 2008)], and is enriched at nuclear
lamina-associated domains (Guelen et al., 2008), CTCF and cohesins could participate in
the high-order chromosome organization that is required to localize gene loci to certain
domains in the nucleus. Loss of CTCF and cohesins may therefore impair this chromatin
organization, leading to elevated expression of imprinted genes. Third, CTCF and
cohesins could function through the imprinted gene network (IGN). The IGN was first
described in a microarray analysis that suggested co-regulation of imprinted genes
(Varrault et al., 2006). Interestingly, H19 , Igf2 and Dlk1 expression levels are co-upregulated or down-regulated in Zac1 -overexpressed or depleted tissues, respectively
(Varrault et al., 2006). Additional evidence for the IGN came from a study showing coordinated down-regulation of 11 imprinted genes (including H19 , Igf2 , Dlk1 and Gtl2 ) independent of DNA methylation during postnatal growth deceleration in various mouse tissues (Lui et al., 2008). Furthermore, Cdkn1c , an imprinted gene in the Kcnq1/Kcnq1ot
127 cluster, was not included in my analysis but identified as a member of the IGN (Varrault
et al., 2006; Lui et al., 2008). It is of interest to examine whether depletion of CTCF and
cohesins would affect Cdkn1c expression levels. CTCF and cohesins may play a role in the IGN because of their binding patterns in various related genes. However, the mechanism underlying IGN gene cross-talk and co-regulation remains to be elucidated.
4.12. Tissue and cell type specificity of CTCF and cohesins functions
The complexity of CTCF and cohesins is further highlighted due to their cell-type
specific binding patterns. ChIP experiments targeting CTCF and cohesins in various cell
types revealed different binding profiles (Essien et al., 2009; Schaaf et al., 2009; Schmidt
et al., 2010; Kagey et al., 2010). As introduced in 1.4.3, analysis by Essien et al.
suggested that CTCF binding sites at the imprinted loci described here had relatively low
affinity (LowOc and MedOc binding sites) for CTCF. The LowOc and MedOc CTCF
binding sites tend to be cell-type specific and developmentally regulated (Essien et al.,
2009). Similarly, cohesins also have different binding profiles between mouse embryonic
stem cells and MEFs (Kagey et al., 2010), as well as among different human tissues
(Schmidt et al., 2010) and Drosophila cell lines (Schaaf et al., 2009). These observations
suggest that CTCF and cohesins have distinct functions in different cell types at a subset
of targets. To illustrate further CTCF and cohesins’ roles in genomic imprinting, more
cell types need to be investigated. By comparing the binding profiles of CTCF and
cohesins at imprinted loci in different cell types (e.g., TS cells versus MEFs), and
subsequently correlating these with the imprinted gene expression patterns, one would be
128 able to determine which binding sites are controlling imprinting, and which are controlling transcriptional levels. The F1 hybrid TS cell lines that we have derived can be a useful tool to study functions of CTCF, cohesins and other protein factors at imprinted loci, especially at extraembryonic-specific imprinted genes.
In conclusion, the data described in this dissertation reveal new and important insights into the roles of protein factors in genomic imprinting (Fig. 4.1). Our data provide the first link between the NuRD corepressor complex and genomic imprinting.
We also showed association of methylated DNA binding proteins MBD3, MBD1 and
Kaiso with imprinted loci, in ES cells and MEFs. These findings suggest a mechanism by which germ-line DMRs can survive the genome-wide reprogramming during preimplantation development: methylated DNA binding proteins and NuRD complexes are involved in the protection of DMRs from DNA demethylation, as well as allelic repression of methylated imprinted genes. Further studies of the cross-talk between these proteins and their functions in postimplantation embryos will provide substantial knowledge of imprinted gene control. I also showed systematic colocalization of cohesins with CTCF at imprinted loci. Depletion of these proteins in MEFs led to similar changes in total expression levels of imprinted genes (typically increased expression), whereas monoallelic expression was maintained. These results uncovered non-allelic repressive roles of CTCF and cohesins in genomic imprinting.
129 Chapter 5. Materials and methods
130 Mouse strains
ES cells, MEFs, embryos and neonatal liver used in this thesis are isolated from crosses between C57BL/6 (B6) (The Jackson Laboratory) females and males with Mus musculus castaneus Chr 7 in a B6 background [B6(CAST7)], or males with Mus musculus castaneus Chr7, distal 12 and X in a B6 background [B6(CAST7P12X)] as previously described (Mann et al., 2003; Reese et al., 2007). Throughout this dissertation, in allele-specific analysis, the B6 allele (B) is the maternal allele, and the CAST (C) allele is the paternal allele. TS cells are isolated from crosses between 129S1 (The Jackson
Laboratory) females and males with Mus musculus castaneus Chr 7 in a B6 background
[B6(CAST7)], or males with Mus musculus castaneus Chr 7, distal 12 and X in a B6 background [B6(CAST7P12X)]. All experiments were conducted with the approval of the
Institutional Animal Care and Use Committee at the University of Pennsylvania.
Culture of ES cells
ES cells are cultured in ES cell medium (Table 2) with supplement of 1000 U/ml LIF.
Inactivated MEFs were used as feeder cells. MEFs were inactivated in Mitomycin
C-containing medium (Table 2) for 2 h at 37 °C and trypsinized and frozen. For ES cell culture, plates/flasks were treated with 0.1% Gelatin (in PBS) for 15 min-2 h prior to cell plating. For a 6-cm plate, 10 6 MEFs were plated prior to ES cell plating.
131 Isolation and culture of F1 hybrid MEFs
F1 hybrid MEFs were isolated from individual mouse E12.5-14.5 embryos generated
∆ from crosses between B6 (for both wildtype and the H19 R/+ mutant) females and
B6(CAST7P12X) males as previously described (Verona et al., 2008). The liver was removed from embryos for genotyping. The remaining of each embryo was placed in a 15 ml conical tube and trimmed into pieces by a sanitize scissor. 2.5 ml 0.1% Trypsin-EDTA was added into each conical tube. After incubation at 37 °C for 30-45 min with occasionally agitation, the digested cells and tissues were transferred into a T75 flask with 12 ml MEF medium (Table 2). When the cells became confluent, MEFs were split
1:8 or frozen as Passage 2 (P2). For siRNA and ChIP experiments, P4~P6 MEFs were used.
Derivation and culture of F1 hybrid TS cells
TS cells were derived from E3.5 blastocysts as described before (Himeno et al., 2008).
Briefly, E3.5 blastocysts were flushed and washed a couple times with PBS. Each blastocyst was placed in a well on a 4-well plate that contained inactivated MEF feeder cells. 0.5 ml of medium (see below) was added to each well. The blastocysts were cultured for 2-5 days until they hatched and had outgrowth (~500 to 1000 µM in diameter). The outgrown embryos were then disaggregated by trypsin-EDTA treatment followed by vigorous pipeting. Disaggregated embryos were cultured for up to two weeks and expanded if colonies were seen. TS cells were cultured on inactivated MEFs (50% of what was used for ES cells, for a 6-cm plate or T25, 0.5 x 10 6 MEFs were used), in TS
132 cell medium (Table 2). For derivation, 70% conditional medium (CM) + 30% TS medium
+ 1.5 x FGF4 and heparin (FGH) were used. To make the CM, inactivated MEFs were
plated at a high density (For a 15-cm plate, 6 x 106 MEFs were plated) and cultured in TS medium without FGH for at least 72 h. The CM was harvested and centrifuged at 2,300 g for 20 min to remove cell debris, and the supernatant was collected and passed through a
0.2 µm filter and frozen. For culturing TS cells, the medium were changed every 48 h, and fresh FGH was added each time. To passage TS cells, cells were treated with 0.1%
Trypsin-EDTA at 37 °C for 3 min and 15 s and were split 1:5 to 1:20. To collect
MEF-free TS cells, cells were trypsinized and re-plated in 70% CM + FGH for 45 min.
The supernatant was transferred to a new plate for another 45 min. The new supernatant contained mostly TS cells and was used for protein and RNA analysis.
PCR programming
Except for the bisulfite mutagenesis assays, all PCRs were performed for 2 min at 95°C;
32-38 cycles of 15 s at 95°C, 15 s at annealing temperature (listed in tables), 20-30 s at
72°C; and a final 2 min extension at 72 °C. PCR conditions are listed in Table 3 (mouse genotyping), Table 4 (Bisulfite assays), Table 5 (Real-time PCRs), Table 6
(Allele-specific PCRs), and Table 7 (All other PCR primers).
Transgenic mice genotyping
The Zp3-dsMbd3 transgenic mouse Line 37, was generated previously (Reese et al.,
133 2007). Transgenic and non-transgenic littermates were genotyped by PCR assays for part
∆ of the transgenic containing the Egfp gene. The H19 R transgenic mouse was generated previously (Engel et al., 2006). The mice carrying Mbd1 -null, Mbd2 -null and Mbd3 -null
alleles were generated by other labs (Hendrich et al., 2001; Zhao et al., 2003). For all
transgenic mice lines, tail clippings, ear punches or other tissues from these mice were
digested 4 h to overnight at 55 °C in 60 µg proteinase K/100 µl Quick Tail Buffer (50mM
KCl, 1.5 mM MgCl 2, 10mM Tris pH 8.5, 0.01% gelatin, 4.5% Tween-20, 4.5% Igepal).
Then proteinase K was heat-inactivated for 10min at 95 °C. One µl of digested tissue was
amplified using primers and PCR conditions listed in Table 3.
Bisulfite mutagenesis
The bisulfite mutagenesis for dsMbd3 RNAi embryos was performed as described (Reese
et al., 2007). For dsMta2-injected and control embryos, the allele-specific DNA
methylation patterns were examined for DMRs. Genomic DNA was isolated from two to
twelve blastocysts using The QIAamp DNA Micro Kit (QIAGEN). The DNA was then
denatured and treated with bisulfite using the Imprint DNA Modification Kit according to
the manufacturer’s protocol (Sigma-Aldrich). Partial mutagenized DNA was amplified by
the Epitech Whole Bisulfitome Kit (QIAGEN) according to manufacturer’s instructions.
Half of a blastocyst equivalent of the mutagenized DNA or 2 µl of the amplified DNA were used for each PCR reaction (with 0.5 µM of each primer and 1.5 mM MgCl 2, Table
4) and the products were first analyzed by combined bisulfite and restriction analysis
(COBRA) assays (Table 4). The products from a successful PCR that contains both
134 unmethylated and methylated strands were subsequently cloned and sequenced as described previously (Reese et al., 2007; Market-Velker et al., 2010b). Six or more clones were sequenced for each PCR reaction. The sequencing data was analyzed using an online tool BDPC [http://biochem.jacobs-university.de/BDPC/index.php; (Rohde et al.,
2008)]. Strands from a PCR that contained an identical pattern of unconverted cytosines and that could not be distinguished from other strands by polymorphisms were only counted once. For all three DMRs analyzed, two rounds of nested PCRs were performed.
Primers, PCR conditions and polymorphisms between B6 and CAST are listed in Table 4.
Chromatin Immunoprecipitation
ChIP assays were carried out using the Chromatin Immunoprecipitation Assay Kit
(Upstate 17-295 and 17-610) according to the manufacturer’s instructions and as previously described (Verona et al., 2008). Approximately 1 X 10 6 MEFs were used in each IP that resulted in 40-100 µg of ChIPed DNA. Cross-linking was achieved by adding 1% Formaldehyde to cell medium or PBS buffer and incubated 15 min at room temperature. All PCR primers and conditions are listed in Table 5 and 6.
Antibodies
The following antibodies were used in chromatin immunoprecipitation (ChIP) and western blot (WB): anti-Histone H3 (Abcam ab1791, ChIP 5µg); anti-Acetylated Histone
H3 (Millipore 06-599, ChIP 5µg); anti-dimethyl-Histone H3 Lysine4 (Millipore 07-030,
ChIP 5ul); anti-MBD3 (Santa Cruz Sc-9402,ChIP 20µl); anti-MBD1 (Santa Cruz
135 SC-25261, ChIP 10 µl); anti-Kaiso (Millipore 05-659, ChIP 10 µl); anti-CTCF (Millipore
07-729, ChIP 5µl, WB 1:1000); anti-SMC1 (Bethyl A300-055A, ChIP 5µl; WB 1:1000);
anti-RAD21 (Bethyl A300-080A, ChIP 5µl; WB 1:500); anti-αTUBULIN (Sigma T6199,
WB 1:2000). All antibodies used and details are listed in Table 8.
RNA extraction and reverse transcription (RT)
Total RNA from ES cells, MEFs and TS cells was extracted using RNeasy Micro Kit
(Qiagen) following manufacturer’s instructions. The reverse transcription reaction, primed with random primers, was performed using Superscript II or III reverse transcriptase (Invitrogen) following the manufacturer’s instructions. For each sample,
-RT control was done without the Superscript II or III and confirmed negative by PCR of
Gapdh.
siRNA treatments
MEFs were cultured in 24-well plates and treated every 48 h for two to three sequential siRNA treatments. Cells were trypsinized and plated at 30-40% confluency. 60-80 picomolar (pM) siRNAs were mixed with 1 µl Lipofectamine 2000 (Invitrogen) and
100µl OPTI low serum medium (Invitrogen) and added to the medium (high glucose
DMEM+10% FBS) upon cell plating. The following siRNAs and amounts were used: control siRNA (Invitrogen 12935-300) (60 or 80 pM), CTCF (Invitrogen Stealth
Ctcf-MSS203343) (60 pM); SMC1 (Invitrogen Smc1a-MSS216158) (60 pM); RAD21
(Dharmacon RAD21 ON-TARGET plus siRNA J-058531-12) (40 pM). Additional
136 siRNAs were also tested: CTCF (Dharmacon CTCF ON-TARGET plus siRNA
J-044693-9, 10, 11 and 12) and RAD21 [Dharmacon RAD21 ON-TARGET plus siRNA
J-058531-9, 10 and 11; Santa Cruz Rad21 siRNA (m): sc-72050], among which, one siRNA targeting CTCF (Dharmacon 10, use 60 pM) and one siRNA targeting RAD21
(Dharmacon 9, use 80 pM) were effective; and other siRNAs were not effective.
