Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Methyl-CpG-Binding domain proteins and histone deacetylases in the stage-specific differentiation of… MacDonald, Jessica 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_macdonald_jessica.pdf [ 7.15MB ]
Metadata
JSON: 24-1.0066189.json
JSON-LD: 24-1.0066189-ld.json
RDF/XML (Pretty): 24-1.0066189-rdf.xml
RDF/JSON: 24-1.0066189-rdf.json
Turtle: 24-1.0066189-turtle.txt
N-Triples: 24-1.0066189-rdf-ntriples.txt
Original Record: 24-1.0066189-source.json
Full Text
24-1.0066189-fulltext.txt
Citation
24-1.0066189.ris

Full Text

METHYL-CPG-BINDING DOMAIN PROTEINS AND HISTONE DEACETYLASES IN THE STAGE-SPECIFIC DIFFERENTIATION OF OLFACTORY RECEPTOR NEURONS by JESSICA LINN MACDONALD B.Sc. (Hons), University of Toronto, 2001  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES (Neuroscience)  THE UNIVERSITY OF BRITISH COLUMBIA December 2007  © Jessica Linn MacDonald, 2007  Abstract DNA methylation-dependent gene silencing, catalyzed by DNA methyltransferases (DNMTs) and mediated by methyl binding domain proteins (MBDs) and histone deacetylases (HDACs), is essential for mammalian development, with the nervous system demonstrating particular sensitivity to perturbations. Little is known, however, about the role of DNA methylation in the stage-specific differentiation of neurons. In the olfactory epithelium (OE), where neurogenesis is continuous and the cells demonstrate a laminar organization with a developmental hierarchy, we identified sequential, transitional stages of differentiation likely mediated by different DNMT, MBD and HDAC family members. Biochemically, HDAC1 and HDAC2 associate with repressor complexes recruited by both MBD2 and MeCP2. HDAC1 and HDAC2, however, are divergently expressed in the OE, a pattern that is recapitulated in the brain. Rather than simultaneous inclusion in a complex, therefore, the individual association of HDAC1 or HDAC2 may provide specificity to a repressor complex in different cell types. Furthermore, distinct transitional stages of differentiation are perturbed in the absence of MBD2 or MeCP2. MeCP2 is expressed in the most apical immature olfactory receptor neurons (ORNs), and is up-regulated with neuronal maturation. In the MeCP2 null OE there is a transient delay in ORN maturation and an increase in neurons of an intermediate developmental stage. Two protein variants of MBD2 are expressed in the OE, with MBD2b expressed in cycling proge nitor cells and MBD2a in the maturing ORNs. MBD2 null ORNs undergo increased apoptotic cell death. There is also a significant increase in proliferating progenitors in the MBD2 null OE, likely due, at least in part, to feedback from the dying ORNs, acting to up-regulate neurogenesis. Increased cell cycling in the MBD2 null is also observed post- lesion, however, in the absence of feedback back from the ORNs, a phenotype that is recapitulated by an acute inhibition of HDACs with valproic acid.  ii  Therefore, disruptions at both transitional stages of ORN differentiation are likely in the MBD2 null mouse. Together, these results provide the first evidence for a sequential recruitment of different MBD proteins and repressor complexes at distinct transitional stages of neuronal differentiation.  iii  Table of Contents Abstract ............................................................................................................................................ii Table of contents.............................................................................................................................iv List of tables ..................................................................................................................................vii List of figures................................................................................................................................viii List of abbreviations ........................................................................................................................x Acknowledgments ..........................................................................................................................xi Chapter 1: Literature Review and Introduction 1.1 Epigenetics ................................................................................................................................1 1.1.1 Chromatin and histone modifications ...............................................................................1 1.1.2 DNA methylation..............................................................................................................3 1.1.3 Methyl binding domain proteins.......................................................................................5 1.1.4 Histone deacetylases .........................................................................................................9 1.2 DNA methylation-dependent gene silencing in the nervous system.......................................10 1.2.1 DNMTs in neural development ......................................................................................10 1.2.2 MeCP2 ............................................................................................................................11 1.2.3 MBD1 .............................................................................................................................14 1.2.4 MBD2 .............................................................................................................................15 1.2.5 MBD3 .............................................................................................................................15 1.2.6 DNMTs and MBDs in the neural response to injury......................................................16 1.2.7 DNA methylation and synaptic plasticity.......................................................................17 1.3 Histone Deacetylase Inhibitors ................................................................................................18 1.3.1 Valproic acid...................................................................................................................19 1.3.2 HDAC inhibitors in the nervous system.........................................................................21 1.4 The olfactory system ...............................................................................................................22 1.4.1 The cell types of the olfactory epithelium ......................................................................22 1.4.2 Odorant receptors............................................................................................................25 1.4.3 Interactions between the olfactory bulb and olfactory epithelium .................................25 1.5 Introduction and project summary...........................................................................................29 Chapter 2: Materials and Methods 2.1 Mice 31 2.1.2 MBD2 null mice.............................................................................................................31 2.2.2 MeCP2 null mice ............................................................................................................31 2.2 Olfactory bulbectomies............................................................................................................32 2.3 BrdU labelling .........................................................................................................................32 2.4 Tissue preparation....................................................................................................................32 2.5 Immunofluorescence................................................................................................................33 2.6 Immunohistochemistry ............................................................................................................34 2.7 ß-galactosidase staining ...........................................................................................................34 2.8 Reverse transcription PCR.......................................................................................................35 2.9 In situ hybridization.................................................................................................................35 2.9.1 Probe preparation............................................................................................................35  iv  2.9.2 In sit u hybridization........................................................................................................36 2.10 Immunoprecipitatio n..............................................................................................................37 2.11 SDS-PAGE and western blotting...........................................................................................38 2.12 Cell counts and measurements...............................................................................................38 2.12.1 Developmental cell counts ...........................................................................................38 2.12.2 Post-bulbectomy IdU and CldU cell counts .................................................................39 2.12.3 Glomerular measurements ............................................................................................39 2.13 Image analysis .......................................................................................................................40 Chapter 3: Stage-Specific Expression of Methyl-CpG-Binding Domain Proteins and Histone Deacetylases During Olfactory Neurogenesis 3.1 Introduction..............................................................................................................................46 3.2 Results......................................................................................................................................49 3.2.1 All MBD proteins are expressed in the developing OE.................................................49 3.2.2 Multiple isoforms of MBD2 are detected in the developing OE....................................49 3.2.3 MBD2a, MBD2b and MeCP2 are sequentially expressed at distinct stages of ORN differentiation........................................................................................................52 3.2.4 HDAC1 and HDAC2 display divergent, sequential expression patterns during the development of ORNs ..............................................................................................56 3.2.5 MBD2 interacts with both HDAC1 and HDAC2 in the postnatal OE...........................59 3.3 Discussion................................................................................................................................62 Chapter 4: HDAC1 and HDAC2 Are Divergently Expressed in Distinct Developmental Stages and Lineages in the Developing and Adult Mouse Brain 4.1 Introduction..............................................................................................................................68 4.2 Results......................................................................................................................................69 4.2.1 HDAC1 is expressed in progenitors while HDAC2 is expressed in post- mitotic neurons in the E13.5 CNS ..............................................................................................69 4.2.2 HDAC1 is expressed in glia and progenitors and HDAC2 is expressed in neurons in the postnatal day 7 brain ............................................................................................71 4.2.3 HDAC1 expression is primarily glial while HDAC2 is expressed in neurons throughout the adult brain ..............................................................................................77 4.2.4 MeCP2 is expressed in mature neurons throughout the postnatal day 7 brain while MBD2 displays a more restricted ne uronal expression..................................................80 4.3 Discussion................................................................................................................................86 Chapter 5: MBD2 and MeCP2 Null Mice Display Stage-Specific Defects in Olfactory Neurogenesis 5.1 Introduction..............................................................................................................................91 5.2 Results......................................................................................................................................92 5.2.1 MBD2 and MeCP2 null mice display stage-specific perturbations in ORN differentiation, corresponding to their observed expression patterns ............................92 5.2.2 Decreased retention of BrdU labelled cells in the epithelium of the MBD2 null ..........98 5.2.3 MBD2 null ORNs display an increased rate of apoptotic cell death............................100 5.2.4 Aberrant glomerular formation in the absence of either MBD2 or MeCP2.................103 5.3 Discussion..............................................................................................................................105  v  Chapter 6: Progenitor Proliferation Is Increased in the MBD2 Null OE and Following an Acute Perturbation of Histone Deacetylation with Valproic Acid 6.1 Introduction............................................................................................................................115 6.2 Results....................................................................................................................................116 6.2.1 Labelling cycling basal cells with iododeoxyuridine and chlorodeoxyuridine ............116 6.2.2 Validation of the bulbectomy lesion model..................................................................117 6.2.3 The number of IdU and CldU labelled cells is increased in the unlesioned MBD2 null and VPA-treated OE .............................................................................................122 6.2.4 Increased proliferation and cell cycle re-entry in the MBD2 null and VPA-treated lesioned olfactory epithelium.......................................................................................122 6.3 Discussion..............................................................................................................................125 Chapter 7: Summary and Future Directions 7.1 Summary of results and conclusions .....................................................................................129 7.2 Final conclusions and general discussion..............................................................................132 7.3 Future directions ....................................................................................................................136 7.3.1 Identification of repressor complexes containing HDAC1 and HDAC2 during neuronal differentiation................................................................................................136 7.3.2 Candidate genes for MBD2 and MeCP2 mediated repression.....................................137 7.3.3 Global analysis of MBD2 and MeCP2 target genes.....................................................140 References...................................................................................................................................142 Appendix.....................................................................................................................................173  vi  List of Tables Table 2.1 Table 2.2 Table 2.3  Primer Sequences .....................................................................................................41 Primary Antibody Dilutions and Suppliers ..............................................................42 Validation of MBD and HDAC Antibodies.............................................................43  vii  List of Figures Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 Figure 1.5 Figure 2.1 Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.8 Figure 3.9 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Figure 4.8 Figure 5.1 Figure 5.2 Figure 5.3 Figure 5.4 Figure 5.5 Figure 5.6 Figure 5.7 Figure 5.8 Figure 5.9  Mechanisms of DNA methylation-dependent gene silencing................................... 2 Methyl CpG binding domain proteins....................................................................... 6 Repressor complexes recruited by MBD proteins ..................................................... 7 Laminar structure of the postnatal olfactory epithelium ......................................... 24 Synaptic pathways of the olfactory system............................................................. 27 Turbinates of the adult olfactory epithelium........................................................... 45 DNMT3b is expressed in olfactory progenitors while DNMT3a is expressed in post-mitotic ORNs................................................................................................... 47 Transcripts of all MBD family members are detected in the developing OE......... 50 Multiple isoforms of MBD2 are detected in the developing OE ............................ 51 MBD2a, MBD2b and MeCP2 are sequentially expressed at distinct stages of ORN differentiation at postnatal day 7 ................................................................... 55 Sub-cellular distribution of MeCP2 overlaps with 5- methyl cytosine .................... 57 HDAC2 is expressed in a subpopulation of DNMT3b positive basal cells and DNMT3a positive immature ORNs in the developing OE..................................... 58 HDAC1 and HDAC2 are exp ressed at divergent developmental stages in the postnatal day 7 olfactory epithelium ....................................................................... 60 MBD2a interacts with both HDAC1 and HDAC2 in the P7 OE ............................ 61 Summary of expression of DNMTs, MBDs and HDACs in the OE ...................... 67 HDAC1 is expressed in progenitors in the E13.5 brain .......................................... 70 HDAC2 is expressed in post- mitotic neurons in the E13.5 brain ........................... 72 HDAC1 is primarily expressed in glial cells in the postnatal day 7 brain .............. 74 HDAC2 is expressed in neurons throughout the postnatal day 7 brain .................. 76 HDAC1 is expressed in glia, progenitors and some neurons in the adult brain...... 79 HDAC2 is expressed in neurons throughout the adult brain................................... 82 MeCP2 is expressed in neurons throughout the postnatal day 7 brain while MBD2 demonstrates a more restricted neuronal expression................................... 85 Summary of HDAC1 and HDAC2 expression in the developing nervous System..................................................................................................................... 89 Decreased body weight of MBD2 and MeCP2 null mice ....................................... 93 Adult male MBD2 and MeCP2 null mice display stage-specific perturbations in olfactory neurogenesis ........................................................................................ 94 Increase in GAP43 positive neurons and disrupted laminar structure in the postnatal MeCP2 null olfactory epithelium ............................................................ 96 A significant increase in cycling progenitors and decrease in mature ORNs is not observed until 7 weeks of age in the MBD2 null OE ....................................... 97 An antibody against PCNA labels more proliferating cells in the postnatal day 7 OE than are detected by BrdU incorporation....................................................... 99 Decreased retention of BrdU labelled cells in the MBD2 null OE ....................... 101 Increased apoptotic cell death of mature ORNs in the MBD2 null OE ................ 102 MBD2 and MeCP2 null mice display aberrant glomerular formation in the postnatal day 21 olfactory bulb ............................................................................. 104 Disrupted glomerular formation in the postnatal day 7 MBD2 null OB............... 106  viii  Figure 5.10 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Figure 7.1  Summary of MBD2 and MeCP2 null olfactory phenotypes................................. 114 Detection of IdU and CldU in the postnatal olfactory epithelium ........................ 118 A varying extent of medial sparing following unilateral bulbectomy .................. 119 The number of IdU and CldU labelled cells is increased in the lesioned OB....... 121 Increased incorporation of IdU and CldU in the unlesioned MBD2 null and VPA-treated OE .................................................................................................... 123 Increased proliferation and cell cycle re-entry in the lesioned MBD2 null and VPA treated OE..................................................................................................... 124 Proposed model for the role of MBD proteins and HDACs in the stage-specific differentiation of ORNs......................................................................................... 134  ix  List of Abbreviations 5-MeC aC3 ß-gal BDNF BrdU CGN CldU CNS CpG DNMT E GABA GAP43 GBC GFAP HAT HBC HDAC IdU IRN KO LP MBD MIZF NST NuRD OB OE OEC OMP OR ORN P PBx PCNA Rb RE1 REST SGZ Sus TRD TSA VPA  5-Methylated Cytosine activated Caspase 3 ß-galactosidase Brain Derived Neurotrophic Factor Bromodeoxyuridine Cerebellar Granule Neurons Chlorodeoxyuridine Central Nervous System Cytosine-phosphate-guanosine dinucleotide DNA Methyltransferase Embryonic day ?-Aminobutyric Acid Growth Associated Protein 43 Globose Basal Cell Glial Fibrillary Acid Protein Histone Acetyltransferase Horizontal Basal Cell Histone Deacetylase Iododeoxyuridine Immature Receptor Neuron Knockout (genetic deletion) Lamina Propria Methyl-CpG-Binding Domain MBD2-binding Zinc Finger Neuron Specific ßIII Tubulin Nucleosome Remodelling and histone Deacetylase Olfactory Bulb Olfactory Epithelium Olfactory Ensheathing Cell Olfactory Marker Protein Odorant Receptor Olfactory Receptor Neuron Postnatal day Post-Bulbectomy Proliferating Cell Nuclear Antigen Retinoblastoma Protein Repressor Element 1 RE1 Silencing Transcription factor Subgranular Zone Sustentacular cell Transcriptional Repression Domain Trichostatin A Valproic Acid  x  Acknowledgements  I would like to thank the many members of the Roskams lab, past and present, that I have had the pleasure of working with over the years. Your endless support and encouragement, scientifically and otherwise, has truly made for a fun and intellectually stimulating work environment. I would also like to thank my supervisor, Dr. Jane Roskams, for encouraging me to set the highest standards possible for myself and pushing me to succeed.  I would also like to acknowledge my family, without whose encouragement and support I would not have made it this far. Yo ur unwavering belief in me has often given me the courage to test my limits and to not accept the defeats I have encountered along the way. And I would like to thank Scott Neal, for always listening, always supporting me and, probably most importantly, always making me laugh.  I would also like to take this opportunity to acknowledge those who have supplied reagents or funding, without which I could not have completed this work. Thanks go to Dr. Frank Margolis, University of Maryland, for his gift OMP antibody, and to Dr. Adrian Bird and Dr. Brian Hendrich, University of Edinburgh, for their generous gift of the MBD2 mice. I would also like to thank the following organizations who have provided me with fellowships: The Cordula and Gunter Paetzold Graduate Fellowship, Natural Sciences and Engineering Research Council, and the Canadian Institutes of Health Research.  “We must have exceptional results by tomorrow” - Bruce Banner, Hulk (Universal Pictures, 2003)  xi  Chapter 1: Literature Review and Introduction 1.1 Epigenetics  The term ‘epigenetics’ was coined to describe events that could not be explained by genetic principles, defined by Conrad Waddington in 1942 as “the branch of biology which studies the causal interactions between genes and their products, which bring the phenotype into being” (Waddington, 1942). As such, cellular differentiation can be considered an epigenetic phenomenon. With a few limited exceptions, all cells in a multicellular organism have an identical genotype. However, development produces a wide range of cell types with divergent cellular functions. This diversity is generated by distinct gene expression profiles, representing both developmental stage-specific and lineage-specific gene activation and gene silencing, rather than changes in DNA sequence. A growing field of research aims to understand the epigenetic regulation of gene expression, under the current definition of epigenetics as “the study of mitotically and/or meiotically heritable changes in gene function that cannot be explained by changes in DNA sequence (Russo et al., 1996).  1.1.1 Chromatin and Histone Modifications  Research on epigenetics has converged on covalent and non-covalent modifications of DNA and histone proteins and the effects such modifications exert on the overall structure of chromatin (Goldberg et al., 2007). Chromatin refers to the state in which DNA is packaged within the cell. The fundamental unit of chromatin is the nucleosome, composed of 147 base pairs of DNA wrapped around an octamer of the four core histone proteins (H3, H4, H2A and H2B). The core histones are predominantly globular in structure, with the exception of their N-terminal ‘tails’ which are unstructured (see Figure 1.1A) (Luger et al., 1997; Strahl and Allis, 2000; Wade, 2001). Histones, and particularly their tails, possess a surprisingly al rge number and type of modifiable residues. There are at least eight distinct types of modifications found on histones (acetylation, lysine methylation, arginine methylation, phosphorylation, ubiquitylation, sumoylation, ADP ribosylation, deimination and proline isomerisation) and over 60 different residues on histones where modifications have been detected (Kouzarides, 2007).  1  Figure 1.1: Mechanisms of DNA Methylation-Dependent Gene Silencing (A) The basic unit of chromatin is the nucleosome, composed of 147 bases of DNA wrapped around an octamer of core histone proteins (H2A, H2B, H3 and H4). The histones are globular in shape, with the exception of their N-terminal tails. (B) DNA methylation mediates gene silencing through two distinct mechanisms. De novo DNA methyltransferases, DNMT3a and DNMT3b, catalyze the methylation of cytosine residues in CpG dinucleotides. The methylated cytosines (red hexagons) can either sterically hinder transcription factors (TF) from binding to the promoter of a gene, or the methylated cytosines can be bound by methyl binding domain proteins (MBD) which recruit histone deacetylase (HDAC) containing repressor complexes, and lead to chromatin condensation.  2  Histone modifications employ two primary mechanisms to influence patterns of gene expression. The first mechanism is to modify higher-order chromatin structure by affecting the contact between histones in adjacent nucleosomes or the interaction of histones with DNA. The second mechanism is an effecter- mediated function, acting to recruit, block or stabilize the localization of non-histone proteins that carry enzymatic activities that further modify chromatin structure (Goldberg et al., 2007; Kouzarides, 2007). In general, histone modifications can be divided into those that correlate with gene activation and those that correlate with repression (Jenuwein and Allis, 2001). Recently, however, bivalent domains possessing both activating and repressive modifications have been reported (Bernstein et al., 2005; Azuara et al., 2006; Berns tein et al., 2006). In addition, a given modification may have the potential to activate or repress under different conditions, including its localization in the coding region versus the promoter (Vakoc et al., 2005; Kouzarides, 2007). While these findings and others call into question the view of a strict histone code defining transcriptionally active and silent chromatin states, it is becoming increasingly clear that significant cross-talk occurs between distinct histone modifications, as well as DNA methylation, to modify chromatin structure and regulate gene expression (Fuks, 2005; Bernstein et al., 2007; Goldberg et al., 2007; Kouzarides, 2007).  1.1.2 DNA Methylation  DNA methylation, one of the best characterized epigenetic modifications, is a stable and heritable component of epigenetic regulation and a chief contributor to the stability of gene expression states (Jaenisch and Bird, 2003). DNA methylation occurs through the covalent addition of a methyl group at the 5 position of the pyrimidine ring of cytosine residues. In vertebrates, this modification occur s almost exclusively within cytosine-phosphate-guanosine (CpG) dinucleotides (Ramsahoye et al., 2000; Bird, 2002; Goll and Bestor, 2005). DNA methylation is catalyzed by DNA methyltransferases, or DNMTs, which transfer methyl groups from S-adenosy-L- methionine to the cytosine residue (Bird, 1992). There are two functionally different classes of DNMTs – maintenance methyltransferases and de novo methyltransferases. DNMT1, the primary maintenance methyltransferase, associates with the replication fork, targeting hemi- methylated CpGs and preserving methylation patterns in daughter cells (Yoder et  3  al., 1997; Pradhan et al., 1999). The de novo methyltransferases, DNMT3a and DNMT3b, target unmethylated CpGs, establishing new methylation patterns (Okano et al., 1999).  In human somatic cells, methylated cytosines account for approximately 1% of total DNA bases, therefore affecting 70-80% of all CpG dinucleotides in the genome (Ehrlich et al., 1982). The extent and pattern of DNA methylation is, however, dynamic during mammalian development. Within hours of fertilization, the paternal genome is actively demethylated while the maternal genome is passively demethylated during subsequent cleavage divisions. This wave of demethylation is followed by genome wide de novo methylation following implantation (Mayer et al., 2000; Oswald et al., 2000; Li, 2002; Santos et al., 2002). DNA methylation is essential for mammalian development to proceed, as evidenced by the lethality of DNMT knockout mice (Li et al., 1992; Okano et al., 1999), and remains indispensable for the survival of differentiated cells (Jackson-Grusby et al., 2001). This absolute requirement for DNA methylation likely reflects the diverse range of cellular functions and pathologies in which it has been implicated, including silencing of repetitive and centromeric sequences, tissue-specific gene expression, genomic imprinting, maintenance of X chromosome inactivation, carcinogenesis and aging (Paulsen and Ferguson-Smith, 2001; Bird, 2002; Jones and Baylin, 2002; Jaenisch and Bird, 2003).  The added methyl group on methyl-CpGs does not affect the base pairing itself, but the methyl group protrudes into the major groove of the DNA helix and can affect DNA-protein interactions (Razin and Riggs, 1980). Mechanistically, the methylated cytosine can repress transcription by either precluding or promoting recruitment of regulatory proteins (summarized in Figure 1.1B) (Bird, 2002). The former case is a direct mechanism, whereby methylated cytosines interfere with the binding of a protein, such as a transcription factor, to its target DNA sequence, as has been shown for the binding of CTCF to the H19 locus (Hark et al., 2000). The second mode of repression is opposite to the first in that it requires the specific binding of regulatory proteins to the methylated cytosine. A family of Methyl-CpG-binding proteins bind to methylated DNA and mediate transcriptional repression through the recruitment of repressor complexes, generally containing histone deacetylases (reviewed in (Nan et al., 1998a; Bird, 2002).  4  1.1.3 Methyl Binding Domain Proteins  The first methyl-CpG binding protein to be cloned was actually the second methyl-CpG binding activity to be discovered and, as such, it was named MeCP2 (Lewis et al., 1992). The methylCpG-binding domain (MBD) was dissected from the multidomain MeCP2 protein in deletion studies (Nan et al., 1993) and was used to search EST clones for other possible methyl-CpG binding proteins. Four more proteins containing the conserved MBD were identified, and named MBD1 – MBD4 (Hendrich and Bird, 1998). Similarity between these proteins is limited to the MBD, with the exception of MBD2 and MBD3 (see Figure 1.2). These two MBD family members show an approximate sequence identity of 70% through the C terminus and are believed to have arisen through a gene duplication event (Hendrich and Bird, 1998; Hendrich and Tweedie, 2003). Four of the MBD protein family members interact with HDAC-associated repressor complexes and are involved in transcriptional repression. The exception is MBD4, which appears to function in DNA repair rather than transcriptional silencing (Hendrich et al., 1999). A novel methyl binding protein, Kaiso, has also been identified. Kaiso lacks the conserved MBD but recognizes methylated cytosines through zinc finger domains and mediates repression by associating with the histone deacetylase containing N-CoR co-repressor complex (Prokhortchouk et al., 2001; Yoon et al., 2003).  The other atypical member of the MBD protein family is MBD3. MBD3 has two amino acid substitutions within its MBD that prevent it from binding directly to DNA, methylated or unmethylated (Hendrich and Bird, 1998; Saito and Ishikawa, 2002; Fraga et al., 2003). MBD3 is, however, a core component of the Mi2/NuRD (Nucleosome Remodelling and Histone Deacetylase) repressor complex, which also contains Mi2, a member of the SWI2/SNF2 family of ATP-dependent chromatin remodelling proteins, and the histone deacetylases 1 and 2 (see Figure 1.3) (Wade et al., 1999; Zhang et al., 1999; Vignali et al., 2000). Mi2/NuRD is an abundant co-repressor complex that can be recruited to DNA by several different repressor proteins (Knoepfler and Eisenman, 1999; Ahringer, 2000), including to methylated DNA through MBD2 (Zhang et al., 1999). MBD2 is a component of the methyl-CpG binding protein 1 (MeCP1; Figure 1.3) complex (Ng et al., 1999), the first methyl CpG binding activity to be described and defined functionally as an activity that could bind methylated DNA in solution  5  Figure 1.2: Methyl-CpG-Binding Domain Proteins There are 5 classic Methyl-CpG-Binding Domain (MBD) protein family members, identified by a conserved MBD (blue). With the exception of MBD2 and MBD3, similarity between the MBD proteins is restricted to the MBD. MBD2 and MBD3 share a 70% sequence identity through the MBD and the Cterminus, however MBD3 has two amino acid substitutions in its MBD that prevent it from binding to DNA. Two variants of MBD2 arise from an alternative start codon, generating a full-length MBD2a isoform and an N-terminal truncated MBD2b isoform. The main domains of each protein are indicated, including the common transcriptional repression domain (TRD) and the zinc binding domains (CxxC) of MBD1.  6  Figure 1.3: Repressor complexes recruited by MBD proteins MBD3 is a core component of the Mi2/NuRD repressor complex, which also contains HDAC1 and HDAC2. The MeCP1 repressor complex is composed of the Mi2/NuRD complex plus MBD2. MeCP2 recruits the Sin3A repressor complex, which also contains HDAC1 and HDAC2.  7  (Meehan et al., 1989). Since then, biochemical studies have shown MeCP1 to be a multiprotein complex composed of 10 major polypeptides, including MBD2 and all of the components of the Mi2/NuRD complex (Feng and Zhang, 2001). Only a small fraction of the total Mi2/NuRD activity is detected in the MeCP1 complex (Zhang et al., 1999), however, while a small fraction of MBD2 is also detected in a complex of 100 kD, distinct from the much larger (1 MD) MeCP1 complex (Feng and Zhang, 2001).  Unlike MeCP1, MeCP2 is composed of a single polypeptide with two functional domains: an 80 amino acid MBD that is necessary and sufficient to target MeCP2 to methylated CpGs (Nan et al., 1993; Nan et al., 1996) and a transcriptional repression domain (TRD) (Nan et al., 1997). While MeCP2 was originally believed capable of binding to a single symmetrically methylated CpG in vitro through its MBD (Lewis et al., 1992), more recent evidence suggests that MeCP2 requires an A/T run of four or more bases ([A/T]=4) adjacent to the methyl-CpG for efficient DNA binding (Klose et al., 2005). Through the TRD, MeCP2 interacts with the Sin3 repressor complex (Figure 1.3), containing HDAC1 and HDAC2, to repress transcription (Jones et al., 1998; Nan et al., 1998b). MeCP2 also associates with histone H3 lysine 9 methylation activity (Fuks et al., 2003) and the Co-Rest repressor complex (Lunyak et al., 2002) which associates with REST (RE1 silencing transcription factor) to silence neuronal genes in non- neuronal cells (Lunyak et al., 2002; Lunyak et al., 2004; Roopra et al., 2004). A novel role for MeCP2 in the regulation of RNA splicing through an interaction with the RNA-binding protein Y box-binding protein 1 has also recently been proposed (Young et al., 2005).  Like MeCP2, MBD1 contains a distinct TRD which represses transcription through the recruitment of histone deacetylases (Fujita et al., 1999; Fujita et al., 2000; Ng et al., 2000). However, unlike MeCP1 and MeCP2, this activity is not dependent on HDAC1 and HDAC2, and the identity of the MBD1 HDAC is not established (Ng et al., 2000). MBD1 is, however, unique among MBD proteins in its ability to repress transcription from both methylated and unmethylated promoters (Fujita et al., 1999; Fujita et al., 2000; Jorgensen et al., 2004). MBD1 interacts with unmethylated DNA and represses transcription through one of its three zinccoordinating CXXC domains (Fujita et al., 2000; Jorgensen et al., 2004). The cysteine-rich  8  CXXC motif is also found in other chromatin associated proteins, including DNMT1 and the histone methyltransferase MLL (Cross et al., 1997).  1.1.4 Histone Deacetylases  HDACs regulate gene expression by reversibly removing the acetyl groups of select lysine residues in the conserved tails of core histone proteins. In general, increased histone acetylation (hyperacetylation) is associated with increased transcriptional activity, whereas decreased acetylation (deacetylation or hypoacetylation) is associated with repression of gene expression (Strahl and Allis, 2000; Wade, 2001). The repressive effects of histone deacetylation are, at least in part, due to its ability to unmask the basic charge of the lysine, leading to higher affinity histone/DNA interactions and diminished accessibility to transcription factors (Hong et al., 1993; Strahl and Allis, 2000; Wade, 2001). The conserved catalytic domain of HDACs is formed by a stretch of approximately 390 amino acids that form a curved tubular pocket with a wider bottom consisting of hydrophobic walls and a Zn2+ cation at the base. An acetylated lysine residue fits in the pocket, where Zn2+ catalyzes the hydrolysis of the acetyl group (Finnin et al., 1999). There are two main protein families with histone deacetylase activity: the Sir2 family of NAD+dependent HDACs and the classical HDAC family (de Ruijter et al., 2003). The classical HDAC family can be further divided into two phylogenetic classes, class I and class II (Bjerling et al., 2002; Fischle et al., 2002). Class I, including HDAC1, 2, 3 and 8, is composed of small proteins (377-488 amino acids) that are most closely related to the yeast transcription regulator RPD3 (Bjerling et al., 2002). Class II includes HDAC4, 5, 6, 7 and 9, proteins of a larger size (6691215 amino acids) that share sequence homology with HDA1, another deacetylase found in yeast (Bjerling et al., 2002; Fischle et al., 2002). Class I HDACs are found almost exclusively within the nucleus, while Class II HDACs are able to shuttle in and out of the nucleus in response to certain cellular signals (de Ruijter et al., 2003). The class I HDACs 1 and 2 are currently two of the best studied family members. They are highly similar proteins with an overall sequence identity of 82% and they are often found in complexes together (Khochbin and Wolffe, 1997; de Ruijter et al., 2003), including the Sin3, Mi2/NuRD, and by extension MeCP1, and Co-Rest  9  repressor complexes (Jones et al., 1998; Nan et al., 1998b; Wade et al., 1999; Zhang et al., 1999; Ballas et al., 2001).  1.2 DNA Methylation-Dependent Gene Silencing in the Nervous System  Disruption of epigenetic mechanisms have specifically been implicated in a number of syndromes associated with mental retardation, including Rett, ICF, Fragile-X, ATRX, Rubinstein-Taybi and Angelman (reviewed in (Egger et al., 2004)), as well as complex psychiatric disorders such as schizophrenia (Veldic et al., 2004; Grayson et al., 2005; Veldic et al., 2005). Why the development and function of the nervous system is so particularly sensitive to disruptions in epigenetic regulation is still being elucidated. The nervous system is unique, however, in the remarkably complex patterning of the brain that must be established, while still maintaining a large degree of plasticity in order for the brain to respond to a changing environment. Epigenetic modifications could provide a coordinated system for regulating gene expression at each stage of neurogenesis, from fate specification through synaptic connectivity and plasticity (Hsieh and Gage, 2005).  