Lentivirus-RNAi
The pLKO.1 vector was used for cloning of short hairpin RNAs (shRNA), as instructed by the RNAi consortium and Addgene online protocol (Addgene). shRNA sequences (21 to 25 bp) were either modified from the siRNAs (see above) or designed using online tools (GenScript; Invitrogen; siRNA Wizard; Yuan et al., 2004). A few nucleotides required for restriction enzyme digestion were added to the 5’ and 3’ ends of shRNA oligos. All forward strands had a 5’-CCGG- end and a –TTTTG-3’ end; all reverse strands had a 5’-AATTCAAAAA end. Each shRNA has a looping sequence (4 to 6 bp) based on the suggestion from online tools used (GenScript; Invitrogen; siRNA Wizard;
Yuan et al., 2004). Moreover, it is suggested that a shRNA hairpin sequence should starts with a nucleotide G and ends with a corresponding C, to enable efficient transcription initiation by RNA Polymerase III U6 (GenScript; Invitrogen). Therefore, an additional G was added to the 5’ end of the forward strand of shRNAs that does not start with a G, following the added sequences for restriction sites; and an additional C was add to the reverse strand of shRNAs that does not end with a C. All target sequences were subjected to NCBI BLAST and tested for off-target effects, only sequences have no more than 15
137 nucleotides that are identical to mouse ESTs other than the target mRNA were used in the following experiments. In addition, a scrambled shRNA that was previously designed and cloned into the pLKO.1 vector by another lab (Sarbassov et al., 2005) and was deposited to Addgene (Addgene plasmid 1864, hairpin sequence:
CCTAAGGTTAAGTCGCCCTCGCTCTAGCGAGGGCGACTTAACCTTAGG) was used as a control. A total number of 13 shRNAs were cloned and sequenced, including a shRNA scrambled control, a Ctcf-siRNA scrambled control, four shRNAs targeting
CTCF, four shRNAs targeting SMC1 and three shRNAs targeting RAD21. The two strands of shRNA oligos were annealed and cloned into the pLKO.1 vector. The plasmids were sequenced and correct clones were expanded. They were subsequently transfected into MEFs and examined for target protein levels by western blot. Only shRNAs that can knock-down target proteins were included in following experiments. All shRNA sequences are listed in Table 9. To produce lentiviruses, 293FT cells (Invitrogen) were transfected with pLKO.1 shRNA plasmid, psPAX2 packaging plasmid and pMD2.G envelop plasmid (Addgene) at a ratio of 4:3:1, using Lipofectamine 2000. For a 15-cm plate, 15 µg of pLKO.1, 11.25 µg of psPAX2 and 3.75 µg of pMD2.G were transfected.
DMEM (high-glucose) + 10% FBS was used as a plating medium for the transfection.
Cells were fed with 293FT medium (Table 2) 6 h to overnight after transfection, and were cultured for another 96 h. The medium containing the lentiviruses were collected every
48 h and combined, centrifuged at 800g for 5 min to remove cell debris, and subsequently passed through a 0.45 µm filter. The purified medium containing the lentiviruses particles were then subjected to ultracentrifugation for ~30,000 g for 2 h at 4 °C. For lentiviruses
138 produced by a 15-cm plate, ~500 µl RPMI 1640 medium were added to the pellets and let
sit at 4 °C overnight. The lentiviruses were subsequently resuspended, aliquoted and
frozen. To titrate the lentiviruses, 293FT cells were infected and treated with geneticin.
Numbers of viable cells after 48 h of drug selection were counted and the virus titer were
calculated. To infect TS cells, lentivirues were added to TS culture at a multiplicity of
infection (MOI) = 5 to 20. To assist in effective infection, the TS cells were centrifuged at
1,000 g for 1 h at room temperature or 37 °C (Golding et al., 2010), and incubated at
37 °C for at least 24 h before drug selection.
Real-time PCRs
For both ChIP and RNA expression assays, real-time PCRs were performed using the
LightCycler Real Time PCR System (Roche). Reactions were set up in duplicates or
triplicates using the Ready-To-Go PCR beads (Amersham). Each PCR bead was
dissolved ddH 2O with 0.38-0.5 µl of TaqStart antibody (Clontech) in a total volume of 15
µl, incubated for 5 min at room temperature. 10 µl was used for each PCR in a 20 µl
system, which contains 0.2 to 0.6 µM primers (Table 5), 1x EvaGreen (Biotium) and
1.25-5 mM MgCl 2 (Table 5). For ChIP assays and relative RNA expression assays, data analysis was performed using the Light Cycler 4.0 software, using the Relative
Quantification program (Monocolor) to determine the ratio of each ChIP sample relative to input or cDNA levels relative to house-keeping genes. For each pairs of PCR primers, a standard curve was generated prior to experiments and used to correct different efficiency between different PCRs (e.g., Igf2 versus Rplp0 ). For allele-specific PCRs
139 (H19 , and Snrpn ), data analysis was performed using Light Cycler 3.0 software. Relative signals from each allele were determined by melting-curve analysis and the relative peak areas for each allele were determined (Reese et al., 2007).
Allele-specific PCRs
For both ChIP and RNA expression assays, allele-specific PCRs were carried out with
Ready-to-go PCR beads (Amersham) or GoTaq master mixture (Promega) as previously described (Verona et al., 2008). For most allele-specific PCRs, 0.3 µM of each primer
(Table 6) and 0.1 µCi of [ 32 P] dCTP were included in each PCR. The PCR products were digested with restriction enzymes listed in Table 6 and resolved in 7% or 12% polyacrylamide gels. These gels were dried, exposed to phosphor screens and scanned on a Typhoon Trio Phosphorimager (GE). The relative band intensities were quantified using Image J (NIH). Band intensities were normalized by numbers of radioactive dCTPs in each digested fragment. H19 and Snrpn RNA assays were conducted on cDNA using the LightCycler Real Time PCR System (Roche) (see above) as described previously
(Reese et al., 2007). Peg3 and Dlk1 RNA assays were conducted with some modifications as previously described (Mann et al., 2004; Hagan et al., 2009). The gels were stained with ethidium bromide. The contribution of each parental allele to the total RNA was determined using the Quantity One software (Bio-Rad). All primers and PCR conditions are listed in Table 6.
140 Western blots
Cell pellets were lysed with TNE (100 mM Tris, pH 7.4; 1% NP-40; 10 mM EDTA) buffer with 1:100 Proteinase inhibitor cocktail (Sigma) and 1 mM DTT. Lysates were mixed with 5X loading buffer, denatured by heating at 95 °C for 10 min, and fractionated on an 8% SDS-PAGE gel. The protein was transferred to a nitrocellulose membrane
(BioRad), blocked in 3% non-fat milk (BioRad) in TBST (0.05% Tween-20 in 1X TBS) and probed with primary and secondary antibodies (HRP-conjugated anti-mouse and anti-rabbit, GE). The blot was visualized using chemiluminescence (ECL Plus; GE).
Quantification was performed using ImageJ (NIH).
Statistics
For real-time PCRs following ChIP experiments, one-tail paired T-test was used. For analysis of bisulfite treated DNA strands, Chi-square test was used. For RNA expression changes, two-tail paired T-test was used. For all tests, differences of p<0.05 were considered significant.
141 Table 2 Tissue culture medium
Media Recipe
15% Fetal bovine serum (FBS)
1% Penicillin (500U/ml) & Streptomycin (5 mg/ml) (Pen/Strep)
1% 200mM L-Glutamine (L-Gln)
1% Non-essential amino acids ES cell 0.1% Gentamicin
55µm β-Mercaptoethanol (BME)
Filled with DMEM (high-glucose) to 100%
Add fresh: 1000 U/ml LIF/ESGRO (Millipore Chemicon)
10% FBS
1% Pen/Strep
MEFs 1% L-Gln
0.1% Gentamicin
Filled with DMEM (high-glucose) to 100%
15%-20% FBS
1% Pen/Strep
1% L-Gln
1% 100mM sodium pyruvate TS cell 100 µM BME
Filled with RPMI 1640 (pH 7.2) to 100%
Add fresh: 0.1% 25 µg/ml FGF4 (Sigma F8424 or R&D 235-F4)
0.1% 1.5 mg/ml heparin
142 Media Recipe
10% FBS
1% Pen/Strep
293FT cells 1% L-Gln
1% 100mM sodium pyruvate
Filled with DMEM (high glucose) to 100%
5% FBS
Mitomycin C medium 10 µg/ml Mitomycin C
Filled with DMEM (high-glucose) to 100%
143 (Engel et (Engel al., 2006) (Reese et et (Reese al., 2007) et (Zhao al., 2003) Reference Reaction Reaction (µl) Primer (µl) enotyping C) ° Temp Temp ( Mutant Mutant (bp) 174 258 55 0.5 15 343 60 0.5 15 300 500 55 1 20 WT (bp) Table 3 Primers 3 conditions PCR and used mousefor g Table H19-2.3F(25): CAATGTTCATAAGGGTCATGGGGTG H19-2.0R(25): CGTAAAGTGTCACAAATGCCTGATCCC GFP-F2 (10): GCACAATCTTCTTCAAGGACGAC GFP-F2GCACAATCTTCTTCAAGGACGAC (10): TCTTTGCTCAGGGCGGACTG(10):GFP-R1 OXZ59(common) (10): TCTTCTCAGACTGAGAAGGGTGA OXZ60(WT) (10): CACTGAACATTGCCCAGAGCACA (10): OXZ61(mutant) AAACGGCGGATTGACCGTAATGG PrimerµM)(stock in concentration sequence
-null R ∆ Zp3-dsMbd3 H19 Mbd1 Transgenic Transgenic mouse
144 Reference (Hendrich (Hendrich al., et 2001) (Hendrich (Hendrich al., et 2001) Reaction Reaction (µl) Primer (µl) C) ° Temp Temp ( Mutant Mutant (bp) WT (bp) ~600 ~200 59 1 20 272 150 55 1 20
PrimerµM)(stock in concentration sequence MBD2.P61(common)(15): ACGCTGGCCTAGTGCCGTGC (WT)TTGTGGTTGTGCTCAGTTC MBD2.P62 (25): TCCGCAAACTTCTATTTCTG ENP1(mutant)(10): MBD3.P26(common)(15): TGTAGCCACCTAGCTCAAGG (WT) MBD3.P27 CACGCTGGCGACTCTTATTC (10): ENP1(mutant) (10): TCCGCAAACTTCTATTTCTG TCCGCAAACTTCTATTTCTG ENP1(mutant)(10): -null -null Transgenic Transgenic mouse Mbd2 Mbd3
145 SNPs: SNPs: position- (B/C) 1425 - T/C -1425 -1485 G/deleted -1566 G/A A/G -1654 2181 -2181 G/A -2241 C/T T/G -2251 Referenceand accession number (Tremblay et (Tremblay 1997),al., and Abramowitz, unpublished;U 19619 (1304-1726) (Lucifero et al.,(Lucifero et 2002), and Abramowitz, te mutagenesis COBRA enzyme COBRA and products (bp) HinfI: HinfI: unmethylated uncut: 423; methylated: ~210, ~200 HinfI: HinfI: unmethylated 451; uncut C C 4 ° C C C C ° ° C C 30s C 2 C min, ° ° C/s), 72 C/s),72 C2min; 35 ° ° 1min (ramping speed (ramping 0.5 30s, 55 30s, 94 94 2 cycles cycles 2 of 94 cycles of cycles 94 min, 55
Table 4 Primers 4 conditions PCR and used for bisulfi Table BMsp2t1: GAGTATTTAGGAGGTATAAGAATT GAGTATTTAGGAGGTATAAGAATT BMsp2t1: ATCAAAAACTAACATAAACCCCT BHha1t3: TATGTAATATGATATAGTTTAGAAATTAG SnrpnA: Bmsp2t2c: GTAAGGAGATTATGTTTATTTTTGG Bmsp2t2c:GTAAGGAGATTATGTTTATTTTTGG BHha1t4ct: CTAACCTCATAAAACCCATAACTAT SnrpnD: AATAAACCCAAATCTAAAATATTTTAATC DMR Primersfor nested PCRs conditions PCR H19 H19 ICR Snrpn DMR
146 SNP :3,683,191-G/ SNP A SNPs: SNPs: position- (B/C) 2259 - T/A -2259 -2268 G/A -2281 C/T -2292 C/T -2348 G/T (Market-Velker (Market-Velker 2010b);al., et NT_039413.7 (3,683,033-3,6 82,588) Referenceand accession number unpublished; AF081460 (2151-2570) unmethylated 446; uncut EcoRV: methylated114, 332; BsrBI: methylated200, 246 COBRA enzyme COBRA and products (bp) methylated262, 351 C ° C C C 1 1 C ° °
C C 30s, C 2 C min, ° ° C1min C2min; 35 Cmin 2 Cmin; 2 35 ° ° ° ° 72 72 30s, 50 30s, 94 94 Annealing53 cycles of cycles 94 72 72 min, 55 72 72 cycles of cycles 95