1.2.1 DNMTs in Neural Development  The overall level of DNA methylation undergoes dynamic changes perinatally in the mo use brain (Tawa et al., 1990), with the level of DNA methylation higher in the adult brain than in other tissues (Wilson et al., 1987; Tawa et al., 1990; Ono et al., 1993). The maintenance methyltransferase DNMT1 is highly expressed within the CNS, not only in the nucleus of cycling cells, but also cytoplasmically in differentiated neurons of the CNS (Goto et al., 1994; Brooks et al., 1996). Conditional deletion of Dnmt1 in the CNS is perinatal lethal, while mosaic animals containing 30% Dnmt1-depleted CNS cells are viable into adulthood. However, the Dnmt1 deficient cells are eliminated from the brain within the first 3 weeks postnatal, indicating that they are not able to functionally mature (Fan et al., 2001). Interestingly, conditional deletion of Dnmt1 in a sub-population of post- mitotic neurons of the postnatal brain had no effect on cell survival (Fan et al., 2001). These results suggest that DNMT1 is essential during the maturation  10  of a neuron, but may not be required for the survival of a mature neuron, at least in the steadystate.  DNA methylation is also a critical cell intrinsic determinant of astrocyte differentiation. The promoter of glial fibrillary acid protein (Gfap), a marker involved in astrocyte differentiation, is methylated at early, neurogenic stages of embryonic development, but demethylated at later stages, as progenitors differentiate into astroglia (Teter et al., 1996). Methylation of the STAT binding element within the Gfap promoter inhibits association of activated STATs with the glial promoter, thereby repressing transcription of Gfap and preventing cells from proceeding down an astroglial lineage during the neurogenic stages of the brain development (Takizawa et al., 2001). DNA hypomethylation within progenitor cells, through the conditional deletion of Dnmt1, results in precocious astroglial differentiation, both in vivo and in vitro, appearing to result from the accelerated developmental demethylation of Gfap, as well as genes encoding the crucial components of the gliogenic JAK-STAT pathway (Fan et al., 2005).  1.2.2 MeCP2  Despite the largely ubiquitous expression of the MBD proteins (Hendrich and Bird, 1998), the phenotype observed when one is absent or mutated appears to be largely neuronal. The most striking example of this is MeCP2. Mutations or genetic rearrangements within the X- linked Mecp2 gene are considered the primary genetic cause of Rett Syndrome (Amir et al., 1999), a progressive neurodevelopmental disorder and one of the most common causes of mental retardation in females, with an incidence of 1 in 10 000 – 15 000 (Hagberg, 1985). Mutations in Mecp2 have also been implicated in cases of non-specific X- linked mental retardation (Meloni et al., 2000; Orrico et al., 2000; Van Esch et al., 2005) as well as Angelman syndrome (Watson et al., 2001). Patients with classic Rett Syndrome appear to develop normally from birth until about 6-18 months of age, at which point development regresses with a loss of speech and purposeful hand movements and the development of microcephaly, autistic features and severe mental retardation, ataxia and apraxia, irregular breathing and hyperventilation, stereotypic hand movements and possible onset of seizures (Rett, 1966; Hagberg et al., 1983). After this initial regression, the condition stabilizes and patients usually survive into adulthood.  11  Neuropathological studies of Rett patients demonstrate normal gross brain morphology with no evidence of abnormal degeneration or neuronal migration. However, microcephaly may be apparent at birth, becoming definite by early childhood (Jellinger et al., 1988; Leonard and Bower, 1998). The reduction in brain size is attributed primarily to a decrease in the size of individual neurons, increased cell packing density and a reduc tion in dendritic complexity with abnormal synaptic morphology (Armstrong et al., 1995; Bauman et al., 1995; Belichenko et al., 1997). Together, these observations suggest that the critical period of postnatal synaptic refinement may be disrupted by the loss of MeCP2 (Armstrong et al., 1995; Belichenko et al., 1997). The severity of the Rett phenotype is, however, variable, depending on the type and position of the Mecp2 mutation, the level of MeCP2 expression, and the relative expression of the disrupted allele based on X-inactivation mosaicism (Amir et al., 2000; Hammer et al., 2003; Miltenberger-Miltenyi and Laccone, 2003; Schanen et al., 2004; Charman et al., 2005; Jian et al., 2005).  Deletion of the murine Mecp2 gene (also X- linked) recapitulates many of the symptoms of the human disorder (Chen et al., 2001; Guy et al., 2001). However, male hemizygous Mecp2 null mice appear to better recapitulate the clinical progression of Rett Syndrome than heterozygous females (Chen et al., 2001; Guy et al., 2001; Shahbazian et al., 2002a). Mecp2 null male mice display no initial phenotype, but by three to eight weeks of age they develop a stiff, uncoordinated gait, exhibit reduced spontaneous activity, hindlimb clasping, irregular breathing and weight loss with most animals dying by 10 weeks of age (Chen et al., 2001; Guy et al., 2001). Mecp2 heterozygous females are viable and fertile and appear phenotypically normal into adulthood. However, at approximately 6 months of age they begin to exhibit neurological symptoms, including ataxic gait, reduced activity and tremors, in addition to weight gain (Chen et al., 2001; Guy et al., 2001). This correlation of time of onset of symptoms in mouse and human female heterozygotes is unexpected based on the vast difference in developmental stage at the age of six months. Mice with a truncated form of MeCP2 also develop a Rett phenotype, albeit at a slightly delayed time course relative to the complete Mecp2 null mice (Shahbazian et al., 2002a). An engineered truncation following codon 308 of Mecp2 (MeCP2308 ) retains the MBD, TRD and nuclear localization signal, but eliminates approximately one third of the C-  12  terminus (Shahbazian et al., 2002a), mimicking mutations observed in classic Rett patients (Shahbazian and Zoghbi, 2001). MeCP2308/y male mice appear phenotypically normal for at least 6 weeks, but then develop progressive neurological defects reminiscent of Rett syndrome, including tremors, seizures, hypoactivity, stereo-typed forelimb motions, kyphosis and motor dysfunction. Approximately 90% of MeCP2308/y mice do, however, survive to one year of age (Shahbazian et al., 2002a).  Mice in which Mecp2 is deleted only in nestin-positive precursors (and therefore their progeny – neurons and glia) are indistinguishable from complete Mecp2 nulls (Chen et al., 2001; Guy et al., 2001), indicating that this complex phenotype is primarily, if not completely, due to MeCP2 deletion within the nervous system. In fact, deleting Mecp2 only in post- mitotic neurons of the forebrain, hippocampus and brainstem using a Cam kinase Cre transgene leads to a phenotype that is similar, albeit delayed and less severe, than that seen in animals with a germline or NestinCre mediated deletion of the gene (Chen et al., 2001; Gemelli et al., 2006). Taken together, the Rett phenotype and Mecp2 conditional null experiments provide strong evidence that MeCP2 is involved in the maturation and/or maintenance of mature neurons, rather than in earlier stages of neurogenesis, such as cell fate decisions and neuronal migration. In fact, expression of MeCP2 is not detected within glial cells or neuroblasts in vivo, its expression within neurons progressively increasing with maturation (Shahbazian et al., 2002b; Jung et al., 2003b; Kishi and Macklis, 2004; Mullaney et al., 2004). Furthermore, MeCP2 null embryonic neural precursors proliferate and differentiate into neurons and glia in a manner indistinguishable from wildtype (Kishi and Macklis, 2004; Smr t et al., 2007). Interestingly, this is in contrast to findings in the frog Xenopus laevis where xMeCP2 was found to be critical for embryonic development (Stancheva et al., 2003). This involvement of MeCP2 only at later stages of mammalian neurogenesis likely explains the late onset of Rett Syndrome. However, the question remains as to exactly what role MeCP2 plays in the maturation or maintenance of neurons.  Whether the Rett phenotype is the result of global transcriptional deregulation or disruption in specific target genes with neuronal function is also a key question in understanding the molecular pathogenesis of Mecp2 disruption. A number of studies have been undertaken to examine global gene expression changes in the absence of MeCP2, including Mecp2-deficient mouse brains  13  (Tudor et al., 2002), Rett patient cell lines (Traynor et al., 2002), and post-mortem Rett brain tissue (Colantuoni et al., 2001). However, only small and non-reproducible differences in gene expression have been observed, failing to produce a unifying molecular picture of MeCP2-null pathogenesis (Caballero and Hendrich, 2005; Bienvenu and Chelly, 2006). However, the lack of dramatic, widespread de-repression observed with global gene expression studies may be a limitation of the experimental design. Such approaches fail to account for the heterogeneity of the brain and the possibility that MeCP2 regulates different genes in different cell types. In addition, the brain contains a large proportion of glial cells which do not express MeCP2 (Shahbazian et al., 2002b; Jung et al., 2003b; Kishi and Macklis, 2004), thus diluting gene expression changes that might be present in the neurons. This is further compounded in heterozygous female samples where less than half of the neurons express the mutant Mecp2 allele due to unbalanced X chromosome inactivation (Braunschweig et al., 2004; Young and Zoghbi, 2004). Furthermore, MeCP2 target genes may not be fully expressed in the adult Mecp2 null, but instead demonstrate incomplete repression, leading to small changes in gene expression that would be difficult to detect by global expression analysis (Caballero and Hendrich, 2005). In order to design more focused experiments to identify the molecular pathogenesis of MeCP2 deficiency, it will be necessary to clarify the developmental stage disrupted by the loss of MeCP2 in a defined neuronal population.  1.2.3 MBD1  MBD1 is also expressed in neurons, but not glial cells, throughout the brain, with the highest concentration in the highly plastic hippocampus (Gage, 2000; Zhao et al., 2003). Within the subgranular layer of the dentate gyrus, MBD1 is also detected within some nestin-positive progenitors (Zhao et al., 2003). In contrast to MeCP2, MBD1 null mice have no obvious developmental defects and appear healthy and viable. However, closer examination reveals defects in adult hippocampal neurogenesis and function. Progenitors isolated from the hippocampus of adult MBD1 null mice form neurospheres more readily than wildtype cells under proliferating conditions; but on induction of differentiation, MBD1 null progenitors display a significant decrease in neuronal differentiation, with no change in the production of glia. The MBD1 null adult-derived neurospheres also contain a significant increase in aneuploid  14  cells, suggesting MBD1 may be important for maintaining genomic stability. In vivo, adult MBD1 null mice have normal rates of cell division in the dentate gyrus, as measured by BrdU incorporation, however significantly fewer BrdU labelled cells remain 4 weeks later. This decreased survival of new neurons in the absence of MBD1 may be due to the increased ge nomic instability. With this decreased adult neurogenesis, MBD1 null mice exhibit impaired spatial learning and a significant reduction in long-term potentiation in the dentate gyrus of the hippocampus (Zhao et al., 2003).  1.2.4 MBD2  Similar to MBD1, MBD2 null mice are viable, fertile and appear to develop normally (Hendrich et al., 2001). However, MBD2 null mice do show a resistance to intestinal cancer (Sansom et al., 2003), and a lack of region-specific gene silencing in the gut (Berger et al., 2007). Helper T cells from MBD2 null mice also exhibit disordered differentiation (Hutchins et al., 2002) and male MBD2 null fibroblasts show a consistent 3- fold increase in Xist expression over wildtype, implicating MBD2 in the DNA methylation-dependent maintenance of Xist repression (Barr et al., 2007). However, the neuronal phenotype of the MBD2 null mice has not been examined in detail. Expression of Mbd2 is detected in the developing and adult brain (Hendrich and Bird, 1998; Jung et al., 2003a) and MBD2 null mice display aberrant maternal behaviour (Hendrich et al., 2001). Litters born to MBD2 null females are reduced in size, with the average weight of the pups also significantly lower than pups born to wildtype mothers. This phenotype is irrespective of the genotype of the fathers, and can be rescued by fostering the pups. A defect in nurturing behaviour, manifested as a failing of MBD2 null mothers to recognize and retrieve their pups, is believed responsible for the phenotype of the pups (Hendrich et al., 2001). While the cause(s) underlying this behavioural abnormality have not been elucidated, the results could suggest an olfactory or pheremonal phenotype (Gandelman et al., 1971).  1.2.5 MBD3  The most severe developmental phenotype is observed with the disruption of Mbd3. MBD3 null mice are not viable and die during early embryogenesis, with no normal- looking MBD3 null  15  embryos recovered after implantation (Hendrich et al., 2001). Furthermore, MBD3 null embryonic stem cells can be maintained in vitro, but fail to commit to developmental lineages, either in differentiation conditions in vitro or in chimeric embryos (Kaji et al., 2006). While these results provide further evidence that MBDs are critical for proper development, they preclude the ability to assess a neuronal specific phenotype in the MBD3 null. However, Mbd3 displays a dynamic spatial and temporal pattern of expression in the embryonic and adult brain. Mbd3 is highly expressed in the developing forebrain, with relative expression decreasing postnatally in the matur ing hippocampus and cortex; non-ubiquitous expression of Mbd3 is observed within these tissues in the adult (Jung et al., 2003a). The expression pattern thus supports a role for MBD3 in neuronal development.  1.2.6 DNMTs and MBDs in the Neuronal Response to Injury  In addition to the neurodevelopmental defects observed in the absence of MBD proteins, DNA methylation-dependent gene silencing may be critical for the response of the nervous system to challenges or injury. Transient forebrain ischemia in the rat rapidly induces the expression of Mbd2 (within 6 hours) throughout the hippocampus, followed by a slower up-regulation of Mbd1 and Mecp2 (at 24 hours) in both the CA1 and CA3 subfields. In contrast, Mbd3 expression is rapidly decreased (within 3 hours) following insult in the CA3 and dentate gyrus regions, with decreased expression remaining in the CA1 subfield (Jung et al., 2002). These post- ischemic changes in MBD expression in the hippocampus occur prior to the delayed death of vulnerable neurons within the CA1 subfield. Furthermore, post-ischemic changes in MBD expression are not evident in the cerebellum, which is not highly vulnerable to this ischemic challenge. Therefore, the changes in MBD expression observed in the hippocampus likely contribute to the balance between vulnerability and protection as these neurons respond to the challenge (Jung et al., 2002). Ischemia and reperfusion generates a three- to four- fold increase in the incorporation of [3 H] labelled methyl groups into DNA of the brain over control baseline levels (Endres et al., 2000). Thus the increased MBD expression (Jung et al., 2002) correlates with the increased DNA methylation triggered by an ischemic insult. This increase in DNA methylation is not observed in  16  the brains of mice he terozygous for a Dnmt1 deletion, which contain about 50% of the DNMT1 protein as wildtype mice (Endres et al., 2000). The reduced DNA methylation response confers resistance to ischemia- induced injury, as evidenced by significantly smaller infarcts compared to wildtype littermates. A similar reduction in ischemic brain injury is also observed following treatment with the histone deacetylase inhibitor trichostatin A (TSA) (Endres et al., 2000), suggesting that the post-ischemic effect of DNA methylation is mediated by HDACs. Reduction of DNMT1 in post-mitotic neurons alone is sufficient to significantly reduce ischemic infarct size (Endres et al., 2001), indicating that the reduced neuronal vulnerability is not the result of a developmental alteration in the DNMT1-reduced neurons or due to systemic physiological parame ters. Interestingly, mice in which both copies of Dnmt1 are deleted in post-mitotic neurons display no protection from ischemic brain damage and have infarcts indistinguishable from wildtype (Endres et al., 2001), highlighting a complex role for DNA methylation-dependent gene silencing in mechanisms of injury and repair in the nervous system.  1.2.7 DNA Methylation and Synaptic Plasticity  There is also growing evidence to indicate that epigenetic modifications may serve as an important control mechanism in memory associated transcriptional regulation (Swank and Sweatt, 2001; Guan et al., 2002; Huang et al., 2002; Alarcon et al., 2004; Korzus et al., 2004; Levenson et al., 2004; Kumar et al., 2005; Wood et al., 2005; Chwang et al., 2006). In addition to a number of histone modifications, more recent evidence has directly implicated DNA methylation in synaptic plasticity and memory formation. DNA methylation inhibitors block the induction of long-term potentiation in the hippocampus and induce acute changes in the methylation status of the Reelin and Bdnf promoters, two genes implicated in the induction of synaptic plasticity in the hippocampus (Levenson et al., 2006). An increase in Dnmt3a expression correlates with the time-course of Reelin promoter methylation (Levenson et al., 2006), while transcripts for both Dnmt3a and Dnmt3b are increased in the CA1 region of the hippocampus by contextual fear conditioning, a hippocampus-dependent associative memory paradigm (Miller and Sweatt, 2007). DNA methylation is necessary for memory formation of the contextual fear conditioning, however, DNMT inhibitors do not permanently block associative memory formation, as mice re-trained immediately after the first testing, this time without  17  DNMT inhibitors, do learn. This ability is reportedly due to dynamic, reversible changes in the methylation status of the promoters of the memory suppressor gene PP1 as well as the synaptic plasticity gene Reelin (Miller and Sweatt, 2007). While these recent findings remain to be validated, by other groups and other experimental approaches, they raise the interesting possibility that changes to DNA methylation in the adult nervous system are not necessarily permanent, but can be dynamic and reversible.  1.3 Histone Deacetylase Inhibitors  The ability to experimentally manipulate epigenetic regulation of gene expression has been greatly enhanced by a growing number of chemical inhibitors of HDACs, the mediators of DNA methylation-dependent ge ne silencing. A number of naturally occurring and synthetic HDAC inhibitors have been identified, many investigated for their effectiveness as anti-cancer agents (Marks et al., 2001; Villar-Garea and Esteller, 2004). HDAC inhibitors are categorized into several different classes, based on their chemical structures. These include hydroxamic acids, short chain fatty acids (carboxylic acids), benzamides and cyclic peptides (epoxides) (de Ruijter et al., 2003). Hydroxamic acids, probably the broadest set of HDAC inhibitors, are very potent inhibitors of class I/II HDACs, functioning at nanomolar concentrations (Villar-Garea and Esteller, 2004). Crystal structure studies have revealed that these inhibitors function by binding to the zinc ion in the conserved HDAC catalytic site, abolishing the deacetylase activity (Finnin et al., 1999). Whether other structural classes of HDAC inhibitor also function through direct binding to the HDAC catalytic domain remains to be determined (Morrison et al., 2007). The hydroxamic acid Trichostatin A (TSA) was one of the first HDAC inhibitors to be described (Yoshida et al., 1990) and remains widely used in in vitro experiments. However, its toxicity to patients and lack of specificity for certain HDACs has prompted the search for other compounds for clinical use (Villar-Garea and Esteller, 2004).  The short chain fatty acid class of inhibitors, including butyrate (Sealy and Chalkley, 1978), phenylbutyrate (Lea and Tulsyan, 1995) and valproic acid (Gottlicher et al., 2001; Phiel et al., 2001), are currently the best studied class of inhibitors for in vivo use, despite being much less effective than hydroxamic acids and requiring high millimolar concentrations in vivo. Valproic  18  acid has already been approved for treatment of epilepsy (see below for more information) while studies of the effectiveness of butyrate and phenylbutyrate at treating some forms of cancer are currently underway (He et al., 2001; Pili et al., 2001; Gore et al., 2002; Patnaik et al., 2002; Zhou et al., 2002). The benzamides MS-275 and N-acetyldinaline are also currently in clinical trials for their ability to arrest tumour cell growth (Pauer et al., 2004; Acharya et al., 2006). While the exact mechanism of action of this class of inhibitors is not yet clear, they have been shown to inhibit HDACs at micromolar concentrations (Saito et al., 1999; Jaboin et al., 2002).  1.3.1 Valproic Acid  Valproic acid (VPA) demonstrates potent effects in the nervous system beyond its function as an HDAC inhibitor and, therefore, deserves special consideration. VPA was originally employed as a solvent, its efficacy as an anticonvulsant discovered serendipitously when it was used to dissolve other compounds for treatment of experimental models of epilepsy (Meunier et al., 1963; Tunnicliff, 1999; Gurvich and Klein, 2002). For 40 years, VPA has been commonly employed as both an antiepileptic and mood stabilizer, although its specific mechanism of action in ameliorating these disorders is not well understood (reviewed in (Emrich et al., 1980; Tunnicliff, 1999; Johannessen, 2000; Gurvich and Klein, 2002). The efficacy of VPA in the treatment of diverse forms of epilepsy and bipolar disorders, as well as chronic neuropathic pain (Covington, 1998), and as a migraine prophylaxis (Silberstein, 1996), suggests that VPA acts through multiple targets within the CNS (Johannessen, 2000).  VPA increases the level of the inhibitory neurotransmitter ?-aminobutyric acid (GABA), with acute administration in rodents eliciting an increase in whole brain GABA levels of 15-45% (Godin et al., 1969; Kukino and Deguchi, 1978; Hariton et al., 1984). The midbrain regions, such as the substantia nigra, thought to be critically involved in seizure generation and propagation, show the largest increase in GABA (Gale, 1986; Loscher, 1989); however, the mechanism of VPA- induced increases in GABA has not been definitively identified (Anlezark et al., 1976; Whittle and Turner, 1978; van der Laan et al., 1979; Larsson et al., 1986; Loscher, 1993; Wikinski et al., 1996). Like other anticonvulsive agents, the predominant mode of action of VPA described in the literature is its action on ion channels (Johannessen, 2000). VPA appears to act  19  at the voltage-dependent Na+ channel, inhibiting high frequency firing of neurons (McLean and Macdonald, 1986; Farrant and Webster, 1989), reducing sodium conductance and retarding recovery from inactivation (Van den Berg et al., 1993). Chronic treatment with VPA has also been suggested to up-regulate cell surface expression of sodium channels, potentially reflecting a compensatory mechanism for acute effects of VPA on sodium influx (Yamamoto et al., 1997). Potassium and calcium homeostasis are disrupted in neurons treated with VPA, although direct effects have only been observed with high, non-clinical concentrations (Slater and Johnston, 1978; Franceschetti et al., 1986). VPA may alternatively, or in conjunction, cause decreased excitatory neurotransmission by inhibiting signalling by excitatory neurotransmitters. VPA potently suppresses N-methyl- D-aspartate (NMDA) receptor-mediated excitation, although the molecular mechanisms underlying this effect are unknown (Czuczwar et al., 1985; Zeise et al., 1991).  VPA is also a potent teratogen in both humans and mouse models (Robert and Guibaud, 1982; Gurvich and Klein, 2002). Strict structural requirements for the teratogenic activity of VPA ha ve been defined, with distinct analogues of VPA demonstrating either teratogenic or antiepileptic activity (Nau et al., 1991; Lampen et al., 1999). This suggests that some of the clinically observed effects of VPA have distinct molecular targets. In fact, the teratogenic activity of VPA is due to its ability to potently inhibit HDACs (Gottlicher et al., 2001; Phiel et al., 2001). VPA directly inhibits Class I and II HDACs, causing increased acetylation of core histones H3 and H4 and gene activation at therapeutically relevant concentrations (Gottlicher et al., 2001; Phiel et al., 2001). VPA is postulated to directly block substrate access to the catalytic centre of the enzyme, with Class I HDACs appearing far more susceptible to this mode of inhibition than Class II HDACs (Gottlicher et al., 2001). In addition to its selectivity toward catalytic inhibition of Class I HDACs, an unusual trait for HDAC inhibitors, VPA has been reported to selectively induce proteasomal degradation of the Class I HDAC2 (Kramer et al., 2003). Non-teratogenic (antiepileptic) VPA analogues, such as valpromide, do not inhibit HDACs, further supporting a distinction between the antiepileptic and HDAC inhibition activities of VPA (Gottlicher et al., 2001; Phiel et al., 2001). Taken together, these multiple modes of action of VPA do provide a caveat in the interpretation of experimental results obtained with its use. However, the affected  20  pathway can be teased apart by comparing VPA and other modes of disrupting HDAC-mediated gene repression.  1.3.2 HDAC Inhibitors in the Nervous System  A growing literature has demonstrated the efficacy of HDAC inhibitors as cancer drugs due to their ability to arrest cell proliferation and induce differentiation and cell death of transformed cells in culture or in tumours in vivo (reviewed in (Marks et al., 2001; Lin et al., 2006; Marchion and Munster, 2007). This includes cancers of the nervous system, such as neuroblastomas (Cinatl et al., 1997; Cinatl et al., 2002; Stockhausen et al., 2005). The effect of HDAC inhibitors on noncancerous cells in the developing and adult nervous system is, however, not as clear. The inhibitor employed, the cell types tested (including different subtypes of neuron), and the developmental stage of the cells all appear to modify the outcome of HDAC inhibition.  The therapeutic potential of HDAC inhibitors in the treatment of various neurodegenerative disorders has also been explored widely in recent years. In cell cultures and mouse models, various HDAC inhibitors have been shown to ameliorate symptoms of Huntington’s Disease, spinal and bulbar muscular atrophy, Alzheimer’s and Parkinson’s Disease, amyotrophic lateral sclerosis, spinal muscular atrophy, and Friedreich’s ataxia (reviewed in (Morrison et al., 2007). In this context, it has been proposed that HDAC inhibitors act to re-establish the balance between histone deacetylation and histone acetylation (HAT) activity. A loss of acetylation homeostasis is a critical mechanism commonly underlying neuronal dysfunction and degeneration, the primary cause of this imbalance likely varying between disorders (Saha and Pahan, 2006). In normal neurons, where the HDAC-HAT balance is presumably intact, HDAC inhibitors have been found to be detrimental, frequently inducing apoptosis (Salminen et al., 1998; Boutillier et al., 2002, 2003; Morrison et al., 2006). In cultured cerebellar granule neurons (CGN), the HDAC inhibitors TSA and sodium butyrate decrease neuronal survival, even under activity induced survival conditions (Morrison et al., 2006), and induce Caspase-3 activation and apoptosis, in a protein synthesis dependent manner (Salminen et al., 1998). Over-expression of the histone acetyltransferase CBP in cultured CGN, again tipping the balance toward acetylation, also leads to cell death (Rouaux et al., 2003).  21  The varied results obtained with HDAC inhibitors indicate that the cellular functions of HDACs in the nervous system are diverse and context dependent, relying on the developmental stage and identity of the cell effected as well as its general health (i.e. cancerous or degenerative). Current HDAC inhibitors are broad-spectrum inhibitors and their use does not take into account the possible tissue and stage-specific expression of different HDACs as well as the possibility of opposing roles of different HDACs in a given cellular context. Identifying the HDACs expressed in a given cell type at a specific developmental stage and targeting inhibition directly will be essential to understanding the critical functions of HDACs and to minimize side-effects of therapeutic use of HDAC inhibitors.  1.4 The Olfactory System  The complexity of the embryonic nervous system – where single regions contain a variety of neuronal subtypes at different developmental stages - is a highly challenging environment in which to test how DNA methylation-dependent gene silencing may affect the development of a single neuronal lineage. Olfactory receptor neurons (ORNs) in the olfactory epithelium (OE) of the nasal cavity are, however, a highly accessible, simple neuronal lineage in which to examine the stage-specific roles of developmentally regulated processes on neurogenesis. The OE is organized in a stratified, developmentally hierarchical manner, allowing for the identification of developmentally restricted gene expression patterns. Furthermore, ORNs undergo complete functional regeneration throughout adulthood, with homeostasis and population density tightly controlled by local signals (Graziadei and Graziadei, 1979a; Carr and Farbman, 1992; Calof et al., 1996; Huard et al., 1998). Thus, at any given time the OE contains cells at all stages of development, including those undergoing cell death, neurogenesis, axon outgrowth and functional maturation. As such, we can examine mechanisms of neuronal differentiation, regeneration, and plasticity in ORNs, in exclusion of other neurons.  1.4.1 The Cell Types of the Olfactory Epithelium  The postnatal OE is a pseudostratified columnar epithelium composed of four main cell types: olfactory receptor neurons (ORNs), globose basal cells (GBCs), horizontal basal cells (HBCs),  22  and sustentacular (Sus) cells (Figure 1.4). The Sus cells comprise a single layer of cell bodies in the apical region of the OE; however they extend throughout the height of the OE, terminating apically with microvilli in the nasal mucosa and extending processes to the basement membrane. Sus cells are supporting cells and are not associated with the neuronal lineage, but rather are thought to regulate the passage of substances between connective tissue and the OE and to maintain the ionic composition of the overlying mucosa (Hempstead and Morgan, 1983; Farbman, 1992).  The middle region of the OE is comprised of the ORNs, the sensory neurons responsible for our sense of smell. The mature ORNs are located apically to the immature ORNs (IRNs) and the developmental stage can be distinguished immunohistochemically. IRNs, still differentiating and in the process of axon outgrowth and targeting, express high levels of growth associated protein (GAP43) (Calof and Chikaraishi, 1989; Verhaagen et al., 1989) and ßIII neuron specific tubulin (NST) (Roskams et al., 1998). As IRNs functionally mature, they down-regulate GAP43 expression and up-regulate the expression of olfactory marker protein (OMP) (Farbman and Margolis, 1980; Calof and Chikaraishi, 1989), which is frequently employed to identify mature ORNs. ORNs are bipolar neurons and extend dendritic processes apically, terminating in dendritic knobs directly exposed to chemical odorants in the overlying mucosa. ORNs extend axons basally, which cross the basement membrane of the OE and form axon bundles within the lamina propria (LP). These axon bundles are ensheathed by olfactory ensheathing cells (OECs), which accompany the axons as they exit the LP, traverse the cribriform plate and target to the olfactory bulb (OB) within the central nervous system (Doucette, 1990, 1991).  The basal compartment of the OE contains two cell types, HBCs and GBCs (Graziadei and Graziadei, 1979b). GBCs are spherical in appearance and are found directly basal to the IRN layer. The GBCs are highly proliferative transit amplifying progenitors and immediate neuronal precursors, which display a limited capacity to self- renew (Graziadei and Graziadei, 1979b; Calof and Chikaraishi, 1989; Schwartz Levey et al., 1991; Huard et al., 1998; Shou et al., 1999). HBCs, on the other hand, have a flattened morphology and form a layer directly adjacent to the basement membrane. HBCs are relatively quiescent during steady-state neurogenesis in vivo, but in vitro have the capacity to generate colonies containing both neurons and glia  23  Figure 1.4: Laminar structure of the postnatal olfactory epithelium Horizontal basal cells (HBCs) form a single cell layer, adhered to the basement membrane, immediately above the lamina propria in the basal compartment of the OE. They can also be distinguished by the expression of cytokeratin 5/6 or ICAM-1. The globose basal cells (GBCs), many of which can be identified by the expression of PCNA, are located immediately apical to the HBCs. The middle OE contains immature olfactory receptor neurons (ORNs), identified by GAP43 or NST expression, with mature ORNs situated in the apical OE and distinguished by expression of OMP. Mature ORNs extend dendrites into the nasal cavity and a single axon through the lamina propria, where they are ensheathed by OECs, and to the olfactory bulb. The very large nuclei of the Sustentacular cells form a single cell layer in the apical OE, with their processes extending throughout the height of the epithelium.  24  (Carter et al., 2004). Following an extensive lesion in vivo, lineage-tracing experiments have also demonstrated a contribution of HBCs to the reconstitution of the neuronal layers as well as the support cells, both Sustentacular cells and Bowman’s Glands (Leung et al., 2007). Thus, HBCs can be considered the multipotent adult progenitors in the postnatal OE.  1.4.2 Odorant Receptors  The mature ORNs are a largely homogeneous population of neurons (Sammeta et al., 2007), with the exception of odorant receptor (OR) expression. The intronless coding regions of OR genes encode putative seven-transmembrane domain structures, characteristic of G protein-coupled receptors (Buck and Axel, 1991). The OR proteins are enriched in the cilia of ORN dendrites, where activation of the ORs by chemical odorants elicits a G protein/cAMP signalling pathway (Reed, 1992). ORs represent the largest mammalian gene family, comprised of 1000-1300 intact open reading frames in the mouse, scattered throughout the genome in clusters of various sizes (Young et al., 2002; Zhang and Firestein, 2002; Zhang et al., 2004). It is generally proposed that each ORN expresses a single OR in a strictly monoallelic manner (Chess et al., 1994; Malnic et al., 1999), although the evidence for the one receptor-one neuron rule has been questioned (Mombaerts, 2004). A given OR is expressed by a few tho usand ORNs, scattered in a defined spatial zone of the OE and interspersed with ORNs expressing other ORs in a seemingly stochastic pattern (Ressler et al., 1993; Vassar et al., 1993; Strotmann et al., 1994a; Strotmann et al., 1994b; Iwema et al., 2004; Miyamichi et al., 2005). The mechanisms underlying the regulation of OR expression, however, remain poorly understood (Fuss et al., 2007).  1.4.3 Interactions Between the Olfactory Bulb and Olfactory Epithelium  Each ORN extends a single axon which traverses the cribriform plate, enters the CNS and coalesces into a spherical, synaptic neuropil structure, termed a glomerulus, in the outer layers of the OB. Here, ORN axons synapse with the mitral/tufted cell projection neurons and the periglomerular cell interneurons (Figure 1.5). Although mosaically distributed within one of the roughly defined zones in the OE, neurons expressing the same OR converge on one of two glomeruli, one medial and one lateral, per OB (Figure 1.5). The topographical position of these  25  two glomeruli is dependent on the OR and is conserved between the left and right OB and between different individuals (Ressler et al., 1994; Vassar et al., 1994; Mombaerts et al., 1996; Treloar et al., 2002). OR proteins are, in fact, detected in the ORN axons and axon terminals (Barnea et al., 2004; Feinstein et al., 2004; Strotmann et al., 2004) and OR proteins directly influence the targeting of ORN axons to the OB (Mombaerts et al., 1996; Wang et al., 1998; Bozza et al., 2002; Feinstein et al., 2004). How OR proteins impact ORN axon wiring remains elusive, however, it is clear that ORs are not the sole determinant of ORN targeting and likely act in concert with other factors (Mombaerts, 2006), including activity-dependent mechanisms (Yu et al., 2004; Zou et al., 2004; Imai et al., 2006; Serizawa et al., 2006; Chesler et al., 2007; Col et al., 2007), and adhesion molecules and guidance receptors such as galectin-1 (Puche et al., 1996), NCAM (Treloar et al., 1997), neuropilin-1 (Pasterkamp et al., 1998) and its ligand Semaphorin 3a (Crandall et al., 2000), as well as neuropilin-2 (Cloutier et al., 2002; Walz et al., 2002).  The OB develops as an out pocketing of the rostral end of the telencephalon. This development begins at E12, as immature ORN axons reach this region. Functional synapses between the ORNs and the OB are not observed, however, until E14 (Hinds and Hinds, 1976). At the time of birth there are only a few glomeruli present, the majority of the growth and maturation of the OB occurring postnatally, in response to cell activity (Farbman, 1992). The interactions between the two main components of the olfactory system, the OE and the OB, are essential for the proper development of each structure. This reciprocal influence is maintained in the adult organism with innervation by ORNs and activity levels playing a role in cell survival in the bulb (FrazierCierpial and Brunjes, 1989; Najbauer and Leon, 1995) and, ultimately, OB size (Brunjes and Borror, 1983). ORNs, in turn, are dependent on the OB for survival. In its absence, they fail to mature and exhibit a reduced life span. The effect of the OB on ORN maturation is thought to be due to physical contact between their axons and the OB cells, rather than on a diffusion of factors (Farbman, 1992).  By severing this essential contact between the OB and OE, a wave of apoptotic cell death and subsequent neuronal regeneration can be induced in a synchronized manner, thus providing a more effective way to study the mechanisms regulating olfactory neurogenesis. This can be done  26  Figure 1.5: Synaptic pathways of the olfactory system (A) The olfactory neural pathway of the rodent. Olfactory receptor neurons (ORNs) reside in the periphery, in the olfactory epithelium of the nasal cavity, where they are directly exposed to chemical odorants. ORNs extend axons through the cribriform plate and to the olfactory bulb (OB) of the central nervous system. Here, the axons synapse with mitral and tufted cells within glomeruli. Mitral and tufted cells extend axons along the lateral olfactory tract (LOT) and synapse within the olfactory cortex. The olfactory cortex then projects to other cortical regions and to the medial dorsal nucleus of the thalamus (green). New interneurons are added to the OB throughout the lifetime of the animal from progenitors that migrate from the lateral ventricle in the rostral migratory stream (RMS; asterisks). (B) The synaptic organization of the olfactory bulb. ORNs extend dendrites into the nasal cavity, where they respond to chemical odorants depending on which odorant receptor (OR) they express. ORNs expressing the same OR target to the same glomerulus within the OB. Within the glomeruli, ORNs synapse with the two main projection neurons of the OB, the Mitral (M) and Tufted (T) cells. Two classes of interneurons are also found in the OB, the periglomerular (PG) and granule cell (Gr) neurons. Adapted with permission from (Carleton et al., 2002; Lledo and Gheusi, 2003).  27  in two ways: an axotomy, which involves severing the olfactory nerve on the OB side of the cribriform plate, or a bulbectomy, which involves surgical removal of the OB. The bulbectomy results in apoptosis of the ORNs ipsilateral to the surgically altered side (Graziadei et al., 1979; Costanzo and Graziadei, 1983), with the wave of apoptosis maximal at approximately 48 hours post-bulbectomy and completing within 3 days (Holcomb et al., 1995; Cowan et al., 2001). The progenitors in the ipsilateral OE respond by increasing proliferation, which peaks at approximately 6 days post-bulbectomy (Schwartz Levey et al., 1991; Gordon et al., 1995; Carter et al., 2004). However, in the absence of the OB, newborn neurons seldom mature and usually have a lifespan of less than 2 weeks (Carr and Farbman, 1992, 1993). The proliferation of progenitors is, therefore, found to be permanently upregulated to compensate for this loss (Schwartz Levey et al., 1991; Carr and Farbman, 1992, 1993; Gordon et al., 1995; Holcomb et al., 1995), suggesting that the number ORNs is tightly regulated.  The olfactory epithelium is thus in a dynamic equilibrium between basal cell birth, neuronal differentiation, and apoptosis of neurons. This equilibrium is presumed to be highly regulated by autocrine and paracrine mechanisms and a number of different factors, including FGF2, EGF, TGF-a, PDGF, TGF-ß, BMP2, 4 and 7, NGF, BDNF, NT-3 and LIF have been implicated in the proliferation, differentiation and survival of ORNs (Calof et al., 1991; Mahanthappa and Schwarting, 1993; Farbman and Buchholz, 1996; Roskams et al., 1996; Goldstein et al., 1997; Shou et al., 1999; Shou et al., 2000; Getchell et al., 2002; Moon et al., 2002; Bauer et al., 2003). Mature ORNs release negative regulators of proliferation, feeding back on progenitors to inhibit neurogenesis as a mechanism of controlling ORN cell number (Calof et al., 1998). In vitro, differentiated ORNs suppress the formation of neuronal progenitor colonies (Mumm et al., 1996) and this anti-neurogenic effect can be mimicked by exogenous BMPs 2, 4 and 7 (Shou et al., 1999; Shou et al., 2000). Furthermore, growth and differentiation factor 11 (GDF11) is expressed by ORNs and acts on progenitors to induce reversible cell-cycle arrest (Wu et al., 2003). Death of ORNs would, therefore, remove this negative feedback inhibition, thus promoting proliferation of progenitors and neurogenesis. The timeline of ORN death and progenitor proliferation post-bulbectomy support this model of negative feedback control of homeostasis.  28  1.5 Introduction and Project Summary  Development of the nervous system is highly complex, requiring stereotyped waves and spatial patterns of lineage commitment (neurons and glia), migration, differentiation and connection of newborn neurons, and the formation of synaptic networks. In so doing, however, the brain must also retain a high degree of plasticity in order to respond to a changing environment and sensory experiences. Epigenetic modifications provide an ideal mechanism for translating dynamic intrinsic and extrinsic stimuli onto a static genome. Substantial evidence exists to indicate that DNA methylation-dependent gene silencing is absolutely critical for the proper development and functioning of the nervous system. The complexity and heterogeneity of the developing brain is, however, a substantial hindrance to the elucidation of the function of distinct epigenetic modifiers during the stage-specific differentiation of a neuron. Olfactory receptor neurons, on the other hand, are a highly accessible, homogeneous population of neurons which undergo neurogenesis throughout the lifetime of an animal. Therefore, they provide an ideal model system in which to examine the expression and function of components of the DNA methylationdependent gene silencing machinery.  I test the hypothesis that epigenetic gene silencing drives stage-specific differentiation of olfactory receptor neurons, likely through the sequential recruitment of distinct repressor complexes, in the following four aims:  Aim 1: Are methyl-CpG-binding domain proteins and associated repressor complex members sequentially expressed at distinct transitional stages of olfactory neurogenesis? I will identify the developmental expression patterns of the MBD protein family members and HDAC1 and HDAC2 in the olfactory epithelium. This will establish potential transitional stages of ORN differentiation that employ epigenetic modifications and identify the likely MBD protein to mediate each stage.  Aim 2: Are the stage-specific expression patterns of MBDs and HDACs paralleled in the developing CNS?  29  The stage- and lineage-specific expression profiles of MBD2, MeCP2, HDAC1 and HDAC2 will be determined in the embryonic, early post-natal and adult brain. This will determine if the sequential, stage-specific expression pattern of these epigenetic modifiers is unique to the olfactory epithelium or a conserved feature of neural development.  Aim 3: Do MBD2 and MeCP2 null mice display stage-specific defects in ORN differentiation corresponding to their observed expression patterns? Using MBD2 and MeCP2 knockout mice, I will determine if either MBD is necessary for ORNs to traverse distinct stages of differentiation. The developmental stages of cells within the OE of each null will be determined using developmental stage-specific antigenic markers. The ability of the null ORNs to target to the olfactory bulb and the stability of the ORNs will also be assessed.  Aim 4: Is there aberrant progenitor cell cycling in the MBD2 null in the absence of ORNs? Mature ORNs feed back on basal cells to regulate proliferation and neurogenesis. I will employ a bulbectomy lesion paradigm to remove the mature ORNs and will examine the cell cycle kinetics of MBD2 null progenitors in their absence. In addition, an acute disruption of HDAC-mediated epigenetic regulation will be performed on wildtype mice using a chemical inhibitor of HDACs, valproic acid. This will be compared to the genetic disruption of Mbd2, where a functional compensation may have occurred.  30  Chapter 2: Materials and Methods 2.1 Mice  2.1.1 MBD2 Null Mice  MBD2 mice were a gift from Dr. Adrian Bird, University of Edinburgh (Hendrich et al., 2001). The mice were maintained in heterozygous breeding pairs and were fed Love Mash (Bio-Serve) to enhance health for breeding. For genotyping, genomic DNA was extracted from tail clips using the Qiagen DNeasy kit, following manufacturer’s instructions. Genotyping reactions were run in parallel to detect the wildtype and disrupted alleles (primers outlined in Table 2.1) with the following conditions: 1X PCR buffer (Invitrogen), 0.2 mM dNTPs, 3 mM MgCl2 , 1 unit Taq DNA polymerase (Invitrogen) and 0.4 µM primers. The PCR cycling protocol was as follows; 94°C for 2 minutes, followed by 5 cycles of 94°C for 30 seconds, 60°C for 30 seconds, 72°C for 30 seconds and 35 cycles of 94°C for 30 seconds, 58°C for 30 seconds, 72°C for 30 seconds and then 72°C for 5 minutes (using a PE Applied Biosystems Geneamp 9700).  2.1.2 MeCP2 Null Mice  MeCP2 mice were purchased from Jackson Laboratories (Strain Name B6.129P2(C)Mecp2tm1.1Bird, Stock Number 003890). Mecp2 is an X- linked gene and hemizygous males (-/y) display internal testis and are infertile. The mice, therefore, were maintained by breeding heterozygous females (+/-) with wildtype males (+/y). Breeding cages were fed Love Mash (BioServe) to enhance breeding health and mating cages were not disturbed for a minimum of three days following birth to reduce the incidence of infanticide. For genotyping, genomic DNA was extracted from tail clips using the Qiagen DNeasy kit, following manufacturer’s instructions. Genotyping reactions were run in parallel to detect the wildtype and disrupted alleles (primers outlined in Table 2.1) with the following conditions: 1X PCR buffer (Invitrogen), 0.2 mM dNTPs, 3 mM MgCl2 , 1 unit Taq DNA polymerase (Invitrogen) and 0.4 µM primers. The PCR cycling protocol was as follows; 94°C for 3 minutes, followed by 35 cycles of 94°C for 30  31  seconds, 65°C for 1 min, 72°C for 1 min and then 72°C for 2 minutes (using a PE Applied Biosystems Geneamp 9700).  2.2 Olfactory Bulbectomies  Unilateral bulbectomies were performed on adult (2-4 months of age) male MBD2 null mice and littermate controls. Mice were anesthetized with Xylaket (25% Ketamine HCL (MTC Pharmaceuticals), 2.5% Xylazine (Bayer Inc.), 15% ethanol, 0.55% NaCl), a 1.4 mm diameter hole was drilled through the skull directly above the right olfactory bulb and the bulb was ablated by suctioning. The wound was filled with Gelfoam (Pharmacia & Upjohn, Kalamazoo MI) and the skin sealed with VetBond (3M, Minneapolis). Half of the littermate control mice were administered 250 mg/kg sodium valproate (Sigma) by IP injection at 12 hour intervals for 48 hours, beginning 6 days after bulbectomy. All mice were injected with equimolar concentrations (65 mM) of iododeoxyuridine (57.5 mg/kg; Sigma) and chlorodeoxyuridine (42.5 mg/kg; Sigma) at 6 and 8 days post-bulbectomy, respectively. The mice were sacrificed 10 days postbulbectomy. 3-4 mice were included in each experimental group - MBD2 nulls, wildtype controls and VPA treated.  2.3 BrdU Labelling  On postnatal day 7, MBD2 and MeCP2 null mice and littermate controls were given two injections of BrdU (Sigma; 50 mg/kg IP), administered 2 hours apart. The mice were then sacrificed at one of three developmental time points: (1) one hour after the second injection (for baseline BrdU incorporation), (2) at postnatal day 21 or (3) at postnatal day 49. 3-5 mice of each genotype (Wildtype, MBD2 null and MeCP2 null) were analyzed at each developmental stage.  2.4 Tissue Preparation  Adult and Postnatal Day 21 Mice: Mice were anaesthetised with Xylaket (25% Ketamine HCL (MTC Pharmaceuticals), 2.5% Xylazine (Bayer Inc.), 15% ethanol, 0.55% NaCl) and rapidly perfused with cold 0.1M phosphate buffered saline (PBS), followed by 4% paraformaldehyde  32  (PFA) in PBS. Brains, olfactory bulbs and olfactory epithelia were dissected out and post- fixed in 4% PFA for 2 hours at 4ºC. Tissue was then equilibrated in 10% sucrose in PBS followed by 30% sucrose for 24 hours each at 4ºC, before proceeding to the embedding step. Postnatal Day 7 Mice: Pups were anaesthetised with AErrane inhalation anaesthetic (Janssen, Toronto ON) and perfused by hand with 3 ml of ice cold PBS followed by 3 ml of ice cold 4% PFA. The mice were decapitated, the skin removed and whole heads were post- fixed for 4 hours in 4% PFA at 4ºC. Heads were then equilibrated in 10% sucrose followed by 30% sucrose for 24 hours each at 4ºC, before proceeding to the embedding step. Embryos: Pregnant dams were anaesthetised as outlined for adults. Embryos were dissected out and immersion-fixed in 4% PFA overnight at 4ºC. They were then equilibrated in 10% sucrose in PBS followed by 30% sucrose for 24 hours each at 4ºC. For staging of embryos, mid-day after the appearance of a vaginal plug was considered E0.5. Embedding and sectioning: All tissues were equilibrated (under suction, with the exception of embryos) in warm Tissue-Tek embedding medium (Sakura Finetek, Torrance, CA) for 5 minutes, and frozen in liquid nitrogen. Coronal or sagittal sequential sections of 10 – 16 µm were taken on a HM 500 cryostat (Micron), mounted onto charged Superfrost glass slides (Fisher, Edmonton AB) and stored at –20°C for subsequent analysis.  2.5 Immunofluorescence  Sections were warmed on a slide warmer at an approximate temperature of 42°C for 10 minutes and post- fixed in 4% PFA for 10 minutes, followed by two 5 minute washes in PBS. Sections were then subjected to antigen retrieval in 0.01% citric acid, pH 6.0, for 10 minutes in the microwave on high power, washed 5 minutes in PBS and then permeabilized in 0.1% Triton-X100 in PBS for 30 minutes. Following 2 washes of 5 minutes in PBS, the sections were blocked with 4% normal serum in PBS for 20 minutes, and then incubated at 4ºC for 12-20 hours with primary antibodies (see Table 2.2) in 2% normal serum. Sections were washed twice for 5 minutes in PBS and then incubated in secondary antibodies in 2% normal serum for one hour at room temperature. Fluorescently labelled secondary antibodies, used at a dilution of 1:100, were Molecular Probes Alexa 488 and Alexa 594, raised in donkey or goat. Sections were washed twice for 5 minutes in PBS and then nuclei were counter-stained with Diaminopyridine  33  imidazole (DAPI; 1:10 000, Sigma) for 5 minutes at room temperature. Sections were washed twice for 5 minutes in PBS and then coverslips were mounted with Vectashield mounting medium (Vector Laboratories). Exceptions : For standard BrdU detection, sections were permeabilized with 4M HCl for 10 minutes instead of the 0.1% Triton-X-100 for 30 minutes. For IdU and CldU detection, sequential immunohistochemistry was performed, probing first for CldU (rat anti-BrdU, Accurate) followed by IdU (mouse anti-BrdU, Becton Dickinson). After each primary antibody incubation, the sections were washed with 0.5% Tween 20 / 0.5M NaCl / TBS to remove nonspecific binding of the antibodies. Expression Analysis : For all analyses of gene expression patterns, a minimum of three animals of each developmental stage were used.  2.6 Immunohistochemistry  This procedure is identical to immunofluorescence through the primary antibody incubation step. After incubation with primary antibody, the sections were washed twice with PBS and then incubated with 1:200 biotinylated secondary antibody in 2% normal serum for 30 minutes at room temperature. The sections were washed twice in PBS, endogenous peroxidase activity was quenched for 10 minutes in 0.5% hydrogen peroxide (Sigma) in PBS, followed by another PBS wash. They were then conjugated to avidin using the Vectastain ABC kit (Vector Laboratories) for 30 minutes at room temperature. Two more washes with PBS followed and then the sections were developed with VIP (Vector Laboratories). Developing was stopped by a 10 minute wash in ddH2O and coverslips were mounted with Aqua Poly/Mount (Polysciences Inc., Warrington PA).  2.7 ß-galactosidase Staining  Sections were post- fixed in 4% PFA for 10 minutes, washed twice for 5 minutes in PBS and then permeabilized in 0.1% Triton-X-100 in PBS for 30 minutes. Following 2 washes in PBS, the sections were incubated in LacZ buffer (2 mM MgCl2 , 0.01% deoxycholate, 0.02% Nonidet-P40, and 100 mM NaPO4 , pH 7.3) containing 1 mg/mL X- gal, 5 mM potassium ferrocyanide, and 5  34  mM potassium ferricyanide, for 30 to 60 minutes at 37°C in the dark. Negative controls did not demonstrate LacZ staining and included CD-1 non-transgenic mice and littermate control wildtype mice from the MBD2 line. Results were confirmed using antibodies to β-galactosidase, however the detection level and reproducibility were not as good as that with histochemistry. When immunohistochemistry was performed with LacZ histochemistry, LacZ staining was performed first, the slides were washed in PBS and immunohistochemistry was performed as described above, beginning with the citric acid permeabilization and omitting further Triton-X permeabilization.  2.8 Reverse Transcription PCR  Olfactory epithelium tissue was microdissected from CD-1 mice of various developmental time points by peeling OE and lamina propria off the cartilaginous turbinates, and homogenizing in lysis buffer with a Fischer Sciences Powergen 125. RNA was extracted from the homogenates using the Qiagen RNeasy kit, following the manufacturer’s instructions. First strand cDNA was generated from 1.0 µg of total RNA using the SuperScript II kit (Invitrogen). Negative control reactions were run with no reverse transcriptase enzyme, and all primers were designed to span introns to control for genomic DNA contamination. 1 µl of the first strand reaction was combined with 1X PCR buffer (Invitrogen), 0.2 mM dNTPs, 1 mM MgCl2 , 2.5 units Taq DNA polymerase (Invitrogen) and 0.4 mM primers. Primers were selected using the Primer3 program and are listed in Table 2.1. The PCR cycling protocol was as follows; 94°C for 5 minutes (first cycle), followed by 94°C for 30 seconds, 60°C for 35 seconds, 72°C for 60 seconds (35 cycles), 72°C for 4 minutes (last cycle) (using a PE Applied Biosystems Geneamp 9700). RT-PCR products were electrophoresed on a 1.5% TBE/agarose gel and the picture was captured using a BioRad Geldoc 1000.  2.9 In Situ Hybridization  2.9.1 Probe Preparation  35  Anti-sense and sense RNA probes for MBD2, MBD3 and MeCP2 were generated from RT-PCR products amplified from the postnatal olfactory epithelium (Chapter 2.8). Purified PCR products were ligated into the pDrive cloning vector (Qiagen) and transformed into EZ Competent E. Coli cells (Qiagen), following the  manufacturer’s instructions. Clones were selected for  transformation by ampicillin resistance and for plasmid inserts by blue/white screening. Three clones were selected for each MBD and used to inoculate 2 ml overnight cultures of Luria Broth (LB) with ampicillin selection. Pla smid DNA was purified from overnight cultures using the Qiagen Mini-Prep kit. The directionality of the PCR product insert was determined by restriction digest mapping and sequencing. For anti-sense probes, the plasmids were linearized with the restriction enzyme HindIII and for sense probes, BamH1. The restriction enzyme digestion was stopped by incubation with 50 µg of Protease K (Invitrogen) for 30 minutes at 37°C and the linearized plasmids were purified using a MinElute DNA purification kit (Qiagen). Anti-sense digoxigenin- labelled RNA probes were synthesized using the T7 RNA polymerase and sense probes were prepared using the SP6 RNA polymerase, using the Roche DIG RNA Labelling Mix. OMP cDNA in the pGEMT vector was a gift from Dr. Frank Margolis (University of Maryland). This plasmid was linearized with the restriction enzyme EagI and the anti-sense probe synthesized with the T7 RNA polymerase. The RNA probes were electrophoresed on a formaldehyde-agarose gel to determine probe integrity and yield. Probes were stored at -80°C.  2.9.2 In Situ Hybridization  All buffers were made with diethyl pyrocarbonate (DEPC) treated ddH2 O or PBS and all glassware was treated to remove RNAse contamination. Tissue sections were thawed on a slide warmer for 10 minutes, post- fixed in 4% PFA and washed 3 times for 3 minutes in PBS. Sections were acetylated for 10 minutes in fresh acetylation mix (29.5 ml H2 O, 400 µl triethanolamine (Sigma), 52.5 µl concentrated hydrochloric acid, 75 µl acetic anhydride). Sections were washed 3 times for 5 minutes in PBS and then equilibrated with hybridization buffer (50% formamide, 5X SSC, 5X Denhardt’s solution, 250 µg/ml ye ast tRNA (Invitrogen), 500 µg/ml salmon sperm DNA (Invitrogen)) for 2 hours at room temperature. DIG- labelled probes were denatured for 5 minutes at 65°C and diluted in hybridization buffer at a concentration of 500-1000 ng/ml. Probe was added to the sections and they were covered with  36  siliconized coverslips, placed in a humidified hybridization chamber and hybridized 12-16 hours at 72°C. Slides were soaked in 5X SSC to remove the coverslips and washed twice for 10 minutes in 0.2X SSC at 72°C. Slides were equilibrated for 5 minutes in RNAse buffer (0.5 M NaCl, 10 mM Tris-HCl pH 7.5, 5 mM EDTA) at 37°C, and then treated for 30 minutes with 2 µg/ml RNAse H at 37°C. Sections were then washed for 5 minutes in RNAse buffer, followed by 2 times 20 minutes in 0.2X SSC at 72°C and 5 minutes in room temperature 0.2X SSC. Slides were then equilibrated for 5 minutes in B1 buffer (0.1 M Tris-HCl pH 7.5, 0.15 M NaCl) and blocked with 5% heat inactivated normal goat serum (hiNGS) in B1 buffer for 1 hour at room temperature. Sections were incubated with anti-digoxigenin antibody (1:5000; Roche) diluted in 4% hiNGS/B1 overnight at 4°C. Slides were washed 3 times 5 minutes in B1 buffer, equilibrated for 5 minutes in B3 buffer (0.1 M Tris-HCl pH 9.5, 0.1 M NaCl, 50 mM MgCl2 ) and developed in the dark with fresh B4 buffer (3.375 µl/ml NBT (Roche), 3.5 µl/ml BCIP (Roche) and 0.24 mg/ml levamisole (Sigma) in B3 buffer). The developing reaction was stopped by washing twice for 10 minutes in ddH2 O and coverslips were mounted with aquapolymount mounting media (Polysciences).  2.10 Immunoprecipitation  Olfactory epithelium was dissected from postnatal day 7 CD-1 pups and snap frozen on dry ice (tissue samples were pooled from all pups in a litter). The tissue was homogenized and a nuclear/cytoplasmic fractionation and protein extraction was performed using the Pierce NEPER Nuclear and Cytoplasmic Extraction Kit (Pierce Biotechnology), following manufacturer’s instructions. The nuclear protein extract was quantified using the BioRad Protein Assay Reagent, as per the manufacturer’s instructions. Immunoprecipitations were performed on 600 µg of nuclear protein diluted to a total volume of 500 µl in IP buffer (150 mM NaCl, 10 mM Tris-HCl pH 7.9, 0.1% IGEPAL CA-630, 10% glycerol) containing protease inhibitors (1µg/mL aprotein, 1 µg/mL leupeptin, 100 µg/mL PMSF). The protein samples were incubated with primary antibodies for 12-16 hours on an oscillator at 4°C. Pre- immune rabbit serum (1:200) was used as a negative control. 30 µl of Protein G sepharose beads (Pierce Biotechnology) were added and the samples incubated for 2 hours on an oscillator at 4°C. The beads were spun down at 6000  37  rpm for 90 seconds and washed with 1 mL of IP buffer with protease inhibitors. The wash step was repeated 5 times after which the beads were resuspended in 20 µl of 2X SDS sample buffer.  2.11 SDS-PAGE and Western Blotting  To determine developmental expression of MBD2, protein homogenates were prepared from olfactory epithelium tissue dissected and snap frozen from a minimum of three mice of each developmental time point and genotype. The tissue samples were homogenized in a lysis buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% Triton X-100), containing protease inhibitors (1µg/mL aprotein, 1 µg/mL leupeptin, 100 µg/mL PMSF), with a Fischer Sciences Powergen 125 and the total protein concentration was determined using the BioRad Protein Assay Reagent, as per the manufacturer’s instructions. The protein homogenates were diluted 1:1 in 2X SDS sample buffer. All protein samples were denatured by heating at 70°C for 20 minutes and immunoprecipitates were further centrifuged at 13000 rpm for 1 minute to pellet the Protein G beads. All protein samples were separated by 10% SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane (Bio-Rad Trans-Blot). Membranes were blocked for 1 hour at room temperature with 5% non-fat milk in Tris-Buffered saline (TBS), incubated 12-20 hours at 4°C in primary antibody in 2% milk/TBS, washed 3 times 10 minutes in 0.05% Tween20 in TBS and incubated for 1 hour at room temperature in peroxidase-conjugated goat antirabbit IgG or goat anti-sheep IgG (BioRad) diluted in 2% milk/TBS. Following three washes of 10 minutes in 0.05% Tween-20/TBS, the membranes were treated with ECL chemiluminescence substrate (Amersham) and the signal detected on X-ray film.  2.12 Cell Counts and Measurements  2.12.1 Developmental Cell Counts  A minimum of three to five animals per genotype and developmental time-point were used for all cell counts. One matched coronal section from the middle of the olfactory epithelium was  38  analyzed from each animal. Four defined regions of the OE were analyzed on each section – the septum, endoturbinate IIa, ectoturbinate 1 and ectoturbinate 2 (see Figure 2.1). The linear length of OE along the basement membrane was measured in each region using Northern Eclipse Software. For OMP, GAP43 and DAPI counts, a minimum of 100 µm of OE was assayed at each region from pictures taken at a magnification of 40X. For all other counts, a minimum of 500 µm of OE was assayed from images taken at a magnification of 20X. Each region of the OE was first analyzed separately and compared between genotypes. If all regions demonstrated a similar pattern, the data was compiled and the total length of OE analyzed was compared between animals. Error bars denote standard error of the mean between animals.  2.12.2 Post-bulbectomy IdU and CldU Cell Counts  A minimum of three animals were analyzed in each experimental group – MBD2 null, wildtype control, and wildtype treated with VPA. One matched coronal section from the middle of the olfactory epithelium was analyzed from each animal. An adjacent section from each animal was assessed for OMP expression to ensure completeness of the lesion. Only regions of the lesioned epithelium containing less than a single cell layer of OMP positive neurons were included in the analysis. The identical regions in the unlesioned epithelium were analyzed in all cases. From each animal, a minimum of 6 mm of linear OE from endoturbinates IIa and IIb and ectoturbinate 2 was analyzed on both the lesioned and unlesioned side. In addition to the total length of OE, the length of OE occupied by IdU and CldU positive cells was measured and the total number of IdU positive, CldU positive and IdU/CldU double positive cells was counted. Error bars denote standard error of the mean.  2.12.3 Glomerular Measurements  OMP immunohistochemistry was performed on one matched coronal section from the middle of the olfactory bulb from each mouse. The structure of the OE was used as a landmark, in addition to distance from the first section containing OB in order to match the sections as closely as possible. One section each from 3-4 mice of each genotype and developmental time point (P7 and P21) was analyzed. The area of every OMP positive glomerulus was measured by tracing the  39  outline of the glomerulus with Northern Eclipse Software. No significant difference in the number of glomeruli measured per animal was found between genotypes at either developmental stage. The average number of glomeruli measured per animal was, however, significantly greater at P7 (91.8 ± 6.8 glomeruli / animal) than at P21 (79.4 ± 4.5 glomeruli / animal). From each animal, the measured glomerular areas were binned in increments of 250 µm 2 , and the number of glomeruli falling into each area range was averaged across the animals of each genotype. Error bars denote standard error of the mean.  2.13 Image Analysis  All images were visualized with an Axioplan 2 Imaging microscope (Zeiss, Jena GER) using a Retiga 1350EX camera (Quantitative Imaging Corporation) with Northern Eclipse software (Empix Imaging Inc., Mississauga, ON) and were compiled using Adobe Photoshop 7.0.  40  Table 2.1 Primer Sequences  Gene  Reverse Primer  AAGAACAAGCAGAGACTCCG  ACGCTGGCCTAGTGCCGTGC  TCCGCAAACTCCTATTTCTG  TTGTGGTTGTGCTCAGTTC  Wildtype Allele  GGTAAAGACCCATGTGACCC  GGCTTGCCACATGACAA  Genotyping Disrupted Allele  GGTAAAGACCCATGTGACCC  TCCACCTAGCCTGCCTGTA C  MBD2  Wildtype Allele  Forward Primer  Genotyping Disrupted Allele MeCP2  RT-PCR  ß-actin  AGCCATGTACGTAGCCATCCAG GGAGTACTTCTAGGACTCGCTCG  Mbd1  CCCCTACTGCCACAGCTCTA  ACTGTGATGCAGGATGTGGA  Mbd2 (splice  GTTACGGAAGAAGCGCAGAG  ATCGCTCTTGCCAGCACT  Mbd2 Full-length  AAGCAGAGACTCCGGAATGA  TTGCGTACTTGTTGGACTCG  Mbd2 testis -  AAGCAGAGACTCCGGAATGA  GCCACAAACACATGGTCATC  Mbd3  AATAAGAGTCGCCAGCGTGT  CTGTGCTACACTCGCTCTGG  Mbd4  GGAGCATCAATCCCGGTTAT  CTGCTCACGATTTGCTTCAC  MeCP2  CCCCTCCAGGAGAGAGCA  TAGGTGGGGAGGAGGCACT  insensitive)  specific  41  Table 2.2 Primary Antibody Dilutions and Suppliers  Antibody  Supplier  Dilution IHC  Rabbit anti-activated caspase 3  BD Pharmingen  IP  IB  1:500  (aC3) Mouse anti-ß-actin  Sigma  1:1000  Mouse anti- BrdU  Roche  1:1000  Rat anti-BrdU (CldU)  Accurate  1:250  Mouse anti- BrdU (IdU)  Becton Dickinson  1:500  Mouse anti-CNPase  Abcam  1:200  Mouse anti-cytokeratin 5/6  Boehringer Mannheim  1:100  Mouse anti-DNMT3a  Imgenex  1:100  Mouse anti-DNMT3b  Imgenex  1:100  Goat anti-doublecortin (Dcx)  Santa Cruz  1:200  Mouse anti-GAP43  Chemicon  1:500  Mouse anti-GFAP (Cy3 conjugate)  Sigma  1:500  Rabbit anti- HDAC1  Abcam  1:2000  1:200  1:5000  Rabbit anti- HDAC2  Abcam  1:5000  1:200  1:5000  Hamster anti-ICAM-1  Pharmingen  1:100  Sheep anti-MBD2  Upstate  1:500  1:50  1:500  Rabbit anti-MeCP2  Upstate  1:500  Rabbit anti-NCAM  Chemicon  1:500  Mouse anti-Nestin  BD Pharmingen  1:200  Mouse anti-NeuN  Chemicon  1:500  Mouse anti-NST  Covance  1:500  Goat anti-OMP  Gift (F. Margolis)  1:5000  Mouse anti-PCNA  Sigma  1:5000  Rabbit anti-phospho-Histone H3  Upstate  1:1000  Note: IHC = Immunohistochemistry, IP = Immunoprecipitation, IB = Immunoblot  42  Table 2.3 Validation of MBD and HDAC Antibodies Antibody MBD2: Upstate Biotech (Cat#07198)  Western Blots Recognizes bands in protein extracts from the OE at 49kD and 30kD, corresponding to MBD2a and MBD2b, respectively. In protein from the MBD2 null OE, recognizes only a ~25kD band, corresponding to exon 1 of the disrupted allele (as reported and validated in Ng et al., 1999).  MeCP2: Upstate Biotech (Cat#07013)  Recognizes a band of approximately 75kD in protein extracts from brain and OE. This band is not detected in extracts from MeCP2 null tissue.  In Vivo Expression In the mouse brain, this antibody recognizes subsets of neurons in a pattern that correlates with the in situ hybridization of the Allen Brain Atlas. Furthermore, in the brain, this antibody displays a similar pattern of detection as ß-galactosidase expressed under the Mbd2 promoter. In the OE, this antibody exhibits partial overlap with in situ hybridization and ßgalactosidase; this is likely due to the expression of MBD2b, with which this antibody has weak crossreactivity. In tissue sections from MBD2 null mice, this antibody weakly detects expression in the ORNs, likely representing exon 1. Similar expression in ORNs of MBD2 null mice is also observed with another antibody gene rated against a peptide from exon 1 (Santa Cruz, Cat#sc-9397) Expression in both the OE and brain with this antibody matches in situ hybridization, as well as many published reports of MeCP2 expression. This antibody does not detect expression above background levels in MeCP2 null OE.  References Generation of antibody and validation by Western blotting: (Ng et al., 1999)  Expression of MeCP2 in the OE, using an alternate antibody: (Cohen et al., 2003)  43  Antibody HDAC1: Abcam (Cat#ab70 28)  Western Blots Recognizes a band of approximately 65kD in protein extracts from both the brain and OE. Does not appear to cross-react with HDAC2 by Western blotting or immunoprecipitation.  HDAC2: Abcam (Cat#ab70 29)  Recognizes a band of approximately 55kD by Western blotting and immunoprecipitation from protein extracts from both OE and postnatal brain.  In Vivo Expression Expression in the brain is largely unpublished. The detection with this antibody matches reported expression in the postnatal rat corpus callosum identified with a different antibody. This antibody also detects a similar pattern of HDAC1 expression as the in situ hybridization reported by the Allen Brain Atlas. Similar to HDAC1, expression of HDAC2 has not been widely analyzed in the brain. I employed a different HDAC2 antibody (Santa Cruz Cat#sc-7899) and demonstrated a similar pattern of detection in both the OE (multiple developmental time points) and the embryonic brain. The Santa Cruz antibody demonstrated a similar pattern of detection in the P7 rat corpus callosum in a previously published report.  References Expression of HDAC1 in the postnatal rat corpus callosum (Western Blotting and immunohistochemistry ): (Shen et al., 2005)  Expression of HDAC2 in the postnatal rat corpus callosum (Western Blotting and immunohistochemistry ): (Shen et al., 2005)  44  Figure 2.1: Turbinates of the adult olfactory epithelium This schematic outlines the structure of the adult olfactory epithelium, depicting a coronal section from the middle of the epithelium, similar to the region sampled. Endoturbinates IIa, IIb and III are shown, as are ectoturbinates 1 and 2. DR indicates the dorsal recess.  45  Chapter 3: Stage-Specific Expression of Methyl-CpG-Binding Domain Proteins and Histone Deacetylases During Olfactory Neurogenesis  Figures 3.6 and 3.7 as well as a modified version of Figure 3.9 were published in MacDonald JL, Gin CSY, Roskams AJ (2005) Stage-specific induction of DNA methyltransferases in olfactory receptor neuron development. Developmental Biology 288:461-473  3.1 Introduction  In an RNA subtraction-hybridization screen, Dnmt3b, Hdac2 and Mbd2 were all identified as genes that are highly upregulated at a peak time of olfactory neurogenesis following a bulbectomy lesion (AJ Roskams, unpublished observations), thus implicating DNA methylationdependent gene silencing in the differentiation of ORNs. To begin to understand how DNA methylatio n-dependent gene silencing impacts ORN differentiation, we must first identify the key proteins involved in this process in the OE and delineate the developmental stages at which each likely function. At the onset of this thesis, a complimentary project within the Roskams laboratory demonstrated that the two de novo DNMTs are expressed at dynamic and successive stages of ORN differentiation (CSY Gin, Masters Thesis, 2004; MacDonald et al., 2005). DNMT3b is expressed in actively cycling, proliferating cell nuclear antigen (PCNA) positive, basal progenitors, and is down-regulated as they exit mitosis and commit to a neuronal lineage (Figure 3.1). The co-expression of DNMT3b and PCNA within the OE could suggest that DNMT3b also functions as a maintenance methyltransferase during mitosis (Okano et al., 1998; Chen et al., 2003). In this context, DNMT3b should co-localize with PCNA at the replication fork during S-phase in mitotic olfactory precursors. However, such co- localization is not detected, indicating that DNMT3b is likely acting after DNA synthesis to establish new methylation marks in progenitors of the olfactory system preparing to differentiate (MacDonald et al., 2005). Thus, DNMT3b may initiate gene silencing to preve nt re-entry into the cell cycle or to prevent a choice of alternative lineages.  46  Figure 3.1: DNMT3b is expressed in olfactory progenitors while DNMT3a is expressed in postmitotic ORNs (A,B) DNMT3b (red) is expressed in the PCNA positive (green) mitotic basal cells, but it is not expressed in the apical PCNA positive progenitors at E18. (B) The expression of DNMT3b is down-regulated concomitant with decreased neurogenesis in the adult. (C) At E17, DNMT3a (red) is expressed throughout the middle of the OE, in the immature receptor neuron population as well as in the majority of the OMP positive (green) mature neurons. (D) Expression of DNMT3a is also down-regulated in the adult, where it is restricted to the immature receptor neuron population and a small subset of the more basally situated mature, OMP positive, neurons. (E) DNMT3b is expressed in the cycling progenitors, as they exit the cell cycle and commit to a lineage. DNMT3a is expressed across the immature-to-mature neuronal transition.  47  In contrast, expression of DNMT3a is initiated in post- mitotic immature receptor neurons and is down-regulated as they functionally mature ((MacDonald et al., 2005); summarized in Figure 3.1). The progressive down-regulation of DNMT3a as ORNs functionally mature (ie express OMP), suggests that DNMT3a may be laying down methylation marks essential for transitioning IRNs from a developing state into one of terminal ORN differentiation. Therefore, it may silence genes that were needed for IRN differentiation (e.g. axon guidance genes), but that are no longer required as ORNs complete targeting and solidify their synapses. In addition, the maintenance methyltransferase DNMT1 is expressed in the nuclei of mitotic olfactory basal cells, as expected, as well as in the cytoplasm of ORNs (MacDonald et al., 2005). The cytoplasmic localization of DNMT1 in ORNs is unusual, but has also been reported in post- mitotic neurons of the CNS (Goto et al., 1994; Brooks et al., 1996; Inano et al., 2000).  The tightly regulated expression of the de novo DNMTs at dynamic and successive stages of olfactory receptor neuron development coincides with defined transitional stages of developmental gene expression patterns during the differentiation of ORNs. DNA methylation can prevent gene expression either by sterically hindering transcription factors from binding to promoter regions or, alternatively, through the recruitment of repressor complexes (Bird, 2002). If coordinated sets of genes no longer required for further development are being silenced at distinct transitional stages in ORN differentiation, we hypothesized that this would likely be mediated by the recruitment of chromatin-remodelling repressor complexes. We further hypothesized that distinct repressor complexes would mediate the DNA methylation catalyzed by DNMT3a and DNMT3b at the different stages of ORN differentiation. Therefore, we determined which key mediators of DNA methylation-dependent gene silencing, the methyl binding domain proteins and histone deacetylases, are expressed at the developmental stages defined by expression of DNMT3b and DNMT3a.  48  3.2 Results  3.2.1 All Methyl Binding Domain (MBD) Proteins are Expressed in the Developing OE  There are five classic MBD protein family members, identified by a shared highly conserved methyl binding domain. The structure of the proteins diverges highly outside of this domain, with the exception of MBD2 and MBD3 which share a 70% amino acid identity downstream of the MBD ((Hendrich and Bird, 1998); Figure 1.2). By reverse transcription PCR (RT-PCR), using splice variant- insensitive primers targeted outside of the highly conserved MBD domain, transcripts corresponding to Mbd1, Mbd2, Mbd3, Mbd4 and Mecp2 are detected in the embryonic (E11, E15) and early postnatal OE. Mbd1, Mbd2, Mbd3 and Mecp2 are also detected in the adult OE (Figure 3.2A). Mecp2 has two known splice variants, e1 and e2, both of which were detected in the developing and adult OE (data not shown). At postnatal day 7, Mbd2 transcripts are detected throughout much of the OE with two stripes of higher intens ity, one in the basal layer and one in the apical OE (Figure 3.2B). Detection of Mbd3 and Mecp2, on the other hand, is maximal in the apical OE, similar to olfactory marker protein (OMP) which is expressed by mature receptor neurons (Figure 3.2B).  3.2.2 Multiple Isoforms of MBD2 are Detected in the Developing OE  Mbd2 contains an alternatively spliced exon, previously identified in the adult rodent as being specific to the testis. Using a forward primer directed against exon 2 and reverse primers in exon 6 and the testis-specific exon (Figure 3.3A), PCR products of 489 and 322 base pairs are detected, corresponding to the full length (F-L) and testis-specific (T-S) transcripts, respectively. We confirmed the specificity of the primers using cDNA derived from adult mouse testis, where the F-L and T-S variants are equally detected (Figure 3.3B). Sequencing of the amplified products confirmed the identity and specificity of the transcripts (data not shown). Interestingly, the T-S variant is developmentally expressed in both OE and embryonic brain. At E11, the T-S and F-L transcripts are equally detected, with relative detection of the T-S transcript decreasing with development. To determine if the expression of the T-S variant is common in the embryo or restricted to the nervous system, cDNA from the brain and the liver were tested. Mbd2 T-S was  49  Figure 3.2: Transcripts of all MBD family members are detected in the developing olfactory epithelium (A) Transcripts of all 5 classic MBD proteins are detected in cDNA generated from the developing olfactory epithelium, using intron-spanning, splice variant insensitive primers. (B) The expression pattern of the MBDs varies in the postnatal (P5) olfactory epithelium by in situ hybridization. Mbd2 is detected throughout the OE, while expression of Mbd3 and Mecp2 is restricted to the post-mitotic neurons. OMP is expressed in mature ORNs only. Arrows indicate the position of the basement membrane. E = embryonic day, P = postnatal day, Ad = adult, ORN = olfactory receptor neuron, IRN = immature receptor neuron, BC = basal cell, LP = lamina propria.  50  Figure 3.3: Multiple isoforms of MBD2 are detected in the developing olfactory epithelium (A) Genomic structure of the murine Mbd2 gene. Mbd2 has a known splice variant, previously identified as being specific to the testis. Mbd2 also has two protein translational start sites, A and B, indicated with bent arrows. Straight arrows indicate PCR primer targets. The gray regions on exons 1 and 2 denote the coding region for the conserved methyl binding domain. (B) Both the full-length (F-L) and testis-specific (T-S) splice variants are detected in the adult testis and the embryonic OE, with detection of the T-S variant decreasing with age in the OE. Both variants are also detected in the embryonic brain but not the liver. (C) Western blot analysis of protein homogenates from the developing OE. MBD2 is detected as an approximately 49kD band, with expression increased in the adult OE compared to the developing OE. A smaller band at approximately 30kD is also detected, most highly at E18. One band of approximately 25kD is detected in protein derived from adult MBD2 null OE. A ß-actin loading control demonstrates consistent total protein between the samples. E = embryonic day, P = postnatal day, Ad = adult , WT = wildtype, KO = MBD2 null.  51  detected in the embryonic brain, but not adult brain, however it was not detected in the liver at either stage (Figure 3.3B). This suggests that, in addition to its expression in the adult testis, the testis-specific variant of Mbd2 is expressed during nervous system development. Whether or not the testis-specific splice variant of Mbd2 is translated, however, producing a C-terminal truncated MBD2 protein, has not been determined (Hendrich and Bird, 1998).  Two protein isoforms, a and b, have been identified for MBD2, arising from alternate start codons within exon 1 (Hendrich and Bird, 1998). MBD2b has an N-terminal truncation, with translation initiated at the start of the me thyl binding domain (Figure 3.3A). An antibody generated against recombinant full- length MBD2 detects an approximately 49 kD band by Western blotting, corresponding to MBD2a, and shows weak cross-reactivity with MBD2b at approximately 30 kD (Ng et al., 1999). Bands at approximately 49 kD and 30 kD are detected in the developing OE, suggesting that both MBD2a and MBD2b are expressed throughout olfactory development (Figure 3.3C). Expression of MBD2a increases in the adult OE, however, while MBD2b is most robustly detected embryonically. In protein extract from adult MBD2 null OE, only a 25 kD band is detected (Figure 3.3C), predicted to be encoded by exon 1 sequences, which are unaffected in the targeted allele (Hendrich et al., 2001). This indicates the specificity of this antibody for MBD2 and suggests that it does not cross-react with MBD3, a common problem for antibodies that recognize the C-terminus of MBD2.  