SnrpnB: AATTTGTGTGATGTTTGTAATTATTTGG SnrpnB: ATAAAATACACTTTCACTACTAAAATCC SnrpnC: TTTTGATAAGGAGGTGTTT Peg3A-BL: ACTCTAATATCCACTATAATAA Peg3D-BL: AGTGTGGGTGTATTAGATT Peg3B-BL: TAACAAAACTTCTACATCATC Peg3C-BL:
DMR Primersfor nested PCRs conditions PCR Peg3 DMR
147 Modified from (Verona et al., from Modified (Verona 2008) et al., from Modified (Verona 2008) et al., from Modified (Verona 2008) et al., from Modified (Verona 2008) et al., from Modified (Verona 2008) et al., from Modified (Verona 2008) Reference
2 MgCl (mM) me PCR quantifications Primer (µM) C) ° 60 0.6 4 60 0.3 4 62 0.3 60 1.25 0.4 64 4 0.3 64 3 0.6 4 Temp Temp ( Table 5 Primers 5 conditions PCR and used for real-ti Table Peg3F3714: GACGGTATCTAAGAGGGTGCAT GACGGTATCTAAGAGGGTGCAT Peg3F3714: Peg3R3838:GGTTCAGTGTGGGTGCACTA H19 4017F: CAGGACTCAAAGGAACATGCTAC 4017F: H19 CAGGACTCAAAGGAACATGCTAC 3594R: H19 GCAATCCGTTTTAGGACTGCG SnUPCHIPs-F: AATCTGTGTGATGCTTGCAATCACTTGG SnUPCHIPs-R: ATAGGATGCACTTTCACTACTAGAATCC TACGGAGATGTGCTGTGGAC IG-DMR207: IG-DMR442:CTCGCTAGTTCACGGAGGTC 5488F:CCCATAGTCCTTCCTGGGTA TGATGTGCCACCTGGATAGA 5357R: TGGGCAGTGAGTCTCCTTCTH19GB374: H19GB675R:GCCACTGTCTCCAAGGACTC DMR DMR promoter ICR Region/Gene Primers H19 (Rep1&2) Snrpn IG-DMR Peg3 DMD_UP H19
148 Modified from (Verona et al., from Modified (Verona 2008) Reference
2 MgCl (mM) Primer (µM) C) ° 60 0.5 60 2.5 0.4 58-64 4 0.3 3 56-64 al., et 2008) (Stedman 0.6 2.5 61-64 0.5 3.5 61 0.5 56 3 0.4 5 Temp Temp ( RT1: GCACTAAGTCGATTGCACTGG GCACTAAGTCGATTGCACTGG RT1: AACACTTTATGATGGAACTGC HE5: GGATCCTGAGTCGCAGTATAAAAGA cMyc-f: CCTCTGTCTCTCGCTGGAATTAC cMyc-r: 2456F: H19IVR GGCAATCCACACCTCCTCT 2664R: H19IVR GTGAGGTCAGCGGTTAGCAT AGGCACTGCAGCTGGATTAT 77F: H19IVR 354R: H19IVR CCCCTAAAGACCCACTCCAT TCAGTTGCAACAGGAACAGC6025F: H19IVR TCTAGCAATGACCCTAACCACA 6261R: H19IVR TGGTTGGGCTATTGGAGTCT 1998F: Gtl2 2158R: Gtl2 CAATGGGAGGGGTACAGATG ACCATGCAGAGAAAAGCACA KvDMR 2644F: KvDMR CTAGCCGTTGTCGCTAGGAG 2844R: Ex5 DMR Region/Gene Primers H19 region Myc CBS IVR Ctrl1 IVR Ctrl2 IVR Gtl2 KvDMR1 CTCFBS1
149 Modified from Modified et (Thorvaldsen 2006)al., from Modified Vigneau, unpublished Modified from Modified Vigneau, unpublished Reference
2 MgCl (mM) Primer (µM) C) ° 55 0.4 4.5 60 0.4 2 62 0.2 58 3 0.6 58 4 0.6 54 2 0.6 2.5 58 0.25 2.4 Temp Temp ( HE2: TGATGGAGAGGACAGAAGGG TGATGGAGAGGACAGAAGGG HE2: TTGATTCAGAACGAGACGGAC HE4: TTGCTGTTGTGCTCAGGTTC Meg3f: ATCCTGGGGTCCTCAGTCTT Meg3r: Igf2f:CGCTTCAGTTTGTCTGTTCG Igf2r:GCAGCACTCTTCCACGATG KvDMR 2258F: CTGAGAAGCCAAGTGGATCG KvDMR 2258F:CTGAGAAGCCAAGTGGATCG KvDMR CCACCAGCCTCAGCATATTT 2553R: KvDMR 1762F:CACTCACCTTGGGACTCGAC AGAAGCAGAGGTGATTCGTG KvDMR 2007R: IUCBS1-961F:GCTCGAGCTTGACATCTGGT TTCCATTCTCAAGGCTTGGT IUCBS1-1072R: ACTATGATAAGCCAGCCCTTC IUCBS2-960F: IUCBS2-1055R:GCCAGAAATGGTTAGTCATGG
Region/Gene Primers KvDMR1 CTCFBS2 KvDMR1Ctrl IUCBS1 IUCBS2 H19 Igf2 Gtl2/Meg3
150
(Mohammad et al., (Mohammadet Modified from Modified Vigneau, unpublished from Modified 2008) Reference
2 MgCl (mM) Primer (µM) C) ° 60 0.4 56 1.25 0.5 55 3.5 0.4 2008) al., et from Modified (Wan 55 2 0.3 55 2.5 0.6 55 2 0.4 4.5 Temp Temp ( Dlk1f: CGGGAAATTCTGCGAAATAG Dlk1f:CGGGAAATTCTGCGAAATAG TGTGCAGGAGCATTCGTACT Dlk1r: CTCF5F1:GCCAGCAGGGACACATACAAG CTCF5R1:GCTTTCGCAAGTGGACACC Smc1CAAGTACCCAGATGCCAACC 3767: Smc1CGATCCATGATAGGGGGTAA 3983: 2815: Rad21 CAAGGCTGCACACTCCTGTA 3045: Rad21 CCCCATAAAAGTGCCAACAC TGTTACCAACTGGGACGACA B-actin-F: B-actin-R:GGGGTGTTGAGGTCTCAAA TCCCACTTACTGAAAAGGTCAAG Arpp0#72L: TCCGACTCTTCCTTTGCTTCArpp0#72R:
Region/Gene Primers Dlk1 Ctcf Smc1 Rad21 Actb Rplp0
151 (Verona et 2008)al., (Verona et 2008)al., (Verona et 2008)al., (Verona et 2008)al., (Verona Reference B 107,29 B uncut C 136 216,89 B uncut C 305 216,89 B uncut C 305 104,95, B 37 199,37 C Products Products (bp) Enzyme canalysis by PCRs SNP SNP (B/C)
C) ° Temp ( Temp 64 G/A HphI 64 G/A HphI 62 A/G NcoI 62 A/T Tsp45I Table 6 Primers 6 enzymes forand used allele-specifi Table H19GB374F: TGGGCAGGTGAGTCTCCTTCTH19GB374F: H19GB675R:GCCACTGTCTCCAAGGACTC GCACTAAGTCGATTGCACTGG RT1: AACACTTTATGATGGAACTGC HE5: TACGGAGATGTGCTGTGGAC IG-DMR207: IG-DMR442:CTCGCTAGTTCACGGAGGTC Rep3F2: CAGTTGTGTTTCTGGAGGGRep3F2: TAGGAGTATGCTGCCACC Rep3R2: Prom Ex5 ICR Region/Gene Primers H19 H19 H19 H19 IG-DMR (Rep3)
152 (Verona et 2008)al., (Verona Reference
(Verona et 2008)al., (Verona
B 253,99 B uncut C 352 204,4 B 109.96, C 4 Products Products (bp) C 155,57 C B uncut 212 uncut B B 278,145 B uncut C 423 Enzyme A/G BseRI SNP SNP (B/C)
C) ° 60 T/G BmgBI add 63, TaqStart 61 C/T AvaII 60 C/T SmlI Temp ( Temp KvDMRfor:GCGGGTTTCTTCTCTGAGTC TGTCCTAGGCCACTCACCTT KvDMRrev: 2456F: H19IVR GGCAATCCACACCTCCTCT 2644R: H19IVR GTGAGGTCAGCGGTTAGCAT Peg3F3627: GCCTTGTCAGTTACCCTTGG GCCTTGTCAGTTACCCTTGG Peg3F3627: Peg3R3838:GGTTCAGTGTGGGTGCACTA SnUPCHIPs-F: AATCTGTGTGATGCTTGCAATCACTTGG SnUPCHIPs-R: ATAGGATGCACTTTCACTACTAGAATCC DMR DMR Region/Gene Primers Snrpn KvDMR1 Peg3 CBS IVR
153 (Thorvaldsen et 2002, al., 2006)
Reference
B 109,70, B 17 48, 97, C 70, 17, 12 48, 398 uncut B Products Products (bp) C 292,106 C 214,82 B uncut C 296
Enzyme SNP SNP (B/C)
C) ° 61 T/C HaeII 62 C/T Hpy188I 62 A/G MfeI/Tsp509I 55 G/A real-time N/A, PCR Temp ( Temp Gtl2 1915F: TGAAACCTGTTGGGCG1915F: Gtl2 2158R: Gtl2 CAATGGGAGGGGTACAGATG KvDMR 2701F:CCCACCGAAGTAATCCAAAA TCAGCTAGGAAGGGATGAGG KvDMR 3098R: KvDMR 2258F:CTGAGAAGCCAAGTGGATCG KvDMR CCACCAGCCTCAGCATATTT 2553R: CCTCAAGATGAAAGAAATGGT HI3: AACACTTTATGATGGAACTGC HE5: -FL Probes: Mut CCACCTGTCGTCCATCTCC TCTGAGGGCAACTGGGTGTGGRED640- Anc
DMR Region/Gene Primers Gtl2 KvDMR1 CTCF BS1 KvDMR1 CTCF BS2 H19
154
(Thorvaldsen et 2006) al., al., (Mann et 2003) Reference B 180,20 B 165,20, C 15 250,87 B uncut C 337 Products Products (bp) B 173,62 B uncut C 235 Enzyme SNP SNP (B/C)
C) ° 58 ? Tsp509I 65 C/T real-time N/A, PCR al., (Mann et 2003) 55 T/? SfcI 55 G/A Cac8I Temp ( Temp Igf18: ATCTGTGACCTCTTGAGCAGG ATCTGTGACCTCTTGAGCAGG Igf18: GGGTTGTTTAGAGCCAATCAA Igf20: CTCCACCAGGAATTAGAGGC Sn1: TATAGTTAATGCAGTAAGAGG3 Sn3: GAAGCATTGTAGGGGAAGAGAA Probes: Mut -FL RED640- Anc GGCTGAGATTTATCAACTGTATCTTAGGGTC CCAAAGCCATCATCTGGAATC Gtl3: CAGCCCTGTGAGGTAGGAAC Gtl4: HE2: TGATGGAGAGGACAGAAGGG TGATGGAGAGGACAGAAGGG HE2: TTGATTCAGAACGAGACGGAC HE4:
Region/Gene Primers Igf2 Snrpn Gtl2
155 Modified from Modified et (Hagan 2009)al., and Jimenez, unpublished Reference Mann et et Mann al. unpublished
B 212,104 B uncut C 316 315,232, B 47 315,279 C Products Products (bp) Enzyme SNP SNP (B/C)
C) ° 54 T/C DraIII 60 ? HpyCh4III Temp ( Temp Dlk2up: CTGGCTTTCTTCCCGCTGGACDlk2up: GACACAGCCAGGGGCAGTTA Dlk317dn: FOR: GCCGCTTCATCTGTCTCTGTAG ATP7A GCAGCACATTAGCAACTTCTAAC REV: ATP7A
Region/Gene Primers Dlk1 Atp7a
156
Reference
C) ° 55 2005) al., et from Modified (Strumpf 55 55 unpublished Vigneau, 55 2005)al., et (Strumpf Temp ( Temp 55 al., (Mann et 2003)
Table 7 Other 7 PCR conditionsprimers and Table Esrrb-c286f: ACTCTGCATCCCGGACCCCC ACTCTGCATCCCGGACCCCC Esrrb-c286f: Esrrb-c473r:GCGTGGGTGCTCAGGGCAAT mEomes-F:GTGACAGAGGACGGTGTGGAGG Cdx2f:CTGCGGTTCTGAAACCAAAT CACCATCAGGAGGAAAAGTGA Cdx2r: Fgfr2-F:GACAAGCCCACCAACTGCACC Fgfr2-R:CGTCCCCTGAAGAACAAGAGC Primers AGAGGAGGCCGTTGGTCTGTGG mEomes-R: Gapdh F1: ATCACTGCCACCCAGAACAC ATCACTGCCACCCAGAACAC F1: Gapdh ATCCACGACGGACACATTGG Gapdh R1:
Gapdh Essrb Eomes Cdx2 Fgfr2 Gene and reference Gene
157 ChIP 000 :5000 t µg 5 (5µl) it µl 5-10 µl 5 bit µl 5 Rabbit µg 5 (5µl) Rabbit µl 5
00-1:10000) Table 8 Antibodies 8 Table Primary Ab Primary Ab Secondary Western dilutions Western 50 kD 50 µg/ml 0.5-1 (1:1000-1:2000) Mouse 1:2000-1:10 -TUBULIN, Clone DM1A (Sigma T T (Sigma -TUBULIN,Clone DM1A Histone H3 (Millipore H3 Histone 06-755) (Abcam H3 Histone ab1791) kD 17 (Millipore AcH3 06-599) H3K4me2(Millipore 07-030) kD 17 1:1000 H3K4me2(Abcam ab11946) kD 17 H3K4me3(Abcam ab8580) 1:1000-1:5000 kD 17 H3K9me2(Millipore 07-441) kD 17 µg/ml 0.01-0.05 (1:20 H3K9me3(Millipore 07-442) 1:2000-1:10000 kD 17 H3K27me3µg/ml 1 (Abcam ab6002) (1:1000) kD 17 Rabbi 1 Rabbit (Millipore AcH4 06-866) kD 17 µg/ml 1 (1:200-1:600) H4K20me3 (Millipore 07-463) 1:500 kD 17 Rab α Goat 1:500 6199) kD 10 kD 10 1:500 1:2000 1:500-1:2000 Rabbit Rabbit Mouse Rabb µl 5 Rabbit µl 5 µl 5 µl 5 Antibody (Ab) Antibody size Protein
158
, , not working -1:5000 µl 5 1:3000 µl 5-10 Sheep 1:500-1:1000 Sheep 5-20µl,not working 1:5000 Goat µl 20 (1:500-1:1000) Mouse 1:5000 Not working 00-1:1000) Mouse 1:2000-1:3000 µl 10 00-1:500 Mouse 1:500-1:1000 µl 10 1-4 µg/ml 1-4 1:250-1:1000 µg/ml 1 (1:100-1:1000) 1:200 1:200 (1:100-1:1000) 1:200 Rabbit 1:1000 1:2500-1:10000 1:500 Rabbit 1:3000-1:5000 Rabbit µl 5-7.5 mg 1 1:5000 Rabbit µl 5-7.5
MBD1 (M-254) (Santa Cruz (M-254) MBD1 (Santa Sc-10751) kD, 83 bands 2-3 (B-5)Cruz(Santa MBD1 Sc-25261) kD, 80 bands 2 1:1 (Millipore MBD2 07-198) (C-18) MBD3 Cruz(Santa Sc-9402) kD, 34 bands 2 kD 48 (Abcam MBD3 ab3755) clone Kaiso, (Millipore 6F 05-659) Cruz,(Santa(12H) Kaiso Sc-23930) kD 97 kD 97 kD 35 (AbcamNP95 ab22235) (Abcam ZFP57 ab45341) 0.5-2µg/ml (1:1 1:200 1:1000 CTCFC39220) (BD kD 90 CTCF07-729) (Millipore kD, 70 bands 2 1:1000 CTCF(Abcam ab70303) 1:1000 A300-055A) (Bethyl SMC1 kD 140 kD, bands 130 2 A300-080A) (Bethyl RAD21 1:1000 Rabbit kD, bands 130 2 1:500 kD 160 kD, bands 130 2 µl 5 Rabbit 1:1000 1:5000 Rabbit Mouse 1:1000 1:2500 Rabbit
159 (GE NA934); anti-Sheep
5-003); anti-Mouse HRP (GE 5-003); NA931);anti-Mouse HRP anti-Rabbit HRP HRP(Millipore 12-342) Secondary Secondary (Jacksonantibodies: anti-Goat HRP 705-03
160 Plasmid,bacteria / Not working Plasmid,bacteria, viruses Plasmid,bacteria, viruses (siRNA Wizard) (siRNA (Yuan et 2004)al., (Yuan (Invitrogen) Plasmid,bacteria Reference/Tool Reference/Tool Stock/Note Invitrogen Stealth siRNA siRNA Invitrogen Stealth Ctcf-MSS2033343 TTCG TTCG CGAA CTCGAG CTCTTGA CGAA CGAA
Table 9 shRNA sequences shRNA 9 Table TCAAGAG CTCGAG C-3’ TTTTTG-3’ -3’ TTTTTG-3’ TTTTTG-3’ -3’ TTTTTG-3' -3' TACGTATGGACTGATTTCCGAAGCC AGATTTGTGGTCATTCTTCGC TGGACCAGCACAGTTATCTGCATGT 5’-AATTCAAAAACATGCAGATAACTGTGCTGGTCCA 5'-AATTCAAAAAGCGAAGAATGACCACAAATCT 5’-CCGGGGCTTCGGAAATCAGTCCATACGTA 5’-CCGGG ACATGCAGATAACTGTGCTGGTCCA 5'-CCGGGCGAAGAATGACCACAAATCT 5’-CCGGGCGGCATCGTCGTTATAAACA TGTTTATAACGACGATGCCGC 5’-AATTCAAAAAGCGGCATCGTCGTTATAAACA 5’-AATTCAAAAAGGCTTCGGAAATCAGTCCATACGTA AGATTTGTGGTCATTCTTCGC TGTTTATAACGACGATGCCGC TGGACCAGCACAGTTATCTGCATG TACGTATGGACTGATTTCCGAAGCC Ctcf Scrambled Ctcf 343 Ctcf 178 Ctcf1323 Name hairpin Oligossequences(the underlined) are
161 Plasmid,bacteria, viruses Plasmid,bacteria, viruses Plasmid,bacteria, viruses Plasmid,bacteria, viruses (GenScript) Invitrogen Stealth siRNA siRNA Invitrogen Stealth Smc1a-MSS216160 Invitrogen Stealth siRNA siRNA Invitrogen Stealth Smc1a-MSS216158 Invitrogen Stealth siRNA siRNA Invitrogen Stealth Smc1a-MSS216159 TTCG TTCG TTCG