3.2.3 MBD2a, MBD2b and MeCP2 Are Sequentially Expressed at Distinct Stages of ORN Differentiation  Mbd2 expression is detected in multiple stages of ORN development, potentially representing the multiple isoforms of Mbd2, while expression of Mecp2 appears restricted to the stage of terminal differentiation of ORNs (Figure 3.2B). As such, MBD2 and MeCP2 may regulate gene expression at distinct, sequential stages of ORN neurogenesis. To more clearly establish the expression pattern of both MBD2 and MeCP2 in the OE, we used double immunohistochemistry on postnatal day 7 (P7) mice. For all patterns of expression, a minimum of three animals were analyzed. As demonstrated in Figure 3.3C, an antibody specific to MBD2 recognizes the fulllength MBD2a protein isoform, but shows only weak cross-reactivity with the N-terminal  52  truncated MBD2b protein isoform. To detect the full spectrum of MBD2 expression, we employed heterozygous mice in which a promoterless ßgeo cassette has replaced exon 2 of the Mbd2 gene (Hendrich et al., 2001) and ß-galactosidase is expressed under the Mbd2 promoter. To first confirm that the ß-galactosidase accurately reflects Mbd2 expression, we compared the ß-gal expression pattern in the adult brain to the Allen Brain Atlas expression pattern of Mbd2 (Lein et al., 2007). ß-gal expression was detected either with an antibody against ß-gal (data not shown), or by LacZ histochemistry (Figure 3.4). Specificity of this staining was confirmed by lack of detection on wildtype tissue.  Within the OE, we analyzed the MBD2 and MeCP2 expression at P7 because this developmental time-point displays a high rate of neurogenesis within the OE, concurrent with a large complement of mature ORNs with an established laminar structure. This provides an enhanced representation of multiple stages of ORN maturation, while still accurately representing a developed epithelium. Using the heterozygous mice, LacZ is detected in two layers of the P7 OE (Figure 3.4A), similar to the pattern observed by in situ hybridization (Figure 3.2B). In the basal region, LacZ is detected in PCNA positive cycling progenitors (Figure 3.4B), with expression extending into the immature receptor neuron population, as labelled by NST (Figure 3.4C). LacZ is down-regulated in most NST-positive neurons (Figure 3.4C) with expression reinstated as the neurons transition to a mature, OMP-positive phenotype (Figure 3.4D). In contrast, the MBD2 antibody only detects MBD2 in the apic al layers of the OE (Figure 3.4E). MBD2 is not detected in the PCNA-positive cycling progenitors (Figure 3.4F) or the majority of NST-positive immature neurons (Figure 3.4G), however it is strongly expressed in the more mature NCAMpositive neurons located apically to NST (Figure 3.4H). This suggests that MBD2a is primarily expressed in mature ORNs, while MBD2b is primarily expressed in cycling progenitors and immediately post-mitotic immature receptor neurons, as they commit to their neuronal lineage.  MeCP2 displays a divergent expression pattern to both MBD2a and MBD2b. MeCP2 is expressed in a restricted subset of Cytokeratin 5/6 positive horizontal basal cells at P7 that have a flattened, adult morphology (Figure 3.4I). MeCP2 is not expressed within actively cycling progenitors (Figure 3.4J). MeCP2 is most highly detected in the more apical maturing neurons  53  Figure 3.4: MBD2a, MBD2b and MeCP2 are sequentially expressed at distinct stages of ORN differentiation at postnatal day 7 (A-D) Using heterozygous mice in which one copy of the MBD2 gene is disrupted by a ß-Galactosidase cassette (Bgal), ß-gal is expressed in frame under the MBD2 promoter. (A) LacZ (blue) is detected at two developmental stages of ORN differentiation (arrowheads). LacZ detection overlaps partly with expression of (B) PCNA (red), but detection also extends apically (arrowhead) and partially overlaps with expression of (C) NST (red). LacZ detection is primarily basal (arrow) and apical (arrowhead) to NST expression (red), and detection partially overlaps also with (D) OMP (red). (E-H) Detection of MBD2a using an antibody that primarily recognizes the full-length MBD2a isoform with weak cross-reactivity with MBD2b. (E) The MBD2a antibody detects a single apical band of expression (arrowheads) in the epithelium. (F) MBD2a (green) is not detected in the PCNA-positive basal cells. MBD2a expression is detected (G) in a subset of apical NST-positive immature neurons (arrowhead), with most MBD2a expression apical to NST (arrow), but overlapping with (H) NCAM. (I) MeCP2 expression (green) is detected in a subpopulation of Cytokeratin 5/6 positive HBCs (arrowhead), but (J) MeCP2 is not detected in PCNA positive mitotic cells (arrowhead). (K) MeCP2 is detected in apical NST positive immature neurons (arrow), with expression increasing in (L) OMP positive neurons. Sus = sustentacular cells, ORN = olfactory receptor neurons, IRN = immature receptor neurons, BC = basal cells, LP = lamina propria. Scale bars = 50µm  54  55  (Figure 3.4K) and increasing with maturity in the OMP-positive neurons (Figure 3.4L). Expression of MeCP2 was detected in more basal maturing receptor neurons than MBD2a (data not shown). The pattern of MeCP2 expression also changes with maturation, from, often, two intense focal points of detection in each immature neuron to more extensive nuclear staining in the OMP positive neurons. In addition, MeCP2 is diffusely expressed within the nuclei of sustentacular cells (Figure 3.4I-L).  The punctate pattern of expression of MeCP2 closely resembles the nuclear pattern of detection of 5- methylated cytosine (5-MeC; Figure 3.5). 5-MeC is detected in nuclei throughout the P7 OE with nuclear distribution changing throughout development (Figure 3.5B). Many small 5-MeC puncta are observed throughout the large nucleus of basal cells, while immature ORNs often have only 2 or 3 larger puncta in a stereotypic pattern in a more condensed region. In mature ORNs, 5-MeC is usually detected as one large condensed focal point within the nucleus. This suggests increasing condensation of heterochromatin during the differentiation of ORNs. Sustentacular cells, on the other hand, contain very diffuse, small 5-MeC puncta throughout very large nuclei. The nuclear distribution of 5-MeC in immature and mature ORNs and Sus cells is very similar to MeCP2 (Figure 3.5A) and detection of the two overlap (Figure 3.5C).  3.2.4  HDAC1 and HDAC2 Display Divergent, Sequential Expression Patterns During the  Development of ORNs  Both MBD2 and MeCP2 have been reported to silence gene expression by recruiting repressor complexes containing HDAC1 and HDAC2 (Jones et al., 1998; Nan et al., 1998; Ng et al., 1999). Hdac2 was identified in the earlier suppression subtraction screen performed in our lab to identify genes up-regulated at the peak of ORN neurogenesis. Therefore, I first compared the expression pattern of HDAC2 to the differentially expressed DNMT3b and DNMT3a to determine if the methylation patterns established by either DNMT are likely mediated by HDAC2 during olfactory development. In the developing OE, HDAC2 is first expressed by a small subpopulation of basal PCNA-expressing progenitors (Figure 3.6A,B). HDAC2 is  56  Figure 3.5: Sub-cellular distribution of MeCP2 overlaps with detection of 5-methyl cytosine (A) In the P7 OE, MeCP2 is detected in the mature ORNs as a large, dense focal point in the nucleus (arrows). Within the sustentacular cells, MeCP2 is detected as diffuse, punctate dots throughout the nucleus (arrowhead). (B) 5-methyl cytosine is also detected as a dense focal point in the nucleus of ORNs (arrows), but diffuse puncta in the sustentacular cells (arrowhead), (C) with overlap observed between MeCP2 (red) and 5-methyl cytosine (green). 5-methyl cytosine, however, is also detected throughout the immature receptor neuron and basal cell layers of the OE, where it appears as multiple focal points. LP = lamina propria, BC = basal cells, IRN = immature receptor neurons, ORN = mature olfactory receptor neurons, Sus = sustentacular cells. Scale bar = 50 µm.  57  Figure 3.6: HDAC2 is expressed in a subpopulation of DNMT3b positive basal cells and DNMT3a positive immature ORNs in the developing olfactory epithelium (A,B) HDAC2 is co-expressed by a small subpopulation of PCNA positive cells in both the basal (arrowhead) and middle (thin arrow) layers of the (A) E17 and (B) P5 OE. (D-F) HDAC2 is co-expressed in a subpopulation of basally situated DNMT3b positive cells (arrowhead) in the (D) E17, (E) P5 and (F) adult OE, however many DNMT3b positive cells do not co-express HDAC2 (arrow). (G-I) HDAC2 is coexpressed in many DNMT3a positive cells in the OE. At (G) E17 and (H) P5, the most apical cells express DNMT3a exclusively (arrow) and the most basal cells express HDAC2 exclusively (arrowhead). (I) In the adult OE, the apical DNMT3a positive / HDAC2 negative population is no longer apparent. (C) HDAC2 is co-expressed in only a small subpopulation of OMP positive cells (arrowhead) in the adult. Scale bars = 50 µm  58  similarly co-expressed with DNMT3b in a small subpopulation of mitotic basal progenitors throughout development and into adulthood (Figure 3.6D-F). Expression of HDAC2 extends into the neuronal layers of the OE, where the more apical HDAC2 expressing neurons overlap with the more basal DNMT3a expressing neurons (Figure 3.6G-I). Rare examples of co-expression of HDAC2 and OMP are found in the more basally situated mature ORNs (Figure 3.6C).  Because expression of HDAC2 is largely restricted to the immature ORN population, we next wished to determine if HDAC1 might be associated with repressor complexes at different stages of ORN development. We found that, despite their commonly reported association in MBDmediated co-repressor complexes, HDAC1 and HDAC2 display divergent expression patterns within developing ORNs at postnatal day 7. HDAC1 is highly expressed within PCNA-positive, NST-negative, cycling basal cell progenitors (Figure 3.7A,B) and occasionally co-expressed with PCNA in the middle and apical layers of the OE (Figure 3.7A). HDAC1 is also expressed in OMP-positive mature ORNs (Figure 3.7C), as well as in NST/OMP-negative Sustentacular cells located apically to ORNs (Figure 3.7A-D). HDAC2, on the other hand, is largely co-expressed in NST-positive immature ORNs (Figure 3.7F), and is only occasionally detected in PCNA-positive cycling progenitors (Figure 3.7E) or OMP positive mature ORNs (Figure 3.7G). The majority of DNMT3b expressing progenitors co-express HDAC1 (Figure 3.7D), while only a few co-express HDAC2 (Figure 3.7H).  3.2.5 MBD2 Associates With Both HDAC1 and HDAC2 in the Postnatal OE  The divergent expression profiles of HDAC1 and HDAC2 during ORN differentiation are unexpected based on the biochemical evidence of their mutual inclusion in multiple repressor complexes. To determine if MBD2 directly associates with either HDAC1 or 2 or both in the developing OE, we first analyzed co-expression at P7. MBD2a and HDAC1 are co-expressed in a subset of the most apical MBD2 a expressing cells (Figure 3.8A). MBD2a co-expression with HDAC2, on the other hand, is found in the more basal MBD2a expressing cells (Figure 3.8B). This suggests that MBD2a could associate with both HDAC1 and 2, although perhaps not within the same complex. To directly test for MBD2a/HDAC complex formation, we performed reciprocal co- immunoprecipitation experiments on nuclear extracts from whole P7 OE (Figure  59  Figure 3.7: HDAC1 and HDAC2 are expressed at divergent developmental stages in the postnatal day 7 olfactory epithelium (A) HDAC1 (green) is highly expressed in both the apical (arrow) and basal cell layers of the post-natal OE with most of the PCNA-positive (red) progenitors in all layers of the OE co-expressing HDAC1 (arrowheads). (B) The majority of HDAC1-expressing cells do not co-express NST (red, arrows), however (C) many OMP positive (red) neurons do co-express HDAC1 (arrowhead). (D) Similar to PCNA, most DNMT3b positive (red) cells also co-express HDAC1 (arrowhead). (E) Only a subset of PCNA positive cells co-express HDAC2 (green; arrowhead), with (F) most HDAC2 expression coinciding with NST and (G) in some cases OMP (arrowhead). In a similar frequency to PCNA, HDAC2 is expressed in a subset of DNMT3b positive cells (arrowhead). Scale bars = 50 µm  60  Figure 3.8: MBD2 interacts with both HDAC1 and HDAC2 in the postnatal day 7 olfactory epithelium (A) MBD2a (top panel, bottom panel green) is co-detected in a subpopulation of HDAC1 (middle panel, bottom panel red) expressing neurons (arrow), but is not detected in the majority of HDAC1 positive basal cells and sustentacular cells (arrowheads). (B) Co-expression of MBD2a (top panel, bottom panel green) and HDAC2 (middle panel, bottom panel red) is detected in a subpopulation of MBD2a expressing neurons (arrows). (C) Nuclear extracts of post-natal day 7 OE were immunoprecipitated with antibodies against MBD2a, HDAC1, HDAC2 or pre-immune serum (No Ab IP). The immunoprecipitates and 10% of the starting nuclear extract (P7 OE) were separated on a 10% SDS-PAGE and the Western blots probed with antibodies against MBD2a, HDAC1 and HDAC2. MBD2 formed immunocomplexes with both HDAC1 and HDAC2. Scale bars = 50 µm  61  3.8C). MBD2a co- immunoprecipitates with both HDAC1 and HDAC2, indicating that it can associate with both HDACs. The expression analysis, however, suggests that MBD2a largely associates with the two HDACs individually, at different developmental stages. Interestingly, HDAC1 co-immunoprecipitated HDAC2 and vice versa (Figure 3.8C). This could suggest that the two HDACs are co-expressed in some cells during the transition in expression from one to the other. Direct analysis of co-expression of HDAC1 and 2 in vivo was not possible, however, due to incompatibility of the antibodies employed.  3.3 Discussion  The first step in understanding the role of DNA methylation-dependent gene silencing in olfactory neuronal differentiation is to identify the likely time windows of action of the key players. Our lab has demonstrated that the de novo DNMTs, which establish new methylation marks, are sequentially expressed at defined transitional stages of ORN differentiation (Figure 3.1); DNMT3b is expressed as progenitors exit the cell cycle and commit to the neuronal lineage, while DNMT3a is expressed as immature neurons transition into a mature state (MacDonald et al., 2005). This dynamic expression pattern of DNMT3b and DNMT3a is largely recapitulated during the development of the CNS (Feng et al., 2005; MacDonald et al., 2005). Dnmt3b expression is detected across the cortical ventricular layer at E10.5 and E13.5, where most cells are proliferating neural progenitors (Feng et al., 2005), and DNMT3b is further detected in migrating neuroblasts at E17 (MacDonald et al., 2005). DNMT3a is detected in post-mitotic maturing neurons in the E17 brain (MacDonald et al., 2005) and the expression of DNMT3a within post- mitotic neurons of the cortex increases during the first two weeks postnatal, after which it declines to a lower level in adulthood (Feng et al., 2005). This is similar to the downregulation of DNMT3a demonstrated with the maturation of ORNs.  Mice in which Dnmt3a has been conditionally deleted in the CNS under a Nestin promoter appear healthy at birth, but become hypoactive and exhibit an abnormal gait and reduced motor coordination as they reach adulthood, dying at approximately 18 weeks of age (Nguyen et al., 2007). This delayed symptomatic onset argues that DNA methylation catalyzed by DNMT3a is critical at later stages of neurogenesis, either for the final maturation of a neuron, its ability to  62  function properly or, possibly, its survival. Dnmt3b null mice, on the other hand, die midgestation with multiple developmental defects, including abnormalities of the rostral neural tube (Okano et al., 1999), indicating that DNMT3b, in contrast to DNMT3a, is critical for early stages of neurogenesis. The expression pattern of DNMT3b and DNMT3a suggests that these methyltransferases act at similar defined developmental stages during ORN differentiation.  The newly methylated cytosines catalyzed by DNMT3a and DNMT3b can silence gene expression either by sterically hindering transcription factors from binding to a promoter, or by serving as a ‘mark’ that is recognized and bound by MBD proteins, which recruit HDACcontaining repressor complexes (reviewed in (Bird, 2002)). Transcripts corresponding to all five MBD-containing proteins are detected in the OE (Figure 3.2), suggesting that MBD- mediated gene silencing may be a critical regulatory mechanism for olfactory neurogenesis. In particular, MBD2 and MeCP2 could mediate distinct transitional stages of ORN differentiation by responding to the sequential methylation patterns established by DNMT3b and DNMT3a. Mbd2 is, in fact, expressed at two distinct stages of ORN differentiation – across the cycling basal cell to immature receptor neuron transition, and again in maturing ORNs; this indicates that MBD2 could respond to methylation catalyzed by both DNMT3b and DNMT3a. Expression of Mbd2 at two developmental stages of ORN differentiation was detected by two different methods, in situ hybridization to detect Mbd2 transcripts (Figure 3.2) and LacZ histochemistry to detect ßgalactosidase transcribed under the Mbd2 promoter in heterozygous transgenic mice (Figure 3.4).  The detection of Mbd2 across multiple developmental stages of ORN differentiation likely represents the expression of different MBD2 isoforms. Two splice variants of Mbd2 are detected developmentally in the OE, in addition to two protein isoforms translated from the full- length transcript (Figure 3.3). To detect the MBD2 protein, we employed an antibody generated against recombinant full- length MBD2 ; this antibody primarily detects the full- length MBD2a isoform, with weak cross-reactivity with the N-terminal truncated MBD2b isoform (Ng et al., 1999). By Western blotting, bands corresponding to both MBD2a and MBD2b are detected in the OE at multiple developmental ages; neither isoform is detected in protein extracts from the MBD2 null OE, however, supporting the specificity of this antibody. Because this antibody does not  63  demonstrate equal cross-reactivity with the two isoforms, we can not draw any conclusions about the relative expression of MBD2a and MBD2b, but can only determine that both are present. In vivo, this antibody detects MBD2 in the most apical IRNs and the mature ORNs in the postnatal OE, but it does not detect MBD2 in the basal cells, where Mbd2 transcripts are detected (Figure 3.5). This suggests that MBD2b is likely the isoform expressed in the basal cells, with the antibody employed failing to detect its expression in vivo due to its weak cross-reactivity with this isoform. However, it is also possible that Mbd2 is transcribed in basal cells, detected by both in situ hybridization and LacZ histochemistry, but that it is not translated. Directly testing for the expression of the MBD2b protein in basal cells is problematic, however, due to a lack of antibodies specific for MBD2b. MBD2b and MBD3 share a high sequence identity and have similar molecular weights (Hendrich and Bird, 1998); antibodies generated against regions of MBD2 outside of the N-terminus (specific to MBD2a) often cross-react with MBD3. We have tested several commercially available C-terminus MBD2 antibodies (data not shown), and all detect bands in the MBD2 null protein extract, with a size corresponding to MBD3. To confirm that MBD2b protein is translated in the olfactory basal cells, it will likely be necessary to purify the basal cells from the MBD2a-expressing neurons and perform Western blots with the antibody generated against the full- length MBD2 protein. I would expect to see a 49kd band, corresponding to MBD2a, in the neuron fraction and a 30kd band, corresponding to MBD2b, in the basal cell fraction. Nevertheless, MBD2b is detected at all developmental ages in protein extracts from the whole OE, despite the weak cross-reactivity of the antibody with this isoform. Therefore, it is likely that MBD2b is highly expressed in the OE, and thus is translated in the basal cells.  MeCP2 expression in the OE recapitulates the neuronal expression of the CNS (Shahbazian et al., 2002; Jung et al., 2003; Kishi and Macklis, 2004; Mullaney et al., 2004). Expression of MeCP2 is initiated in the post- mitotic immature ORNs, preceding MBD2a expression, and it is up-regulated as the ORNs functionally mature (Figure 3.5). MeCP2, therefore, likely mediates methylation catalyzed by DNMT3a. In addition, the pattern of nuclear distribution of MeCP2 changes with differentiation. Within immature ORNs, MeCP2 is detected as two to three large puncta, increasing in expression to one large focal point filling the nucleus of mature ORNs, a distribution that has previously been documented in neurons (Shahbazian et al., 2002; Cohen et  64  al., 2003; Kishi and Macklis, 2004; Jugloff et al., 2005). Sustentacular cells, on the other hand, contain very diffuse, small MeCP2 puncta throughout their very large nuclei (Figures 3.4,3.5). The pattern of MeCP2 nuclear distribution correlates with the detection of 5- methylated cytosine (Figure 3.5), which indicates regions of DNA that might be targeted by MeCP2.  5-MeC is detected in the nuclei throughout the postnatal OE, but demonstrates a change in pattern with ORN differentiation, from diffuse in the GBCs to a very condensed region in the mature ORNs. This pattern may depict an increasing condensation of heterochromatin during the differentiation of ORNs, fitting the view of cellular differentiation as a progressive silencing of genes, leading to an increasing inactivation of the genome (reviewed in Fisher and Merkenschlager, 2002). Changes in chromatin structure have been reported to underlie neuronal differentiation (reviewed in Meshorer, 2007), with chromatin remodelling postulated to be a prominent process for adult neurogenesis (Lim et al., 2006) and oligodendrocyte differentiation (Shen et al., 2005; Liu et al., 2007; Lyssiotis et al., 2007). Genes involved in silencing and chromatin remodelling, such as Cbx4, H2afy (macroH2A), Sap18 and Sirt2, are expressed throughout the ORN lineage (Sammeta et al., 2007), with many enriched in mature ORNs (Shetty et al., 2005). MBD2 and MeCP2 can induce the aggregation of pericentric heterochromatin during the differentiation of myotubes (Brero et al., 2005) and MeCP2 has been implicated in the assembly of secondary chromatin structure (Georgel et al., 2003), suggesting that MBD2 and MeCP2 may play a direct role in the increasing chromatin compaction with ORN differentiation. This, however, was not directly analyzed within the scope of this thesis.  It is interesting to note that MeCP2 is also expressed within Sustentacular cells as well as a subpopulation of HBCs. In addition to its commonly explored role in neurons, MeCP2 also associates with the REST (RE1 Silencing Transcription factor) / CoREST repressor complex (Lunyak et al., 2002), which silences neuronal genes in fully differentiated non- neuronal cell types (reviewed in Ballas and Mandel, 2005). It is likely that MeCP2 is functioning in that role within Sustentacular cells. In addition, the REST complex maintains neuronal genes in an off, but poised state within embryonic stem cells and neural stem cells (Ballas et al., 2005), thus likely explaining the detection of MeCP2 within a subpopulation of P7 HBCs. HBCs are the largely quiescent, multi-potent adult progenitors of the OE (Carter et al., 2004; Leung et al.,  65  2007). To determine if MeCP2 is functioning as part of the CoREST complex in Sustentacular cells and HBCs, the first step would be to determine if components of the CoREST complex are expressed in those cell types. To the best of my knowledge, the expression of REST or CoREST has not been reported in the OE. However, HDAC1, a component of the CoREST repressor complex (Ballas and Mandel, 2005), is expressed in Sustentacular cells (Figure 3.7). Coimmunoprecipitation experiments could be used to determine if MeCP2 directly associates with CoREST in the OE.  The sequential and alternating expression patterns of different members of the DNMT, HDAC and MBD protein families (summarized in Figure 3.9) suggests a model of successive transitional checkpoints at defined stages of olfactory neurogenesis, requiring the silencing of genes no longer necessary in future development in order to traverse or maintain the next differentiation stage. These checkpoints would occur (1) when quiescent stem cells shift into mitosis, (2) as dividing progenitors exit mitosis and commit to the neuronal lineage and (3) as immature receptor neurons lose developmental plasticity and transition into a mature, reinforced and stable stage. Each of these developmental stages would likely be facilitated by different repressor complexes. Whether or not the repressor complexes are necessary and/or sufficient for ORNs to traverse these transitional checkpoints in differentiation must now be determined.  66  Figure 3.9: Summary of expression of DNMTs, MBDs and HDACs in the olfactory epithelium The de novo DNMTs, HDACs 1 and 2 and MBD2 and MeCP2 display distinct, stage-specific expression patterns in the OE, traversing potential checkpoints in olfactory neurogenesis that may be mediated by methylation-dependent gene silencing. Checkpoint 1 occurs when quiescent neuro-glial progenitors (HBCs) shift into mitotic state. Checkpoint 2 occurs when dividing progenitors exit the cell cycle and commit to a lineage. Checkpoint 3 occurs as immature receptor neurons lose developmental plasticity and transition into a mature, reinforced stable state.  67  Chapter 4: HDAC1 and HDAC2 Are Divergently Expressed in Distinct Developmental Stages and Lineages in the Developing and Adult Mouse Brain 4.1 Introduction  Within the developing OE, HDAC1 and HDAC2 display opposing expression patterns. HDAC1 is expressed in progenitors, non-neuronal support cells and a subset of mature neurons. HDAC2, on the other hand, is primarily expressed in post- mitotic immature neurons, and is downregulated as they functionally mature (Chapter 3). This nearly complete divergence in HDAC1 and HDAC2 expression was unexpected based on the commonly reported association of both HDACs in the same repressor complexes, including the Sin3A, Mi-2/NuRD and co-REST repressor complexes (Jones et al., 1998; Nan et al., 1998; Wade et al., 1999; Zhang et al., 1999; Ballas et al., 2001). While HDAC1 and HDAC2 may be able to associate with the same repressor complex members, these findings would suggest that they do not frequently do so simultaneously in vivo, at least in the OE.  This, of course, raises the question of whether or not the OE is unique in this regard. Despite the common usage of HDAC inhibitors, very little is known about the expression pattern of the HDACs they target. HDAC1 and HDAC2 are the deacetylases most potently inhibited by VPA (Gottlicher et al., 2001; Phiel et al., 2001; Kramer et al., 2003), an inhibitor commonly used clinically on both young children and adults (reviewed in Emrich et al., 1980; Tunnicliff, 1999; Johannessen, 2000; Gurvich and Klein, 2002). Identifying the cell types and developmental stages that express HDAC1 and HDAC2 within the brain is an important step, not only for understanding the therapeutic mode of action of VPA, but also the potential side effects of such inhibitors. We, therefore, examined the expression pattern of HDAC1 and HDAC2 in the murine brain, (1) during early neurogenesis at E13.5, (2) during early postna tal growth and refinement at P7 and (3) in the mature, adult brain. Furthermore, we compared the expression of MBD2 and MeCP2 to determine if the same sequential, stage-specific expression of HDACs and MBDs observed in the postnatal OE is recapitulated in the CNS.  68  4.2 Results  4.2.1 HDAC1 is Expressed in Progenitors While HDAC2 is Expressed in Post-Mitotic Neurons in the E13.5 CNS  In the early developing mouse brain at E13.5, HDAC1 is expressed in defined regions, with its highest detection in the neurogenic zones around the ventricles. HDAC1 lines the lateral ventricle (LV), but it is only expressed in a few scattered cells in the midbrain (Figure 4.1A). The expression of HDAC1 is highly similar to PCNA, expressed by proliferating progenitors (Figure 4.1B), and exp ression of HDAC1 and PCNA demonstrate nearly complete overlap (Figure 4.1C). HDAC1 is expressed in proliferating progenitors extending along the rostral tip of the LV and into the developing OB; however, expression of HDAC1 extends only minimally into the doublecortin positive cells of the developing OB (Figure 4.1E-E). Doublecortin, a protein that promotes microtubule polymerization, labels late mitotic neuronal precursors and early postmitotic neurons (Francis et al., 1999; Gleeson et al., 1999). HDAC1 is also expressed in PCNA positive proliferating progenitors around the central canal of the spinal cord, but it is excluded from the NeuN positive neurons (Figure 4.1F-G). NeuN is a soluble nuclear protein localized to the cell nucleus and neuronal cytoplasm of most late post- mitotic neurons of the CNS (Mullen et al., 1992; Lind et al., 2005).  As is the case for the lateral ventricle, HDAC1 is highly expressed in the proliferative zone around the 4th ventricle in the hindbrain. HDAC1 is expressed in the doublecortin negative, nestin positive progenitors lining the 4th ventricle at the pontine flexure and it is also detected in cells radiating out into the pons and rostral aspect of the medulla oblongata (Figure 4.1H-I). These chains of HDAC1 positive cells have elongated nuclei, characteristic of migrating cells, and appear to be migrating between doublecortin positive cells, and along nestin positive fibres (Figure 4.1J-K). The fact that these apparently migrating HDAC1 expressing cells are doublecortin negative could suggest either that they are early post- mitotic neuroblasts that have not yet initiated doublecortin expression, or it could indicate that they are non-neuronal progenitors.  69  Figure 4.1: HDAC1 is expressed in progenitors in the E13.5 brain (A) HDAC1 is highly expressed around the lateral ventricle (LV), extending along the rostral tip into the developing olfactory bulb (OB). HDAC1 is also highly expressed in the developing olfactory epithelium (OE), but is only expressed in scattered cells in the midbrain. (B) Proliferating cell nuclear antigen (PCNA) is also highly expressed around the lateral ventricle, into the OB, and in the OE. (C) HDAC1 (red) and PCNA (green) are co-expressed throughout the developing forebrain. (C-K) HDAC1 is red. (DE) HDAC1 is highly expressed around the rostral tip of the lateral ventricle in the developing OB where (D) it is co-expressed with PCNA, but (E) shows only minimal co-expression with doublecortin (Dcx). (F-G) HDAC1 is expressed around the central canal (CC) of the developing spinal cord where (F) it is coexpressed with PCNA. (G) HDAC1 is not detected, however, in the NeuN positive neuronal nuclei of the spinal cord. (H-K) HDAC1 is highly expressed around the 4th Ventricle (4V) at the pontine flexure and HDAC1 positive cells appear to migrate out in chains from the ventricle into the pons and the rostral aspect of the medulla oblongata. (J-K) are higher magnification images of the regions outlined in (H-I) respectively. (H,J) HDAC1 positive cells appear to migrate between Dcx positive cells (arrowheads). (I,K) HDAC1 cells around the ventricle co-express Nestin with HDAC1 positive cells further away from the ventricle appearing to migrate along Nestin positive fibres. (A-C) Scale bars = 200 µm. (D-K) Scale bars = 100 µm.  70  HDAC2 is more extensively expressed in the E13.5 brain than HDAC1. HDAC2 is expressed in the proliferating progenitors around the lateral ventricle. However, in contrast to HDAC1, expression of HDAC2 extends throughout the developing neocortex and the midbrain (Figure 4.2A-C); in fact, relative detection of HDAC2 increases in the non-proliferative zones of the developing brain. This can be seen in the developing neocortex, where HDAC2 is co-expressed with PCNA in the proliferative ventricular and subventricular zones (Figure 4.2D), but is more highly detected in the doublecortin positive neuroblasts and immature neurons of the developing cortical plate and marginal zones (Figure 4.2E). Along the rostral tip of the lateral ventricle, HDAC2 is detected within the ne stin positive progenitor region; however, the expression of HDAC2 again increases in the outer layers of the developing OB, where HDAC2 is co-expressed with doublecortin (Figure 4.2F-G). HDAC2 is also detected in the proliferative zone around the 4th ventricle, at the pontine flexure. Unlike HDAC1, however, the expression of HDAC2 does not extend right to the 4th ventricle within the pons and HDAC2 is also expressed within the immature neurons of the pons and medulla oblongata (Figure 4.2H). Within the medulla oblongata, HDAC2 is primarily expressed in the rounded, doublecortin positive nuclei of immature neurons, interspersed between nestin positive fibres, and it is largely absent from migrating nuclei (Figure 4.2I-J), which express HDAC1 (Figure 4.1H-K). Also in contrast to HDAC1, HDAC2 is not detected in the proliferating progenitors around the central canal of the spinal cord, but it is expressed in the mature neurons of the spinal cord (Figure 4.2K-L).  4.2.2 HDAC1 is Expressed in Glia and Progenitors and HDAC2 in Neurons in the Postnatal Day 7 Brain  We next looked at the expression of HDAC1 and HDAC2 in the P7 brain. The P7 brain contains predominately maturing neurons and glia, with some regions still actively undergoing neurogenesis and, in particular, gliogenesis. HDAC1 is expressed througho ut most of the P7 brain, including in scattered cells throughout the cortex, corpus callosum and hippocampus (Figure 4.3A). Similar to the embryo, HDAC1 is expressed along the ventricles, as well as throughout the rostral migratory stream (RMS), into the OB (Figure 4.3B). HDAC1 is also expressed in multiple layers of the developing cerebellum, a structure that was not yet  71  Figure 4.2: HDAC2 is expressed in post-mitotic neurons in the E13.5 brain (A) HDAC2 is expressed throughout the developing brain, with expression around the lateral ventricle (LV) appearing to increase in the outer layers of the neocortex and in the outer layers of the developing olfactory bulb (OB). (B) Proliferating cell nuclear antigen (PCNA) is expressed around the LV, extending along the lateral tip into the OB. (C) Expression of HDAC2 (red) partially overlaps with PCNA (green), but HDAC2 is most highly expressed in the PCNA negative cells. The box indicates the area magnified in (D,E). (D) HDAC2 is expressed in the PCNA (green) positive cells in the proliferative ventricular (VZ) and subventric ular (SVZ) zones of the developing neocortex, but (E) HDAC2 expression increases in the outer developing cortical plate (CP) and marginal zones (MZ) of the neocortex, where it is co-expressed with Doublecortin (Dcx; green). (F) HDAC2 is weakly expressed in the Nestin positive (green) cells lining the rostral tip of the LV, with (G) expression of HDAC2 increasing in the Doublecortin (green) post-mitotic neurons in the OB. (H,I) (I) is a higher magnification image of the area indicated in (H). HDAC2 is expressed around the 4th Ventricle (4V) at the pontine flexure, as well as in the Dcx positive neurons of the pons and medulla oblongata (arrow), but HDAC2 is not detected within the Dcx negative migrating cells (arrowheads). (J) HDAC2 positive cells are located between nestin positive fibres around the 4th ventricle. (K) HDAC2 is not detected in the PCNA positive cells around the central canal (CC) of the spinal cord, but (L) it is expressed throughout the developing spinal cord, including the ventrally located NeuN positive mature neurons. (A-C) Scale bars = 200 µm. (D-L) Scale bars = 100 µm.  72  Figure 4.3: HDAC1 is primarily expressed in glial cells in the postnatal day 7 brain (A) HDAC1 is expressed in scattered cells throughout the cortex, in the corpus callosum (CC), and in the CA2, CA3 and dentate gyrus (DG) regions of the hippocampus. (B) HDAC1 is detected in cells clustered around the lateral ventricle (LV) and along the rostral migratory stream (RMS), extending into the olfactory bulb (OB). It is also detected in scattered cells throughout the OB. (C) HDAC1 is detected throughout many layers of the developing cerebellum. Boxes indicate regions magnified in (G,H,M). (DI,K-M) HDAC1 is red. (D) HDAC1 is detected in NeuN positive (green) neurons in the CA2, CA3 and DG regions of the hippocampus, but does not extend into the CA1 region (arrow). HDAC1 is also detected in non-neuronal cells in the corpus callosum and around the DG (arrowheads). (E) HDAC1 is not detected within NeuN positive nuclei of the cortex, but it is detected in small nuclei immediately adjacent to the neurons (arrowheads) as well as (F) in PCNA positive cells. (G-H) Within the developing cerebellum, HDAC1 is detected in the white matter (WM), is largely absent from (G) the NeuN positive nuclei of the internal granule layer (IGL) as well as the molecular layer (ML), but is expressed in the Purkinje cell layer (PCL; arrowheads) and external granule layer (EGL). (H) HDAC1 is co-expressed with PCNA (green) in the white matter and EGL (arrowheads). (I) HDAC1 is expressed in GFAP positive (green) glial cells around the DG (arrows) as well as in blood vessels surrounded by GFAP positive astrocytic endfeet (asterisks). (J-K) HDAC1 is expressed in cells arranged in a linear pattern along the corpus callosum which (K) express GFAP (green). (L) HDAC1 is also expressed in GFAP positive glia within the granule cell layer (Gr; arrows) and glomerular layer (Gl) of the OB, as well as in the mitral cells (M; arrowhead). (M) HDAC1 is also co-expressed with GFAP in the white matter of the cerebellum. Scale bars (A-C) = 200 µm, (D-M) = 100 µm.  73  74  established at E13.5 (Figure 4.3C). Within the hippocampus, HDAC1 is co-expressed in a subpopulation of NeuN positive neurons within the CA2 and CA3 sub fields of Ammon’s horn, as well as in some of the granule neurons of the dentate gyrus; however, HDAC1 is not detected within the CA1 region (Figure 4.3D). Within the cortex, on the other hand, HDAC1 is excluded from NeuN positive neuronal nuclei and is instead expressed in small nuclei immediately adjacent to neurons (Figure 4.3E). HDAC1 is also co-expressed in some PCNA positive proliferating cells within the cortex, some of which may be part of blood vessels (Figure 4.3F). Similarly within the developing cerebellum, HDAC1 is not detected within the NeuN positive neuronal nuclei of the internal granule layer, but it is expressed in the white matter as well as in the proliferating progenitors in the externa l granule layer (Figure 4.3G,H). The extensive expression of HDAC1 in non-neuronal, non-proliferating cells throughout the brain suggests that HDAC1 is expressed in glia, in addition to its expression in progenitors. To directly examine this, we examined co-expression of HDAC1 and glial fibrillary acid protein (GFAP). HDAC1 is co-expressed with GFAP in the white matter surrounding the hippocampal formation, as well as in the corpus callosum (Figure 4.3I-K). HDAC1 is also detected in glial cells in the olfactory bulb and the white matter of the cerebellum (Figure 4.3L,M). In addition to glia, HDAC1 is detected long chains of cells associated with GFAP positive fibres (Figure 4.3I). These are likely blood vessels, surrounded by GFAP positive astrocytic endfeet.  As in the embryo, the expression of HDAC2 diverges substantially from HDAC1 in the P7 mouse brain. HDAC2 is highly expressed throughout the brain, including the cortex, hippocampus, olfactory bulb and cerebellum, but it is not detected in the corpus callosum or RMS (Figure 4.4A-C). Within the cortex, the majority of NeuN positive neuronal nuclei coexpress HDAC2; HDAC2 is only occasionally detected in nuclei that do not highly express NeuN, likely indicating immature neurons (Figure 4.4D-E). HDAC2 is also detected in the NeuN positive neuronal nuclei throughout the CA1, CA2 and CA3 regions of the hippocampus, as well as in the granule cells of the dentate gyrus, both the NeuN positive mature neurons and the NeuN negative immature neurons (Figure 4.4F). Within the olfactory bulb, HDAC2 is expressed within the NeuN positive neurons of the granule cell layer, as well as the Mitral/Tufted cells and periglomerular neurons (Figure 4.4G), which do not express NeuN (Mullen et al., 1992; Winner et al., 2002). Similarly within the cerebellum, HDAC2 is detected in the NeuN  75  Figure 4.4: HDAC2 is expressed in neurons throughout the postnatal day 7 brain (A) HDAC2 is highly expressed throughout the cortex and hippocampus, but is not detected in the corpus callosum (CC). (B) HDAC2 is also not detected around the lateral ventricle (LV) or in the rostral migratory stream (RMS), but it is highly expressed throughout the olfactory bulb (OB). (C) HDAC2 is expressed throughout the developing cerebellum. The box indicates the region magnified in (H,I). (DI,K) HDAC2 is in red. (D-E) HDAC2 is expressed in the nuclei of NeuN positive (green) neurons throughout the cortex. (E) In addition to the co-expression of HDAC2 and NeuN in large neuronal nuclei (arrow), HDAC2 is detected in rare NeuN negative nuclei (arrowhead). (F) HDAC2 is expressed in NeuN positive neurons throughout the hippocampus, as well as in NeuN negative immature granule neurons in the dentate gyrus (DG). (G) Within the OB, HDAC2 is co-expressed with NeuN in the neurons of the granule cell layer (Gr), as well as in the NeuN negative mitral cells (M; arrowheads) and periglomerular neurons (arrows) of the glomerular layer (Gl). (H) HDAC2 is weakly expressed in the white matter (WM) of the cerebellum, and is highly expressed in the NeuN positive neurons of the internal granule layer as well as the NeuN negative neurons of the Purkinje cell layer (PCL; arrowheads), molecular layer (ML) and external granule layer (EGL). (I) HDAC2 is co-expressed with PCNA (green) in the EGL (arrows). (J-K) HDAC2 is not detected in the GFAP positive (K; green) glia of the corpus callosum (arrowheads). Scale bars (A-C) = 200 µm, (D-K) = 100 µm.  76  positive neuronal nuclei of the internal granule layer, as well as in the large, NeuN negative, Purkinje cells (Mullen et al., 1992) (Figure 4.4H). In the developing cerebellum, HDAC2 is also expressed in the proliferating progenitors of the external granule layer, as well as some PCNA positive cells within the white matter (Figure 4.4I). HDAC2 is not, however, detected within the white matter of the corpus callosum (Figure 4.4J,K).  4.2.3 HDAC1 Expression is Primarily Glial While HDAC2 is Expressed in Neurons Throughout the Adult Brain  HDAC1 and HDAC2 are both highly expressed in the developing nervous system, albeit it in divergent developmental stages and lineages. To determine if the roles of HDAC1 and HDAC2 are limited to the development of the brain, or if they are likely necessary for the maintenance and/or function of the mature neurons and glia, we examined their expression in the adult mouse brain. Both HDAC1 and HDAC2 maintain high expression in restricted cell types throughout the adult brain, in a pattern similar to that observed during development. In the adult brain, HDAC1 is again expressed in scattered cells within the cortex, corpus callosum and hippocampus (Figure 4.5A). Within the hippocampus, HDAC1 is detected within the dentate gyrus and CA1 and CA2 subfields, but is largely absent from the CA3 subfield. This differs from P7, where HDAC1 is expressed in CA3, but not CA1. HDAC1 is co-expressed in a subpopulation of the NeuN positive neuronal nuclei of the CA1 pyramidal neurons as well some granule neurons of the dentate gyrus (Figure 4.5B-C). HDAC1 is also expressed in the NeuN negative cells of the subgranular zone (SGZ) of the dentate gyrus, which are likely progenitors (Figure 4.5C). Some HDAC1 expressing cells adjacent to the CA1 neurons co-express GFAP, and HDAC1 is also expressed within glia of the corpus callosum (Figure 4.5D,E). Within the cortex, HDAC1 expression is restricted to non- neuronal cells, often found in small nuclei immediately adjacent to the large nuclei of the pyramidal neurons (Figure 4.5F). HDAC1 is also expressed in nonneuronal cells throughout the OB, some of which co-express the oligodendrocyte marker CNPase (Figure 4.5G-J). In the cerebellum, HDAC1 is also detected in scattered NeuN negative cells within the granule cell layer; however, it is most highly detected in a layer of cells bordering the granule and Purkinje cell layers, likely the Bergmann Glia (Figure 4.5K-M). HDAC1 is also detected in a subset of the PCNA posit ive, proliferating cells in the SVZ of the  77  Figure 4.5: HDAC1 is expressed in glia, progenitors, and some neurons in the adult brain (A) HDAC1 is expressed in scattered cells throughout the cortex as well as the CA1, CA2 and dentate gyrus (DG) regions of the hippocampus and the white matter around the hippocampus. (B-F,H-J,L-N,PR) HDAC1 is in red. (B-C) HDAC1 is detected in some NeuN positive neurons (green; arrows) in both the (B) CA1 and (C) DG regions of the hippocampus, as well as in NeuN negative cells around the neurons (arrowheads). (D) HDAC1 is expressed in GFAP positive (green) glia around the CA1 region of the hippocampus (E) as well as in the corpus callosum (CC). (F) HDAC1 is not expressed in most NeuN positive (green) neurons of the cortex, but is detected in small nuclei adjacent to the neurons (arrowhead). (G) HDAC1 is expressed in scattered cells in the granule (Gr), mitral (M) and glomerular (Gl) cell layers as well as the external plexiform layer (EPL) of the OB. (H) The most highly HDAC1 expressing cells in the granule and mitral cell layers of the OB do not co-express NeuN (green; arrowheads). (I-J) A subpopulation of HDAC1 expressing cells in the (I) granule cell layer and (J) glomerular layer of the OB co-express the oligodendrocyte marker CNPase (green; arrows). (K) HDAC1 is weakly detected in the white matter (WM) and granule cell layer (Gr) of the cerebellum. It is not detected in the molecular layer (ML), but is highly expressed in the Purkinje cell layer (PCL). (L) HDAC1 is not co-expressed with NeuN (green) in the granule cell layer (arrowhead) or Purkinje cell layer of the cerebellum. (M) Higher magnification of the box in (L). HDAC1 positive nuclei are adjacent to the NeuN positive cells, and they appear to surround the NeuN negative purkinje neurons (arrow). (N) HDAC1 is co-expressed with PCNA (green) in some cells lining the lateral ventricle (LV) and along the rostral migratory stream (RMS). Inset is a higher magnification of the boxed region of the LV. (O) HDAC1 cells are expressed along the RMS and (P) some of the HDAC1 positive cells co-express PCNA (green; arrow), but not all HDAC1 positive cells are PCNA positive (arrowhead). (Q-R) Many HDAC1 positive cells (Q, white; R, red) along the RMS, and in the corpus callosum, co-express GFAP (green; arrow).However, there are many DAPI (blue) positive nuclei in the RMS that do not express either HDAC1 or GFAP (arrowhead). Scale bars (A,K,N) = 200 µm, (B-J,L-M, N inset, O-R) = 100 µm.  78  79  lateral ventricle and extending along the RMS (Figure 4.5N). Within the RMS, a subpopulation of HDAC1 positive cells do not co-express PCNA (Figure 4.5O,P). Some of these nonproliferating HDAC1 positive cells lining the RMS co-express GFAP (Figure 4.5Q,R); however, there are many nuc lei in the RMS that express neither HDAC1 nor GFAP (Figure 4.5R).  As is the case developmentally, HDAC2 is expressed in most NeuN positive neurons of the adult brain. HDAC2 is highly expressed in the cortex and all regions of the hippocampus, but is not detected in the NeuN negative white matter (Figure 4.6A,B). Within both the cortex and the hippocampus, the majority of NeuN positive neurons co-express HDAC2; however, HDAC2 is also detected in a small number of cells in which NeuN is not detectable (Figure 4.6C-E). This includes a small number of HDAC2 expressing cells in the SGZ of the dentate gyrus, some of which may co-express GFAP (Figure 4.5E-F). HDAC2 is not, however, commonly detected within GFAP positive glia in the white matter around the hippocampus (Figure 4.6G-H). HDAC2 is expressed throughout both the cerebellum and OB, detected within the NeuN positive granule cells of both, as well as the large, NeuN negative Purkinje cell and Mitral cell neurons. HDAC2 is also expressed in the neurons of the external layers of both the OB and cerebellum, the periglomerular neurons, and basket cells and satellite ne urons, respectively (Figure 4.6I-N). In addition to its strong expression in neurons, HDAC2 is weakly detected in the SVZ and RMS (Figure 4.6O). HDAC2 is detected within some GFAP positive glial cells along the RMS, but expression is low relative to the nearby ne urons (Figure 4.6P-Q).  4.2.4 MeCP2 is Expressed in NeuN Positive Neurons Throughout the Postnatal Day 7 Brain, While MBD2 Displays a More Restricted Neuronal Expression  Both HDAC1 and HDAC2 are members of co-repressor complexes recruited by MeCP2 (Sin3A) and MBD2 (Mi-2/NuRD) (Jones et al., 1998; Nan et al., 1998; Ng et al., 1999). Within the OE, expression of MBD2 and MeCP2 is initiated sequentially, with co-expression of MBD2a and MeCP2 observed in the mature ORNs. The expression of both MBD2 and MeCP2 partially overlaps with the divergently expressed HDAC1 and HDAC2, suggesting that the MBDs do not always associate with HDAC1 and HDAC2 simultaneously within the OE. HDAC1 and HDAC2  80  Figure 4.6: HDAC2 is expressed in neurons throughout the adult brain (A) HDAC2 is expressed in the cortex and throughout the regions of the hippocampus, including the dentate gyrus (DG). (B-H,J-K,M-Q) HDAC2 is in red. (B) HDAC2 is co-expressed with NeuN (green) in the cortex and throughout the hippocampus, but is not detected in the NeuN negative white matter of the corpus callosum and around the hippocampus. (C) HDAC2 is co-expressed with NeuN in neuronal nuclei (arrow) throughout the cortex, with rare NeuN negative, HDAC2 positive nuclei (arrowhead) observed. (D-E) Occasional HDAC2 positive, NeuN negative cells are also observed in (D) the CA3 region and (E) dentate gyrus of the hippocampus (arrowheads). (F) Most GFAP positive (green) glia around the dentate gyrus do not co-express HDAC2 (arrowhead), but occasional cells in the subgranular zone appear to coexpress HDAC2 and GFAP (arrow). (G-H) Nuclei are counterstained with DAPI in blue. (G) HDAC2 positive cells (arrowhead) in the white matter around the hippocampus do not express GFAP and (H) HDAC2 is not detected in blood vessels, surrounded by GFAP positive astrocytic endfeet. (I) HDAC2 is detected in the molecular layer (ML), purkinje cell layer (PCL) and granule cell layer (Gr) of the cerebellum, but not the white matter (WM). Box indicates area magnified in (J). (J) HDAC2 is coexpressed with NeuN in the granule cells and is also expressed in the NeuN negative Purkinje cells (arrowhead), as well as (K) within neurons of the molecular layer (arrow). (K) is a higher magnification of the boxed region in (J). (L) HDAC2 is expressed throughout the granule (Gr), mitral (M) and glomerular (Gl) layers of the OB, with scattered cells apparent in the external plexiform layer (EPL). (M) HDAC2 is co-expressed with NeuN in granule neurons and is also expressed in the large, NeuN negative mitral cells (arrowhead) as well as (N) the periglomerular neurons. (O) Weak HDAC2 expression is detected in the cells around the lateral ventricle (LV), with most HDAC2 expression restricted to the nearby NeuN positive neurons. (P-Q) HDAC2 is weakly detected within some GFAP positive (Q; green) glia along the rostral migratory stream (RMS). Scale bars (A,I) = 200 µm, (B-H,J-Q) = 100 µm.  81  82  display a similar divergence in expression in the CNS; therefore, we next examined the expression of MeCP2 and MBD2 in the postnatal brain to determine if either MBD might associate with HDAC1 and HDAC2.  In agreement with previously published findings and our observations within the OE, we found MeCP2 expression to be largely restricted to neurons, where it is upregulated with maturation (Shahbazian et al., 2002; Jung et al., 2003; Kishi and Macklis, 2004; Mullaney et al., 2004). MeCP2 is detected throughout the hippocampal formation at postnatal day 7, but it is not detected in the corpus callosum (Figure 4.7A). Within the developing dentate gyrus, MeCP2 is most highly expressed in the mature, NeuN positive neurons with much weaker detection in the immature granule neurons (Figure 4.7B-C). MeCP2 expression is restricted to the Purkinje cell layer in the developing cerebellum; it is not detected in the NeuN positive neurons of the internal granule layer (Figure 4.7D-E). MeCP2 is expressed in the NeuN positive neurons throughout the cortex (Figure 4.7F), as well as all layers of the OB, including the granule cells, Mitral cells and periglomerular neurons (Figure 4.7H-I).  While little is known about MBD2 expression in the postnatal brain, Mbd2 transcripts have been detected in the adult hippocampus, where Mbd2 is up-regulated following ischemia (Jung et al., 2002). In transgenic mice in which ß-galactosidase is expressed under the Mbd2 promoter, expression is also detected throughout the hippocampus at P7 (Figure 4.7J). Many NeuN positive neurons in the neuronal layers of the CA1, CA2 and CA3 subfields express MBD2 (Figure 4.7K), but detection of MBD2 in the dentate gyrus is restricted to a subset of NeuN positive granule neurons (Figure 4.7L). Mbd2 is not highly detected in the cortex, but Mbd2 is expressed in many subcortical regions, including thalamic and habenular nuclei, as well as many projection neuron nuclei within the brain stem (Figure 4.7M,N). Mbd2 is also detected in the Mitral cells of the OB, as well as in occasional granule neurons and some periglomerular neurons (Figure 4.7O). Similar to MeCP2, expression of Mbd2 in the cerebellum is restricted to the Purkinje cells. This restricted expression is observed both with an antibody against MBD2 and by ßgalactosidase detection (Figure 4.7P-R).  83  Figure 4.7: MeCP2 is expressed in neurons througho ut the postnatal day 7 brain while MBD2 demonstrates a more restricted neuronal expression (A) MeCP2 is expressed throughout the hippocampus, but it is not detected in the corpus callosum (CC). (B,C) Higher magnification of the dentate gyrus (DG) region boxed in (A). MeCP2 (red) is most highly expressed in the NeuN positive (green) mature granule neurons (arrow), but it is also expressed in the NeuN negative immature neurons (arrowhead). (D) MeCP2 is expressed in the Purkinje cell layer (PCL) of the cerebellum, but not the granule cell (Gr) or molecular layers (ML). (E) Higher magnification of the indicated region in (D). Expression of MeCP2 is restricted to the NeuN negative Purkinje neurons (arrowhead). (F) MeCP2 is expressed in NeuN positive neuronal nuclei throughout the cortex. (G) MeCP2 is expressed throughout the granule (Gr), mitral (M), external plexiform (EPL) and glomerular (Gl) layers of the OB. (H) MeCP2 is co-expressed with NeuN in the granule neurons and is also expressed in the large, NeuN negative mitral cell nuclei (arrowhead), as well as (I) the periglomerular neurons, some of which do co-express NeuN (arrow). (J-O,R) ß-galactosidase is expressed under the Mbd2 promoter and LacZ histochemistry (blue) detects Mbd2 expression. (J) Mbd2 is expressed throughout the neuronal layers of the hippocampus. (K) Expression of MBD2 (red) is detected in some NeuN (green) positive mature neurons within the dentate gyrus (arrow), but many NeuN positive neurons do no co-express MBD2 (arrowhead). (L) MBD2 (red) is detected within most NeuN positive (green) neurons of the CA2 region of the hippocampus. (M) Mbd2 is not detected in the cortex, but it is detected (N) in many projection neuron nuclei throughout the brain stem. (O) Within the OB, Mbd2 is not detected in the granule neurons (arrowhead), but it is expressed in the mitral cells and some periglomerular neurons (arrow). (P-Q) An antibody against MBD2 only detects cerebeller MBD2 expression in the large nuclei of the Purkinje neurons, which (Q) are NeuN negative. (Q) is a higher magnification of the area indicated in (P). (R) LacZ histochemistry also detects Mbd2 expression restricted to the Purkinje neurons (arrow) within the cerebellum. Scale bars = 100 µm.  84  85  4.3 Discussion  HDAC1 and HDAC2 are commonly reported to simultaneously associate with the same repressor complexes, including the Sin3A, Mi-2/NuRD and co-REST repressor complexes (Jones et al., 1998; Nan et al., 1998; Wade et al., 1999; Zhang et al., 1999; Ballas et al., 2001). However, HDAC1 and HDAC2 display divergent expression patterns; both in the OE (Figures 3.6-3.8) and the CNS (Figures 4.1-4.6). In general, HDAC1 is expressed in progenitors, glia and a small subset of neurons; HDAC2 is detected in some progenitors, but it is most highly expressed in neurons. This argues against a simultaneous requirement for HDAC1 and HDAC2 within certain repressor complexes, suggesting instead that HDAC1 and HDAC2 may have different activities and their individual recruitment may provide specificity to a given repressor complex. HDAC1 is not essential for HDAC2 enzymatic function in embryonic stem (ES) cells, and over-expression of HDAC2 cannot compensate for loss of HDAC1 in the regulation of ES cell proliferation (Lagger et al., 2002), further supporting different functions for these two HDACs.  HDAC1 is also highly expressed in proliferating cells of the haematopoietic system (Bartl et al., 1997) and it accounts for much of the HDAC activity in mouse embryonic stem cells (Lagger et al., 2002). HDAC1 null mice die by E9.5, indicating that HDAC1 is absolutely essential for early mammalian development (Lagger et al., 2002). The role HDAC1 plays in cell cycle regulation is, however, not as straightforward. The HDAC1 null mouse embryos display severe growth retardation wit h a significant reduction in the number of proliferating cells (Lagger et al., 2002). An HDAC1 mutation in the zebrafish retina, on the other hand, leads to a failure of retinal progenitors to exit the cell cycle, causing an increase in the number of proliferating cells (Yamaguchi et al., 2005). The effect of HDAC1 on proliferation is, therefore, context dependent, likely relying directly on the protein complex with which it is associated. HDAC1 can associate with many different repressor complexes, including, but not restricted to, those that bind methylated DNA. For example, HDAC1 is recruited by the retinoblastoma protein (Rb) to repress proliferation-associated genes (reviewed in (Cress and Seto, 2000; Zhang and Dean, 2001)), a process critical for proper neuronal development (Ferguson and Slack, 2001).  86  HDAC1 is also highly expressed in glia throughout the postnatal brain (Figures 4.3,4.5). Deacetylation is critical for glial differentiation (Hsieh et al., 2004; Laeng et al., 2004; Shen et al., 2005; Siebzehnrubl et al., 2007) and glia differentiated in vitro from adult hippocampal progenitors have lower levels of acetylated histones than progenitors or neurons, suggesting that deacetylation is most prominent in the glial lineage (Hsieh et al., 2004). Furthermore, histone deacetylation is required for oligodendrocyte differentiation in the rat corpus callosum (Shen et al., 2005). Treatment of rats with VPA during the first 2 weeks postnatal induces a stalling of oligodendrocyte differentiation and hypomyelination in the corpus callosum. This stall is not permanent, however, as the precursors differentiate following a two day recovery from VPA. Treatment with VPA after the third week postnatal, when differentiation is largely complete, has no effect on the expression of stage-specific markers or myelination (Shen et al., 2005). This suggests a critical time window and transitional function for histone deacetylation in the differentiation of oligodendrocytes. HDAC1 likely plays a critical role, therefore, in the differentiation and function of glia.  The functio n of HDAC1 in the differentiation of glia is unlikely mediated by MBD proteins, as MBD2, MeCP2 (Figure 4.7) and MBD1 (Zhao et al., 2003) are not expressed in glia. HDAC1 could, however, be recruited by the REST complex in glia. In terminally differentiated, nonneuronal cells, REST represses transcription of neuronal genes by binding to a 23 bp conserved motif, known as RE1 (Repressor Element 1) (Chong et al., 1995; Schoenherr and Anderson, 1995), and recruiting repressor complexes containing HDAC1 and HDAC2 (reviewed in Ballas and Mandel, 2005). HDAC1 may, therefore, be the primary deacetylase component of the REST repressor complex within glia. As mentioned previously, the REST complex also functions to maintain neuronal genes in an inactive state, but nonetheless poised for expression in neural progenitors (Ballas et al., 2005). Therefore, HDAC1 may also be the primary deacetylase within the REST complex in the neural progenitor population. HDAC2 could also be a component of the REST complex in progenitors (Figure 4.2), but not likely within most glia (Figures 4.4,4.6). Interestingly, REST itself must be silenced in mature neurons (Paquette et al., 2000; Ballas et al., 2001), and this occurs in an HDAC-dependent manner, through the recruitment of a repressor complex containing HDACs, mSin3A, CoREST and MeCP2 (Ballas et al., 2005). With its high expression in neurons, HDAC2 is likely the primary deacetylase recruited by the CoREST  87  complex to silence REST in mature neurons, while HDAC1 associates with REST to silence neuronal genes in glia.  Despite the largely divergent developmental expression of HDAC1 and HDAC2, they can associate with many of the same repressor complexes, further evidenced by their coimmunoprecipitation in the OE (Figure 3.8). Within the developing nervous system, they could be simultaneously recruited by a repressor complex in some neural progenitors, where expression of HDAC2 appears to be initiated as HDAC1 is down-regulated. This is most clearly observed in the OE where most cycling basal cells express HDAC1, while HDAC2 is detected in a subpopulation of cycling cells, with the expression of HDAC2 up-regulated in the post-mitotic, immature neurons. Thus, HDAC1 alone might be a component of repressor complexes regulating the cell cycle or maintaining an undifferentiated state, with HDAC1 and HDAC2 acting together in a distinct repressor complex to mediate lineage commitment and differentiation (see Figure 4.8). Alternatively, HDAC2 might oppose or over-ride the action of HDAC1 in progenitors to promote neuronal differentiation. Following lineage commitment, HDAC1 is likely necessary for glial differentiation and/or function, potentially restricting neuronal gene expression as part of the REST complex. HDAC2, on the other hand, is highly expressed in neurons; therefore, HDAC2 is likely involved in neuronal differentiation and function. In this capacity, HDAC2 could play multiple roles through its potential interaction with different repressor complexes, including MeCP2-Sin3A, MBD2-MeCP1, Mi2/NuRD and CoREST.  The work presented here highlights the divergent expression of HDAC1 and HDAC2 during the development of the nervous system and identifies key developmental stages and transitions that may be mediated by each HDAC, likely through association with distinct repressor complexes. A major limitation of this study is, however, its reliance on a single experimental approach – immunohistochemistry – and the use of a single antibody against each HDAC. The HDAC1 and HDAC2 antibodies employed were validated by Western blotting and immunoprecipitation (see Table 2.3 and Figure 3.8), and both were found to cleanly and robustly recognize band s of the appropriate size (HDAC1 = 65kD, HDAC2 = 55kD) in protein extracts from both the OE and the brain (data not shown). A different antibody against HDAC2 (rabbit polyclonal, 1:200; Santa  88  Figure 4.8: Summary of HDAC1 and HDAC2 expression in the developing nervous system In both the OE and the brain, HDAC1 is highly expressed in stem cells and progenitors, where it likely acts to regulate self-renewal and cell cycle, possibly through association with different repressor complexes. Expression of HDAC2 is initiated in some proliferating progenitors. Here, it might associate with HDAC1 in a distinct repressor complex to inhibit cell-cycle re-entry and promote neuronal differentiation. Alternatively, HDAC2 might oppose or over-ride the action of HDAC1 in progenitors in order to promote neuronal differentiation. HDAC2 might then associate with separate repressor complexes, including MBD2 and MeCP2, to promote neuronal differentiation and maintain mature neurons. HDAC1, on the other hand, likely mediates glial differentiation and function, potentially through an association with the REST complex.  89  Cruz), was also tested by immunohistochemistry in the OE (data not shown), and demonstrated a similar pattern of expression as was observed with the rabbit polyclonal anti-HDAC2 antibody (1:5000; Abcam). Furthermore, a similar pattern of HDAC1 and HDAC2 expression was also identified in the early postnatal rat corpus callosum using different antibodies from those we employed (Shen et al., 2005). In the P5 rat corpus callosum, the majority of HDAC2 expression is restricted to neighbouring NeuN positive neurons, while most oligodendrocytes in the corpus callosum express HDAC1. Taken together, I am confident that this study accurately reflects the primary expression pattern of HDAC1 and HDAC2 in the developing mouse brain. However, it cannot be ruled out that the expression of either HDAC is simply below the detection threshold of these antibodies in the cell types in which it is not identified. A more sensitive experimental approach, such as quantitative reverse-transcription PCR on purified cell populations, would be needed to definitively determine if HDAC1 is expressed at low levels in cortical pyramidal neurons, for example.  HDAC1 and HDAC2 are only two me mbers of a large family of deacetylase proteins, but their expression alone encompasses most developmental stages and cell types within the nervous system, including progenitors, neurons, and glia. Not only does this highlight the significance of epigenetic regulation in neuronal development, it should serve as a caution when employing broad HDAC inhibitors as therapeutic agents. HDAC inhibitors are likely to have a devastating effect not only on ne ural development, but also adult brain function and postnatal neurogenesis. Therefore, the function of epigenetic gene silencing at distinct stages of lineage commitment and neural development must be understood. While a few subtle differences in MBD and HDAC expression do exist between the OE and the CNS, a sequential expression of different repressor complex members at distinct transitional stages of development is evident both. Thus, the OE provides an excellent model system in which to elucidate the role of DNA methylationdependent gene silencing in the stage-specific differentiation of a neuron.  90  Chapter 5: MBD2 and MeCP2 Null Mice Display Stage-Specific Defects in Olfactory Neurogenesis 5.1 Introduction  The expression and biochemistry data suggest defined transitional stages of ORN differentiation that may utilize DNA methylation-dependent gene silencing, either to traverse that developmental step or to maintain a neuron at the next developmental stage and prevent reexpression of genes from earlier developmental stages. Again, these transitions occur (1) when HBCs transition into mitosis (MeCP2), (2) as mitotic GBCs exit mitosis and commit to the neuronal lineage (MBD2b), and (3) as immature ORNs lose developmental plasticity and transition into a mature, stable state (MBD2a and MeCP2). Therefore, if MBD2 or MeCP2 expression is perturbed, there should be disruptions in ORN differentiation corresponding to these developmental transitions. Furthermore, the homeostasis of the olfactory epithelium may be disrupted. Within the adult OE, tightly controlled regulatory feedback mechanisms maintain a stereotyped complement of cells at different developmental stages, including cycling progenitors, immature neurons, mature ORNs as well as degenerating ORNs, all in a defined laminar structure with a developmental hierarchy (Figure 1.4). In the absence of MBD2 or MeCP2, this organization may also be perturbed.  To determine if MBD2 and MeCP2 are necessary and/or sufficient for ORNs to traverse the distinct developmental stages of differentiation outlined above, we employed Mbd2 and Mecp2 knockout (KO) mice. The Mbd2 knockout allele was generated by replacing exon 2 of the Mbd2 gene with the promoterless ßgeo cassette (Hendrich et al., 2001). Transcription proceeds as normal through exon 1 and intron 1 (see Figure 3.3a), but terminates at the transcription stop site located in the ßgeo cassette. The resulting transcript can encode the N-terminal 183 amino acids of MBD2, with translation stopping in the middle of the conserved methyl binding domain (MBD). Any function of MBD2 encoded solely by exon 1 is unlikely to involve targeted repression of methylated DNA as only the N-terminal half of the MBD is encoded by exon 1, and the repression domain and putative coiled coil domain are encoded by downstream exons (Hendrich et al., 1999; Boeke et al., 2000; Hendrich et al., 2001). The resulting MBD2 null mice  91  are reported to be viable, fertile and of normal appearance, but display defective maternal behaviour (Hendric h et al., 2001). Within our colony, adult male MBD2 null mice have an approximate 15% reduction in body weight compared to wildtype and heterozygous littermates (Figure 5.1a).  To generate an Mecp2 knockout allele, exons 3 and 4 of the Mecp2 gene were replaced in ES cells with the same exons flanked by loxP sites (Guy et al., 2001). Early embryonic deletion of the gene was then achieved by crossing Mecp2lox/lox females with deleter mice, which express Cre ubiquitously. Recombination between loxP sites subsequently deleted all but the N-terminal eight amino acids of MeCP2 (Guy et al., 2001). The phenotype of both hemizygous Mecp2 null males and heterozygous females is outlined extensively in Chapter 1. For our analysis, we chose to focus on the hemizygous males as they display a much more consistent and severe neurological phenotype, with a much earlier onset. The MeCP2 null males develop symptoms between 3 and 8 weeks, with a variable progression of symptoms leading to death at approximately 54 days (Guy et al., 2001). The MeCP2 null males are significantly smaller than wildtype littermates by three weeks of age, with the discrepancy in body weight increasing at 7 weeks of age (Figure 5.1b), likely due to the symptomatic progression.  5.2 Results  5.2.1  MBD2 and MeCP2 Null Mice Display Stage-Specific Perturbations in ORN  Differentiatio n, Corresponding to Their Observed Expression Patterns  We first examined the adult OE as the laminar structure should be well established and the homeostasis between ORN death and regeneration under tight regulation. Consistent populations of dividing, PCNA positive basal cells, GAP43 positive immature ORNs, and OMP positive mature ORNs, are present in the wildtype adult olfactory epithelium (Figure 5.2). The MBD2 null OE contains a significantly higher number of PCNA positive, dividing basal cells compared to wildtype (Figure 5.2A,B,G), in addition to a small, but significant decrease in the number of OMP positive mature ORNs (Figure 5.2D,E,I). In contrast, MeCP2 null mice have no significant  92  Figure 5.1: Decreased body weight of MBD2 and MeCP2 null mice (A) Adult (7 weeks old) male MBD2 null mice are approximately 15% smaller than wildtype and heterozygous littermates. (B) MeCP2 null male mice are significantly smaller than wildtype littermates as of 3 weeks of age, with the discrepancy in weight increasing by 7 weeks of age. No difference in body weight is observed at 1 week of age. Error bars denote standard error of the mean. * = p < 0.01, ** = p < 0.001  93  Figure 5.2: Adult male MBD2 and MeCP2 null mice display stage-specific perturbations in olfactory neurogenesis (A-F) The olfactory epithelium of adult (B) MBD2 null mice displays an increased number of PCNA positive basal cells compared to (A) wildtype and (C) MeCP2 null mice. (E) MBD2 nulls appear to have ratios of OMP+/GAP-43+ neurons comparable to (D) wildtype while (F) MeCP2 nulls display an increased number of GAP-43 positive neurons. (G-I) The total number of cells expressing (G) PCNA, (H) GAP-43 and (I) OMP was quantified per linear mm of OE in 3-4 male animals of each genotype. The number of PCNA positive basal cells is significantly greater than wildtype in MBD2 null mice, while the number of GAP-43 positive neurons is significantly increased in MeCP2 nulls. The number of OMP positive neurons is significantly decreased in MBD2 nulls. * = p‹0.05, ** = p<0.01. BC = basal cells, IRN = immature receptor neurons, ORN = olfactory receptor neuron. Scale bars = 50 µm  94  change in the number of either PCNA positive cycling progenitors or OMP positive ORNs, compared to wildtype. However, MeCP2 nulls contain a significantly higher number of GAP43 positive immature receptor neurons compared to wildtype (Figure 5.2A,D,H). In addition to its expression in neurons, MeCP2 is expressed in a subpopulation of HBCs (Figure 3.4I). No significant change in the number of HBCs is detected in MeCP2 null mice, however (WT = 144.4 ± 10.2, MeCP2 KO = 150.4 ± 9.4 ICAM1 positive HBCs per linear mm of OE).  The total number of cells, as measured by DAPI positive nuclei per linear millimetre of OE, is not increased in the OE of MeCP2 nulls (WT = 1802.9 ± 102.4, MBD2 null = 1821.9 ± 73.8, MeCP2 null = 1789 ± 62.3). This suggests that rather than a build- up of immature receptor neurons with a normal number of mature ORNs, MeCP2 null mice have an increase in transitional neurons, which still maintain expression of the immature marker GAP43 but have also initiated expression of the mature neuronal marker OMP. The co-expression of GAP43 and OMP is, in fact, commonly observed in the MeCP2 null OE at the juncture between GAP43 and OMP expressing layers of the OE. The co-expression of GAP43 and OMP is particularly evident during postnatal development, concomitant with disrupted laminar positioning of neurons expressing GAP43 and OMP (Figure 5.3C-H). A significant increase in GAP43 positive neurons is first evident at three weeks of age in the OE of MeCP2 null males (Figure 5.3A), the age at which neurological symptoms are first exhibited (Guy et al., 2001). No difference in the number of OMP positive neurons is observed at any of the ages examined (1, 3 and 7 weeks postnatal; Figure 5.3B), indicating that MeCP2 null ORNs do not display a developmental delay in expressing mature neuronal markers. MeCP2 null ORNs do, however, appear to demonstrate a transient delay in down-regulating the expression of the immature neuronal marker GAP43.  The increase in PCNA positive globose basal cells relative to mature ORNs observed in the adult MBD2 null OE could indicate either a stalling of progenitors in the cell cycle, or a heightened state of neurogenesis, likely due to an increased death of ORNs feeding back on the progenitors to up-regulate neurogenesis. To test if this imbalance is present developmentally, we quantified the number of PCNA positive cells at increasing postnatal time points. We found a trend towards increased PCNA positive cells in the MBD2 null OE compared to wildtype at 1 and 3 weeks postnatal, but this increase in proliferation is not significant until 7 weeks of age (Figure 5.4A).  95  Figure 5.3: Increase in GAP43 positive neurons and disrupted laminar structure in the postnatal MeCP2 null olfactory epithelium (A) A significant increase in GAP43-positive neurons , compared to wildtype mice, is first observed at 3 weeks postnatal in the MeCP2 null OE, and this increase is maintained into adulthood. (B) No significant difference in the number of OMP positive mature ORNs is evident between wildtype and MeCP2 null at any of the developmental time points examined. (C) In the wildtype OE, distinct laminar positioning is observed between OMP positive (red) mature ORNs and GAP43 positive (green) immature receptor neurons (arrowhead). (D) In the MeCP2 null OE, GAP43 positive (green) immature neurons extend apically into the OMP positive (red) regions of the OE (arrow), and scattered OMP positive neurons are observed co-expressing GAP43 in the more basal regions of the OE (arrowheads). (E-F) GAP43 expression appears more extensive in the (F) MeCP2 null than the (E) wildtype, while (G-H) laminar positioning of OMP positive cells appears disrupted in the (H) MeCP2 nulls compared to (G) wildtype. * = p‹0.05 Scale bars = 50 µm.  96  Figure 5.4: A significant increase in cycling progenitors and decrease in mature ORNs is not observed until 7 weeks of age in the MBD2 null OE (A) A trend towards increased PCNA positive basal cells is observed at 1 week of age, with a significant increase found at 7 weeks of age. (B) There is a trend towards decreased GAP43 positive immature ORNs in the 1 week MBD2 null OE, but no apparent difference at either 3 or 7 weeks. (C) A trend towards decreased OMP positive mature ORNs is observed in the MBD2 null OE at 1 week of age. There is no apparent difference at 3 weeks, but a significant decrease in OMP positive neurons is observed in the 7 week MBD2 null OE. * = p‹0.05  97  This pattern of increasing divergence developmentally in the number of PCNA positive basal cells between wildtype and MBD2 null suggests that there is an increasing up-regulation of neurogenesis in the MBD2 null OE, likely to compensate for a postnatal loss of neurons. A significant decrease in OMP positive neurons is also not observed in the MBD2 null OE until 7 weeks of age (Figure 5.4C). However, there is a trend towards decreased GAP43 positive immature neurons (Figure 5.4B) and OMP positive mature neurons in the 1 week old MBD2 null OE, concomitant with the trend towards increased PCNA positive progenitors. This could suggest a developmental delay in exit from cell cycle and ORN differentiation, in addition to increased neuronal turnover and feedback up-regulation of neurogenesis in the adult, correlating with the observed expression patterns of MBD2b and MBD2a, respectively.  5.2.2 Decreased Retention of BrdU Labelled Cells in the Epithelium of the MBD2 Null Mouse  To directly examine the rate of cell turnover in the OE of MBD2 null mice, we employed a longterm bromodeoxyuridine (BrdU) labelling experiment. We first compared the extent of BrdU incorporation with other markers of actively cycling cells to ensure that a sufficient number of cells would be labelled in this way. Two BrdU injections, two hours apart, label approximately one half of the number of cycling cells as are detected by an antibody against PCNA in the P7 OE (Figure 5.5). This is likely due to the more extensive expression of PCNA during the cell cycle. PCNA is first synthesized during the late G1 to early S-phase of the cell cycle, immediately preceding the onset of DNA synthesis, is most abundant during S-phase and then declines during G2/M-phase (Kurki et al., 1986; Kurki et al., 1988). BrdU is only incorporated into DNA during S-phase, as the DNA is synthesized. Phosphorylated histone H3, a marker of cells in mitosis (Lake and Salzman, 1972; Gurley et al., 1978), is detected in very few cells of the OE, labelling approximately 3% of the number of cells detected by PCNA (Figure 5.5C,D). The majority of cells in the OE detected by PCNA, therefore, are likely in S-phase and two BrdU injections are sufficient to label a high percentage of these cells.  MBD2 and MeCP2 null mice and wildtype littermates received two injections of BrdU, two hours apart, at one week postnatal. The animals were then sacrificed one hour after the second injection (baseline BrdU incorporation), at 3 weeks postnatal, or at 7 weeks postnatal (Figure  98  Figure 5.5: An antibody against PCNA labels more proliferating cells in the postnatal day 7 OE than are detected by BrdU incorporation (A-C) Examples of detection of (A) PCNA, (B) BrdU and (C) phospho-Histone H3 (pH3) in endoturbinate IIa of a postnatal day 7 wildtype mouse. (D) The number of cells expressing PCNA, phospho-Histone H3 or labelled with BrdU was quantified in the OE of a minimum of 3 wildtype and 3 MBD2 null P7 mice. The mice received two BrdU injections, two hours apart, one hour before sacrifice. No significant differences are observed between wildtype and MBD2 null mice. PCNA expression is detected in approximately twice as many cells as are labelled with BrdU. Very few cells expressing the mitotic marker pH3 are detected. Scale bar = 100 µm.  99  5.6A) and the distribution (Figure 5.6B) and number (Figure 5.6C) of BrdU positive cells was examined. MeCP2 nulls, which display no alteration in the cyc ling progenitor or mature ORN populations, were included in the analysis as a comparison. The MBD2 null mice exhibit a trend towards increased BrdU incorporation at 1 week (Figure 5.6B,C), similar to the increase in PCNA positive cells (Figure 5.4A). However, significantly fewer BrdU labelled cells are detected in the OE of the MBD2 null mice by 3 weeks and 7 weeks of age compared to both wildtype and MeCP2 null mice (Figure 5.6B,C). Approximately 50% of the number of cells labelled with BrdU at one week are detected by 7 weeks in the MBD2 null, compared to approximately 87% in the wildtype. The MeCP2 mice showed no significant difference from wildtype in the number of BrdU- labelled cells detected at any time point examined.  5.2.3 MBD2 Null ORNs Display an Increased Rate of Apoptotic Cell Death  To determine if the decreased number of BrdU labelled cells detected in the adult MBD2 null OE is the result of increased cell death, we analyzed the expression of activated caspase 3 (aC3) as an early indicator of apoptotic cell death in ORNs (Cowan et al., 2001). MBD2 nulls have a significant increase in the number of cells that have activated C3 in the olfactory epithelium at all developmental time points analyzed, compared to both wildtype and MeCP2 nulls (Figure 5.7A). This increased cell death could be due to progenitors that are unable to exit the cell cycle and commit to a cell lineage in the absence of MBD2, or it could represent increased death of neurons, thus leading to an overall feedback up-regulation of neurogenesis. To identify the developmental stage of the aC3 positive cells, we used double immunohistochemistry with aC3 and developmental stage-specific ma rkers (Figure 5.7C-H) at three weeks of age. We chose this developmental time point as it contains the highest number of aC3 positive cells and it is a time when the OE is functionally maturing and reaching an adult- like homeostasis. Co-detection of aC3 and PCNA in cycling basal cells is rare in both the wildtype and knockout (Figure 5.7B-D), with most aC3 expression found within the neuronal (both mature and immature) layers of both the wildtype and null OE (Figure 5.7B), as reported previously (Cowan and Roskams, 2004). Frequently, the aC3 expressing cells are found right at the junction of GAP43 positive immature neurons and OMP positive mature ORNs (Figure 5.7E-H), a point of vulnerability in ORN development. However, the number of mature neurons that activate C3 is significantly increased  100  Figure 5.6: Decreased retention of BrdU labelled cells in the MBD2 null OE (A) Wildtype, MBD2 null and MeCP2 null mice received two BrdU injections 2 hours apart on postnatal day 7. The mice were then sacrificed one hour following the second injection, at 3 weeks of age, and at 7 weeks of age. (B) A similar pattern and number of cells appear to incorporate BrdU at 1 week of age in the wildtype, MBD2 null and MeCP2 null OE. Fewer BrdU labelled cells, however, are apparent in the 3 week and 7 week MBD2 null OE, compared to both wildtype and MeCP2 null. (C) The number of BrdU labelled cells in the wildtype, MBD2 null and MeCP2 null OE is quantified at 1, 3 and 7 weeks of age. The MBD2 null shows a trend towards increased BrdU labelling at 1 week. However, significantly fewer BrdU labelled cells remain at 3 and 7 weeks in the MBD2 null compared to wildtype or MeCP2 nulls. * = p‹0.05, ** = p<0.01  101  Figure 5.7: Increased apoptotic cell death of mature ORNs in the MBD2 null OE (A) Using the expression of activated caspase 3 (aC3) as a measure of cell death, the MBD2 null mice display increased cell death, in comparison to wildtype and MeCP2 nulls, at all ages analyzed, peaking at 3 weeks of age. (B) The number of aC3 cells per millimetre of OE co-expressing PCNA, GAP43 and OMP was quantified in MBD2 wildtype and null mice at 3 weeks of age. (C,D) Very few PCNA positive (red) cycling progenitors co-express active caspase 3 (aC3, green), with the majority of aC3 cells found within (E,F) the GAP-43 positive (red) immature neuronal layers or (G,H) the OMP positive (red) mature neurons, with many located right at the border between the two (arrowheads). BC = basal cells, IRNs = immature receptor neurons, ORNs = olfactory receptor neurons, Scale bars = 50 µm. * = p‹0.05, ** = p<0.01  102  in the MBD2 null OE (Figure 5.7B). In the wildtype, the percentage of aC3 positive cells that express OMP is similar to the percent that express GAP43. In the MBD2 null, on the other hand, approximately 61% of the aC3 cells co-express OMP compared to only 37% that co-express GAP43. This suggests that loss of MBD2 results in an unstable mature ORN population.  