CGAA CGAA CGAA CGAA CGAA TTCAAGAGA TTTTTG-3' C-3' TTTTTG-3' -3' TTTTTG-3' -3' TTTTTG-3’ C-3’ AAGACGAGGAAGTTACGAGCTTTGA AAGACGAGGAAGTTACGAGCTTTGA CAGCATCACAGTAGCGACA 5’-CCGGG TGTCGCTACTGTGATGCTGTTCAAGAGA 5'-CCGGGGCCAAGCTAGTGATTGATGTAATT 5'-CCGGG TCAAAGCTCGTAACTTCCTCGTCTT 5'-CCGGGATGCACGAATTGACCGACAGGAGA AATTACATCAATCACTAGCTTGGCC TCTCCTGTCGGTCAATTCGTGCATC 5’-AATTCAAAAATGTCGCTACTGTGATGCTG AATTACATCAATCACTAGCTTGGCC TCTCCTGTCGGTCAATTCGTGCATC 5'-AATTCAAAAAGATGCACGAATTGACCGACAGGAGA 5'-AATTCAAAAAGGCCAAGCTAGTGATTGATGTAATT 5'-AATTCAAAAATCAAAGCTCGTAACTTCCTCGTCTT CAGCATCACAGTAGCGACA Ctcf1717 Smc1 158 Smc1 159 Smc1 160
162 Plasmid,bacteria, viruses Plasmid,bacteria, viruses Plasmid,bacteria / Not working Plasmid,bacteria / Not working Dharmacon RAD21 ON-TARGET DharmaconON-TARGET RAD21 siRNA-J-058531-09 plus DharmaconON-TARGET RAD21 siRNA-J-058531-12 plus (Yuan et 2004)al., (Yuan (Yuan et 2004)al., (Yuan CTCGAG CTCGAG CTCGAG CTCGAG CTCGAG CTCGAG CTCGAG CTCGAG TTTTTG-3’ -3’ TTTTTG-3’ TTTTTG-3' C-3' TTTTTG-3' -3' AGAATAATCGCTAAGCTGGGC GATATCATTAATGCGAGGCCC TTCACGGTCATCCATTCCG TTCACGGTCATCCATTCCG CAACGCAAAAGCTTCTTCC 5’-CCGGGGGCCTCGCATTAATGATATC 5'-CCGGG CGGAATGGATGACCGTGAA 5'-CCGGGGAAGAAGCTTTTGCGTTG 5'-AATTCAAAAACGGAATGGATGACCGTGAA 5'-AATTCAAAAAGGAAGAAGCTTTTGCGTTG 5’-CCGGGCCCAGCTTAGCGATTATTCT CAACGCAAAAGCTTCTTCC 5’-AATTCAAAAAGGGCCTCGCATTAATGATATC 5’-AATTCAAAAAGCCCAGCTTAGCGATTATTCT AGAATAATCGCTAAGCTGGGC GATATCATTAATGCGAGGCCC
Smc12337 9Rad21 12Rad21 1060 Rad21
163
References
Addgene. pLKO.1 Protocol. Available from: http://www.addgene.org/pgvec1?f=c&cmd=showcol&colid=170&page=2#appendix Ainscough JF, John RM, Barton SC, and Surani MA. 2000. A skeletal muscle-specific mouse Igf2 repressor lies 40 kb downstream of the gene. Development 127:3923-3930. Ainscough JF, John RM, and Surani MA. 1998. Mechanism of imprinting on mouse distal chromosome 7. Genet. Res. 72:237-245. Allen ND, Norris ML, and Surani MA. 1990. Epigenetic control of transgene expression and imprinting by genotype-specific modifiers. Cell 61:853-861. Arnold R, Mäueler W, Bassili G, Lutz M, Burke L, Epplen TJ, and Renkawitz R. 2000. The insulator protein CTCF represses transcription on binding to the (gt)(22)(ga)(15) microsatellite in intron 2 of the HLA-DRB1(*)0401 gene. Gene 253:209-214. Balmer D, Arredondo J, Samaco RC, and LaSalle JM. 2002. MECP2 mutations in Rett syndrome adversely affect lymphocyte growth, but do not affect imprinted gene expression in blood or brain. Hum. Genet. 110:545-552. Barbero JL. 2009. Cohesins: chromatin architects in chromosome segregation, control of gene expression and much more. Cell. Mol. Life Sci. 66:2025-2035. Barlow DP, and Bartolomei MS. 2007. Genomic imprinting in mammals. In: Epigenetics. Cold Spring Harbor, NY: Cold Spring Harbor Lab. Press. p 357-376. Barlow DP. 1993. Methylation and imprinting: from host defense to gene regulation? Science 260:309-310. Barski A, Cuddapah S, Cui K, Roh T, Schones DE, Wang Z, Wei G, Chepelev I, and Zhao K. 2007. High-resolution profiling of histone methylations in the human genome. Cell 129:823-837. Bartolomei MS, Zemel S, and Tilghman SM. 1991. Parental imprinting of the mouse H19 gene. Nature 351:153-155. Bartolomei MS. 2009. Genomic imprinting: employing and avoiding epigenetic processes. Genes Dev. 23:2124-2133. Barton SC, Surani MA, and Norris ML. 1984. Role of paternal and maternal genomes in mouse development. Nature 311:374-376. Bell AC, and Felsenfeld G. 2000. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature 405:482-485. Bell AC, West AG, and Felsenfeld G. 1999. The protein CTCF is required for the
164
enhancer blocking activity of vertebrate insulators. Cell 98:387-396. Bestor T, Laudano A, Mattaliano R, and Ingram V. 1988. Cloning and sequencing of a cDNA encoding DNA methyltransferase of mouse cells. The carboxyl-terminal domain of the mammalian enzymes is related to bacterial restriction methyltransferases. J. Mol. Biol. 203:971-983. Bhogal B, Arnaudo A, Dymkowski A, Best A, and Davis TL. 2004. Methylation at mouse Cdkn1c is acquired during postimplantation development and functions to maintain imprinted expression. Genomics 84:961-970. Birchler JA. 1993. Dosage analysis of maize endosperm development. Annu. Rev. Genet. 27:181-204. Bird AP, and Wolffe AP. 1999. Methylation-induced repression--belts, braces, and chromatin. Cell 99:451-454. Boeke J, Ammerpohl O, Kegel S, Moehren U, and Renkawitz R. 2000. The minimal repression domain of MBD2b overlaps with the methyl-CpG-binding domain and binds directly to Sin3A. J. Biol. Chem. 275:34963-34967. Bogdanovi ć O, and Veenstra GJC. 2009. DNA methylation and methyl-CpG binding proteins: developmental requirements and function. Chromosoma 118:549-565. Bose T, and Gerton JL. 2010. Cohesinopathies, gene expression, and chromatin organization. J. Cell Biol. 189:201-210. Boumil RM, Ogawa Y, Sun BK, Huynh KD, and Lee JT. 2006. Differential methylation of Xite and CTCF sites in Tsix mirrors the pattern of X-inactivation choice in mice. Mol. Cell. Biol. 26:2109-2017. Bourc'his D, Xu GL, Lin CS, Bollman B, and Bestor TH. 2001. Dnmt3L and the establishment of maternal genomic imprints. Science 294:2536-2539. Bourc'his D, and Bestor TH. 2004. Meiotic catastrophe and retrotransposon reactivation in male germ cells lacking Dnmt3L. Nature 431:96-99. Bowen NJ, Palmer MB, and Wade PA. 2004a. Chromosomal regulation by MeCP2: structural and enzymatic considerations. Cell Mol. Life Sci. 61:2163-2167. Bowen NJ, Fujita N, Kajita M, and Wade PA. 2004b. Mi-2/NuRD: multiple complexes for many purposes. Biochim. Biophys. Acta 1677:52-57. Brackertz M, Boeke J, Zhang R, and Renkawitz R. 2002. Two highly related p66 proteins comprise a new family of potent transcriptional repressors interacting with MBD2 and MBD3. J. Biol. Chem. 277:40958-40966. Brandeis M, Kafri T, Ariel M, Chaillet JR, McCarrey J, Razin A, and Cedar H. 1993. The ontogeny of allele-specific methylation associated with imprinted genes in the mouse.
165
EMBO J. 12:3669-3677. Burcin M, Arnold R, Lutz M, Kaiser B, Runge D, Lottspeich F, Filippova GN, Lobanenkov VV, and Renkawitz R. 1997. Negative protein 1, which is required for function of the chicken lysozyme gene silencer in conjunction with hormone receptors, is identical to the multivalent zinc finger repressor CTCF. Mol. Cell. Biol. 17:1281-1288. Burke LJ, Hollemann T, Pieler T, and Renkawitz R. 2002. Molecular cloning and expression of the chromatin insulator protein CTCF in Xenopus laevis. Mech. Dev. 113:95-98. Butler MG. 2009. Genomic imprinting disorders in humans: a mini-review. J. Assist. Reprod. Genet. 26:477-486. Cai X, and Cullen BR. 2007. The imprinted H19 noncoding RNA is a primary microRNA precursor. RNA 13:313-316. Carr MS, Yevtodiyenko A, Schmidt CL, and Schmidt JV. 2007. Allele-specific histone modifications regulate expression of the Dlk1-Gtl2 imprinted domain. Genomics 89:280-290. Caspary T, Cleary MA, Baker CC, Guan XJ, and Tilghman SM. 1998. Multiple mechanisms regulate imprinting of the mouse distal chromosome 7 gene cluster. Mol. Cell. Biol. 18:3466-3474. Cattanach BM, and Kirk M. 1985. Differential activity of maternally and paternally derived chromosome regions in mice. Nature 315:496-498. Chao W, Huynh KD, Spencer RJ, Davidow LS, and Lee JT. 2002. CTCF, a candidate trans-acting factor for X-inactivation choice. Science 295:345-347. Chen T, and Li E. 2006. Establishment and maintenance of DNA methylation patterns in mammals. Curr. Top. Microbiol. Immunol. 301:179-201. Chen T, Ueda Y, Dodge JE, Wang Z, and Li E. 2003. Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b. Mol. Cell. Biol. 23:5594-5605. Chen X, Xu H, Yuan P, Fang F, Huss M, Vega VB, Wong E, Orlov YL, Zhang W, and Jiang J. 2008. Integration of External Signaling Pathways with the Core Transcriptional Network in Embryonic Stem Cells. Cell 133:1106-1117. Chernukhin IV, Shamsuddin S, Robinson AF, Carne AF, Paul A, El-Kady AI, Lobanenkov VV, and Klenova EM. 2000. Physical and functional interaction between two pluripotent proteins, the Y-box DNA/RNA-binding factor, YB-1, and the multivalent zinc finger factor, CTCF. J. Biol. Chem. 275:29915-29921. Chernukhin I, Shamsuddin S, Kang SY, Bergstrom R, Kwon Y, Yu W, Whitehead J, Mukhopadhyay R, Docquier F, Farrar D, et al. 2007. CTCF Interacts with and Recruits
166
the Largest Subunit of RNA Polymerase II to CTCF Target Sites Genome-Wide. Mol. Cell. Biol. 27:1631-1648. Chotalia M, Smallwood SA, Ruf N, Dawson C, Lucifero D, Frontera M, James K, Dean W, and Kelsey G. 2009. Transcription is required for establishment of germline methylation marks at imprinted genes. Genes Dev. 23:105-117. Chung JH, Whiteley M, and Felsenfeld G. 1993. A 5' element of the chicken beta-globin domain serves as an insulator in human erythroid cells and protects against position effect in Drosophila. Cell 74:505-514. Ciccone DN, Su H, Hevi S, Gay F, Lei H, Bajko J, Xu G, Li E, and Chen T. 2009. KDM1B is a histone H3K4 demethylase required to establish maternal genomic imprints. Nature 461:415-418. Ciosk R, Shirayama M, Shevchenko A, Tanaka T, Toth A, Shevchenko A, and Nasmyth K. 2000. Cohesin's binding to chromosomes depends on a separate complex consisting of Scc2 and Scc4 proteins. Mol. Cell 5:243-254. Cirio MC, Martel J, Mann M, Toppings M, Bartolomei M, Trasler J, and Chaillet JR. 2008. DNA methyltransferase 1o functions during preimplantation development to preclude a profound level of epigenetic variation. Dev. Biol. 324:139-150. Collins T, Stone JR, and Williams AJ. 2001. All in the family: the BTB/POZ, KRAB, and SCAN domains. Mol. Cell. Biol. 21:3609-3615. Constância M, Dean W, Lopes S, Moore T, Kelsey G, and Reik W. 2000. Deletion of a silencer element in Igf2 results in loss of imprinting independent of H19. Nat. Genet 26:203-206. Cook PR. 2003. Nongenic transcription, gene regulation and action at a distance. J. Cell. Sci. 116:4483-4491. Cross SH, Meehan RR, Nan X, and Bird A. 1997. A component of the transcriptional repressor MeCP1 shares a motif with DNA methyltransferase and HRX proteins. Nat. Genet. 16:256-259. Cuddapah S, Jothi R, Schones DE, Roh T, Cui K, and Zhao K. 2009. Global analysis of the insulator binding protein CTCF in chromatin barrier regions reveals demarcation of active and repressive domains. Genome Res. 19:24-32. De La Rosa-Velázquez IA, Rincón-Arano H, Benítez-Bribiesca L, and Recillas-Targa F. 2007. Epigenetic regulation of the human retinoblastoma tumor suppressor gene promoter by CTCF. Cancer Res. 67:2577-2585. Deardorff M, Kaur M, Yaeger D, Rampuria A, Korolev S, Pie J, Gilrodriguez C, Arnedo M, Loeys B, and Kline A. 2007. Mutations in Cohesin Complex Members SMC3 and SMC1A Cause a Mild Variant of Cornelia de Lange Syndrome with Predominant Mental Retardation. Am. J. Hum. Genet. 80:485-494.
167
DeChiara TM, Efstratiadis A, and Robertson EJ. 1990. A growth-deficiency phenotype in heterozygous mice carrying an insulin-like growth factor II gene disrupted by targeting. Nature 345:78-80. DeChiara TM, Robertson EJ, and Efstratiadis A. 1991. Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64:849-859. Defossez P, Kelly KF, Filion GJP, Pérez-Torrado R, Magdinier F, Menoni H, Nordgaard CL, Daniel JM, and Gilson E. 2005. The human enhancer blocker CTC-binding factor interacts with the transcription factor Kaiso. J. Biol. Chem. 280:43017-43023. Donohoe ME, Silva SS, Pinter SF, Xu N, and Lee JT. 2009. The pluripotency factor Oct4 interacts with Ctcf and also controls X-chromosome pairing and counting. Nature 460:128-132. Donohoe ME, Zhang L, Xu N, Shi Y, and Lee JT. 2007. Identification of a Ctcf cofactor, Yy1, for the X chromosome binary switch. Mol. Cell 25:43-56. Dorsett D. 2007. Roles of the sister chromatid cohesion apparatus in gene expression, development, and human syndromes. Chromosoma 116:1-13. Dorsett D, Eissenberg JC, Misulovin Z, Martens A, Redding B, and McKim K. 2005. Effects of sister chromatid cohesion proteins on cut gene expression during wing development in Drosophila. Development 132:4743-4753. Drewell RA, Goddard CJ, Thomas JO, and Surani MA. 2002. Methylation-dependent silencing at the H19 imprinting control region by MeCP2. Nucleic Acids Res. 30:1139-1144. Du M, Beatty LG, Zhou W, Lew J, Schoenherr C, Weksberg R, and Sadowski PD. 2003. Insulator and silencer sequences in the imprinted region of human chromosome 11p15.5. Hum. Mol. Genet. 12:1927-1939. Engel N, Raval AK, Thorvaldsen JL, and Bartolomei SM. 2008. Three-dimensional conformation at the H19/Igf2 locus supports a model of enhancer tracking. Hum. Mol. Genet. 17:3021-3029. Engel N, Thorvaldsen JL, and Bartolomei MS. 2006. CTCF binding sites promote transcription initiation and prevent DNA methylation on the maternal allele at the imprinted H19/Igf2 locus. Hum. Mol. Genet. 15:2945-2954. Engel N, West AG, Felsenfeld G, and Bartolomei MS. 2004. Antagonism between DNA hypermethylation and enhancer-blocking activity at the H19 DMD is uncovered by CpG mutations. Nat. Genet. 36:883-888. Engemann S, Strödicke M, Paulsen M, Franck O, Reinhardt R, Lane N, Reik W, and Walter J. 2000. Sequence and functional comparison in the Beckwith-Wiedemann region: implications for a novel imprinting centre and extended imprinting. Hum. Mol. Genet 9:2691-2706.
168
Engström W, Shokrai A, Otte K, Granérus M, Gessbo A, Bierke P, Madej A, Sjölund M, and Ward A. 1998. Transcriptional regulation and biological significance of the insulin like growth factor II gene. Cell Prolif. 31:173-189. Essien K, Vigneau S, Apreleva S, Singh LN, Bartolomei MS, and Hannenhalli S. 2009. CTCF binding site classes exhibit distinct evolutionary, genomic, epigenomic and transcriptomic features. Genome Biol. 10:R131. Evans MJ, and Kaufman MH. 1981. Establishment in culture of pluripotential cells from mouse embryos. Nature 292:154-156. Farrell CM, West AG, and Felsenfeld G. 2002. Conserved CTCF insulator elements flank the mouse and human beta-globin loci. Mol. Cell. Biol. 22:3820-3831. Fedoriw AM, Stein P, Svoboda P, Schultz RM, and Bartolomei MS. 2004. Transgenic RNAi reveals essential function for CTCF in H19 gene imprinting. Science 303:238-240. Feil R, Walter J, Allen ND, and Reik W. 1994. Developmental control of allelic methylation in the imprinted mouse Igf2 and H19 genes. Development 120:2933-2943. Filion GJP, Zhenilo S, Salozhin S, Yamada D, Prokhortchouk E, and Defossez P. 2006. A family of human zinc finger proteins that bind methylated DNA and repress transcription. Mol. Cell. Biol. 26:169-181. Filippova GN, Fagerlie S, Klenova EM, Myers C, Dehner Y, Goodwin G, Neiman PE, Collins SJ, and Lobanenkov VV. 1996. An exceptionally conserved transcriptional repressor, CTCF, employs different combinations of zinc fingers to bind diverged promoter sequences of avian and mammalian c-myc oncogenes. Mol. Cell. Biol 16:2802-2813. Filippova GN. 2008. Genetics and epigenetics of the multifunctional protein CTCF. Curr. Top. Dev. Biol. 80:337-360. Filippova GN, Cheng MK, Moore JM, Truong J, Hu YJ, Nguyen DK, Tsuchiya KD, and Disteche CM. 2005. Boundaries between chromosomal domains of X inactivation and escape bind CTCF and lack CpG methylation during early development. Dev. Cell 8:31-42. Fitzpatrick GV, Pugacheva EM, Shin J, Abdullaev Z, Yang Y, Khatod K, Lobanenkov VV, and Higgins MJ. 2007. Allele-specific binding of CTCF to the multipartite imprinting control region KvDMR1. Mol. Cell. Biol. 27:2636-2647. Fitzpatrick GV, Soloway PD, and Higgins MJ. 2002. Regional loss of imprinting and growth deficiency in mice with a targeted deletion of KvDMR1. Nat. Genet. 32:426-431. Fournier C, Goto Y, Ballestar E, Delaval K, Hever AM, Esteller M, and Feil R. 2002. Allele-specific histone lysine methylation marks regulatory regions at imprinted mouse genes. EMBO J. 21:6560-6570.
169
Freyer C, Zeller U, and Renfree MB. 2003. The marsupial placenta: a phylogenetic analysis. J. Exp. Zoolog. Part A Comp. Exp. Biol. 299:59-77. Fu Y, Sinha M, Peterson CL, and Weng Z. 2008. The insulator binding protein CTCF positions 20 nucleosomes around its binding sites across the human genome. PLoS Genet. 4:e1000138. Fujita N, Takebayashi S, Okumura K, Kudo S, Chiba T, Saya H, and Nakao M. 1999. Methylation-mediated transcriptional silencing in euchromatin by methyl-CpG binding protein MBD1 isoforms. Mol. Cell. Biol. 19:6415-6426. Fujita N, Watanabe S, Ichimura T, Tsuruzoe S, Shinkai Y, Tachibana M, Chiba T, and Nakao M. 2003. Methyl-CpG binding domain 1 (MBD1) interacts with the Suv39h1-HP1 heterochromatic complex for DNA methylation-based transcriptional repression. J. Biol. Chem. 278:24132-24138. Fujita N, Kajita M, Taysavang P, and Wade PA. 2004. Hormonal regulation of metastasis-associated protein 3 transcription in breast cancer cells. Mol. Endocrinol. 18:2937-2949. Gehring M, Bubb KL, and Henikoff S. 2009. Extensive demethylation of repetitive elements during seed development underlies gene imprinting. Science 324:1447-1451. GenScript. siRNA Construct Services - siRNA Design Center - GenScript. Available from: http://www.genscript.com/design_center.html Georgel PT, Horowitz-Scherer RA, Adkins N, Woodcock CL, Wade PA, and Hansen JC. 2003. Chromatin compaction by human MeCP2. Assembly of novel secondary chromatin structures in the absence of DNA methylation. J. Biol. Chem. 278:32181-32188. Ginsburg M, Snow MH, and McLaren A. 1990. Primordial germ cells in the mouse embryo during gastrulation. Development 110:521-528. Glenn CC, Saitoh S, Jong MT, Filbrandt MM, Surti U, Driscoll DJ, and Nicholls RD. 1996. Gene structure, DNA methylation, and imprinted expression of the human SNRPN gene. Am J Hum Genet 58:335-346. Golding MC, Zhang L, and Mann MRW. 2010. Multiple epigenetic modifiers induce aggressive viral extinction in extraembryonic endoderm stem cells. Cell Stem Cell 6:457-467. Gombert WM, Farris SD, Rubio ED, Morey-Rosler KM, Schubach WH, and Krumm A. 2003. The c-myc insulator element and matrix attachment regions define the c-myc chromosomal domain. Mol. Cell. Biol. 23:9338-9348. Gonzales ML, and LaSalle JM. 2010. The role of MeCP2 in brain development and neurodevelopmental disorders. Curr. Psychiatry Rep. 12:127-134. Goren A, Ozsolak F, Shoresh N, Ku M, Adli M, Hart C, Gymrek M, Zuk O, Regev A,
170
Milos PM, et al. 2009. Chromatin profiling by directly sequencing small quantities of immunoprecipitated DNA. Nat. Methods 7:47-49. Gruber S, Haering CH, and Nasmyth K. 2003. Chromosomal cohesin forms a ring. Cell 112:765-777. Guelen L, Pagie L, Brasset E, Meuleman W, Faza MB, Talhout W, Eussen BH, de Klein A, Wessels L, de Laat W, et al. 2008. Domain organization of human chromosomes revealed by mapping of nuclear lamina interactions. Nature 453:948-951. Gullerova M, and Proudfoot NJ. 2008. Cohesin complex promotes transcriptional termination between convergent genes in S. pombe. Cell 132:983-995. Haering CH, Farcas A, Arumugam P, Metson J, and Nasmyth K. 2008. The cohesin ring concatenates sister DNA molecules. Nature 454:297-301. Hagan JP, O'Neill BL, Stewart CL, Kozlov SV, and Croce CM. 2009. At least ten genes define the imprinted Dlk1-Dio3 cluster on mouse chromosome 12qF1. PLoS One 4:e4352. Haig D, and Westoby M. 1989. Parent specific gene expression and the triploid endosperm. Am. Nat. 134:147-155. Hark AT, Schoenherr CJ, Katz DJ, Ingram RS, Levorse JM, and Tilghman SM. 2000. CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature 405:486-489. Hata K, Okano M, Lei H, and Li E. 2002. Dnmt3L cooperates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal imprints in mice. Development 129:1983-1993. Heintzman ND, Hon GC, Hawkins RD, Kheradpour P, Stark A, Harp LF, Ye Z, Lee LK, Stuart RK, Ching CW, et al. 2009. Histone modifications at human enhancers reflect global cell-type-specific gene expression. Nature 459:108-112. Hendrich B, and Bird A. 1998. Identification and characterization of a family of mammalian methyl-CpG binding proteins. Mol. Cell. Biol. 18:6538-6547. Hendrich B, Hardeland U, Ng HH, Jiricny J, and Bird A. 1999. The thymine glycosylase MBD4 can bind to the product of deamination at methylated CpG sites. Nature 401:301-304. Hendrich B, Guy J, Ramsahoye B, Wilson VA, and Bird A. 2001. Closely related proteins MBD2 and MBD3 play distinctive but interacting roles in mouse development. Genes Dev. 15:710-723. Hikichi T, Kohda T, Kaneko-Ishino T, and Ishino F. 2003. Imprinting regulation of the murine Meg1/Grb10 and human GRB10 genes; roles of brain-specific promoters and mouse-specific CTCF-binding sites. Nucleic Acids Res. 31:1398-1406.
171
Himeno E, Tanaka S, and Kunath T. 2008. Isolation and manipulation of mouse trophoblast stem cells. Curr. Protoc. Stem Cell Biol. Chapter 1:Unit 1E.4. Hirasawa R, Chiba H, Kaneda M, Tajima S, Li E, Jaenisch R, and Sasaki H. 2008. Maternal and zygotic Dnmt1 are necessary and sufficient for the maintenance of DNA methylation imprints during preimplantation development. Genes Dev. 22:1607-1616. Howell CY, Bestor TH, Ding F, Latham KE, Mertineit C, Trasler JM, and Chaillet JR. 2001. Genomic imprinting disrupted by a maternal effect mutation in the Dnmt1 gene. Cell 104:829-838. Hsieh T, Ibarra CA, Silva P, Zemach A, Eshed-Williams L, Fischer RL, and Zilberman D. 2009. Genome-wide demethylation of Arabidopsis endosperm. Science 324:1451-1454. Huynh KD, and Lee JT. 2003. Inheritance of a pre-inactivated paternal X chromosome in early mouse embryos. Nature 426:857-862. Ideraabdullah FY, Vigneau S, and Bartolomei MS. 2008. Genomic imprinting mechanisms in mammals. Mutat. Res. 647:77-85. Invitrogen. Invitrogen Block-iT RNAi Designer. Available from: https://rnaidesigner.invitrogen.com/rnaiexpress/? Ishihara K, Hatano N, Furuumi H, Kato R, Iwaki T, Miura K, Jinno Y, and Sasaki H. 2000. Comparative genomic sequencing identifies novel tissue-specific enhancers and sequence elements for methylation-sensitive factors implicated in Igf2/H19 imprinting. Genome Res. 10:664-671. Ishihara K, Oshimura M, and Nakao M. 2006. CTCF-dependent chromatin insulator is linked to epigenetic remodeling. Mol. Cell 23:733-742. Ivanov D, and Nasmyth K. 2005. A topological interaction between cohesin rings and a circular minichromosome. Cell 122:849-860. Jones BK, Levorse J, and Tilghman SM. 2001. Deletion of a nuclease-sensitive region between the Igf2 and H19 genes leads to Igf2 misregulation and increased adiposity. Hum. Mol. Genet. 10:807-814. Jørgensen HF, Ben-Porath I, and Bird AP. 2004. Mbd1 is recruited to both methylated and nonmethylated CpGs via distinct DNA binding domains. Mol. Cell. Biol. 24:3387-3395. Jørgensen HF, and Bird A. 2002. MeCP2 and other methyl-CpG binding proteins. Ment. Retard. Dev. Disabil. Res. Rev. 8:87-93. Kaffer CR, Grinberg A, and Pfeifer K. 2001. Regulatory mechanisms at the mouse Igf2/H19 locus. Mol. Cell. Biol. 21:8189-8196. Kaffer CR, Srivastava M, Park KY, Ives E, Hsieh S, Batlle J, Grinberg A, Huang SP, and Pfeifer K. 2000. A transcriptional insulator at the imprinted H19/Igf2 locus. Genes Dev.