5.2.4 Aberrant Glomerular Formation in the Absence of Either MBD2 or MeCP2  ORN axons traverse the cribiform plate and target to the OB, where they coalesce into synaptic complexes (glomeruli) according to common odorant receptor expression. Many glomeruli form postnatally, and they are refined in the first three weeks postnatal in an activity-dependent manner. To determine if loss of either MBD2 or MeCP2 results in a disruption of ORN axon targeting and synapse formation, we examined the pattern of glomerular formation of P21 OBs of null mice in comparison to littermate controls. By three weeks of age, the OB has developed a mature laminar structure. Therefore, a developmental delay would be evident at this time point, in addition to structural abnormalities. The OBs of both the MBD2 and MeCP2 null mice appear to be grossly normal and laminar structure is intact. However, the formation of OMP positive glomeruli appears aberrant in both the MBD2 and MeCP2 null mice, albeit with different phenotypic outcomes (Figure 5.8).  In the wildtype OB at P21, most glomeruli have been refined to a mature, adult-like morphology. In a coronal cross-section of the middle of the OB, the majority of OMP positive glomeruli (~75%) measure between 500 and 2500 µm 2 in size, producing a bell curve distribution of average areas (Figure 5.8A,D,G). This distribution of areas likely represents the different glomerular cross-sections detected in a single section of the OB. In the MBD2 null mice, on the other hand, the OMP positive glomeruli are disorganized in appearance and approximately half of the glomeruli measure less than 500 µm2 or greater than 2500 µm2 (Figure 5.8B,E), producing a distribution curve of average glomerular areas that is skewed to both the smallest and largest areas (Figure 5.8H). The OMP positive glomeruli in the MeCP2 knockout mice are also disorganized in appearance, with many large and diffuse glomeruli (Figure 5.8C,F). While the average glomerular area in the MBD2 nulls (1450.1 ± 117.7 µm2 ) does not differ from wildtype  103  Figure 5.8: MBD2 and MeCP2 null mice display aberrant glomerular formation in the postnatal day 21 olfactory bulb (A,D,G) In the wildtype OB at postnatal day 21, the OMP positive glomeruli have been refined to a mature, adult morphology with the majority (~75%) measuring between 500 and 2500 µm2 in size. (B,E,H) In the MBD2 null mice the OMP positive glomeruli are disorganized in appearance with approximately half of the glomeruli measuring less than 500 µm 2 or greater than 2500 µm2 . (C,F,I) The OMP positive glomeruli in the MeCP2 null mice are also disorganized in appearance, with many glomeruli large and diffuse in appearance. (D-F) Higher magnification images of the boxed areas in (AC) respectively. (D-F) Arrowheads indicate abnormally small glomeruli while arrows point to large glomeruli. (G-I) Frequency distribution curves of glomerular sizes in (G) wildtype, (H) MBD2 null and (I) MeCP2 null mice. The area of each OMP positive glomerulus in a matched coronal section of the middle of the OB was measured in 3-4 animals of each genotype. The areas were binned in increments of 250 µm2 , and the number of glomeruli falling into each area range was averaged across the animals of each genotype. Error bars denote SEM. (A-F) Scale bars = 100 µm.  104  (due to a balance between very small and very large glomeruli), the average glomerular area of the MeCP2 nulls (1740.1 ± 66.4 µm 2 ) is significantly larger than wildtype (1398.8 ± 66.4 µm 2 ; p < 0.01) and the average glomerular area distribution curve of the MeCP2 null mice is skewed towards the larger glomeruli (Figure 5.8I). There is no significant difference in the total number of glomeruli represented, and therefore measured, in the matched coronal cross-sections from mice of each genotype (WT = 80.2 ± 6.2, MBD2 KO = 80.3 ± 2.9, MeCP2 KO = 76.7 ± 5.1 glomeruli samp led).  The skewed distribution of very small and very large glomeruli observed in the MBD2 null OBs is already apparent at postnatal day 7 (Figure 5.9G). While a number of large, diffuse glomeruli and many small, forming glomeruli are also apparent in wildtype OBs at P7 (Figure 5.9A,C,E), the MBD2 null OBs appear to contain many more aberrantly shaped glomeruli of extreme sizes, demonstrating immature morphologies (Figure 5.9B,D,F). Coalesced OMP positive axons extending past the glomerular layer and into the external plexiform layer of the OB are also prominent in the MBD2 null OB (Figure 5.9F). In addition, the nerve fibre layer, the layer in which ORN axons travel around the OB until targeting to a glomerulus, appears thicker in the MBD2 null mice (Figure 5.9A,B). Taken together, many MBD2 null ORNs extend axons to the OB and form synaptic contacts. However, a subpopulation of MBD2 null ORNs do not appear to target their axons appropriately or, are delayed in doing so.  5.3 Discussion  The expression patterns of MBD2 and MeCP2 in the OE suggest that they may act sequentially to drive defined transitions in ORN differentiation. This, of course, raises the question – are MBD2 and MeCP2 necessary for ORNs to either traverse these transitional checkpoints in differentiation, or to maintain a cell at the next developmental stage? To explore this question, we first analyzed the OE of the adult MBD2 and MeCP2 null mice. In the wildtype, a tightly controlled homeostasis maintains a highly stereotyped complement of cells at each developmental stage. In the absence of MBD2 or MeCP2, this balance is perturbed. MeCP2 null mice display no disruption in the number of proliferating basal cells, and no significant change in the number of OMP positive mature ORNs. However, the MeCP2 null OE contains a significant  105  Figure 5.9: Disrupted glomerular formation in the postnatal day 7 MBD2 null olfactory bulb (A,B) Both (A) wildtype and (B) MBD2 knockout P7 olfactory bulbs (OB) contain OMP positive glomeruli (Gl) at varied stages of developmental refinement. Small, glomeruli (arrows) and large, diffuse glomeruli (arrowheads) are evident in both. MBD2 null OBs appear to have a thicker nerve fibre layer (NFL) and more immature, aberrantly shaped glomeruli. (C-D) Higher magnification images of the boxed areas in (A-B), respectively. (C) GAP43 positive (red) immature ORN axons are evident throughout the NFL of the wildtype bulb, and extending into the glomeruli (arrow). OMP positive (green) mature axons are evident throughout the glomeruli. (D) OMP positive axons have a diffuse appearance throughout the poorly defined MBD2 null glomeruli and GAP43 positive immature axons appear restricted to the very edges of the glomeruli (arrows). (E) Many refined, mature glomeruli are evident in the wildtype OB, expressing OMP (green) throughout with GAP43 positive (red) immature ORN axons just entering the glomeruli. (F) In the MBD2 null OB, many wandering OMP positive axons (green; arrows) and poorly defined glomeruli are still evident at P7. (G) Frequency distribution curves of glomerular sizes in wildtype and MBD2 null mice at P7. The area of each OMP positive glomerulus in a matched coronal section of the middle of the OB was measured in 3-4 animals of each genotype. The areas were binned in increments of 500 µm2 , and the number of glomeruli falling into each area range was averaged across the animals of each genotype. Error bars denote SEM. (A-F) Scale bars = 100 µm.  106  increase in the number of cells expressing GAP43, a marker of immature receptor neurons (Figure 5.2). The increase in the number of GAP43 positive neurons is first significant at 3 weeks of age (Figure 5.3), concurrent with onset of symptoms (Guy et al., 2001). The divergence between wildtype and MeCP2 null in the number of GAP43 positive neurons widens with symptomatic progression; at 3 weeks of age, MeCP2 null mice display an approximate 1.4 fold increase in the number of GAP43 positive neurons, and by 7 weeks of age, an approximate 1.8 fold difference is evident.  Interestingly, a previous study observed an increase in GAP43 positive immature ORNs only at 2 weeks postnatal, with no difference at 4 weeks and a decrease in GAP43 positive neurons at 7 weeks (Matarazzo et al., 2004). There are several possible explanations for this discrepancy. One possibility is a strain difference. Matarazzo and colleagues employed the MeCP2 null line generated by the Jaenisch laboratory (Chen et al., 2001), while we used a model produced by the Bird laboratory (Guy et al., 2001). Both MeCP2 null lines were generated by Cre- lox technology and develop similar phenotypes, but some differences in severity and temporal progression of symptoms have been observed (Chen et al., 2001; Guy et al., 2001). Another possible explanation is the sampling strategy employed. Neurogenesis is not consistent throughout the OE, with patches of basal cells undergoing proliferation at any one time, interspersed with quiescent regions (Weiler and Farbman, 1997). This pattern is particularly evident in the adult OE, where neurogenesis is reduced, and leads to variability in the number of immature neurons in any given region throughout the OE. Therefore, multiple areas throughout the OE must be sampled to ensure a true representation of neurogenesis. While Matarazzo and colleagues counted two to three fields per section (Matarazzo et al., 2004), the extent of the total OE sampled and the regions analyzed are not made clear, but could account for the variability in their GAP43 findings.  The increase in GAP43 positive neurons does not likely represent increased neurogenesis as there is no change in cell turnover in the MeCP2 null OE (Figures 5.6,5.7), nor an up-regulation in the number of cycling basal cells (Figure 5.2), results that are consistent with previous findings in the MeCP2 null CNS (Kishi and Macklis, 2004; Smrt et al., 2007). Furthermore, there is no significant change in the total number of cells in the adult MeCP2 null OE and no decrease  107  in the number of OMP positive mature neurons (Figure 5.3). This indicates that, rather than a build-up of immature neurons, the MeCP2 null OE exhibits an increase in the number of transitional neurons – neurons that have initiated expression of the mature neuronal marker OMP, but that have not yet down-regulated the immature neuronal marker GAP43. An increase in the number of ORNs appearing to co-express OMP and GAP43 is apparent in the developing MeCP2 null OE, as well as a disruption in the laminar distinction between mature and immature neurons (Figure 5.3).  It is possible that the perpetuation of GAP43 expression does not indicate a general retention of immature neuronal gene expression or characteristics, but rather that GAP43 is a direct target of MeCP2-mediated silencing. The HDAC inhibitor VPA increases the expression of GAP43 (Yuan et al., 2001), indicating that, directly or indirectly, GAP43 can be regulated by HDACs. A similar increase in transitional neurons is also seen in the postnatal MeCP2 null hippocampus, however, using different markers of neuronal maturity. An increase in the number of cells coexpressing the immature neuronal marker doublecortin and the mature neuronal marker NeuN is observed within the MeCP2 null dentate gyrus, with no change in the number of mature, NeuN positive neurons (Smrt et al., 2007). Again, the increase in transitional neurons is more pronounced at 8 weeks than at 4 weeks of age, correlating with symptomatic progression (Chen et al., 2001; Guy et al., 2001). Therefore, it is likely that the prolonged expression of GAP43 in MeCP2 null ORNs is indicative of a general retention of immature neuronal gene expression.  While Rett Syndrome has generally been viewed as a neurodevelopmental disorder, recent evidence that symptomatic progression can be attenuated, or even reversed, by postnatal rescue of MeCP2 expression in the mouse (Giacometti et al., 2007; Guy et al., 2007) suggests that developmental absence of MeCP2 does not irreversibly damage neurons. Thus, Rett may not be strictly a neurodevelopmental disorder, with MeCP2 instead required to stabilize and maintain the mature neuronal state (Guy et al., 2007). The increase in MeCP2 expression as neurons functionally mature (Shahbazian et al., 2002; Cohen et al., 2003; Jung et al., 2003; Kishi and Macklis, 2004; Mullaney et al., 2004), could support this hypothesis. This raises the question, does the increase in transitional ORNs represent a temporary delay in terminal differentiation in the absence of MeCP2, or rather an inability of mature ORNs to maintain a fully differentiated,  108  functionally mature state? While the latter scenario cannot be ruled out, there is substantial evidence to suggest disrupted neuronal development in the absence of MeCP2. First of all, disrupting MeCP2 expression postnatally leads to a much delayed phenotype compared to either a complete or embryonic deletion (Chen et al., 2001; Guy et al., 2001; Gemelli et al., 2006). Conversely, there is a direct correlation between age of onset of a Mecp2 rescue transgene and phenotypic rescue, with embryonic expression, when most neurons are immature, by far the most successful at attenuating the phenotype (Giacometti et al., 2007). MeCP2 has also been implicated in dendritic development and, potentially, synaptogenesis (Kishi and Macklis, 2004; Fukuda et al., 2005; Jugloff et al., 2005; Bienvenu and Chelly, 2006; Moretti et al., 2006; Smrt et al., 2007). Furthermore, the GAP43/OMP double positive ORNs are typically found around the disrupted laminar division between immature and mature ORNs, rather than in the apical layers of the most mature OMP positive ORNs (Figure 5.3).  Taken together, it is likely that the increase in transitional ORNs in the MeCP2 null OE is the result of a transitional delay in terminal differentiation and not a reversion of mature ORNs to an immature phenotype. This delay in differentiation appears to be transient, as the markers of neuronal immaturity are eventually down-regulated and normal numbers of neurons expressing mature neuronal markers are identified, both in the OE (Figures 5.2,5.3) and the hippocampus (Smrt et al., 2007). This suggests that, in the absence of MeCP2, genes associated with neuronal immaturity are not immediately silenced as the neurons transition to a mature state. This does not necessarily imply that expression of these genes remains high within the mature neurons, as lack of repression does not equal gene activation. Rather than being immediately repressed, expression of these genes could gradually drop to basal levels of transcription in the MeCP2 null. For example, the absence of MeCP2 leads to a 2- fold increase in the basal transcription levels of Bdnf in resting neurons. However, this does not constitute activation of the Bdnf gene as transcript levels are still approximately 100-fold lower than found in activated neurons (Chen et al., 2003). Other repressive mechanisms may also contribute to the eventual silencing of these genes within mature ORNs. It should be noted, however, that expression of mature neuronal markers and down-regulation of immature neuronal markers does not necessarily indicate that the MeCP2 null neurons are functionally mature.  109  The absence of MBD2 leads to a completely different olfactory phenotype than the absence of MeCP2. The adult male MBD2 null OE contains almost 2- fold more cycling basal cells than wildtype, concomitant with an approximately 20% reduction in the number of OMP positive mature neurons (Figure 5.2). As MBD2 is detected both in mature neurons (MBD2a) and across the cycling basal cell to immature neuron transition (MBD2b), this perturbation at both stages of ORN differentiation is not unexpected. However, what these findings do not distinguish is if the MBD2 nulls display two independent transitional defects or, more likely, perturbations at two stages of ORN differentiation which regulate one another. Inappropriate gene expression during the exit of cell cycle and commitment to the neuronal lineage could lead to developmentally compromised neurons, which would likely undergo apoptotic cell death. Increased neuronal death would remove feedback inhibition on the basal cells, causing increased cycling and neurogenesis. We directly analyzed cell turnover and apoptotic cell death in the OE of MBD2 null mice and found a decreased retention of postnatally- generated neurons in the MBD2 null OE, concomitant with increased apoptotic cell death of mature ORNs (Figure 5.6,5.7). This indicates that MBD2 null ORNs are unstable and turnover at a higher rate than wildtype ORNs.  An increased loss of neurons following the disruption of DNA methylation-dependent ge ne silencing has also been found in mosaic animals containing 30% DNMT1-depleted neurons; the neurons that do not express DNMT1 are eliminated from the brain within the first three weeks postnatal (Fan et al., 2001). Why the MBD2 null ORNs are functionally unstable, however, remains to be determined, but several possible explanations present themselves. One such cause could be genomic instability. MBD1 null adult hippocampal neural progenitors display increased rates of aneuploidy and increased expression of the endogenous retrovirus IAP (Zhao et al., 2003), both hallmarks of genomic instability observed in cancer aetiology (Laird and Jaenisch, 1996; Chen et al., 1998; van Gent et al., 2001). However, no change in IAP expression is found in the absence of MBD2 (Hendrich et al., 2001) and deficiency of MBD2 actually suppresses intestinal tumourigenesis (Sansom et al., 2003). Furthermore, MeCP2 null ORNs do not turn over at an increased rate. Together, these data suggest that the functional instability of MBD2 null ORNs is not likely due to a general genomic instability in the absence of an MBD protein.  110  A loss of acetylation homeostasis in the absence of MBD2 could also underlie the increased ORN degeneration. A disruption in the balance between histone acetylation (HAT) activity and HDAC activity has been postulated to commonly underlie neuronal dysfunction and degeneration; the primary cause of this imbalance likely varying between neurodegenerative disorders (Saha and Pahan, 2006; Morrison et al., 2007). A global decrease in histone acetylation levels and a loss of HATs, such as CBP and p300, are observed during various neurodegenerative challenges, preceding apoptosis (Jin et al., 2001; Jiang et al., 2003; Rouaux et al., 2003). As such, treatment with HDAC inhibitors can act to re-establish the balance between acetylation and deacetylation and promote cell survival (reviewed in (Saha and Pahan, 2006; Morrison et al., 2007). A similar neuroprotective effect of both HDAC inhibition or a disruption in DNA methylation is observed in neurons undergoing an ischemic challenge (Endres et al., 2000; Endres et al., 2001), where an up-regulation in DNA methylation and MBD expression is observed (Jung et al., 2002). In healthy neurons, on the other hand, HDAC inhibitors act to disrupt the acetylation balance, in this case towards increased acetylation. This also proves detrimental and frequently results in apoptotic cell death (Salminen et al., 1998; Boutillier et al., 2002, 2003; Morrison et al., 2006). Thus, an exquisite balance in acetylation regulation appears necessary for ne uronal viability. The loss of MBD2 prevents recruitment of HDAC-containing repressor complexes to specific loci, therefore acting in a similar manner to HDAC inhibitors to disrupt the acetylation homeostasis. Because MBD2 is only one of a number of mediators that recruit HDACs, its disruption would not completely inhibit HDAC activity, as HDAC inhibitors do, thus resulting in the more subtle rate of neurodegeneration observed in the MBD2 null OE.  An inability of some MBD2 null ORNs to appropriately target to the OB and form synapses could also result in the increased ORN turnover. While many MBD2 null ORNs do extend axons to the OB and synaptic glomeruli are formed, the structure of the glomeruli is aberrant, displaying a disorganized, immature phenotype (Figures 5.8,5.9). This could indicate a general lack of postnatal activity-dependent synaptic refinement (Yu et al., 2004; Zou et al., 2004; Chesler et al., 2007; Col et al., 2007), potentially due to functional disruptions in the ORNs and/or their postsynaptic targets in the OB. MBD2 is also expressed in the Mitral cell projection neurons of the OB and some periglomerular interneurons (Figure 4.7). Within the MBD2 null glomeruli, OMP expression appears largely uniform and axonal subcompartments are not as  111  distinct as wildtype (Figures 5.8, 5.9). Wildtype glomeruli contain islands of dense ORN axon terminals interspersed with subcompartments composed primarily of dendritic connections between Mitral cells and periglomerular cells, an organization that emerges during the first weeks postnatal (Kasowski et al., 1999; Treloar et al., 1999; Kim and Greer, 2000).  The apparent loss or delay in the formation of these subcompartments may suggest aberrant synaptic connections in the absence of MBD2. Activity-dependent mechanisms are essential for ORN axon targeting and a disruption in transcriptional regulation may alter the activity of an ORN, increasing the likelihood it will fail to appropriately target. The disruption in transcriptional regulation could be at the level of odorant receptor (OR) expression. Each neuron is believed to express only one allele of one OR gene (Chess et al., 1994; Malnic et al., 1999), and the OR protein impacts ORN axon targeting (Mombaerts et al., 1996; Wang et al., 1998; Bozza et al., 2002; Feinstein et al., 2004). How OR expression is regulated is still a quagmire; however, OR gene choice is reset in mice cloned from a mature ORN (Eggan et al., 2004; Li et al., 2004), ruling out changes in the DNA sequence and suggesting epigenetic regulation. A disruption in OR regulation, either in zonal expression pattern or the number of ORs a given ORN expresses, could therefore lead to disrupted ORN targeting and an increased probability of ORN death.  Multiple disruptions could therefore underlie the increased instability of MBD2 null ORNs, possibly raising the question of why more ORNs are not degenerating at any given time. This is likely due to redundant repressive mechanisms working in tandem to silence gene expression, as has been shown for X chromosome inactivation (Csankovszki et al., 2001). By ‘stacking’ silencing mechanisms, loss of any one component would leave repression largely intact, but could cause partial gene activation in a stochastic manner (Barr et al., 2007; Berger et al., 2007). For example, there is a 2-fold increase of Xist expression in MBD2 null cells, but when these cells are treated with the HDAC inhibitor TSA, Xist expression is induced 20- fold (Barr et al., 2007). Incomplete penetrance of inappropriate gene expression is also seen in the MBD2 null colon; genes that are normally restricted in expression to the exocrine pancreas and duodenum and that are silenced in the colon are significantly over-expressed in the colons of approximately half of MBD2 null mice. Furthermore, the level of over-expression varies widely between mice,  112  demonstrating a continuum of over-expression in the absence of MBD2 (Berger et al., 2007). Variability in the MBD2 null phenotype is also observed in the differentiation of helper T cells, with only some cells inappropriately expressing interleukin 4 (Hutchins et al., 2002). Redundant transcription repression mechanisms, leading to only stochastic transcriptional leakiness when one component is lost, could therefore explain the relatively mild olfactory phenotype of an MBD disruption.  Taken together, the MBD proteins MBD2 and MeCP2 facilitate sequential transitions in ORN development, but are not absolutely necessary for ORNs to traverse these developmental stages. In the absence of MeCP2, immature ORNs display a transient delay in terminal differentiation, resulting in an accumulation of ORNs of an intermediate developmental stage, expressing markers of both immature and mature ORNs (Figure 5.10B). A disruption in MBD2, on the other hand, does not affect the transition from immature to mature ORN. It does, however, perturb the stability of mature ORNs, leading to increased neuronal turnover, as well as an increase in progenitor proliferation (Figure 5.10C). Whether the increase in MBD2 null progenitor proliferation is solely a secondary effect of increased ORN death, feeding back to up-regulate neurogenesis, remains to be determined.  113  Figure 5.10: Summary of MBD2 and MeCP2 null olfactory phenotypes (A) In the adult wildtype OE, consistent populations of horizontal basal cells (HBCs), globose basal cells (GBCs), immature receptor neurons (IRNs) and mature olfactory receptor neurons (ORNs) are present, organized in a laminar structure with a developmental hierarchy. ORN axons are ensheathed by olfactory ensheathing cells (OECs) as they enter the lamina propria on route to the olfactory bulb. Curved arrows denote an ability to proliferate. (B) In the MeCP2 null adult OE, there is a significant increase in the number of transitional neurons (yellow), which express markers of IRNs (green) and mature ORNs (red). (C) In the MBD2 null adult OE, on the other hand, there is no disruption in the IRN population. There is, however, a significant increase in the number ORNs undergoing apoptotic cell death as well as an increase in the number of proliferating GBCs. The increase in GBC proliferation is due, at least in part, to feedback (white arrows) from increased ORN turnover, signalling to up-regulate neurogenesis.  114  Chapter 6: Progenitor Proliferation Is Increased in the Lesioned MBD2 Null OE and Following an Acute Perturbation of Histone Deacetylation with Valproic Acid 6.1 Introduction  The adult MBD2 null OE exhibits a significant increase in the number of proliferating basal cells, concomitant with a significant increase in the number of apoptotic ORNs. The increased death of ORNs in the MBD2 null OE would cause a feedback up-regulation of neurogenesis, by decreasing the number of ORNs secreting inhibitors of proliferation and increasing the number of degenerating neurons releasing positive regulators of proliferation (Mumm et al., 1996; Calof et al., 1998; Shou et al., 1999; Shou et al., 2000; Wu et al., 2003). This would account, at least in part, for the increased number of cycling GBCs observed in the MBD2 null OE. However, MBD2b is expressed in GBCs, in addition to the expression of MBD2a in mature ORNs; therefore, a primary defect in the regulation of progenitor proliferation itself cannot be ruled out. The fact that there is only an approximately 20% reduction in mature ORNs in the adult MBD2 null OE, but an almost 2- fold increase in the number of proliferating basal cells (Figure 5.2) suggests that feedback regulation does not account for the full extent of MBD2 null progenitor proliferation. Furthermore, in the early postnatal MBD2 null OE there is a trend towards increased basal cell proliferation concomitant with a trend towards decreased immature and mature ORNs (Figure 5.4). While a larger sample size will be necessary to determine if these trends are significant, they do suggest a delay in exit from cell cycle and neuronal differentiation.  We therefore wished to determine if MBD2 null basal cells also proliferate at an increased rate in the absence of direct feedback from ORNs. To this end, we performed unilateral bulbectomies on MBD2 null adult male mice and littermate controls. Following the surgical removal of the OB, and associated severing of ORN axons, there is a wave of neuronal degeneration with ORNs ipsilateral to the lesioned OB depleted by 3 days post-bulbectomy (PBx) (Graziadei and Graziadei, 1979; Costanzo and Graziadei, 1983; Holcomb et al., 1995; Cowan et al., 2001). The progenitors in the ipsilateral OE respond by increasing proliferation, which peaks at approximately 6 – 8 days PBx (Schwartz Levey et al., 1991; Gordon et al., 1995; Carter et al., 115  2004). We therefore focused our analysis beginning at 6 days PBx, three days after the depletion of the ORNs, when proliferation is already at an apparent maximum. This lesion model has the additional advantage that the contralateral OE remains intact, thereby serving as an internal control for steady-state progenitor proliferation. If the increased progenitor proliferation in the MBD2 null OE is solely the result of increased neurogenesis to compensate for a postnatal loss of ORNs, there should be no significant difference between MBD2 null and wildtype OE postbulbectomy, when progenitor proliferation is already up-regulated. A significant increase in progenitor proliferation should, however, be observed in the MBD2 null in the steady-state, unlesioned OE.  The multiple levels of feedback control that maintain the homeostasis of cell populations within the OE may also have led to a functional compensation during development to keep neurogenesis in check in the MBD2 null OE. If a functional compensation has developed in the MBD2 null OE, an acute inhibition of HDAC activity, the effectors of MBD2-mediated repression, should cause a more severe disruption of the regulation of neurogenesis than the genetic disruption of MBD2. We therefore treated wildtype littermates with the HDAC inhibitor VPA, beginning six days after bulbectomy. In this way, the rate of progenitor proliferation in both the steady-state OE and in the absence of ORNs can be directly compared between wildtype, MBD2 null and VPA-treated mice.  6.2 Results  6.2.1 Labelling cycling basal cells with iododeoxyuridine and chlorodeoxyuridine  In order to measure the extent of progenitor proliferation at two separate time points in the same set  of  experimental  animals,  we  employed  the  halogenated thymidine  analogues  iododeoxyuridine (IdU) and chlorodeoxyuridine (CldU). These two BrdU ana logues can be distinguished by different anti-BrdU antibodies (Aten et al., 1992; Vega and Peterson, 2005), allowing for a temporal discrimination of cell proliferation. Furthermore, by administering the two analogues at defined intervals, it is possible to identify cells that re-enter the cell cycle and continue to proliferate. We first tested olfactory progenitor labelling with IdU and CldU and  116  optimized the detection conditions using early postnatal wildtype mice, which contain high numbers of cycling basal cells. P7 pups were injected with an equimolar concentration of IdU or CldU and sacrificed one hour after the single injection. A third group of pups was injected with IdU on P7, CldU on P8 and sacrificed on P9. Equimolar concentrations of IdU and CldU were used rather than equivalent weight/volume concentrations as simultaneous equimolar delivery of IdU and CldU equally labels dividing neural progenitors while co-administration of equivalent wt/vol concentrations does not detect proliferating cells with equal probability (Vega and Peterson, 2005).  Stringent conditions must be employed to prevent cross-reactivity of the IdU and CldU antibodies with the other analogue. In the adult mouse brain, limiting dilutions of the antibodies has been reported to be sufficient to ensure specificity of antibody detection (Vega and Peterson, 2005). However, this protocol did not fully prevent cross-reactivity in the postnatal OE, as evidenced by detection with both antibodies in the tissue that was injected with only one of the analogues (data not shown). A combination of sequential incubation with a limiting dilution of each antibody and a high salt wash after each primary antibody, to remove non-specific binding, was sufficient to ensure specificity of the IdU and CldU antibody binding in the OE (Figure 6.1A,B). IdU and CldU are equally detected in the P7 OE (Figure 6.1A,B). In the P9 tissue, most cells labelled with IdU at P7 are situated apically to cells labelled with CldU at P8, with few cells co-labelled with IdU and CldU (Figure 6.1C-F). This pattern is expected as progenitors are rapidly exiting the cell cycle, differentiating and moving apically during this highly neurogenic period.  6.2.2 Validation of the bulbectomy lesion model  Unilateral bulbectomies were performed on adult male MBD2 null mice and wildtype littermates. Six days PBx, all mice were given a single injection of IdU, followed by a single injection of CldU eight days PBx. Half of the wildtype mice were also treated with VPA during days 6-8 PBx. The mice were allowed to recover for a further two days, and then sacrificed at day 10 (Figure 6.2A). Upon sacrifice, we sectioned through the OB to confirm completeness of  117  Figure 6.1: Detection of IdU and CldU in the postnatal olfactory epithelium Postnatal day (P) 7 pups were injected with equimolar concentrations of (A) IdU or (B) CldU and sacrificed one hour post-injection or (C-F) injected with IdU on P7, CldU on P8 and sacrificed at P9. (AF) The sections were probed sequentially with rat anti-BrdU from Accurate to detect CldU and mouse anti-BrdU from Becton Dickinson to detect IdU. The singly injected sections demonstrate no crossreactivity of the antibodies. (D-F) are a higher magnification of the marked box in (C). (D-F) Cells labelled with either IdU or CldU (arrowheads) are detected, as well as several cells labelled with both (arrows). (A-C) scale bar = 50 µm, (D-F) scale bar = 200 µm  118  Figure 6.2: A varying extent of medial sparing following unilateral bulbectomy (A) Unilateral bulbectomies were performed on adult male MBD2 null mice and littermate controls. 6 days post-bulbectomy (PBx), all mice were injected with IdU, followed by CldU on day 8. Half of the wildtype mice were also treated with valproic acid (VPA) between days 6 and 8. The mice were sacrificed at day 10. (B-C) Some medial sparing was observed in the lesioned OBs, ranging from (B) a small amount of the nerve fibre layer to (C) some medial glomeruli (arrows) and external plexiform layer remaining intact. (D-E) depict the lesioned OEs corresponding to the lesioned OBs shown in (B-C), respectively. (D) The largely complete lesion depleted OMP positive neurons from all but the ventral half of the septum (arrow). (E) The sparing of a limited medial glomerular and external plexiform layer of the OB left an intact OMP positive ORN population along the septum, dorsal recess and the tip of endoturbinate IIa. UL = unlesioned, Les = lesioned, NFL = nerve fibre layer, EPL = external plexiform layer, S = septum, DR = dorsal recess, 1 and 2 mark ectoturbinates 1 and 2, and IIa and IIb mark endoturbinates. Scale bars = 400 µm  119  the bulbectomy. We observed varying extents of incomplete removal of tissue along the medial aspect of the lesioned OB, ranging from a small amount of nerve fibre layer (Figure 6.2B) to regions of the medial glomerular layer left intact (Figure 6.2C). The MBD2 nulls, in particular, demonstrated incomplete removal of some medial glomeruli. Within the OE, this translated to varying degrees of depletion of OMP positive ORNs along the septum, dorsal recess and the tip of endoturbinate IIa (Figure 6.2D,E). Therefore, we excluded the septum and dorsal recess from our analysis and focused on endoturbinates IIa (excluding tip) and IIb and ectoturbinates 1 and 2, where a more consistent depletion of OMP positive neurons was recorded. For all animals, an adjacent section was assessed for OMP expression and only those regions containing less than a single cell layer of OMP positive neurons were included in the analysis.  We next compared IdU and CldU labelling in the wildtype lesioned and unlesioned OE to ensure that this experimental approach is sensitive enough to detect the known up-regulation of basal cell proliferation in the lesioned OE. IdU and CldU labelled cells can be detected throughout the unlesioned (Figure 6.3A) and lesioned (Figure 6.3B) OE, but both appear more abundant in the lesioned OE. In both the lesioned and unlesioned OE, few IdU and CldU labelled cells are detected along the septum, where OMP positive neurons are rarely depleted by the bulbectomy. IdU and CldU labelled cells are detected in largely discrete patches, with only occasional overlap between the two, in both the lesioned and unlesioned OE (Figure 6.3A-F). This pattern of incorporation demonstrates the non-uniform timing of neurogenesis throughout the OE (Weiler and Farbman, 1997), both in the steady-state and post-lesion. Furthermore, the low frequency of detection of cells that have incorporated both IdU and CldU suggests that it is rare for a cell that cycled on day 6 to re-enter the cell cycle on day 8. Using the sampling strategy outlined above, a significant increase in the number of both IdU and CldU labelled cells is detected between the unlesioned and lesioned OE. Between the lesioned and unlesioned, steady-state, OE, there is an approximate 2- fold increase in both the number of cells labelled with IdU and the number of cells labelled with CldU (Figure 6.3G). There is, however, no significant difference in the number of cells labelled with IdU versus the number of cells labelled with CldU in either the lesioned or unlesioned OE (Figure 6.3G). Taken together, the bulbectomy lesion with IdU and CldU labelling should allow for the comparison of progenitor cell cycle kinetics, both in the  120  Figure 6.3: The number of IdU and CldU labelled cells is increased in the lesioned OE (A-B) More IdU (green) and CldU (red) labelled cells are apparent in (B) the lesioned OE compared to (A) the unlesioned OE. In both the lesioned and unlesioned OE, few labelled cells are observed along the septum and IdU and CldU labelled cells are usually detected in discrete patches (asterisks), with only occasional overlap between the two (arrows). (C-F) Higher magnification pictures demonstrating (C-D) the distinct stretches of OE labelled with either IdU or CldU and (E-F) the occasional overlap of IdU and CldU expressing regions with occasional cells labelled with both. (G) The number of IdU and CldU labelled cells were counted in the lesioned and unlesioned OE of 3 wildtype mice. The number of both IdU and CldU labelled cells is significantly increased in the lesioned OE compared to unlesioned. The number of cells that incorporate IdU is not significantly different from the number that incorporate CldU in either the lesioned or unlesioned OE. Error bars denote SEM. (A-B) Scale bars = 200 µm, UL = unlesioned, Les = lesioned, S = septum, DR = dorsal recess, 1 and 2 mark ectoturbinates 1 and 2, and IIa and IIb mark endoturbinates. (C-F) Scale bars = 50 µm. (G) * = p‹0.05, ** = p<0.01  121  presence and absence of feedback from ORNs, between wildtype and epigenetically-disrupted mice.  6.2.3 The number of IdU and CldU labelled cells is increased in the unlesioned, steady-state MBD2 null and VPA-treated OE  The number of IdU labelled cells is significantly increased in the unlesioned MBD2 null OE compared to wildtype (Figure 6.4A,B,D). There is also a trend towards an increased number of CldU labelled cells in the MBD2 null OE, with the total number of labelled cells (IdU + CldU) significantly increased in the MBD2 null OE (Figure 6.4E). This supports the earlier observation of increased progenitor proliferation in the steady-state, unlesioned MBD2 null adult OE, as measured by the expression of PCNA (Figure 5.2). Wildtype mice treated with VPA (Figure 6.4C) also contain a significant increase in the number of both IdU and CldU labelled cells in the unlesioned OE (Figure 6.4D,E). This indicates that an acute inhibition of HDAC activity can recapitulate the increase in progenitor proliferation exhibited in the MBD2 knockout OE. In addition, the immediate increase in cell proliferation in response to VPA (IdU and CldU were injected concurrently with VPA treatment) suggests that VPA is acting directly on the progenitors to up-regulate proliferation, rather than through a secondary feedback mechanism from ORNs.  6.2.4 Increased proliferation and cell cycle re-entry in the MBD2 null and VPA-treated lesioned olfactory epithelium  An increase in the number of IdU and CldU labelled cells in the MBD2 null (Figure 6.5B), compared to wildtype (Figure 6.5A), is also found in the lesioned OE (Figure 6.5D). This indicates that MBD2 null basal cells proliferate at increased rate both 6 and 8 days PBx. This increased proliferation occurs in the absence of direct feedback from ORNs, when proliferation is already up-regulated. Furthermore, the VPA treated OE (Figure 6.5C) also exhibits a significant increase in the number of IdU and CldU labelled cells in the lesioned OE (Figure 6.5D). In addition, both the VPA treated mice and the MBD2 null mice display a significant increase in the number of IdU/CldU double labelled cells in the lesioned OE compared to  122  Figure 6.4: Increased incorporation of IdU and CldU in the unlesioned MBD2 null and VPAtreated OE (A-C) IdU (green) and CldU (red) labelled cells are usually detected in small, discrete patches throughout the unlesioned OE with only occasional overlap observed. Both (B) the MBD2 null and (C) the VPAtreated OE appear to contain more IdU and CldU labelled cells than (A) wildtype. (D) The number of IdU and CldU labelled cells were quantified per linear mm in the unlesioned OE of a minimum of three animals of each experimental group. The number of IdU cells is significantly increased in both the MBD2 null and VPA treated mice, compared to wildtype. There is a trend towards an increased number of CldU cells in the MBD2 null, while the number of CldU labelled cells is significantly increased in the VPA treated OE compared to wildtype. (E) The total number of labelled cells (IdU + CldU) is significantly increased in both the MBD2 null and VPA treated OE, compared to wildtype. (A-C) Scale bar = 200 µm, NC = nasal cavity, T IIa = endoturbinate IIa, T IIb = endoturbinate IIb. (D-E) * = p‹0.05, ** = p<0.01  123  Figure 6.5: Increased proliferation and cell cycle re -entry in the lesioned MBD2 null and VPA treated OE (A-C) The number of IdU (green) and CldU (red) labelled cells is increased in both (B) the MBD2 null and (C) the VPA-treated lesioned OE compared to (A) wildtype. Under all conditions, IdU and CldU labelled cells are usually detected in discrete patches, the size of which appear to increase in the MBD2 null and VPA treated lesioned OE (asterisks). The number of cells labelled with both IdU and CldU (arrows) also appear to be increased in the MBD2 null and VPA treated mice. (D) The number of IdU and CldU labelled cells was quantified per linear mm in the lesioned OE of a minimum of three animals of each experimental condition. The number of IdU and CldU labelled cells are both significantly greater in the MBD2 null and VPA treated OE compared to wildtype. (E) The number of IdU/CldU double labelled cells is significantly increased in both the unlesioned and lesioned VPA treated OE, while the MBD2 null OE displays increased double labelling only in the lesioned OE. (A-C) Scale bar = 200 µm, NC = nasal cavity, T IIa = endoturbinate IIa, T IIb = endoturbinate IIb. (D-E) * = p‹0.05, ** = p<0.01  124  wildtype (Figure 6.5E), suggesting an increase in the number of basal cells that re-enter the cell cycle between 6 and 8 days PBx. Interestingly, the VPA treated mice also display an increase in the number of IdU/CldU double labelled cells in the unlesioned OE (Figure 6.5E). The MBD2 null mice, on the other hand, demonstrate no significant increase in the number of IdU/CldU double labelled cells in the steady-state, unlesioned OE (Figure 6.5F). Thus, an acute inhibition of HDACs may have a more disruptive effect on the cell cycle regulation of olfactory progenitors than the MBD2 null.  6.3 Discussion  The MBD2 null adult OE exhibits a significant increase in the number of proliferating basal cells. This is due, at least in part, to an increased turnover of mature ORNs, feeding back to upregulate neurogenesis. If the increase in MBD2 null basal cell proliferation is solely the result of feedback from an increased number of dying ORNs, we hypothesized that the increase would not be apparent in the lesioned OE, where the majority of mature ORNs have already died and proliferation is at an apparent maximum (Schwartz Levey et al., 1991; Gordon et al., 1995; Carter et al., 2004). However, a significant increase in both IdU and CldU labelled cells is found in the lesioned MBD2 null OE (Figure 6.5), indicating that the increased progenitor proliferation exhibited by the MBD2 nulls is not entirely reliant on direct feedback from ORNs. Furthermore, IdU and CldU are incorporated in an increased number of cells in the unlesioned MBD2 null OE (Figure 6.4). This confirms the up-regulation of steady-state MBD2 null basal cell proliferation identified by the expression of PCNA (Figure 5.2).  