172
14:1908-1919. Kagami M, Nishimura G, Okuyama T, Hayashidani M, Takeuchi T, Tanaka S, Ishino F, Kurosawa K, and Ogata T. 2005. Segmental and full paternal isodisomy for chromosome 14 in three patients: narrowing the critical region and implication for the clinical features. Am. J. Med. Genet. A 138A:127-132. Kagami M, O'Sullivan MJ, Green AJ, Watabe Y, Arisaka O, Masawa N, Matsuoka K, Fukami M, Matsubara K, Kato F, et al. 2010. The IG-DMR and the MEG3-DMR at Human Chromosome 14q32.2: Hierarchical Interaction and Distinct Functional Properties as Imprinting Control Centers. PLoS Genet. 6:e1000992. Kagami M, Sekita Y, Nishimura G, Irie M, Kato F, Okada M, Yamamori S, Kishimoto H, Nakayama M, Tanaka Y, et al. 2008. Deletions and epimutations affecting the human 14q32.2 imprinted region in individuals with paternal and maternal upd(14)-like phenotypes. Nat. Genet. 40:237-242. Kagey MH, Newman JJ, Bilodeau S, Zhan Y, Orlando DA, van Berkum NL, Ebmeier CC, Goossens J, Rahl PB, Levine SS, et al. 2010. Mediator and cohesin connect gene expression and chromatin architecture. Nature 467:430-435. Kaji K, Caballero IM, MacLeod R, Nichols J, Wilson VA, and Hendrich B. 2006. The NuRD component Mbd3 is required for pluripotency of embryonic stem cells. Nat. Cell Biol. 8:285-292. Kaji K, Nichols J, and Hendrich B. 2007. Mbd3, a component of the NuRD co-repressor complex, is required for development of pluripotent cells. Development 134:1123-1132. Kanduri C, Pant V, Loukinov D, Pugacheva E, Qi C, Wolffe A, Ohlsson R, and Lobanenkov VV. 2000. Functional association of CTCF with the insulator upstream of the H19 gene is parent of origin-specific and methylation-sensitive. Curr. Biol. 10:853-856. Kanduri M, Kanduri C, Mariano P, Vostrov AA, Quitschke W, Lobanenkov V, and Ohlsson R. 2002. Multiple nucleosome positioning sites regulate the CTCF-mediated insulator function of the H19 imprinting control region. Mol. Cell. Biol. 22:3339-3344. Kaneda M, Hirasawa R, Chiba H, Okano M, Li E, and Sasaki H. 2010. Genetic evidence for Dnmt3a-dependent imprinting during oocyte growth obtained by conditional knockout with Zp3-Cre and complete exclusion of Dnmt3b by chimera formation. Genes Cells [Internet]. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20132320 Kaneda M, Okano M, Hata K, Sado T, Tsujimoto N, Li E, and Sasaki H. 2004. Essential role for de novo DNA methyltransferase Dnmt3a in paternal and maternal imprinting. Nature 429:900-903. Kang K, Chung JH, and Kim J. 2009. Evolutionary Conserved Motif Finder (ECMFinder) for genome-wide identification of clustered YY1- and CTCF-binding sites. Nucleic Acids
173
Res. 37:2003-2013. Kato Y, Kaneda M, Hata K, Kumaki K, Hisano M, Kohara Y, Okano M, Li E, Nozaki M, and Sasaki H. 2007. Role of the Dnmt3 family in de novo methylation of imprinted and repetitive sequences during male germ cell development in the mouse. Hum. Mol. Genet. 16:2272-2280. Kaur M, DeScipio C, McCallum J, Yaeger D, Devoto M, Jackson LG, Spinner NB, and Krantz ID. 2005. Precocious sister chromatid separation (PSCS) in Cornelia de Lange syndrome. Am. J. Med. Genet. A 138:27-31. Kernohan KD, Jiang Y, Tremblay DC, Bonvissuto AC, Eubanks JH, Mann MRW, and Bérubé NG. 2010. ATRX Partners with Cohesin and MeCP2 and Contributes to Developmental Silencing of Imprinted Genes in the Brain. Dev. Cell 18:191-202. Kim J. 2008. Multiple YY1 and CTCF binding sites in imprinting control regions. Epigenetics 3:115-118. Kim TH, Abdullaev ZK, Smith AD, Ching KA, Loukinov DI, Green RD, Zhang MQ, Lobanenkov VV, and Ren B. 2007. Analysis of the vertebrate insulator protein CTCF-binding sites in the human genome. Cell 128:1231-1245. Klenova EM, Nicolas RH, Paterson HF, Carne AF, Heath CM, Goodwin GH, Neiman PE, and Lobanenkov VV. 1993. CTCF, a conserved nuclear factor required for optimal transcriptional activity of the chicken c-myc gene, is an 11-Zn-finger protein differentially expressed in multiple forms. Mol. Cell. Biol. 13:7612-7624. Klose RJ, and Bird AP. 2006. Genomic DNA methylation: the mark and its mediators. Trends Biochem. Sci. 31:89-97. Köhler C, and Weinhofer-Molisch I. 2010. Mechanisms and evolution of genomic imprinting in plants. Heredity 105:57-63. Koide T, Ainscough J, Wijgerde M, and Surani MA. 1994. Comparative analysis of Igf-2/H19 imprinted domain: identification of a highly conserved intergenic DNase I hypersensitive region. Genomics 24:1-8. Kondo E, Gu Z, Horii A, and Fukushige S. 2005. The thymine DNA glycosylase MBD4 represses transcription and is associated with methylated p16(INK4a) and hMLH1 genes. Mol. Cell. Biol. 25:4388-4396. Kota SK, and Feil R. 2010. Epigenetic transitions in germ cell development and meiosis. Dev. Cell 19:675-686. Krantz ID, McCallum J, DeScipio C, Kaur M, Gillis LA, Yaeger D, Jukofsky L, Wasserman N, Bottani A, Morris CA, et al. 2004. Cornelia de Lange syndrome is caused by mutations in NIPBL, the human homolog of Drosophila melanogaster Nipped-B. Nat. Genet. 36:631-635.
174
Kunarso G, Chia N, Jeyakani J, Hwang C, Lu X, Chan Y, Ng H, and Bourque G. 2010. Transposable elements have rewired the core regulatory network of human embryonic stem cells. Nature Genetics 42:631-634. Kurukuti S, Tiwari VK, Tavoosidana G, Pugacheva E, Murrell A, Zhao Z, Lobanenkov V, Reik W, and Ohlsson R. 2006. CTCF binding at the H19 imprinting control region mediates maternally inherited higher-order chromatin conformation to restrict enhancer access to Igf2. Proc. Natl. Acad. Sci. U. S. A. 103:10684-10689. Kuzmin I, Geil L, Gibson L, Cavinato T, Loukinov D, Lobanenkov V, and Lerman MI. 2005. Transcriptional regulator CTCF controls human interleukin 1 receptor-associated kinase 2 promoter. J. Mol. Biol. 346:411-422. Latham KE, Doherty AS, Scott CD, and Schultz RM. 1994. Igf2r and Igf2 gene expression in androgenetic, gynogenetic, and parthenogenetic preimplantation mouse embryos: absence of regulation by genomic imprinting. Genes Dev. 8:290-299. Le Guezennec X, Vermeulen M, Brinkman AB, Hoeijmakers WAM, Cohen A, Lasonder E, and Stunnenberg HG. 2006. MBD2/NuRD and MBD3/NuRD, two distinct complexes with different biochemical and functional properties. Mol. Cell. Biol. 26:843-851. Lee JE, Pintar J, and Efstratiadis A. 1990. Pattern of the insulin-like growth factor II gene expression during early mouse embryogenesis. Development 110:151-159. Leighton PA, Saam JR, Ingram RS, Stewart CL, and Tilghman SM. 1995. An enhancer deletion affects both H19 and Igf2 expression. Genes Dev. 9:2079-2089. Lewis A, Mitsuya K, Umlauf D, Smith P, Dean W, Walter J, Higgins M, Feil R, and Reik W. 2004. Imprinting on distal chromosome 7 in the placenta involves repressive histone methylation independent of DNA methylation. Nat. Genet. 36:1291-1295. Li E, Beard C, and Jaenisch R. 1993. Role for DNA methylation in genomic imprinting. Nature 366:362-365. Li E, Bestor TH, and Jaenisch R. 1992. Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell 69:915-926. Li E. 2002. Chromatin modification and epigenetic reprogramming in mammalian development. Nat. Rev. Genet. 3:662-673. Li LL, Szeto IY, Cattanach BM, Ishino F, and Surani MA. 2000. Organization and parent-of-origin-specific methylation of imprinted Peg3 gene on mouse proximal chromosome 7. Genomics 63:333-340. Li Q, Barkess G, and Qian H. 2006. Chromatin looping and the probability of transcription. Trends Genet. 22:197-202. Li T, Hu J, Qiu X, Ling J, Chen H, Wang S, Hou A, Vu TH, and Hoffman AR. 2008. CTCF regulates allelic expression of Igf2 by orchestrating a promoter-polycomb
175
repressive complex 2 intrachromosomal loop. Mol. Cell. Biol. 28:6473-6482. Li T, Lu Z, and Lu L. 2004. Regulation of eye development by transcription control of CCCTC binding factor (CTCF). J. Biol. Chem. 279:27575-27583. Li X, Ito M, Zhou F, Youngson N, Zuo X, Leder P, and Ferguson-Smith AC. 2008. A maternal-zygotic effect gene, Zfp57, maintains both maternal and paternal imprints. Dev Cell 15:547-557. Lim DHK, and Maher ER. 2010. Genomic imprinting syndromes and cancer. Adv. Genet. 70:145-175. Lin S, Coan P, da Rocha ST, Seitz H, Cavaille J, Teng P, Takada S, and Ferguson-Smith AC. 2007. Differential regulation of imprinting in the murine embryo and placenta by the Dlk1-Dio3 imprinting control region. Development 134:417-426. Lin S, Youngson N, Takada S, Seitz H, Reik W, Paulsen M, Cavaille J, and Ferguson-Smith AC. 2003. Asymmetric regulation of imprinting on the maternal and paternal chromosomes at the Dlk1-Gtl2 imprinted cluster on mouse chromosome 12. Nat. Genet. 35:97-102. Ling JQ, Li T, Hu JF, Vu TH, Chen HL, Qiu XW, Cherry AM, and Hoffman AR. 2006. CTCF mediates interchromosomal colocalization between Igf2/H19 and Wsb1/Nf1. Science 312:269-272. Lobanenkov VV, Nicolas RH, Adler VV, Paterson H, Klenova EM, Polotskaja AV, and Goodwin GH. 1990. A novel sequence-specific DNA binding protein which interacts with three regularly spaced direct repeats of the CCCTC-motif in the 5'-flanking sequence of the chicken c-myc gene. Oncogene 5:1743-1753. Lu X, Kovalev GI, Chang H, Kallin E, Knudsen G, Xia L, Mishra N, Ruiz P, Li E, Su L, et al. 2008. Inactivation of NuRD component Mta2 causes abnormal T cell activation and lupus-like autoimmune disease in mice. J. Biol. Chem. 283:13825-13833. Lucifero D, Mertineit C, Clarke HJ, Bestor TH, and Trasler JM. 2002. Methylation dynamics of imprinted genes in mouse germ cells. Genomics 79:530-538. Lui JC, Finkielstain GP, Barnes KM, and Baron J. 2008. An imprinted gene network that controls mammalian somatic growth is down-regulated during postnatal growth deceleration in multiple organs. Am. J. Physiol. Regul. Integr. Comp. Physiol. 295:R189-R196. Lutz M, Burke LJ, Barreto G, Goeman F, Greb H, Arnold R, Schultheiss H, Brehm A, Kouzarides T, Lobanenkov V, et al. 2000. Transcriptional repression by the insulator protein CTCF involves histone deacetylases. Nucleic Acids Res. 28:1707-1713. Lyle R, Watanabe D, te Vruchte D, Lerchner W, Smrzka OW, Wutz A, Schageman J, Hahner L, Davies C, and Barlow DP. 2000. The imprinted antisense RNA at the Igf2r locus overlaps but does not imprint Mas1. Nat. Genet. 25:19-21.
176
Ma P, Lin S, Bartolomei MS, and Schultz RM. 2010. Metastasis Tumor Antigen 2 (MTA2) is involved in proper imprinted expression of H19 and Peg3 during mouse preimplantation development. Biol. Reprod. 83:1027-1035. Mackay DJG, Callaway JLA, Marks SM, White HE, Acerini CL, Boonen SE, Dayanikli P, Firth HV, Goodship JA, Haemers AP, et al. 2008. Hypomethylation of multiple imprinted loci in individuals with transient neonatal diabetes is associated with mutations in ZFP57. Nat. Genet. 40:949-951. Mak W, Baxter J, Silva J, Newall AE, Otte AP, and Brockdorff N. 2002. Mitotically stable association of polycomb group proteins eed and enx1 with the inactive x chromosome in trophoblast stem cells. Curr Biol 12:1016-1020. Makedonski K, Abuhatzira L, Kaufman Y, Razin A, and Shemer R. 2005. MeCP2 deficiency in Rett syndrome causes epigenetic aberrations at the PWS/AS imprinting center that affects UBE3A expression. Hum. Mol. Genet. 14:1049-1058. Manavathi B, and Kumar R. 2007. Metastasis tumor antigens, an emerging family of multifaceted master coregulators. J. Biol. Chem. 282:1529-1533. Mancini-DiNardo D, Steele SJS, Ingram RS, and Tilghman SM. 2003. A differentially methylated region within the gene Kcnq1 functions as an imprinted promoter and silencer. Hum. Mol. Genet. 12:283-294. Mancini-Dinardo D, Steele SJS, Levorse JM, Ingram RS, and Tilghman SM. 2006. Elongation of the Kcnq1ot1 transcript is required for genomic imprinting of neighboring genes. Genes Dev. 20:1268-1282. Mann MRW, and Bartolomei MS. 2002. Epigenetic reprogramming in the mammalian embryo: struggle of the clones. Genome Biol. 3:REVIEWS1003. Mann MRW, Chung YG, Nolen LD, Verona RI, Latham KE, and Bartolomei MS. 2003. Disruption of imprinted gene methylation and expression in cloned preimplantation stage mouse embryos. Biol. Reprod. 69:902-914. Mann MRW, Lee SS, Doherty AS, Verona RI, Nolen LD, Schultz RM, and Bartolomei MS. 2004. Selective loss of imprinting in the placenta following preimplantation development in culture. Development 131:3727-3735. Market-Velker BA, Fernandes AD, and Mann MRW. 2010a. Side-by-side comparison of five commercial media systems in a mouse model: suboptimal in vitro culture interferes with imprint maintenance. Biol. Reprod. 83:938-950. Market-Velker BA, Zhang L, Magri LS, Bonvissuto AC, and Mann MR. 2010b. Dual effects of superovulation: loss of maternal and paternal imprinted methylation in a dose-dependent manner. Hum. Mol. Genet. 19:36-51. Martin GR. 1981. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci U S A
177
78:7634-7638. Mayer W, Niveleau A, Walter J, Fundele R, and Haaf T. 2000. Demethylation of the zygotic paternal genome. Nature 403:501-502. McDonel P, Costello I, and Hendrich B. 2009. Keeping things quiet: roles of NuRD and Sin3 co-repressor complexes during mammalian development. Int. J. Biochem. Cell Biol. 41:108-116. McGrath J, and Solter D. 1984. Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell 37:179-183. Millar CB, Guy J, Sansom OJ, Selfridge J, MacDougall E, Hendrich B, Keightley PD, Bishop SM, Clarke AR, and Bird A. 2002. Enhanced CpG mutability and tumorigenesis in MBD4-deficient mice. Science 297:403-405. Mineno J, Okamoto S, Ando T, Sato M, Chono H, Izu H, Takayama M, Asada K, Mirochnitchenko O, Inouye M, et al. 2006. The expression profile of microRNAs in mouse embryos. Nucleic Acids Res. 34:1765-1771. Mohammad F, Pandey RR, Nagano T, Chakalova L, Mondal T, Fraser P, and Kanduri C. 2008. Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by targeting to the perinucleolar region. Mol. Cell. Biol. 28:3713-3728. Moon H, Filippova G, Loukinov D, Pugacheva E, Chen Q, Smith ST, Munhall A, Grewe B, Bartkuhn M, Arnold R, et al. 2005. CTCF is conserved from Drosophila to humans and confers enhancer blocking of the Fab-8 insulator. EMBO Rep. 6:165-170. Moore T, Constancia M, Zubair M, Bailleul B, Feil R, Sasaki H, and Reik W. 1997. Multiple imprinted sense and antisense transcripts, differential methylation and tandem repeats in a putative imprinting control region upstream of mouse Igf2. Proc. Natl. Acad. Sci. U.S.A 94:12509-12514. Mukhopadhyay R, Yu W, Whitehead J, Xu J, Lezcano M, Pack S, Kanduri C, Kanduri M, Ginjala V, Vostrov A, et al. 2004. The binding sites for the chromatin insulator protein CTCF map to DNA methylation-free domains genome-wide. Genome Res. 14:1594-1602. Murrell A, Heeson S, and Reik W. 2004. Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 into parent-specific chromatin loops. Nat. Genet. 36:889-893. Nakamura T, Arai Y, Umehara H, Masuhara M, Kimura T, Taniguchi H, Sekimoto T, Ikawa M, Yoneda Y, Okabe M, et al. 2007. PGC7/Stella protects against DNA demethylation in early embryogenesis. Nat. Cell Biol. 9:64-71. Nan X, Campoy FJ, and Bird A. 1997. MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin. Cell 88:471-481.