There is also an increase in the number of MBD2 null cells that have incorporated both IdU six days after bulbectomy and CldU eight days post-bulbectomy, indicating a re-entry into the cell cycle. The average cell cycle time for a GBC is approximately 20 hours (DeHamer et al., 1994; Huard and Schwob, 1995) and GBCs divide a maximum of twice before differentiating, at least in vitro (DeHamer et al., 1994). Therefore, cells double labelled with IdU and CldU are a rare event in this experimental paradigm and their increase suggests an enhanced prevalence of cell cycle re-entry in the absence of MBD2. Interestingly, there is no increase in the number of IdU/CldU double labelled cells in the unlesioned MBD2 null OE, perhaps indicating that the  125  extrinsic regulatory mechanisms controlling steady-state neurogenesis are able to prevent the intrinsic tendency of MBD2 null basal cells to re-enter the cell cycle.  This may represent a functional compensation that has developed to keep neurogenesis largely in check in the MBD2 null OE and to maintain the homeostasis of cell populations. If a functional compensation has arisen during development, one would expect an acute disruption of the DNA methylation-dependent gene silencing pathway to have a more profound effect on the cell cycle regulation than the genetic disruption of Mbd2. This is, in fact, the case. Inhibition of HDAC activity with VPA not only increases the number of cells labelled with IdU and with CldU in both the lesioned and unlesioned OE, it also significantly increases the number of IdU/CldU double labelled cells in both the lesioned and unlesioned OE (Figures 6.4,6.5). The VPA- induced increase in progenitor proliferation and cell cycle re-entry in the lesioned OE recapitulates the increase in cell cycling observed in the MBD2 null phenotype. It further supports a defect in the regulation of proliferation at the level of the progenitor, and not through a residual effect of the unstable ORN population, as VPA treatment was initiated only after the ORNs were largely depleted. The increase in IdU/CldU double labelled cells even in the presence of feedback regulation from ORNs in the VPA treated mice indicates that a sudden disruption in HDAC activity has a more profound effect on cell cycle exit and commitment to the ORN lineage than the genetic disruption of MBD2. This could, as mentioned, be due to a functional compensation derived during development to maintain homeostasis in the MBD2 null OE. It also could indicate that MBD2 is not the only protein capable of recruiting HDACs during the regulation of the cell cycle and may instead be one of several mediators of HDAC repression.  Interestingly, a genetic disruption of HDAC1 can also yield a developmental increase in neural progenitor proliferation in another sensory system, the zebrafish retina (Yamaguchi et al., 2005). HDAC1, potently inhibited by VPA (Phiel et al., 2001), is required for the switch from proliferation to differentiation in the zebrafish retina, functioning as a dual switch that suppresses both cell-cycle progression and inhibition of neurogenesis by inhibiting both the Wnt and Notch/Hes pathways (Yamaguchi et al., 2005). Treatment with the HDAC inhibitor TSA produces an even more severe hyperproliferation in the retina than the HDAC1 mutant, suggesting that other HDACs might also be involved in the regulation of cell cycle exit  126  (Yamaguchi et al., 2005). This is similar to the increased proliferation defect observed in the VPA treated OE, compared to the MBD2 null. VPA also elicits an increase in proliferation, as measured by BrdU incorporation, in neurospheres cultured from the rat E14 lateral ganglionic eminence or E15 neocortex (Laeng et al., 2004), as well as in E18 primary cortical cell cultures (Hao et al., 2004).  The increased progenitor proliferation demonstrated in the OE in response to both VPA and MBD2 deletion is, however, in contrast to some previous studies in the CNS. For example, in the adult rat hippocampus a 14 day treatment with VPA decreases the number of cells in the SGZ that incorporate BrdU over the first 6 days of treatment (Hsieh et al., 2004). This discrepancy may result from different HDAC repressor complex regulation of cell cycling in the relatively quiescent adult hippocampus compared to the highly neurogenic OE. The proliferating basal cells of the OE highly express HDAC1 and MBD2b, with HDAC2 initiated as the cells begin to differentiate (Figures 3.4,3.7). MBD2b can interact with a zinc finger binding protein known as MIZF (MBD2-binding zinc finger) (Sekimata et al., 2001) which binds to a conserved consensus sequence and recruits MBD2b (Sekimata and Homma, 2004a). A number of cell cycle regulatory genes, including Rb, p57Kip2 and Cyclin D1, contain the MIZF consensus binding sequence within their promoters, and MIZF, through the recruitment of MBD2, can silence Rb during myogenic differentiation (Sekimata and Homma, 2004a, b). Therefore, MBD2b may regulate the proliferation of olfactory basal cells through an association with MIZF. Within the hippocampus, however, MBD2 expression is only detected in the mature neurons (Figure 4.7), indicating that HDACs regulate cell cycle by a mechanism independent of MBD2b within hippocampal progenitors.  The length of VPA treatment may also be a factor in the divergent VPA treatment results. A secondary effect on the progenitors due to disruption of the neurons and glia is possible with a prolonged treatment. In fact, a single BrdU injection administered at the onset of a chronic VPA treatment yielded a greater number of labelled cells in the hippocampus, while a single BrdU injection at the end of the VPA treatment found no difference (Hao et al., 2004). This suggests that the length of the VPA treatment can alter the outcome of the study, likely due to different time-courses for the multiple mechanisms of action of VPA. While VPA does have multiple  127  effects on neurons outside of its ability to inhibit HDACs, it is unlikely that these factor into the experimental paradigm employed in the OE. First of all, VPA has the same proliferative effect on the unlesioned OE and the lesioned OE, where the neurons were depleted before treatment began. Secondly, the increase in proliferation is immediate and, therefore, is unlikely a secondary effect of a disruption in the neurons. Taken together, disrupting either MBD2 or components of the repressor complex it recruits, leads to an increase in the number of olfactory progenitors that are cycling at any given time, and an increased prevalence of cell cycle re-entry. While MBD2 and HDACs aren’t absolutely necessary for basal cells to exit the cell cycle and begin to differentiate, they clearly regulate this process.  128  Chapter 7: General Discussion 7.1 Summary of Results and Conclusions  Covalent modification of cytosine residues of DNA with a methyl group is one of the best characterized epigenetic modifications. It is a stable and heritable component of epigenetic ge ne silencing and a chief contributor to the stability of gene expression states. DNA methylation silences gene expression either by sterically hindering transcription factors from binding to promoter sequences, or, more commonly, by serving as a mark that is bound by methyl CpG binding domain proteins and associated repressor complexes. DNA methylation is critical for mammalian development and there is growing evidence to suggest that the nervous system is particularly sensitive to perturbations in this mode of regulation. However, the complexity of the developing CNS, where single regions contain a variety of different neuronal subtypes at different developmental stages, has proven a highly challenging environment in which to elucidate how DNA methylation-dependent gene silencing impacts the development of a single neuronal lineage. To begin to understand this process, therefore, I employed the olfactory epithelium as a model of neurogenesis. Olfactory receptor neurons, an accessible and largely homogeneous neuronal lineage, undergo continuous neurogenesis throughout adulthood. Furthermore, the OE is organized in a stratified, developmentally hierarchical manner, simplifying the identification of developmentally restricted gene expression patterns.  Using the olfactory epithelium as my primary model system, I set out to test the hypothesis that epigenetic gene silencing is essential for defined, successive stages of neuronal differentiation, likely through the sequential recruitment of distinct repressor comp lexes. To this end, I asked, and attempted to answer, the following four questions:  Aim 1: Are methyl-CpG-binding domain proteins and associated repressor complex members sequentially expressed at distinct transitional stages of olfactory neurogenesis?  The N-terminal truncated MBD2b protein isoform is expressed in cycling basal cells, its expression down-regulated as the progenitors exit mitosis and commit to the neuronal lineage.  129  The full- length MBD2a is expressed in a subpopulation of apical immature receptor neurons, and its expression is up-regulated as ORNs functionally mature. MeCP2, on the other hand, is first detected in a subpopulation of the neuro-glial progenitors, HBCs; its expression is re- induced in the immature receptor neurons, and highly up-regulated as they functionally mature. The histone deacetylases HDAC1 and HDAC2, both identified co-repressors of both MBD2 and MeCP2, also display divergent developmental expression patterns within the OE. HDAC1 is expressed in the progenitor cells and the support cells of the OE, (including OECs, Bowmans Glands, and Sustentacular cells) as well as some mature ORNs. HDAC2, on the other hand, is first detected in a small subpopulation of cycling basal cells, but is mainly induced as the cells commit to a neuronal lineage and begin to develop as immature receptor neurons. The sequential and alternating expression patterns of MBD2b, MBD2a, MeCP2 and the HDACs, combined with the stage-specific expression of the de novo DNMTs, suggests a model of sequential transitions in olfactory neurogenesis that are regulated by MBD- mediated gene silencing. These transitions would occur (1) when HBCs transition into mitosis (MeCP2), (2) as mitotic GBCs exit mitosis and commit to the neuronal lineage (MBD2b), and (3) as immature ORNs lose developmental plasticity and transition into a mature, stable state (MBD2a and MeCP2).  Aim 2: Are the stage-specific expression patterns of MBDs and HDACs paralleled in he developing CNS?  Expression of MeCP2 in the CNS is initiated in post- mitotic neurons and is up-regulated as the neurons mature, in a pattern similar to our observations in the OE and consistent with previously published reports in the CNS. ß-Galactosidase expressed under the Mbd2 promoter and antibody detection of MBD2a demonstrate convergent patterns of expression in restricted subsets of mature CNS neurons, including in the hippocampus, cerebellum, OB, and projection nuclei of the brain stem. This suggests that abundant expression of the MBD2b isoform in proliferating progenitors might be a unique feature of the OE. Consistent with the pattern of expression we observed in the OE, HDAC1 and HDAC2 are divergently expressed at distinct developmental stages and in defined cellular lineages within the developing brain. In the embryonic developing brain, HDAC1 expression is largely restricted to the proliferating and migrating progenitors. HDAC2 is detected within the proliferative zones at this age; however, the highest expression of  130  HDAC2 is found in the post- mitotic neurons throughout the brain. In the postnatal brain, HDAC1 continues to be expressed by progenitors, in addition to expression in glial cells. HDAC2, on the other hand, is expressed in the majority of NeuN positive neurons as well as some postnatal progenitors. This includes the PCNA positive progenitors in the external granule layer of the P7 cerebellum, with weak expression of HDAC2 also detected in the SVZ and RMS of the adult. Altogether, MBD2, MeCP2, HDAC1 and HDAC2 display similar restricted expression patterns in the CNS and OE. Furthermore, the divergent expression of HDAC1 and HDAC2 suggests that inclusion of either HDAC1 or HDAC2, or both, may provide specificity to a given repressor complex, including those recruited by MBD2 and MeCP2.  Aim 3: Do MBD2 and MeCP2 null mice display stage-specific defects in ORN differentiation corresponding to their observed expression patterns?  The adult MBD2 null OE contains an approximately 2 fold increase in PCNA positive cycling basal cells compared to wildtype, concomitant with an approximate 20% reduction in OMP positive mature neurons. Significantly fewer cells that incorporate BrdU at one week of age are maintained for 6 weeks in the MBD2 null OE compared to wildtype. In addition, significantly more mature ORNs undergo apoptosis at all postnatal stages analyzed in the MBD2 null OE. Therefore, ORNs turn over at a higher rate in the MBD2 null OE, likely leading to an overall upregulation of neurogenesis. In addition, the synaptic target of ORN axons, glomeruli, are aberrantly formed in the MBD2 null OB, displaying a developmentally immature phenotype, possibly due to a defect in activity-dependent refinement. Taken together, ORNs differentiate and extend axons to the OB in the absence of MBD2, but many MBD2 null ORNs appear to be compromised.  In contrast, the adult MeCP2 null OE contains an approximately 1.7 fold increase in the number of neurons expressing GAP43, a marker of immature neurons. This increase, evident by 3 weeks of age, is likely the result of neurons maintaining expression of GAP43 after they have begun to express the mature neuronal marker OMP, as no increase in the total number of cells in the OE or decrease in the number of OMP positive neurons are observed. Neurons appearing to coexpress GAP43 and OMP are more evident in the developing MeCP2 null OE than wildtype and  131  the laminar distinction between immature and mature neuronal markers is disrupted. Therefore, loss of MeCP2 results in an increase in transitional neurons, expressing genes of both immature and mature ORNs. This appears to be a transitional delay, and these neurons do not turn over at an increased rate in the MeCP2 null OE.  Aim 4: Is there aberrant progenitor cell cycling in the MBD2 null in the absence of ORNs?  The halogenated thymidine analogues IdU and CldU can be antigenically distinguished in the postnatal OE and, thus, can be used to label cycling progenitors at defined temporal intervals within the same experimental animals, and to identify cells that have re-entered the cell cycle. MBD2 null and VPA treated mice incorporate IdU and CldU in an increased number of basal cells, in both the bulbectomy lesioned and steady-state, unlesioned OE. The increased IdU and CldU incorporation in the unlesioned MBD2 null OE confirms the earlier identification of an increased number of cycling basal cells in the adult MBD2 null OE. The increased IdU and CldU incorporation in the lesioned MBD2 null OE indicates that the up-regulation of MBD2 null progenitor proliferation is not solely a secondary effect of feedback from an unstable ORN population. A primary disruption in the regulation of basal cell proliferation in the absence of MBD2/HDAC activity is further evidenced by the increased IdU and CldU incorporation concurrent with VPA treatment, which occurs both in the presence and absence of ORNs. In addition, the number of IdU/CldU double labelled cells is increased in the lesioned MBD2 null and VPA treated OE, indicating an increase in the number of cells that re-enter the cell cycle between days 6 and 8 PBx and continue to proliferate. This increase in double labelled cells is limited to the lesioned side of the MBD2 null OE, but occurs in both the lesioned and unlesioned OE upon treatment with VPA. This suggests a more severe disruption of the regulation of cell cycle re-entry with an acute inhibition of HDACs.  7.2 Final Conclusions and General Discussion  The work presented in this thesis demonstrates that different MBD and HDAC protein family members are expressed at defined, sequential stages of lineage commitment and differentiation, both in the OE and CNS, and suggests that distinct repressor complexes likely silence expression  132  of specific classes of genes at successive transitional steps in neuronal differentiation (Figure 7.1). Within the postnatal OE, both HDAC1 and MeCP2 are expressed in a subpopulation of HBCs. At this developmental stage, HDAC1 and MeCP2 might associate with REST (Ballas et al., 2005) to maintain neuronal genes in an inactive state and to sustain the multipotent HBCs in an undifferentiated state. For example, REST inhibits expression of Mash1, a proneural gene necessary for the generation of ORN progenitors (Cau et al., 2002). HDAC1 is also expressed in the actively cycling progenitor population of the OE, where it likely acts independently of REST and MeCP2. HDAC1 is highly expressed in most, if not all, PCNA positive basal cells and it is down-regulated before or as the progenitors commit to the neuronal lineage. Therefore, HDAC1 likely acts to promote proliferation of GBCs, possibly by silencing cyclin-dependent kinase (CDK) inhibitors, such as p21WAF/CIP1 and p27KIP1 (Lagger et al., 2002).  Expression of HDAC2 is initiated in a subpopulation of PCNA positive putative GBCs and it is maintained as the progenitors commit to the neuronal lineage and differentia te into ORNs. In the GBCs, HDAC2 is likely recruited by MBD2b to repress cell cycle genes, thus promoting cell cycle exit and neuronal differentiation. Expression of HDAC2 is maintained in immature ORNs, as they transition into a mature state. During this developmental transition, HDAC2 likely complexes with MeCP2 to silence genes associated with neuronal differentiation, such as those necessary for axon outgrowth and targeting. MBD2a is also expressed as ORNs terminally differentiate. Unlike MeCP2, however, MBD2 does not appear to be necessary for the transition from immature to mature ORN. MBD2 is necessary, however, to maintain or stabilize mature ORNs, its absence causing increased apoptotic cell death of ORNs. MBD2a may, therefore, silence pro-apoptotic genes and regulate the response of ORNs to injury. MBD2a might regulate different genes by complexing either with HDAC2 or HDAC1 in different subpopulations of ORNs.  The absence of MBD2 or MeCP2 perturbs distinct stages of ORN development and function, indicating that MBD- mediated gene silencing is necessary for defined, successive stages of neuronal differentiation. MBD2 and MeCP2 are not, however, sufficient for ORN development as their absence leads to transitional delays and disruptions, but does not prevent ORN differentiation. The relatively subtle phenotypes of both the MBD2 and MeCP2 null mice are  133  Figure 7.1: Proposed model of the role of MBD proteins and HDACs in the stage -specific differentiation of ORNs HDAC1 and MeCP2 are expressed in HBCs, where they act to regulate self-renewal and to maintain an undifferentiated state, potentially through an association with the REST complex. HDAC1 is also highly expressed in proliferating basal cells, where it regulates cell cycle re-entry, in a complex independent of MeCP2. MBD2b recruits HDAC2 to silence cell cycle genes and promote cell cycle exit and commitment to the neuronal lineage. HDAC2 then associates with MeCP2 to regulate terminal differentiation of ORNs by silencing genes involved in neuronal development, such as those associated with axon outgrowth and targeting. MBD2a is necessary to maintain or stabilize mature ORNs, possibly by repressing proapoptotic genes.  134  likely a result of redundant repressive mechanisms, with the loss of one MBD leaving repression largely intact, but increasing the likeliness of transcriptional leakiness. MBD2 and MeCP2 are unlikely, however, to functionally compensate for one another. MBD2 and MeCP2 have different binding specificities for methylated DNA (Fraga et al., 2003; Klose et al., 2005), the MeCP2 null phenotype is not exacerbated in a MeCP2/MBD2 double null (Guy et al., 2001) and deletion of MBD2 does not lead to a compensatory up-regulation of MeCP2 expression (Berger et al., 2007). Taken together, MBD2 and MeCP2 have distinct repressive targets and function sequentially to mediate defined transitions in olfactory neurogenesis. Furthermore, the Class I deacetylases HDAC1 and HDAC2, are divergently expressed in the OE and brain, thus challenging the biochemical evidence that they are recruited together within multiple repressor complexes. The inclusion of one or the other, or both, may instead add specificity to a given repressor complex, functioning within different cell types. To my knowledge, this is the first demonstration of a sequential recruitment of different MBD and HDAC protein family members in the progressive lineage restriction and differentiation of a given cell type.  The major limitation of this thesis is that the work is largely descriptive in nature. I demonstrated, in detail, the distinct developmental stages and lineages of the olfactory epithelium that express MBD2, MeCP2, HDAC1 and HDAC2; I correlated these observations with the developing CNS to validate the use of the olfactory epithelium as a model of neurogenesis; and I described the phenotypic outcome, in the olfactory system, of disrupting the expression or function of these proteins, through transgenic mouse models for MBD2 and MeCP2 and with a chemical inhibitor of HDAC1 and HDAC2. Within the scope of this thesis, however, I was not able to address the extremely important question of how MBD2, MeCP2 and the HDACs regulate the distinct stages of ORN differentiation. To understand this, I believe that there are two main questions that must be addressed: (1) what repressor complexes do HDAC1 and HDAC2 associate with at each developmental stage, and (2) what are the genes silenced by MBD2b, MBD2a and MeCP2? The work presented in this thesis lays the groundwork to now ask these mechanistic questions in a directed and focused manner.  135  7.3 Future Directions  7.3.1 Identification of Repressor Complexes Containing HDAC1 and HDAC2 During Neuronal Differentiation  The divergent expression pattern of HDAC1 and HDAC2, both in the OE and the brain, is an intriguing, if unexpected, finding. It would be of interest to identify the repressor complexes with which they associate in progenitors versus differentiated cells. Both HDAC1 and HDAC2 are expressed in progenitors as they prepare to differentiate, but in differentiated cells, expression of HDAC1 is glial (or Sustentacular) and HDAC2 neuronal. Does inclusion of HDAC1 versus HDAC2 within a repressor complex govern this lineage decision? Also, HDAC1 is expressed in a small subpopulation of neurons, including some mature ORNs and some neurons within the hippocampus. Are HDAC1 and HDAC2 recruited within the same complex in these neurons? What would be the functional significance of this? Despite the common employment of HDAC inhibitors as therapeutic agents, very little is known about the role of histone deacetylation in neuronal development and even less about the functional diversity of the HDACs. Isolating the repressor complexes that contain HDAC1 and HDAC2 will help to identify the pathways they target for repression.  A number of repressor complexes that contain HDAC1 and HDAC2 have been identified, including Mi2/NuRD, MeCP1, Sin3A and REST/CoREST. It should first be determined if HDAC1, HDAC2 or both interact with other components of these known repressor complexes in the OE, before attempting to identify novel co-repressors. Co-immunoprecipitation experiments with HDAC1 and HDAC2 could first be performed on protein extracts from whole OE to determine which complexes each can associate with. A focused set of immunoprecipitation experiments could then be performed on purified cell populations from the OE to determine, for example, if HDAC2 interacts with MBD2b in basal cells and MeCP2 in neurons. Some cell populations will be easier to extract for these experiments than others. HBCs can be enriched by fluorescence activated cell sorting (FACS) for the cell surface antigen ICAM1 (Carter et al., 2004). Mature neurons can also be enriched by FACS from transgenic mice expressing GFP under the OMP promoter (Sammeta et al., 2007). The representation of cycling basal cells, on  136  the other hand, can be greatly enhanced by performing a bulbectomy lesion and harvesting the tissue 6-8 days, when the neurons have been depleted and proliferation is maximal.  7.3.2 Candidate Genes for MBD2 and MeCP2 Mediated Repression  The MBD2 and MeCP2 null olfactory phenotypes suggest distinct functional classes of genes that are targeted for repression by each MBD protein. MeCP2 likely silences the expression of genes that are necessary for developmental steps such as axon outgrowth and targeting, but that are not needed for the function of a mature neuron. MBD2, on the other hand, likely represses cell cycle associated genes and, possibly, pro-apoptotic factors. Following this model, a number of potential target genes for MeCP2 and MBD2 mediated silencing can be identified within the literature. Directly testing the regulation of these candidate genes in the wildtype versus MBD2/MeCP2 null olfactory epithelia will provide a first step in ascertaining the genes that must be silenced at each stage of ORN differentiation.  MBD2  Very few direct target genes for MBD2 have been identified; however, MBD2b has been postulated to repress a number of cell cycle regulatory genes, including p57/kip2, cyclin D1 and Rb, through an interaction with MIZF (Sekimata and Homma, 2004). A continuous overexpression of cell cycle regulatory proteins in the MBD2 null basal cells could severely compromise the ability of a cell to regulate the cell cycle, leading to aberrant cell cycle re-entry. The increased apoptotic death of MBD2 null ORNs could also be the result of an inability to fully repress the expression of cell cycle regulatory genes. There is extensive overlap in the factors that regulate the cell cycle and that promote apoptotic cell death, with ectopic re-entry of post-mitotic neurons into the cell cycle leading to apoptosis (reviewed in Greene et al., 2004). Among the cell cycle molecules directly linked to neuronal apoptosis is E2F1. E2F1 deficiency confers neuroprotection, including in vivo post- ischemia, and over-expression of E2F1 in neurons in vitro is sufficient to induce apoptosis (MacManus et al., 1999; Hou et al., 2000; O'Hare et al., 2000). HDAC inhibitors induce the expression of E2F1 in neurons, leading to apoptotic cell death (Boutillier et al., 2002). Which HDAC-containing complex represses E2F1  137  in neurons has not been determined, but the possibility exists that MBD2 mediates this silencing in ORNs. Therefore, increased basal expression levels of E2F1 in MBD2 null ORNs could increase the likelihood of ectopic cell cycle re-entry and apoptotic cell death. Increased E2F1 expression could also be responsible, at least in part, for the increased proliferation of MBD2 null GBCs.  MeCP2  BDNF is the most extensively studied target of MeCP2 mediated repression (Chen et al., 2003; Martinowich et al., 2003; Chang et al., 2006; Wang et al., 2006; Zhou et al., 2006). Could a disruption in the regulation of BDNF expression explain, at least in part, the increase in transitional neurons found in the MeCP2 null OE? BDNF is crucial for promoting the maturation and survival of ORNs in vitro (Mahanthappa and Schwarting, 1993) and the addition of BDNF to primary cultures of neonatal OE increases the number of GAP43 positive neurons (Roskams et al., 1996). The BDNF receptor, receptor tyrosine kinase B (TrkB) (Glass et al., 1991; Soppet et al., 1991; Ip et al., 1992; Ip et al., 1993a; Ip et al., 1993b), is expressed throughout the cell body and dendritic knobs of immature ORNs, shifting developmentally to expression in the axons of mature ORNs (Roskams et al., 1996). Therefore, BDNF could potentially modulate the differentiation of immature neurons, as well as the targeting, synapse formation and stability of mature ORNs. These steps in differentiation could be perturbed by two distinct disruptions in BDNF regulation in the MeCP2 null OE. In resting (immature) neurons, lack of MeCP2mediated repression directly causes a two- fold increase in basal Bdnf expression (Chen et al., 2003; Zhou et al., 2006); however, the overall expression of BDNF in the MeCP2 null brain is reduced by about 30% (Chang et al., 2006), likely due to the reduced neuronal activity (Dani et al., 2005). Over-expression of BDNF could disrupt the tight control of ORN differentiation within the OE, while reduced expression of BDNF within mature neurons of the OB could alter ORN synapse formation and, thus, terminal differentiation. BDNF is not detected in ORNs and their axon terminals, but it is present in the soma and dendrites of mitral and periglomerular neurons (McLean et al., 2001), which also express MeCP2 (Figure 4.7). BDNF is, in fact, critical for the activity-dependent pruning of ORN axon arbours (Cao et al., 2007), and its potential  138  reduction in the MeCP2 null OB could underlie the increased size and diffuse appearance of OMP positive glomeruli (Figure 5.8).  The homeobox transcription factor Dlx5 has also been identified as a direct target of MeCP2 mediated repression (Horike et al., 2005), and Dlx5 is necessary for the development of both the olfactory bulb and epithelium (Levi et al., 2003; Long et al., 2003). The OBs of Dlx5 null mice lack glomeruli, exhibit a disorganized laminar structure, and have reduced numbers of periglomerular and granule cell interneurons. Within the OE, ORNs do differentiate and extend some axons in the absence of Dlx5; however the axons do not reach the OB or form synapses (Levi et al., 2003; Long et al., 2003). Therefore, expression of Dlx5 within developing ORNs appears necessary for axon outgrowth and targeting. An inability to silence expression of Dlx5 when ORNs have completed the stage of axon outgrowth and targeting may, thus, underlie the transitional delay in terminal differentiation observed in the MeCP2 null ORNs. An inability to retain repression of REST could also potentially underlie the differentiation defects in the MeCP2 null OE. As outlined previously, REST acts to repress expression of neuronal genes and, therefore, is normally silenced in neurons by a repressor complex including MeCP2 (Chong et al., 1995; Schoenherr and Anderson, 1995; Ballas et al., 2005). Over-expression of REST in developing chick spinal cord neurons significantly increases the frequency of axon guidance errors (Paquette et al., 2000) and induced expression of REST in cortical neurons inhibits differentiation, blocking sodium channel currents and neurite outgrowth (Ballas et al., 2001).  All four known members of the ID (inhibitors of differentiation or inhibitors DNA binding) subfamily of helix- loop-helix (HLH) transcription factors are also primary targets of MeCP2 regulation in neurons (Peddada et al., 2006). The ID proteins block the function of tissue-specific basic helix- loop-helix (bHLH) transcription factors involved in regulating neuronal differentiation (Nakashima et al., 2001; Jogi et al., 2002; Bai et al., 2007). Expression of the ID proteins decreases upon neuronal differentiation (Riechmann et al., 1994; Lopez-Carballo et al., 2002), but this down-regulation is blocked by the DNA methylation inhibitor 5-azacytidine (Persengiev and Kilpatrick, 1997). Failure to repress the ID family of proteins could, therefore, further contribute to the delayed differentiation of MeCP2 null ORNs. The involvement of ID proteins in ORN differentiation has not been examined, but the expression of Id1 is increased  139  post-bulbectomy in the OE, at a time corresponding to early progenitor proliferation and ORN differentiation (Shetty et al., 2005). Within the granule neurons of the MeCP2 null dentate gyrus, where a transient delay in terminal differentiation is also observed, genes involved in cytoskeletal structure formation, such as Prefoldin 5, and genes critical for synaptogenesis, including Syndecan 2, are up-regulated (Smrt et al., 2007). Over-expression of such genes in the OE could also potentially contribute to the delayed differentiation of MeCP2 null ORNs.  7.3.3 Global Analysis of MBD2 and MeCP2 Target Genes  Most global analyses of gene expression have failed to identify many significant differences between the wildtype and MBD null animals. This applies in particular to the MeCP2 null brain, the focus of many such examinations due to its clinical relevance. Most of these experiments, however, have been performed on adult brains, a heterogeneous sample likely containing few neurons at the developmental stage at which MeCP2 function is most crucial. Having elucidated the distinct stages of ORN differentiation perturbed by a loss of MBD2 and MeCP2, however, we can now begin a more focused approach to identify the genes targeted for silencing at each transitional stage. The olfactory epithelium is an ideal system in which to do this because each developmental stage of the ORN population is represented within the postnatal animal and can be manipulated with relative ease.  Within the postnatal OE, neurogenesis can be up-regulated and largely synchronized through lesion models, thus enhancing the representation of a given developmental stage and facilitating the identification of genes mis-regulated at that stage. Because of the potential for redundant silencing mechanisms preventing a large up-regulation of gene expression in the absence of MBD2 or MeCP2, identifying the binding sites of the MBD, with a ChIP-on-Chip approach, for example, might prove more informative than a general comparison of gene expression between wildtype and null animals. Alternatively, preliminary experiments could first be done to globally identify genes aberrantly expressed in the OE in the absence of silencing mechanis ms; direct repression of these newly identified candidate genes by MBD2 or MeCP2 could then be tested. Repression could be perturbed at multiple stages of deve lopment with HDAC inhibitors or by disrupting DNA methylation. The de novo DNMTs, 3a and 3b, are also expressed at defined  140  transitional stages of ORN differentiation (MacDonald et al., 2005). However, the early lethality of the DNMT3a and DNMT3b null mice (Okano et al., 1999) has precluded examination of their role in neuronal differentiation. Conditional null alleles have now been generated (Kaneda et al., 2004; Lin et al., 2006; Nguyen et al., 2007), which could be used to knockout the DNMTs at different stages of ORN differentiation and compare the effect on gene expression.  Identifying the genes targeted for repression by MBD2, MeCP2 and the HDACs in the OE will help to reveal the function of gene silencing at distinct stages of neuronal development, a mechanism that is far less understood than stage-specific gene activation. By identifying the stages of differentiation mediated by each of these MBDs and HDACs, and determining the phenotypic consequences of the MBD2 or MeCP2 disruption, the results presented here will hopefully facilitate this next important step.  141  References  Acharya MR, Karp JE, Sausville EA, Hwang K, Ryan Q, Gojo I, Venitz J, Figg WD, Sparreboom A (2006) Factors affecting the pharmacokinetic profile of MS-275, a novel histone deacetylase inhibitor, in patients with cancer. Invest New Drugs 24:367-375. Ahringer J (2000) NuRD and SIN3 histone deacetylase complexes in development. Trends Genet 16:351-356. Alarcon JM, Malleret G, Touzani K, Vronskaya S, Ishii S, Kandel ER, Barco A (2004) Chromatin acetylation, memory, and LTP are impaired in CBP+/- mice: a model for the cognitive deficit in Rubinstein-Taybi syndrome and its amelioration. Neuron 42:947-959. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY (1999) Rett syndrome is caused by mutations in X- linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23:185-188. Amir RE, Van den Veyver IB, Schultz R, Malicki DM, Tran CQ, Dahle EJ, Philippi A, Timar L, Percy AK, Motil KJ, Lichtarge O, Smith EO, Glaze DG, Zoghbi HY (2000) Influence of mutation type and X chromosome inactivation on Rett syndrome phenotypes. Ann Neurol 47:670-679. Anlezark G, Horton RW, Meldrium BS, Sawaya CB (1976) Anticonvulsant action of ethanolamine-O-sulphate and di- n-propylacetate and the metabolism of gammaaminobutyric acid (GABA) in mice with audiogenic seizures. Biochem Pharmacol 25:413-417. Armstrong D, Dunn JK, Antalffy B, Trivedi R (1995) Selective dendritic alterations in the cortex of Rett syndrome. J Neuropathol Exp Neurol 54:195-201. Aten JA, Bakker PJ, Stap J, Boschman GA, Veenhof CH (1992) DNA double labelling with IdUrd and CldUrd for spatial and temporal analysis of cell proliferation and DNA replication. Histochem J 24:251-259. Azuara V, Perry P, Sauer S, Spivakov M, Jorgensen HF, John RM, Gouti M, Casanova M, Warnes G, Merkenschlager M, Fisher AG (2006) Chromatin signatures of pluripotent cell lines. Nat Cell Biol 8:532-538. Bai G, Sheng N, Xie Z, Bian W, Yokota Y, Benezra R, Kageyama R, Guillemot F, Jing N (2007) Id sustains Hes1 expression to inhibit precocious neurogenesis by releasing negative autoregulation of Hes1. Dev Cell 13:283-297. Ballas N, Mandel G (2005) The many faces of REST oversee epigenetic programming of neuronal genes. Curr Opin Neurobiol 15:500-506.  