178
Nan X, Meehan RR, and Bird A. 1993. Dissection of the methyl-CpG binding domain from the chromosomal protein MeCP2. Nucleic Acids Res. 21:4886-4892. Nativio R, Wendt KS, Ito Y, Huddleston JE, Uribe-Lewis S, Woodfine K, Krueger C, Reik W, Peters J, and Murrell A. 2009. Cohesin Is Required for Higher-Order Chromatin Conformation at the Imprinted IGF2-H19 Locus. PLoS Genet. 5:e1000739. Navarro P, Page DR, Avner P, and Rougeulle C. 2006. Tsix-mediated epigenetic switch of a CTCF-flanked region of the Xist promoter determines the Xist transcription program. Genes Dev. 20:2787-2792. Ng HH, Jeppesen P, and Bird A. 2000. Active repression of methylated genes by the chromosomal protein MBD1. Mol. Cell. Biol. 20:1394-1406. Ohlsson R, Renkawitz R, and Lobanenkov V. 2001. CTCF is a uniquely versatile transcription regulator linked to epigenetics and disease. Trends Genet. 17:520-527. Ohlsson R, Bartkuhn M, and Renkawitz R. 2010. CTCF shapes chromatin by multiple mechanisms: the impact of 20 years of CTCF research on understanding the workings of chromatin. Chromosoma 119:351-360. Okano M, Bell DW, Haber DA, and Li E. 1999. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99:247-257. Okano M, Xie S, and Li E. 1998. Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nat. Genet. 19:219-220. Ono R, Shiura H, Aburatani H, Kohda T, Kaneko-Ishino T, and Ishino F. 2003. Identification of a large novel imprinted gene cluster on mouse proximal chromosome 6. Genome Res. 13:1696-1705. Ooi SKT, and Bestor TH. 2008. The colorful history of active DNA demethylation. Cell 133:1145-1148. Ooi SKT, Qiu C, Bernstein E, Li K, Jia D, Yang Z, Erdjument-Bromage H, Tempst P, Lin S, Allis CD, et al. 2007. DNMT3L connects unmethylated lysine 4 of histone H3 to de novo methylation of DNA. Nature 448:714-717. Oswald J, Engemann S, Lane N, Mayer W, Olek A, Fundele R, Dean W, Reik W, and Walter J. 2000. Active demethylation of the paternal genome in the mouse zygote. Curr. Biol. 10:475-478. Owen CM, and Segars JH. 2009. Imprinting disorders and assisted reproductive technology. Semin. Reprod. Med. 27:417-428. Pant V, Kurukuti S, Pugacheva E, Shamsuddin S, Mariano P, Renkawitz R, Klenova E, Lobanenkov V, and Ohlsson R. 2004. Mutation of a single CTCF target site within the H19 imprinting control region leads to loss of Igf2 imprinting and complex patterns of de
179
novo methylation upon maternal inheritance. Mol. Cell. Biol. 24:3497-3504. Pant V, Mariano P, Kanduri C, Mattsson A, Lobanenkov V, Heuchel R, and Ohlsson R. 2003. The nucleotides responsible for the direct physical contact between the chromatin insulator protein CTCF and the H19 imprinting control region manifest parent of origin-specific long-distance insulation and methylation-free domains. Genes Dev. 17:586-590. Parelho V, Hadjur S, Spivakov M, Leleu M, Sauer S, Gregson HC, Jarmuz A, Canzonetta C, Webster Z, Nesterova T, et al. 2008. Cohesins functionally associate with CTCF on mammalian chromosome arms. Cell 132:422-433. Pask AJ, Papenfuss AT, Ager EI, McColl KA, Speed TP, and Renfree MB. 2009. Analysis of the platypus genome suggests a transposon origin for mammalian imprinting. Genome Biol 10:R1. Paulsen M, El-Maarri O, Engemann S, Strödicke M, Franck O, Davies K, Reinhardt R, Reik W, and Walter J. 2000. Sequence conservation and variability of imprinting in the Beckwith-Wiedemann syndrome gene cluster in human and mouse. Hum. Mol. Genet. 9:1829-1841. Paulsen M, Takada S, Youngson NA, Benchaib M, Charlier C, Segers K, Georges M, and Ferguson-Smith AC. 2001. Comparative sequence analysis of the imprinted Dlk1-Gtl2 locus in three mammalian species reveals highly conserved genomic elements and refines comparison with the Igf2-H19 region. Genome Res. 11:2085-2094. Poirier F, Chan CT, Timmons PM, Robertson EJ, Evans MJ, and Rigby PW. 1991. The murine H19 gene is activated during embryonic stem cell differentiation in vitro and at the time of implantation in the developing embryo. Development 113:1105-1114. Popp C, Dean W, Feng S, Cokus SJ, Andrews S, Pellegrini M, Jacobsen SE, and Reik W. 2010. Genome-wide erasure of DNA methylation in mouse primordial germ cells is affected by AID deficiency. Nature 463:1101-1105. Prokhortchouk A, Hendrich B, Jørgensen H, Ruzov A, Wilm M, Georgiev G, Bird A, and Prokhortchouk E. 2001. The p120 catenin partner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes Dev. 15:1613-1618. Prokhortchouk A, Sansom O, Selfridge J, Caballero IM, Salozhin S, Aithozhina D, Cerchietti L, Meng FG, Augenlicht LH, Mariadason JM, et al. 2006. Kaiso-deficient mice show resistance to intestinal cancer. Mol. Cell. Biol. 26:199-208. Pugacheva EM, Kwon Y, Hukriede NA, Pack S, Flanagan PT, Ahn J, Park JA, Choi K, Kim K, Loukinov D, et al. 2006. Cloning and characterization of zebrafish CTCF: Developmental expression patterns, regulation of the promoter region, and evolutionary aspects of gene organization. Gene 375:26-36. Pugacheva EM, Tiwari VK, Abdullaev Z, Vostrov AA, Flanagan PT, Quitschke WW,
180
Loukinov DI, Ohlsson R, and Lobanenkov VV. 2005. Familial cases of point mutations in the XIST promoter reveal a correlation between CTCF binding and pre-emptive choices of X chromosome inactivation. Hum. Mol. Genet. 14:953-965. Qiu X, Vu TH, Lu Q, Ling JQ, Li T, Hou A, Wang SK, Chen HL, Hu JF, and Hoffman AR. 2008. A complex deoxyribonucleic acid looping configuration associated with the silencing of the maternal Igf2 allele. Mol. Endocrinol. 22:1476-1488. Reese KJ, Lin S, Verona RI, Schultz RM, and Bartolomei MS. 2007. Maintenance of paternal methylation and repression of the imprinted H19 gene requires MBD3. PLoS Genet. 3:e137. Reik W, Dean W, and Walter J. 2001. Epigenetic reprogramming in mammalian development. Science 293:1089-1093. Renfree MB, Hore TA, Shaw G, Graves JAM, and Pask AJ. 2009. Evolution of genomic imprinting: insights from marsupials and monotremes. Annu. Rev. Genomics Hum. Genet. 10:241-262. Rivera RM, Stein P, Weaver JR, Mager J, Schultz RM, and Bartolomei MS. 2008. Manipulations of mouse embryos prior to implantation result in aberrant expression of imprinted genes on day 9.5 of development. Hum. Mol. Genet. 17:1-14. Rohde C, Zhang Y, Jurkowski TP, Stamerjohanns H, Reinhardt R, and Jeltsch A. 2008. Bisulfite sequencing Data Presentation and Compilation (BDPC) web server--a useful tool for DNA methylation analysis. Nucleic Acids Res. 36:e34. Rollins RA, Korom M, Aulner N, Martens A, and Dorsett D. 2004. Drosophila nipped-B protein supports sister chromatid cohesion and opposes the stromalin/Scc3 cohesion factor to facilitate long-range activation of the cut gene. Mol. Cell. Biol. 24:3100-3111. Rougier N, Bourc'his D, Gomes DM, Niveleau A, Plachot M, Pàldi A, and Viegas-Péquignot E. 1998. Chromosome methylation patterns during mammalian preimplantation development. Genes Dev. 12:2108-2113. Rubio ED, Reiss DJ, Welcsh PL, Disteche CM, Filippova GN, Baliga NS, Aebersold R, Ranish JA, and Krumm A. 2008. CTCF physically links cohesin to chromatin. Proc. Natl. Acad. Sci. U. S. A. 105:8309-8314. Saito M, and Ishikawa F. 2002. The mCpG-binding domain of human MBD3 does not bind to mCpG but interacts with NuRD/Mi2 components HDAC1 and MTA2. J. Biol. Chem. 277:35434-35439. Samaco RC, Hogart A, and LaSalle JM. 2005. Epigenetic overlap in autism-spectrum neurodevelopmental disorders: MECP2 deficiency causes reduced expression of UBE3A and GABRB3. Hum. Mol. Genet. 14:483-492. Santos F, Hendrich B, Reik W, and Dean W. 2002. Dynamic reprogramming of DNA methylation in the early mouse embryo. Dev. Biol. 241:172-182.
181
Sapienza C, Paquette J, Tran TH, and Peterson A. 1989. Epigenetic and genetic factors affect transgene methylation imprinting. Development 107:165-168. Sarbassov DD, Guertin DA, Ali SM, and Sabatini DM. 2005. Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307:1098-1101. Sarraf SA, and Stancheva I. 2004. Methyl-CpG binding protein MBD1 couples histone H3 methylation at lysine 9 by SETDB1 to DNA replication and chromatin assembly. Mol. Cell 15:595-605. Sasaki H, Jones PA, Chaillet JR, Ferguson-Smith AC, Barton SC, Reik W, and Surani MA. 1992. Parental imprinting: potentially active chromatin of the repressed maternal allele of the mouse insulin-like growth factor II (Igf2) gene. Genes Dev. 6:1843-1856. Schaaf CA, Misulovin Z, Sahota G, Siddiqui AM, Schwartz YB, Kahn TG, Pirrotta V, Gause M, and Dorsett D. 2009. Regulation of the Drosophila Enhancer of split and invected-engrailed Gene Complexes by Sister Chromatid Cohesion Proteins. PLoS One 4:e6202. Schmidt D, Schwalie PC, Ross-Innes CS, Hurtado A, Brown GD, Carroll JS, Flicek P, and Odom DT. 2010. A CTCF-independent role for cohesin in tissue-specific transcription. Genome Res. 20:578-588. Schoenherr CJ, Levorse JM, and Tilghman SM. 2003. CTCF maintains differential methylation at the Igf2/H19 locus. Nat. Genet. 33:66-69. Schultz DC, Friedman JR, and Rauscher FJ. 2001. Targeting histone deacetylase complexes via KRAB-zinc finger proteins: the PHD and bromodomains of KAP-1 form a cooperative unit that recruits a novel isoform of the Mi-2alpha subunit of NuRD. Genes Dev. 15:428-443. Schulz R, Proudhon C, Bestor TH, Woodfine K, Lin C, Lin S, Prissette M, Oakey RJ, and Bourc'his D. 2010. The parental non-equivalence of imprinting control regions during mammalian development and evolution. PLoS Genet. 6:e1001214. Schuster-Gossler K, Bilinski P, Sado T, Ferguson-Smith A, and Gossler A. 1998. The mouse Gtl2 gene is differentially expressed during embryonic development, encodes multiple alternatively spliced transcripts, and may act as an RNA. Dev. Dyn. 212:214-228. Scott RJ, Spielman M, Bailey J, and Dickinson HG. 1998. Parent-of-origin effects on seed development in Arabidopsis thaliana. Development 125:3329-3341. Seki Y, Hayashi K, Itoh K, Mizugaki M, Saitou M, and Matsui Y. 2005. Extensive and orderly reprogramming of genome-wide chromatin modifications associated with specification and early development of germ cells in mice. Dev. Biol. 278:440-458. Sharif J, Muto M, Takebayashi S, Suetake I, Iwamatsu A, Endo TA, Shinga J, Mizutani-Koseki Y, Toyoda T, Okamura K, et al. 2007. The SRA protein Np95 mediates
182
epigenetic inheritance by recruiting Dnmt1 to methylated DNA. Nature 450:908-912. Shin J, Fitzpatrick GV, and Higgins MJ. 2008. Two distinct mechanisms of silencing by the KvDMR1 imprinting control region. EMBO J. 27:168-178. siRNA Wizard. siRNA Wizard - The Online Tool to Choose and Design short hairpin RNAs. Available from: http://www.sirnawizard.com/ Sleutels F, Zwart R, and Barlow DP. 2002. The non-coding Air RNA is required for silencing autosomal imprinted genes. Nature 415:810-813. Smilinich NJ, Day CD, Fitzpatrick GV, Caldwell GM, Lossie AC, Cooper PR, Smallwood AC, Joyce JA, Schofield PN, Reik W, et al. 1999. A maternally methylated CpG island in KvLQT1 is associated with an antisense paternal transcript and loss of imprinting in Beckwith–Wiedemann syndrome. Proc. Natl. Acad. Sci. U. S. A. 96:8064-8069. Stadnick MP, Pieracci FM, Cranston MJ, Taksel E, Thorvaldsen JL, and Bartolomei MS. 1999. Role of a 461-bp G-rich repetitive element in H19 transgene imprinting. Dev Genes Evol 209:239-248. Stedman W, Kang H, Lin S, Kissil JL, Bartolomei MS, and Lieberman PM. 2008. Cohesins localize with CTCF at the KSHV latency control region and at cellular c-myc and H19/Igf2 insulators. EMBO J. 27:654-666. Stöger R, Kubicka P, Liu CG, Kafri T, Razin A, Cedar H, and Barlow DP. 1993. Maternal-specific methylation of the imprinted mouse Igf2r locus identifies the expressed locus as carrying the imprinting signal. Cell 73:61-71. Strumpf D, Mao C, Yamanaka Y, Ralston A, Chawengsaksophak K, Beck F, and Rossant J. 2005. Cdx2 is required for correct cell fate specification and differentiation of trophectoderm in the mouse blastocyst. Development 132:2093-2102. Suetake I, Shinozaki F, Miyagawa J, Takeshima H, and Tajima S. 2004. DNMT3L stimulates the DNA methylation activity of Dnmt3a and Dnmt3b through a direct interaction. J. Biol. Chem. 279:27816-27823. Sumara I, Vorlaufer E, Gieffers C, Peters BH, and Peters JM. 2000. Characterization of vertebrate cohesin complexes and their regulation in prophase. J. Cell Biol. 151:749-762. Surani MA, Barton SC, and Norris ML. 1984. Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature 308:548-550. Suzuki S, Ono R, Narita T, Pask AJ, Shaw G, Wang C, Kohda T, Alsop AE, Marshall Graves JA, Kohara Y, et al. 2007. Retrotransposon silencing by DNA methylation can drive mammalian genomic imprinting. PLoS Genet. 3:e55. Szabo PE, Tang SE, Silva FJ, Tsark WMK, and Mann JR. 2004. Role of CTCF binding sites in the Igf2/H19 imprinting control region. Mol. Cell. Biol. 24:4791-4800.
183
Szabó PE, Tang SE, Rentsendorj A, Pfeifer GP, and Mann JR. 2000. Maternal-specific footprints at putative CTCF sites in the H19 imprinting control region give evidence for insulator function. Curr. Biol. 10:607-610. Szeto IYY, Barton SC, Keverne EB, and Surani AM. 2004. Analysis of imprinted murine Peg3 locus in transgenic mice. Mamm. Genome 15:284-295. Takada S, Tevendale M, Baker J, Georgiades P, Campbell E, Freeman T, Johnson MH, Paulsen M, and Ferguson-Smith AC. 2000. Delta-like and gtl2 are reciprocally expressed, differentially methylated linked imprinted genes on mouse chromosome 12. Curr. Biol. 10:1135-1138. Tanaka M, Puchyr M, Gertsenstein M, Harpal K, Jaenisch R, Rossant J, and Nagy A. 1999. Parental origin-specific expression of Mash2 is established at the time of implantation with its imprinting mechanism highly resistant to genome-wide demethylation. Mech. Dev. 87:129-142. Tanaka S, Kunath T, Hadjantonakis AK, Nagy A, and Rossant J. 1998. Promotion of trophoblast stem cell proliferation by FGF4. Science 282:2072-2075. Tanaka S. 2006. Derivation and culture of mouse trophoblast stem cells in vitro. Methods Mol. Biol 329:35-44. Tatematsu KI, Yamazaki T, and Ishikawa F. 2000. MBD2-MBD3 complex binds to hemi-methylated DNA and forms a complex containing DNMT1 at the replication foci in late S phase. Genes Cells 5:677-688. Thakur N, Kanduri M, Holmgren C, Mukhopadhyay R, and Kanduri C. 2003. Bidirectional silencing and DNA methylation-sensitive methylation-spreading properties of the Kcnq1 imprinting control region map to the same regions. J. Biol. Chem. 278:9514-9519. Thakur N, Tiwari VK, Thomassin H, Pandey RR, Kanduri M, Göndör A, Grange T, Ohlsson R, and Kanduri C. 2004. An antisense RNA regulates the bidirectional silencing property of the Kcnq1 imprinting control region. Mol. Cell. Biol. 24:7855-7862. Thiagalingam S, Cheng K, Lee HJ, Mineva N, Thiagalingam A, and Ponte JF. 2003. Histone deacetylases: unique players in shaping the epigenetic histone code. Ann. N. Y. Acad. Sci. 983:84-100. Thorvaldsen JL, Duran KL, and Bartolomei MS. 1998. Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2. Genes Dev. 12:3693-3702. Thorvaldsen JL, Fedoriw AM, Nguyen S, and Bartolomei MS. 2006. Developmental profile of H19 differentially methylated domain (DMD) deletion alleles reveals multiple roles of the DMD in regulating allelic expression and DNA methylation at the imprinted H19/Igf2 locus. Mol. Cell. Biol. 26:1245-1258.
184
Thorvaldsen JL, Mann MRW, Nwoko O, Duran KL, and Bartolomei MS. 2002. Analysis of sequence upstream of the endogenous H19 gene reveals elements both essential and dispensable for imprinting. Mol. Cell. Biol. 22:2450-2462. Tierling S, Dalbert S, Schoppenhorst S, Tsai C, Oliger S, Ferguson-Smith AC, Paulsen M, and Walter J. 2006. High-resolution map and imprinting analysis of the Gtl2-Dnchc1 domain on mouse chromosome 12. Genomics 87:225-235. Toh Y, Oki E, Oda S, Tokunaga E, Ohno S, Maehara Y, Nicolson GL, and Sugimachi K. 1997. Overexpression of the MTA1 gene in gastrointestinal carcinomas: correlation with invasion and metastasis. Int. J. Cancer 74:459-463. Tolhuis B, Palstra RJ, Splinter E, Grosveld F, and de Laat W. 2002. Looping and interaction between hypersensitive sites in the active beta-globin locus. Mol. Cell 10:1453-1465. Tonkin ET, Wang T, Lisgo S, Bamshad MJ, and Strachan T. 2004. NIPBL, encoding a homolog of fungal Scc2-type sister chromatid cohesion proteins and fly Nipped-B, is mutated in Cornelia de Lange syndrome. Nat. Genet. 36:636-641. Tremblay KD, Duran KL, and Bartolomei MS. 1997. A 5' 2-kilobase-pair region of the imprinted mouse H19 gene exhibits exclusive paternal methylation throughout development. Mol. Cell. Biol. 17:4322-4329. Tremblay KD, Saam JR, Ingram RS, Tilghman SM, and Bartolomei MS. 1995. A paternal-specific methylation imprint marks the alleles of the mouse H19 gene. Nat. Genet. 9:407-413. Trivers R, and Burt A. 1999. Kinship and genomic imprinting. Results Probl. Cell Differ. 25:1-21. Varrault A, Gueydan C, Delalbre A, Bellmann A, Houssami S, Aknin C, Severac D, Chotard L, Kahli M, Le Digarcher A, et al. 2006. Zac1 regulates an imprinted gene network critically involved in the control of embryonic growth. Dev. Cell 11:711-722. Verona RI, Thorvaldsen JL, Reese KJ, and Bartolomei MS. 2008. The transcriptional status but not the imprinting control region determines allele-specific histone modifications at the imprinted H19 locus. Mol. Cell. Biol. 28:71-82. Vostrov AA, and Quitschke WW. 1997. The zinc finger protein CTCF binds to the APBbeta domain of the amyloid beta-protein precursor promoter. Evidence for a role in transcriptional activation. J. Biol. Chem. 272:33353-33359. Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, and Wolffe AP. 1999. Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat. Genet. 23:62-66. Wan L, and Bartolomei MS. 2008. Regulation of Imprinting in Clusters: Noncoding RNAs Versus Insulators. Adv. Genet. 61:207-223.
185
Wan L, Pan H, Hannenhalli S, Cheng Y, Ma J, Fedoriw A, Lobanenkov V, Latham KE, Schultz RM, and Bartolomei MS. 2008. Maternal depletion of CTCF reveals multiple functions during oocyte and preimplantation embryo development. Development 135:2729-2738. Wang Y, Zhang H, Chen Y, Sun Y, Yang F, Yu W, Liang J, Sun L, Yang X, Shi L, et al. 2009. LSD1 is a subunit of the NuRD complex and targets the metastasis programs in breast cancer. Cell 138:660-672. Watrin E, and Peters J. 2006. Cohesin and DNA damage repair. Exp. Cell Res. 312:2687-2693. Watrin E, Schleiffer A, Tanaka K, Eisenhaber F, Nasmyth K, and Peters J. 2006. Human Scc4 is required for cohesin binding to chromatin, sister-chromatid cohesion, and mitotic progression. Curr. Biol. 16:863-874. Weaver JR, Sarkisian G, Krapp C, Mager J, Mann MRW, and Bartolomei MS. 2010. Domain-specific response of imprinted genes to reduced DNMT1. Mol. Cell. Biol. 30:3916-3928. Weaver JR, Susiarjo M, and Bartolomei MS. 2009. Imprinting and epigenetic changes in the early embryo. Mamm. Genome 20:532-543. Webber AL, Ingram RS, Levorse JM, and Tilghman SM. 1998. Location of enhancers is essential for the imprinting of H19 and Igf2 genes. Nature 391:711-715. Weitzel JM, Buhrmester H, and Strätling WH. 1997. Chicken MAR-binding protein ARBP is homologous to rat methyl-CpG-binding protein MeCP2. Mol. Cell. Biol. 17:5656-5666. Wendt KS, Yoshida K, Itoh T, Bando M, Koch B, Schirghuber E, Tsutsumi S, Nagae G, Ishihara K, Mishiro T, et al. 2008. Cohesin mediates transcriptional insulation by CCCTC-binding factor. Nature 451:796-801. Williams A, and Flavell RA. 2008. The role of CTCF in regulating nuclear organization. J. Exp. Med. 205:747-750. Williamson CM, Turner MD, Ball ST, Nottingham WT, Glenister P, Fray M, Tymowska-Lalanne Z, Plagge A, Powles-Glover N, Kelsey G, et al. 2006. Identification of an imprinting control region affecting the expression of all transcripts in the Gnas cluster. Nat. Genet. 38:350-355. Wong E, Yang K, Kuraguchi M, Werling U, Avdievich E, Fan K, Fazzari M, Jin B, Brown AMC, Lipkin M, et al. 2002. Mbd4 inactivation increases Cright-arrowT transition mutations and promotes gastrointestinal tumor formation. Proc. Natl. Acad. Sci. U. S. A. 99:14937-14942. Wood AJ, Schulz R, Woodfine K, Koltowska K, Beechey CV, Peters J, Bourc'his D, and Oakey RJ. 2008. Regulation of alternative polyadenylation by genomic imprinting. Genes
186
Dev. 22:1141-1146. Wu D, Li T, Lu Z, Dai W, Xu M, and Lu L. 2006a. Effect of CTCF-binding motif on regulation of PAX6 transcription. Invest. Ophthalmol. Vis. Sci. 47:2422-2429. Wu M, Tsai T, and Beaudet AL. 2006b. Deficiency of Rbbp1/Arid4a and Rbbp1l1/Arid4b alters epigenetic modifications and suppresses an imprinting defect in the PWS/AS domain. Genes Dev. 20:2859-2870. Wutz A, Smrzka OW, Schweifer N, Schellander K, Wagner EF, and Barlow DP. 1997. Imprinted expression of the Igf2r gene depends on an intronic CpG island. Nature 389:745-749. Xi H, Shulha HP, Lin JM, Vales TR, Fu Y, Bodine DM, McKay RDG, Chenoweth JG, Tesar PJ, Furey TS, et al. 2007. Identification and characterization of cell type-specific and ubiquitous chromatin regulatory structures in the human genome. PLoS Genet. 3:e136. Xie X, Mikkelsen TS, Gnirke A, Lindblad-Toh K, Kellis M, and Lander ES. 2007. Systematic discovery of regulatory motifs in conserved regions of the human genome, including thousands of CTCF insulator sites. Proc. Natl. Acad. Sci. U. S. A. 104:7145-7150. Xu N, Donohoe ME, Silva SS, and Lee JT. 2007. Evidence that homologous X-chromosome pairing requires transcription and Ctcf protein. Nat. Genet. 39:1390-1396. Yang Y, Quitschke WW, Vostrov AA, and Brewer GJ. 1999. CTCF is essential for up-regulating expression from the amyloid precursor protein promoter during differentiation of primary hippocampal neurons. J. Neurochem. 73:2286-2298. Yao H, Brick K, Evrard Y, Xiao T, Camerini-Otero RD, and Felsenfeld G. 2010. Mediation of CTCF transcriptional insulation by DEAD-box RNA-binding protein p68 and steroid receptor RNA activator SRA. Genes Dev. 24:2543-2555. Yao Y, and Yang W. 2003. The metastasis-associated proteins 1 and 2 form distinct protein complexes with histone deacetylase activity. J. Biol. Chem. 278:42560-42568. Yoon B, Herman H, Hu B, Park YJ, Lindroth A, Bell A, West AG, Chang Y, Stablewski A, Piel JC, et al. 2005. Rasgrf1 imprinting is regulated by a CTCF-dependent methylation-sensitive enhancer blocker. Mol. Cell. Biol. 25:11184-11190. Yoon YS, Jeong S, Rong Q, Park K, Chung JH, and Pfeifer K. 2007. Analysis of the H19ICR insulator. Mol. Cell. Biol. 27:3499-3510. Yuan B, Latek R, Hossbach M, Tuschl T, and Lewitter F. 2004. siRNA Selection Server: an automated siRNA oligonucleotide prediction server. Nucleic Acids Res 32:W130-134. Yusufzai TM, Tagami H, Nakatani Y, and Felsenfeld G. 2004. CTCF tethers an insulator
187
to subnuclear sites, suggesting shared insulator mechanisms across species. Mol. Cell 13:291-298. Zeng F, Baldwin DA, and Schultz RM. 2004. Transcript profiling during preimplantation mouse development. Dev. Biol. 272:483-496. Zhang Y, LeRoy G, Seelig HP, Lane WS, and Reinberg D. 1998. The dermatomyositis-specific autoantigen Mi2 is a component of a complex containing histone deacetylase and nucleosome remodeling activities. Cell 95:279-289. Zhang Y, Ng HH, Erdjument-Bromage H, Tempst P, Bird A, and Reinberg D. 1999. Analysis of the NuRD subunits reveals a histone deacetylase core complex and a connection with DNA methylation. Genes Dev. 13:1924-1935. Zhao H, and Dean A. 2004. An insulator blocks spreading of histone acetylation and interferes with RNA polymerase II transfer between an enhancer and gene. Nucleic Acids Res. 32:4903-4919. Zhao X, Ueba T, Christie BR, Barkho B, McConnell MJ, Nakashima K, Lein ES, Eadie BD, Willhoite AR, Muotri AR, et al. 2003. Mice lacking methyl-CpG binding protein 1 have deficits in adult neurogenesis and hippocampal function. Proc. Natl. Acad. Sci. U. S. A. 100:6777-6782. Zhu D, Fang J, Li Y, and Zhang J. 2009. Mbd3, a component of NuRD/Mi-2 complex, helps maintain pluripotency of mouse embryonic stem cells by repressing trophectoderm differentiation. PLoS ONE 4:e7684. Zwart R, Sleutels F, Wutz A, Schinkel AH, and Barlow DP. 2001. Bidirectional action of the Igf2r imprint control element on upstream and downstream imprinted genes. Genes Dev. 15:2361-2366.
188