142  Ballas N, Grunseich C, Lu DD, Speh JC, Mandel G (2005) REST and its corepressors mediate plasticity of neuronal gene chromatin throughout neurogenesis. Cell 121:645-657. Ballas N, Battaglioli E, Atouf F, Andres ME, Chenoweth J, Anderson ME, Burger C, Moniwa M, Davie JR, Bowers WJ, Federoff HJ, Rose DW, Rosenfeld MG, Brehm P, Mandel G (2001) Regulation of neuronal traits by a novel transcriptional complex. Neuron 31:353365. Barnea G, O'Donnell S, Mancia F, Sun X, Nemes A, Mendelsohn M, Axel R (2004) Odorant receptors on axon termini in the brain. Science 304:1468. Barr H, Hermann A, Berger J, Tsai HH, Adie K, Prokhortchouk A, Hendrich B, Bird A (2007) Mbd2 contributes to DNA methylation-directed repression of the Xist gene. Mol Cell Biol 27:3750-3757. Bartl S, Taplick J, Lagger G, Khier H, Kuchler K, Seiser C (1997) Identification of mouse histone deacetylase 1 as a growth factor- inducible gene. Mol Cell Biol 17:5033-5043. Bauer S, Rasika S, Han J, Mauduit C, Raccurt M, Morel G, Jourdan F, Benahmed M, Moyse E, Patterson PH (2003) Leukemia inhibitory factor is a key signal for injury- induced neurogenesis in the adult mouse olfactory epithelium. J Neurosci 23:1792-1803. Bauman ML, Kemper TL, Arin DM (1995) Microscopic observations of the brain in Rett syndrome. Neuropediatrics 26:105-108. Belichenko PV, Hagberg B, Dahlstrom A (1997) Morphological study of neocortical areas in Rett syndrome. Acta Neuropathol (Berl) 93:50-61. Berger J, Sansom O, Clarke A, Bird A (2007) MBD2 is required for correct spatial gene expression in the gut. Mol Cell Biol 27:4049-4057. Bernstein BE, Meissner A, Lander ES (2007) The mammalian epigenome. Cell 128:669-681. Bernstein BE, Kamal M, Lindblad-Toh K, Bekiranov S, Bailey DK, Huebert DJ, McMahon S, Karlsson EK, Kulbokas EJ, 3rd, Gingeras TR, Schreiber SL, Lander ES (2005) Genomic maps and comparative analysis of histone modifications in human and mouse. Cell 120:169-181. Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert DJ, Cuff J, Fry B, Meissner A, Wernig M, Plath K, Jaenisch R, Wagschal A, Feil R, Schreiber SL, Lander ES (2006) A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125:315-326. Bienvenu T, Chelly J (2006) Molecular genetics of Rett syndrome: when DNA methylation goes unrecognized. Nat Rev Genet 7:415-426.  143  Bird A (1992) The essentials of DNA methylation. Cell 70:5-8. Bird A (2002) DNA methylation patterns and epigenetic memory. Genes Dev 16:6-21. Bjerling P, Silverstein RA, Thon G, Caudy A, Grewal S, Ekwall K (2002) Functional divergence between histone deacetylases in fission yeast by distinct cellular localization and in vivo specificity. Mol Cell Biol 22:2170-2181. Boeke J, Ammerpohl O, Kegel S, Moehren U, 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. Boutillier AL, Trinh E, Loeffler JP (2002) Constitutive repression of E2F1 transcriptional activity through HDAC proteins is essential for neuronal survival. Ann N Y Acad Sci 973:438-442. Boutillier AL, Trinh E, Loeffler JP (2003) Selective E2F-dependent gene transcription is controlled by histone deacetylase activity during neuronal apoptosis. J Neurochem 84:814-828. Bozza T, Feinstein P, Zheng C, Mombaerts P (2002) Odorant receptor expression defines functional units in the mouse olfactory system. J Neurosci 22:3033-3043. Braunschweig D, Simcox T, Samaco RC, LaSalle JM (2004) X-Chromosome inactivation ratios affect wild-type MeCP2 expression within mosaic Rett syndrome and Mecp2-/+ mouse brain. Hum Mol Genet 13:1275-1286. Brero A, Easwaran HP, Nowak D, Grunewald I, Cremer T, Leonhardt H, Cardoso MC (2005) Methyl CpG-binding proteins induce large-scale chromatin reorganization during terminal differentiation. J Cell Biol 169:733-743. Brooks PJ, Marietta C, Goldman D (1996) DNA mismatch repair and DNA methylation in adult brain neurons. J Neurosci 16:939-945. Brunjes PC, Borror MJ (1983) Unilateral odor deprivation: differential effects due to time of treatment. Brain Res Bull 11:501-503. Buck L, Axel R (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65:175-187. Caballero IM, Hendrich B (2005) MeCP2 in neurons: closing in on the causes of Rett syndrome. Hum Mol Genet 14 Spec No 1:R19-26. Calof AL, Chikaraishi DM (1989) Analysis of neurogenesis in a mammalian neuroepithelium: proliferation and differentiation of an olfactory neuron precursor in vitro. Neuron 3:115127.  144  Calof AL, Lander AD, Chikaraishi DM (1991) Regulation of neurogenesis and neuronal differentiation in primary and immortalized cells from mouse olfactory epithelium. Ciba Found Symp 160:249-265; discussion 265-276. Calof AL, Mumm JS, Rim PC, Shou J (1998a) The neuronal stem cell of the olfactory epithelium. J Neurobiol 36:190-205. Calof AL, Hagiwara N, Holcomb JD, Mumm JS, Shou J (1996) Neurogenesis and cell death in olfactory epithelium. J Neurobiol 30:67-81. Calof AL, Rim PC, Askins KJ, Mumm JS, Gordon MK, Iannuzzelli P, Shou J (1998b) Factors regulating neurogenesis and programmed cell death in mouse olfactory epithelium. Ann N Y Acad Sci 855:226-229. Cao L, Dhilla A, Mukai J, Blazeski R, Lodovichi C, Mason CA, Gogos JA (2007) Genetic modulation of BDNF signaling affects the outcome of axonal competition in vivo. Curr Biol 17:911-921. Carleton A, Rochefort C, Morante-Oria J, Desmaisons D, Vincent JD, Gheusi G, Lledo PM (2002) Making scents of olfactory neurogenesis. J Physiol Paris 96:115-122. Carr VM, Farbman AI (1992) Ablation of the olfactory bulb up-regulates the rate of neurogenesis and induces precocious cell death in olfactory epithelium. Exp Neurol 115:55-59. Carr VM, Farbman AI (1993) The dynamics of cell death in the olfactory epithelium. Exp Neurol 124:308-314. Carter LA, MacDonald JL, Roskams AJ (2004) Olfactory horizontal basal cells demonstrate a conserved multipotent progenitor phenotype. J Neurosci 24:5670-5683. Cau E, Casarosa S, Guillemot F (2002) Mash1 and Ngn1 control distinct steps of determination and differentiation in the olfactory sensory neuron lineage. Development 129:1871-1880. Chang Q, Khare G, Dani V, Nelson S, Jaenisch R (2006) The disease progression of Mecp2 mutant mice is affected by the level of BDNF expression. Neuron 49:341-348. Charman T, Neilson TC, Mash V, Archer H, Gardiner MT, Knudsen GP, McDonnell A, Perry J, Whatley SD, Bunyan DJ, Ravn K, Mount RH, Hastings RP, Hulten M, Orstavik KH, Reilly S, Cass H, Clarke A, Kerr AM, Bailey ME (2005) Dimensional phenotypic analysis and functional categorisation of mutations reveal novel genotype-phenotype associations in Rett syndrome. Eur J Hum Genet 13:1121-1130. Chen RZ, Akbarian S, Tudor M, Jaenisch R (2001) Deficiency of methyl-CpG binding protein-2 in CNS neurons results in a Rett- like phenotype in mice. Nat Genet 27:327-331.  145  Chen RZ, Pettersson U, Beard C, Jackson-Grusby L, Jaenisch R (1998) DNA hypomethylation leads to elevated mutation rates. Nature 395:89-93. Chen T, Ueda Y, Dodge JE, Wang Z, Li E (2003a) Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b. Mol Cell Biol 23:5594-5605. Chen WG, Chang Q, Lin Y, Meissner A, West AE, Griffith EC, Jaenisch R, Greenberg ME (2003b) Derepression of BDNF transcription involves calcium-dependent phosphorylation of MeCP2. Science 302:885-889. Chesler AT, Zou DJ, Le Pichon CE, Peterlin ZA, Matthews GA, Pei X, Miller MC, Firestein S (2007) A G protein/cAMP signal cascade is required for axonal convergence into olfactory glomeruli. Proc Natl Acad Sci U S A 104:1039-1044. Chess A, Simon I, Cedar H, Axel R (1994) Allelic inactivation regulates olfactory receptor gene expression. Cell 78:823-834. Chong JA, Tapia-Ramirez J, Kim S, Toledo-Aral JJ, Zheng Y, Boutros MC, Altshuller YM, Frohman MA, Kraner SD, Mandel G (1995) REST: a mammalian silencer protein that restricts sodium channel gene expression to neurons. Cell 80:949-957. Chwang WB, O'Riordan KJ, Levenson JM, Sweatt JD (2006) ERK/MAPK regulates hippocampal histone phosphorylation following contextual fear conditioning. Learn Mem 13:322-328. Cinatl J, Jr., Kotchetkov R, Blaheta R, Drie ver PH, Vogel JU, Cinatl J (2002) Induction of differentiation and suppression of malignant phenotype of human neuroblastoma BE(2)C cells by valproic acid: enhancement by combination with interferon-alpha. Int J Oncol 20:97-106. Cinatl J, Jr., Cinatl J, Driever PH, Kotchetkov R, Pouckova P, Kornhuber B, Schwabe D (1997) Sodium valproate inhibits in vivo growth of human neuroblastoma cells. Anticancer Drugs 8:958-963. Cloutier JF, Giger RJ, Koentges G, Dulac C, Kolodkin AL, Ginty DD (2002) Neuropilin-2 mediates axonal fasciculation, zonal segregation, but not axonal convergence, of primary accessory olfactory neurons. Neuron 33:877-892. Cohen DR, Matarazzo V, Palmer AM, Tu Y, Jeon OH, Pevsner J, Ronnett GV (2003) Expression of MeCP2 in olfactory receptor neurons is developmentally regulated and occurs before synaptogenesis. Mol Cell Neurosci 22:417-429. Col JA, Matsuo T, Storm DR, Rodriguez I (2007) Adenylyl cyclase-dependent axonal targeting in the olfactory system. Development 134:2481-2489.  146  Colantuoni C, Jeon OH, Hyder K, Chenchik A, Khimani AH, Narayanan V, Hoffman EP, Kaufmann WE, Naidu S, Pevsner J (2001) Gene expression profiling in postmortem Rett Syndrome brain: differential gene expression and patient classification. Neurobiol Dis 8:847-865. Costanzo RM, Graziadei PP (1983) A quantitative analysis of changes in the olfactory epithelium following bulbectomy in hamster. J Comp Neurol 215:370-381. Covington EC (1998) Anticonvulsants for neuropathic pain and detoxification. Cleve Clin J Med 65 Suppl 1:SI21-29; discussion SI45-27. Cowan CM, Roskams AJ (2004) Caspase-3 and caspase-9 mediate developmental apoptosis in the mouse olfactory system. J Comp Neurol 474:136-148. Cowan CM, Thai J, Krajewski S, Reed JC, Nicholson DW, Kaufmann SH, Roskams AJ (2001) Caspases 3 and 9 send a pro-apoptotic signal from synapse to cell body in olfactory receptor neurons. J Neurosci 21:7099-7109. Crandall JE, Dibble C, Butler D, Pays L, Ahmad N, Kostek C, Puschel AW, Schwarting GA (2000) Patterning of olfactory sensory connections is mediated by extracellular matrix proteins in the nerve layer of the olfactory bulb. Journal of Neurobiology 45:195-206. Cress WD, Seto E (2000) Histone deacetylases, transcriptional control, and cancer. J Cell Physiol 184:1-16. Cross SH, Meehan RR, Nan X, Bird A (1997) A component of the transcriptional repressor MeCP1 shares a motif with DNA methyltransferase and HRX proteins. Nat Genet 16:256-259. Csankovszki G, Nagy A, Jaenisch R (2001) Synergism of Xist RNA, DNA methylation, and histone hypoacetylation in maintaining X chromosome inactivation. J Cell Biol 153:773784. Czuczwar SJ, Frey HH, Loscher W (1985) Antagonism of N-methyl- D,L-aspartic acid- induced convulsions by antiepileptic drugs and other agents. Eur J Pharmacol 108:273-280. Dani VS, Chang Q, Maffei A, Turrigiano GG, Jaenisch R, Nelson SB (2005) Reduced cortical activity due to a shift in the balance between excitation and inhibition in a mouse model of Rett syndrome. Proc Natl Acad Sci U S A 102:12560-12565. de Ruijter AJ, van Gennip AH, Caron HN, Kemp S, van Kuilenburg AB (2003) Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem J 370:737-749.  147  DeHamer MK, Guevara JL, Hannon K, Olwin BB, Calof AL (1994) Genesis of olfactory receptor neurons in vitro: regulation of progenitor cell divisions by fibroblast growth factors. Neuron 13:1083-1097. Doucette R (1990) Glial influences on axonal growth in the primary olfactory system. Glia 3:433-449. Doucette R (1991) PNS-CNS transitional zone of the first cranial nerve. J Comp Neurol 312:451-466. Eggan K, Baldwin K, Tackett M, Osborne J, Gogos J, Chess A, Axel R, Jaenisch R (2004) Mice cloned from olfactory sensory neurons. Nature 428:44-49. Egger G, Liang G, Aparicio A, Jones PA (2004) Epigenetics in human disease and prospects for epigenetic therapy. Nature 429:457-463. Ehrlich M, Gama-Sosa MA, Huang LH, Midgett RM, Kuo KC, McCune RA, Gehrke C (1982) Amount and distribution of 5- methylcytosine in human DNA from different types of tissues of cells. Nucleic Acids Res 10:2709-2721. Emrich HM, von Zerssen D, Kissling W, Moller HJ, Windorfer A (1980) Effect of sodium valproate on mania. The GABA-hypothesis of affective disorders. Arch Psychiatr Nervenkr 229:1-16. Endres M, Fan G, Meisel A, Dirnagl U, Jaenisch R (2001) Effects of cerebral ischemia in mice lacking DNA methyltransferase 1 in post- mitotic neurons. Neuroreport 12:3763-3766. Endres M, Meisel A, Biniszkiewicz D, Namura S, Prass K, Ruscher K, Lipski A, Jaenisch R, Moskowitz MA, Dirnagl U (2000) DNA methyltransferase contributes to delayed ischemic brain injury. J Neurosci 20:3175-3181. Fan G, Martinowich K, Chin MH, He F, Fouse SD, Hutnick L, Hattori D, Ge W, Shen Y, Wu H, ten Hoeve J, Shuai K, Sun YE (2005) DNA methylation controls the timing of astrogliogenesis through regulation of JAK-STAT signaling. Development 132:33453356. Fan G, Beard C, Chen RZ, Csankovszki G, Sun Y, Siniaia M, Biniszkiewicz D, Bates B, Lee PP, Kuhn R, Trumpp A, Poon C, Wilson CB, Jaenisch R (2001) DNA hypomethylation perturbs the function and survival of CNS neurons in postnatal animals. J Neurosci 21:788-797. Farbman AI (1992) Cell Biology of Olfaction. Cambridge: Cambridge University Press. Farbman AI, Margolis FL (1980) Olfactory marker protein during ontogeny: immunohistochemical localization. Dev Biol 74:205-215.  148  Farbman AI, Buchholz JA (1996) Transforming growth factor-alpha and other growth factors stimulate cell division in olfactory epithelium in vitro. J Neurobiol 30:267-280. Farrant M, Webster RA (1989) Neuronal activity, amino acid concentration and amino acid release in the substantia nigra of the rat after sodium valproate. Brain Res 504:49-56. Feinstein P, Bozza T, Rodriguez I, Vassalli A, Mombaerts P (2004) Axon guidance of mouse olfactory sensory neurons by odorant receptors and the beta2 adrenergic receptor. Cell 117:833-846. Feng J, Chang H, Li E, Fan G (2005) Dynamic expression of de novo DNA methyltransferases Dnmt3a and Dnmt3b in the central nervous system. J Neurosci Res 79:734-746. Feng Q, Zhang Y (2001) The MeCP1 comp lex represses transcription through preferential binding, remodeling, and deacetylating methylated nucleosomes. Genes Dev 15:827-832. Ferguson KL, Slack RS (2001) The Rb pathway in neurogenesis. Neuroreport 12:A55-62. Finnin MS, Donigian JR, Cohen A, Richo n VM, Rifkind RA, Marks PA, Breslow R, Pavletich NP (1999) Structures of a histone deacetylase homologue bound to the TSA and SAHA inhibitors. Nature 401:188-193. Fischle W, Dequiedt F, Hendzel MJ, Guenther MG, Lazar MA, Voelter W, Verdin E (2002) Enzymatic activity associated with class II HDACs is dependent on a multiprotein complex containing HDAC3 and SMRT/N-CoR. Mol Cell 9:45-57. Fisher AG, Merkenschlager M (2002) Gene silencing, cell fate and nuclear organisation. Curr Opin Genet Dev 12:193-197. Fraga MF, Ballestar E, Montoya G, Taysavang P, Wade PA, Esteller M (2003) The affinity of different MBD proteins for a specific methylated locus depends on their intrinsic binding properties. Nucleic Acids Res 31:1765-1774. Franceschetti S, Hamon B, Heinemann U (1986) The action of valproate on spontaneous epileptiform activity in the absence of synaptic transmission and on evoked changes in [Ca2+]o and [K+]o in the hippocampal slice. Brain Res 386:1-11. Francis F, Koulakoff A, Boucher D, Chafey P, Schaar B, Vinet MC, Friocourt G, McDonnell N, Reiner O, Kahn A, McConnell SK, Berwald-Netter Y, Denoulet P, Chelly J (1999) Doublecortin is a developmentally regulated, microtubule-associated protein expressed in migrating and differentiating neurons. Neuron 23:247-256. Frazier-Cierpial L, Brunjes PC (1989) Early postnatal cellular proliferation and survival in the olfactory bulb and rostral migratory stream of normal and unilaterally odor-deprived rats. J Comp Neurol 289:481-492.  149  Fujita N, Takebayashi S, Okumura K, Kudo S, Chiba T, Saya H, Nakao M (1999) Methylationmediated transcriptional silencing in euchromatin by methyl-CpG binding protein MBD1 isoforms. Mol Cell Biol 19:6415-6426. Fujita N, Shimotake N, Ohki I, Chiba T, Saya H, Shirakawa M, Nakao M (2000) Mechanism of transcriptional regulation by methyl-CpG binding protein MBD1. Mol Cell Biol 20:51075118. Fuks F (2005) DNA methylation and histone modifications: teaming up to silence genes. Curr Opin Genet Dev 15:490-495. Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T (2003) The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J Biol Chem 278:40354040. Fukuda T, Itoh M, Ichikawa T, Washiyama K, Goto Y (2005) Delayed maturation of neuronal architecture and synaptogenesis in cerebral cortex of Mecp2-deficient mice. J Neuropathol Exp Neurol 64:537-544. Fuss SH, Omura M, Mombaerts P (2007) Local and cis effects of the H element on expression of odorant receptor genes in mouse. Cell 130:373-384. Gage FH (2000) Mammalian Neural Stem Cells. Science 287:1433-1439. Gale K (1986) Role of the substantia nigra in GABA- mediated anticonvulsant actions. Adv Neurol 44:343-364. Gandelman R, Zarrow MX, Denenberg VH, Myers M (1971) Olfactory bulb removal eliminates maternal behavior in the mouse. Science 171:210-211. Gemelli T, Berton O, Nelson ED, Perrotti LI, Jaenisch R, Monteggia LM (2006) Postnatal loss of methyl-CpG binding protein 2 in the forebrain is sufficient to mediate behavioral aspects of Rett syndrome in mice. Biol Psychiatry 59:468-476. Georgel PT, Horowitz-Scherer RA, Adkins N, Woodcock CL, Wade PA, 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. Getchell TV, Shah DS, Partin JV, Subhedar NK, Getchell ML (2002) Leukemia inhibitory factor mRNA expression is upregulated in macrophages and olfactory receptor neurons after target ablation. J Neurosci Res 67:246-254. Giacometti E, Luikenhuis S, Beard C, Jaenisch R (2007) Partial rescue of MeCP2 deficiency by postnatal activation of MeCP2. Proc Natl Acad Sci U S A 104:1931-1936.  150  Glass DJ, Nye SH, Hantzopoulos P, Macchi MJ, Squinto SP, Goldfarb M, Yancopoulos GD (1991) TrkB mediates BDNF/NT-3-dependent survival and proliferatio n in fibroblasts lacking the low affinity NGF receptor. Cell 66:405-413. Gleeson JG, Lin PT, Flanagan LA, Walsh CA (1999) Doublecortin is a microtubule-associated protein and is expressed widely by migrating neurons. Neuron 23:257-271. Godin Y, Heiner L, Mark J, Mandel P (1969) Effects of DI-n-propylacetate, and anticonvulsive compound, on GABA metabolism. J Neurochem 16:869-873. Goldberg AD, Allis CD, Bernstein E (2007) Epigenetics: a landscape takes shape. Cell 128:635638. Goldstein BJ, Wolozin BL, Schwob JE (1997) FGF2 suppresses neuronogenesis of a cell line derived from rat olfactory epithelium. J Neurobiol 33:411-428. Goll MG, Bestor TH (2005) Eukaryotic cytosine methyltransferases. Annu Rev Biochem 74:481-514. Gordon MK, Mumm JS, Davis RA, Holcomb JD, Calof AL (1995) Dynamics of MASH1 expression in vitro and in vivo suggest a non-stem cell site of MASH1 action in the olfactory receptor neuron lineage. Mol Cell Neurosci 6:363-379. Gore SD, Weng LJ, Figg WD, Zhai S, Donehower RC, Dover G, Grever MR, Griffin C, Grochow LB, Hawkins A, Burks K, Zabelena Y, Miller CB (2002) Impact of prolonged infusions of the putative differentiating agent sodium phenylbutyrate on myelodysplastic syndromes and acute myeloid leukemia. Clin Cancer Res 8:963-970. Goto K, Numata M, Komura JI, Ono T, Bestor TH, Kondo H (1994) Expression of DNA methyltransferase gene in mature and immature neurons as well as proliferating cells in mice. Differentiation 56:39-44. Gottlicher M, Minucci S, Zhu P, Kramer OH, Schimpf A, Giavara S, Sleeman JP, Lo Coco F, Nervi C, Pelicci PG, Heinzel T (2001) Valproic acid defines a novel class of HDAC inhibitors inducing differentiation of transformed cells. Embo J 20:6969-6978. Grayson DR, Jia X, Chen Y, Sharma RP, Mitchell CP, Guidotti A, Costa E (2005) Reelin promoter hypermethylation in schizophrenia. Proc Natl Acad Sci U S A 102:9341-9346. Graziadei GA, Graziadei PP (1979a) Neurogenesis and neuron regeneration in the olfactory system of mammals. II. Degeneration and reconstitution of the olfactory sensory neurons after axotomy. J Neurocytol 8:197-213. Graziadei PP, Graziadei GA (1979b) Neurogenesis and neuron regeneration in the olfactory system of mammals. I. Morphological aspects of differentiation and structural organization of the olfactory sensory neurons. J Neurocytol 8:1-18.  151  Graziadei PP, Levine RR, Monti Graziadei GA (1979) Plasticity of connections of the olfactory sensory neuron: regeneration into the forebrain following bulbectomy in the neonatal mouse. Neuroscience 4:713-727. Greene LA, Biswas SC, Liu DX (2004) Cell cycle molecules and vertebrate neuron death: E2F at the hub. Cell Death Differ 11:49-60. Guan Z, Giustetto M, Lomvardas S, Kim JH, Miniaci MC, Schwartz JH, Thanos D, Kandel ER (2002) Integration of long-term- memory-related synaptic plasticity involves bidirectional regulation of gene expression and chromatin structure. Cell 111:483-493. Gurley LR, D'Anna JA, Barham SS, Deaven LL, Tobey RA (1978) Histone phosphorylation and chromatin structure during mitosis in Chinese hamster cells. Eur J Biochem 84:1-15. Gurvich N, Klein PS (2002) Lithium and valproic acid: parallels and contrasts in diverse signaling contexts. Pharmacol Ther 96:45-66. Guy J, Hendrich B, Holmes M, Martin JE, Bird A (2001) A mouse Mecp2- null mutation causes neurological symptoms that mimic Rett syndrome. Nat Genet 27:322-326. Guy J, Gan J, Selfridge J, Cobb S, Bird A (2007) Reversal of neurological defects in a mouse model of Rett syndrome. Science 315:1143-1147. Hagberg B (1985) Rett's syndrome: prevalence and impact on progressive severe mental retardation in girls. Acta Paediatr Scand 74:405-408. Hagberg B, Aicardi J, Dias K, Ramos O (1983) A progressive syndrome of autism, dementia, ataxia, and loss of purposeful hand use in girls: Rett's syndrome: report of 35 cases. Ann Neurol 14:471-479. Hammer S, Dorrani N, Hartiala J, Stein S, Schanen NC (2003) Rett syndrome in a 47,XXX patient with a de novo MECP2 mutation. Am J Med Genet A 122:223-226. Hao Y, Creson T, Zhang L, Li P, Du F, Yuan P, Gould TD, Manji HK, Chen G (2004) Mood stabilizer valproate promotes ERK pathway-dependent cortical neuronal growth and neurogenesis. J Neurosci 24:6590-6599. Hariton C, Ciesielski L, Simler S, Valli M, Jadot G, Gobaille S, Mesdjian E, Mandel P (1984) Distribution of sodium valproate and GABA metabolism in CNS of the rat. Biopharm Drug Dispos 5:409-414. Hark AT, Schoenherr CJ, Katz DJ, Ingram RS, Levorse JM, Tilghman SM (2000) CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature 405:486-489.  152  He LZ, Tolentino T, Grayson P, Zhong S, Warrell RP, Jr., Rifkind RA, Marks PA, Richon VM, Pandolfi PP (2001) Histone deacetylase inhibitors induce remission in transgenic models of therapy-resistant acute promyelocytic leukemia. J Clin Invest 108:1321-1330. Hempstead JL, Morgan JI (1983) Monoclonal antibodies to the rat olfactory sustentacular cell. Brain Res 288:289-295. Hendrich B, Bird A (1998) Identification and characterization of a family of mammalian methylCpG binding proteins. Mol Cell Biol 18:6538-6547. Hendrich B, Tweedie S (2003) The methyl-CpG binding domain and the evolving role of DNA methylation in animals. Trends Genet 19:269-277. Hendrich B, Hardeland U, Ng HH, Jiricny J, Bird A (1999a) 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, Bird A (2001) Closely related proteins MBD2 and MBD3 play distinctive but interacting roles in mouse development. Genes Dev 15:710-723. Hendrich B, Abbott C, McQueen H, Chambers D, Cross S, Bird A (1999b) Genomic structure and chromosomal mapping of the murine and human Mbd1, Mbd2, Mbd3, and Mbd4 genes. Mamm Genome 10:906-912. Hinds JW, Hinds PL (1976) Synapse formation in the mouse olfactory bulb. II. Morphogenesis. J Comp Neurol 169:41-61. Holcomb JD, Mumm JS, Calof AL (1995) Apoptosis in the neuronal lineage of the mouse olfactory epithelium: regulation in vivo and in vitro. Dev Biol 172:307-323. Hong L, Schroth GP, Matthews HR, Yau P, Bradbury EM (1993) Studies of the DNA binding properties of histone H4 amino terminus. Thermal denaturation studies reveal that acetylation markedly reduces the binding constant of the H4 "tail" to DNA. J Biol Chem 268:305-314. Horike S, Cai S, Miyano M, Cheng JF, Kohwi-Shigematsu T (2005) Loss of silent-chromatin looping and impaired imprinting of DLX5 in Rett syndrome. Nat Genet 37:31-40. Hou ST, Callaghan D, Fournier MC, Hill I, Kang L, Massie B, Morley P, Murray C, Rasquinha I, Slack R, MacManus JP (2000) The transcription factor E2F1 modulates apoptosis of neurons. J Neurochem 75:91-100. Hsieh J, Gage FH (2005) Chromatin remodeling in neural development and plasticity. Curr Opin Cell Biol 17:664-671.  153  Hsieh J, Nakashima K, Kuwabara T, Mejia E, Gage FH (2004) Histone deacetylase inhibitionmediated neuronal differentiation of multipotent adult neural progenitor cells. Proc Natl Acad Sci U S A 101:16659-16664. Huang Y, Doherty JJ, Dingledine R (2002) Altered histone acetylation at glutamate receptor 2 and brain-derived neurotrophic factor genes is an early event triggered by status epilepticus. J Neurosci 22:8422-8428. Huard JM, Schwob JE (1995) Cell cycle of globose basal cells in rat olfactory epithelium. Dev Dyn 203:17-26. Huard JM, Youngentob SL, Goldstein BJ, Luskin MB, Schwob JE (1998) Adult olfactory epithelium contains multipotent progenitors that give rise to neurons and non- neural cells. J Comp Neurol 400:469-486. Hutchins AS, Mullen AC, Lee HW, Sykes KJ, High FA, Hendrich BD, Bird AP, Reiner SL (2002) Gene silencing quantitatively controls the function of a developmental transactivator. Mol Cell 10:81-91. Imai T, Suzuki M, Sakano H (2006) Odorant receptor-derived cAMP signals direct axonal targeting. Science 314:657-661. Inano K, Suetake I, Ueda T, Miyake Y, Nakamura M, Okada M, Tajima S (2000) Maintenancetype DNA methyltransferase is highly expressed in post- mitotic neurons and localized in the cytoplasmic compartment. J Biochem (Tokyo) 128:315-321. Ip NY, Li Y, Yancopoulos GD, Lindsay RM (1993a) Cultured hippocampal neurons show responses to BDNF, NT-3, and NT-4, but not NGF. J Neurosci 13:3394-3405. Ip NY, Stitt TN, Tapley P, Klein R, Glass DJ, Fandl J, Greene LA, Barbacid M, Yancopoulos GD (1993b) Similarities and differences in the way neurotrophins interact with the Trk receptors in neuronal and nonneuronal cells. Neuron 10:137-149. Ip NY, Ibanez CF, Nye SH, McClain J, Jones PF, Gies DR, Belluscio L, Le Beau MM, Espinosa Rd, Squinto SP, et al. (1992) Mammalian neurotrophin-4: structure, chromosomal localization, tissue distribution, and receptor specificity. Proc Natl Acad Sci U S A 89:3060-3064. Iwema CL, Fang H, Kurtz DB, Youngentob SL, Schwob JE (2004) Odorant receptor expression patterns are restored in lesion-recovered rat olfactory epithelium. J Neurosci 24:356-369. Jaboin J, Wild J, Hamidi H, Khanna C, Kim CJ, Robey R, Bates SE, Thiele CJ (2002) MS-27275, an inhibitor of histone deacetylase, has marked in vitro and in vivo antitumor activity against pediatric solid tumors. Cancer Res 62:6108-6115.  154  Jackson-Grusby L, Beard C, Possemato R, Tudor M, Fambrough D, Csankovszki G, Dausman J, Lee P, Wilson C, Lander E, Jaenisch R (2001) Loss of genomic methylation causes p53dependent apoptosis and epigenetic deregulation. Nat Genet 27:31-39. Jaenisch R, Bird A (2003) Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat Genet 33 Suppl:245-254. Jellinger K, Armstrong D, Zoghbi HY, Percy AK (1988) Neuropathology of Rett syndrome. Acta Neuropathol (Berl) 76:142-158. Jenuwein T, Allis CD (2001) Translating the histone code. Science 293:1074-1080. Jian L, Archer HL, Ravine D, Kerr A, de Klerk N, Christodoulou J, Bailey ME, Laurvick C, Leonard H (2005) p.R270X MECP2 mutation and mortality in Rett syndrome. Eur J Hum Genet 13:1235-1238. Jiang H, Nucifora FC, Jr., Ross CA, DeFranco DB (2003) Cell death triggered by polyglutamineexpanded huntingtin in a neuronal cell line is associated with degradation of CREBbinding protein. Hum Mol Genet 12:1-12. Jin K, Mao XO, Simon RP, Greenberg DA (2001) Cyclic AMP response element binding protein (CREB) and CREB binding protein (CBP) in global cerebral ischemia. J Mol Neurosci 16:49-56. Jogi A, Persson P, Grynfeld A, Pahlman S, Axelson H (2002) Modulation of basic helix- loophelix transcription complex formation by Id proteins during neuronal differentiation. J Biol Chem 277:9118-9126. Johannessen CU (2000) Mechanisms of action of valproate: a commentatory. Neurochem Int 37:103-110. Jones PA, Baylin SB (2002) The fundamental role of epigenetic events in cancer. Nat Rev Genet 3:415-428. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP (1998) Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet 19:187-191. Jorgensen HF, Ben-Porath I, Bird AP (2004) Mbd1 is recruited to both methylated and nonmethylated CpGs via distinct DNA binding domains. Mol Cell Biol 24:3387-3395. Jugloff DG, Jung BP, Purushotham D, Logan R, Eubanks JH (2005) Increased dendritic complexity and axonal length in cultured mouse cortical neurons overexpressing methylCpG-binding protein MeCP2. Neurobiol Dis 19:18-27.  155  Jung BP, Zhang G, Ho W, Francis J, Eubanks JH (2002) Transient forebrain ischemia alters the mRNA expression of methyl DNA-binding factors in the adult rat hippocampus. Neuroscience 115:515-524. Jung BP, Zhang G, Nitsch R, Trogadis J, Nag S, Eubanks JH (2003a) Differential expression of methyl CpG-binding domain containing factor MBD3 in the developing and adult rat brain. J Neurobiol 55:220-232. Jung BP, Jugloff DG, Zhang G, Logan R, Brown S, Eubanks JH (2003b) The expression of methyl CpG binding factor MeCP2 correlates with cellular differentiation in the developing rat brain and in cultured cells. J Neurobiol 55:86-96. Kaji K, Caballero IM, MacLeod R, Nichols J, Wilson VA, Hendrich B (2006) The NuRD component Mbd3 is required for pluripotency of embryonic stem cells. Nat Cell Biol 8:285-292. Kaneda M, Okano M, Hata K, Sado T, Tsujimoto N, Li E, Sasaki H (2004) Essential role for de novo DNA methyltransferase Dnmt3a in paternal and maternal imprinting. Nature 429:900-903. Kasowski HJ, Kim H, Greer CA (1999) Compartmental organization of the olfactory bulb glomerulus. J Comp Neurol 407:261-274. Khochbin S, Wolffe AP (1997) The origin and utility of histone deacetylases. FEBS Lett 419:157-160. Kim H, Greer CA (2000) The emergence of compartmental organization in olfactory bulb glomeruli during postnatal development. J Comp Neurol 422:297-311. Kishi N, Macklis JD (2004) MECP2 is progressively expressed in post- migratory neurons and is involved in neuronal maturation rather than cell fate decisions. Mol Cell Neurosci 27:306-321. Klose RJ, Sarraf SA, Schmiedeberg L, McDermott SM, Stancheva I, Bird AP (2005) DNA binding selectivity of MeCP2 due to a requirement for A/T sequences adjacent to methylCpG. Mol Cell 19:667-678. Knoepfler PS, Eisenman RN (1999) Sin meets NuRD and other tails of repression. Cell 99:447450. Korzus E, Rosenfeld MG, Mayford M (2004) CBP histone acetyltransferase activity is a critical component of memory consolidation. Neuron 42:961-972. Kouzarides T (2007) Chromatin modifications and their function. Cell 128:693-705.  156  Kramer OH, Zhu P, Ostendorff HP, Golebiewski M, Tiefenbach J, Peters MA, Brill B, Groner B, Bach I, Heinzel T, Gottlicher M (2003) The histone deacetylase inhibitor valproic acid selectively induces proteasomal degradation of HDAC2. Embo J 22:3411-3420. Kukino K, Deguchi T (1978) Anticonvulsant activity and effects of sodium dipropylacetate on cerebral 5-hydroxytryptamine and gamma-aminobutyric acid in reserpinized mice. Chem Pharm Bull (Tokyo) 26:3551-3555. Kumar A, Choi KH, Renthal W, Tsankova NM, Theobald DE, Truong HT, Russo SJ, Laplant Q, Sasaki TS, Whistler KN, Neve RL, Self DW, Nestler EJ (2005) Chromatin remodeling is a key mechanism underlying cocaine- induced plasticity in striatum. Neuron 48:303-314. Kurki P, Ogata K, Tan EM (1988) Monoclonal antibodies to proliferating cell nuclear antigen (PCNA)/cyclin as probes for proliferating cells by immunofluorescence microscopy and flow cytometry. J Immunol Methods 109:49-59. Kurki P, Vanderlaan M, Dolbeare F, Gray J, Tan EM (1986) Expression of proliferating cell nuclear antigen (PCNA)/cyclin during the cell cycle. Exp Cell Res 166:209-219. Laeng P, Pitts RL, Lemire AL, Drabik CE, Weiner A, Tang H, Thyagarajan R, Mallon BS, Altar CA (2004) The mood stabilizer valproic acid stimulates GABA neurogenesis from rat forebrain stem cells. J Neurochem 91:238-251. Lagger G, O'Carroll D, Rembold M, Khier H, Tischler J, Weitzer G, Schuettengruber B, Hauser C, Brunmeir R, Jenuwein T, Seiser C (2002) Essential function of histone deacetylase 1 in proliferation control and CDK inhibitor repression. Embo J 21:2672-2681. Laird PW, Jaenisch R (1996) The role of DNA methylation in cancer genetic and epigenetics. Annu Rev Genet 30:441-464. Lake RS, Salzman NP (1972) Occurrence and properties of a chromatin-associated F1-histone phosphokinase in mitotic Chinese hamster cells. Biochemistry 11:4817-4826. Lampen A, Siehler S, Ellerbeck U, Gottlicher M, Nau H (1999) New molecular bioassays for the estimation of the teratogenic potency of valproic acid derivatives in vitro: activation of the peroxisomal proliferator-activated receptor (PPARdelta). Toxicol Appl Pharmacol 160:238-249. Larsson OM, Gram L, Schousboe I, Schousboe A (1986) Differential effect of gamma-vinyl GABA and valproate on GABA-transaminase from cultured neurones and astrocytes. Neuropharmacology 25:617-625. Lea MA, Tulsyan N (1995) Discordant effects of butyrate analogues on erythroleukemia cell proliferation, differentiation and histone deacetylase. Anticancer Res 15:879-883.  157  Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS, Brockway KS, Byrnes EJ, Chen L, Chen TM, Chin MC, Chong J, Crook BE, Czaplinska A, Dang CN, Datta S, Dee NR, Desaki AL, Desta T, Diep E, Dolbeare TA, Donelan MJ, Dong HW, Dougherty JG, Duncan BJ, Ebbert AJ, Eichele G, Estin LK, Faber C, Facer BA, Fields R, Fischer SR, Fliss TP, Frensley C, Gates SN, Glattfelder KJ, Halverson KR, Hart MR, Hohmann JG, Howell MP, Jeung DP, Johnson RA, Karr PT, Kawal R, Kidney JM, Knapik RH, Kuan CL, Lake JH, Laramee AR, Larsen KD, Lau C, Lemon TA, Liang AJ, Liu Y, Luong LT, Michaels J, Morgan JJ, Morgan RJ, Mortrud MT, Mosqueda NF, Ng LL, Ng R, Orta GJ, Overly CC, Pak TH, Parry SE, Pathak SD, Pearson OC, Puchalski RB, Riley ZL, Rockett HR, Rowland SA, Royall JJ, Ruiz MJ, Sarno NR, Schaffnit K, Shapovalova NV, Sivisay T, Slaughterbeck CR, Smith SC, Smith KA, Smith BI, Sodt AJ, Stewart NN, Stumpf KR, Sunkin SM, Sutram M, Tam A, Teemer CD, Thaller C, Thompson CL, Varnam LR, Visel A, Whitlock RM, Wohnoutka PE, Wolkey CK, Wong VY, Wood M, et al. (2007) Genome-wide atlas of gene expression in the adult mouse brain. Nature 445:168-176. Leonard H, Bower C (1998) Is the girl with Rett syndrome normal at birth? Dev Med Child Neurol 40:115-121. Leung CT, Coulombe PA, Reed RR (2007) Contribution of olfactory neural stem cells to tissue maintenance and regeneration. Nat Neurosci 10:720-726. Levenson JM, O'Riordan KJ, Brown KD, Trinh MA, Molfese DL, Sweatt JD (2004) Regulation of histone acetylation during memory formation in the hippocampus. J Biol Chem 279:40545-40559. Levenson JM, Roth TL, Lubin FD, Miller CA, Huang IC, Desai P, Malone LM, Sweatt JD (2006) Evidence that DNA (cytosine-5) methyltransferase regulates synaptic plasticity in the hippocampus. J Biol Chem 281:15763-15773. Levi G, Puche AC, Mantero S, Barbieri O, Trombino S, Paleari L, Egeo A, Merlo GR (2003) The Dlx5 homeodomain gene is essential for olfactory development and connectivity in the mouse. Mol Cell Neurosci 22:530-543. Lewis JD, Meehan RR, Henzel WJ, Maurer-Fogy I, Jeppesen P, Klein F, Bird A (1992) Purification, sequence, and cellular localization of a novel chromosomal protein tha t binds to methylated DNA. Cell 69:905-914. Li E (2002) Chromatin modification and epigenetic reprogramming in mammalian development. Nat Rev Genet 3:662-673. Li E, Bestor TH, Jaenisch R (1992) Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell 69:915-926. Li J, Ishii T, Feinstein P, Mombaerts P (2004) Odorant receptor gene choice is reset by nuclear transfer from mouse olfactory sensory neurons. Nature 428:393-399.  158  Lim DA, Suarez-Farinas M, Naef F, Hacker CR, Menn B, Takebayashi H, Magnasco M, Patil N, Alvarez-Buylla A (2006) In vivo transcriptional profile analysis reveals RNA splicing and chromatin remodeling as prominent processes for adult neurogenesis. Mol Cell Neurosci 31:131-148. Lin H, Yamada Y, Nguyen S, Linhart H, Jackson-Grusby L, Meissner A, Meletis K, Lo G, Jaenisch R (2006a) Suppression of intestinal neoplasia by deletion of Dnmt3b. Mol Cell Biol 26:2976-2983. Lin HY, Chen CS, Lin SP, Weng JR (2006b) Targeting histone deacetylase in cancer therapy. Med Res Rev 26:397-413. Lind D, Franken S, Kappler J, Jankowski J, Schilling K (2005) Characterization of the neuronal marker NeuN as a multiply phosphorylated antigen with discrete subcellular localization. J Neurosci Res 79:295-302. Liu A, Han YR, Li J, Sun D, Ouyang M, Plummer MR, Casaccia-Bonnefil P (2007) The glial or neuronal fate choice of oligodendrocyte progenitors is modulated by their ability to acquire an epigenetic memory. J Neurosci 27:7339-7343. Lledo PM, Gheusi G (2003) Olfactory processing in a cha nging brain. Neuroreport 14:16551663. Long JE, Garel S, Depew MJ, Tobet S, Rubenstein JL (2003) DLX5 regulates development of peripheral and central components of the olfactory system. J Neurosci 23:568-578. Lopez-Carballo G, Moreno L, Masia S, Perez P, Barettino D (2002) Activation of the phosphatidylinositol 3-kinase/Akt signaling pathway by retinoic acid is required for neural differentiation of SH-SY5Y human neuroblastoma cells. J Biol Chem 277:2529725304. Loscher W (1989) Valproate enhances GABA turnover in the substantia nigra. Brain Res 501:198-203. Loscher W (1993) Effects of the antiepileptic drug valproate on metabolism and function of inhibitory and excitatory amino acids in the brain. Neurochem Res 18:485-502. Luger K, Mader AW, Richmond RK, Sargent DF, Richmond TJ (1997) Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389:251-260. Lunyak VV, Prefontaine GG, Rosenfeld MG (2004) REST and peace for the neuronal-specific transcriptional program. Ann N Y Acad Sci 1014:110-120. Lunyak VV, Burgess R, Prefontaine GG, Nelson C, Sze SH, Chenoweth J, Schwartz P, Pevzner PA, Glass C, Mandel G, Rosenfeld MG (2002) Corepressor-dependent silencing of chromosomal regions encoding neuronal genes. Science 298:1747-1752.  159  Lyssiotis CA, Walker J, Wu C, Kondo T, Schultz PG, Wu X (2007) Inhibition of histone deacetylase activity induces developmental plasticity in oligodendrocyte precursor cells. Proc Natl Acad Sci U S A 104:14982-14987. MacDonald JL, Gin CS, Roskams AJ (2005) Stage-specific induction of DNA methyltransferases in olfactory receptor neuron development. Dev Biol 288:461-473. MacManus JP, Koch CJ, Jian M, Walker T, Zurakowski B (1999) Decreased brain infarct following focal ischemia in mice lacking the transcription factor E2F1. Neuroreport 10:2711-2714. Mahanthappa NK, Schwarting GA (1993) Peptide growth factor control of olfactory neurogenesis and neuron survival in vitro: roles of EGF and TGF-beta s. Neuron 10:293305. Malnic B, Hirono J, Sato T, Buck LB (1999) Combinatorial receptor codes for odors. Cell 96:713-723. Marchion D, Munster P (2007) Development of histone deacetylase inhibitors for cancer treatment. Expert Rev Anticancer Ther 7:583-598. Marks PA, Richon VM, Breslow R, Rifkind RA (2001) Histone deacetylase inhibitors as new cancer drugs. Curr Opin Oncol 13:477-483. Martinowich K, Hattori D, Wu H, Fouse S, He F, Hu Y, Fan G, Sun YE (2003) DNA methylation-related chromatin remodeling in activity-dependent BDNF gene regulation. Science 302:890-893. Matarazzo V, Cohen D, Palmer AM, Simpson PJ, Khokhar B, Pan SJ, Ronnett GV (2004) The transcriptional repressor Mecp2 regulates terminal neuronal differentiation. Mol Cell Neurosci 27:44-58. Mayer W, Niveleau A, Walter J, Fundele R, Haaf T (2000) Demethylation of the zygotic paternal genome. Nature 403:501-502. McLean JH, Darby-King A, Bonnell WS (2001) Neonatal olfactory sensory deprivation decreases BDNF in the olfactory bulb of the rat. Brain Res Dev Brain Res 128:17-24. McLean MJ, Macdonald RL (1986) Sodium valproate, but not ethosuximide, produces use- and voltage-dependent limitation of high frequency repetitive firing of action potentials of mouse central neurons in cell culture. J Pharmacol Exp Ther 237:1001-1011. Meehan RR, Lewis JD, McKay S, Kleiner EL, Bird AP (1989) Identification of a mammalian protein that binds specifically to DNA containing methylated CpGs. Cell 58:499-507.  160  Meloni I, Bruttini M, Longo I, Mari F, Rizzolio F, D'Adamo P, Denvriendt K, Fryns JP, Toniolo D, Renieri A (2000) A mutation in the rett syndrome gene, MECP2, causes X- linked mental retardation and progressive spasticity in males. Am J Hum Genet 67:982-985. Meshorer E (2007) Chromatin in embryonic stem cell neuronal differentiation. Histol Histopathol 22:311-319. Meunier H, Carraz G, Neunier Y, Eymard P, Aimard M (1963) [Pharmacodynamic properties of N-dipropylacetic acid.]. Therapie 18:435-438. Miller CA, Sweatt JD (2007) Covalent modification of DNA regulates memory formation. Neuron 53:857-869. Miltenberger-Miltenyi G, Laccone F (2003) Mutations and polymorphisms in the human methyl CpG-binding protein MECP2. Hum Mutat 22:107-115. Miyamichi K, Serizawa S, Kimura HM, Sakano H (2005) Continuous and overlapping expression domains of odorant receptor genes in the olfactory epithelium determine the dorsal/ventral positioning of glomeruli in the olfactory bulb. J Neurosci 25:3586-3592. Mombaerts P (2004) Odorant receptor gene choice in olfactory sensory neurons: the one receptor-one neuron hypothesis revisited. Curr Opin Neurobiol 14:31-36. Mombaerts P (2006) Axonal wiring in the mouse olfactory system. Annu Rev Cell Dev Biol 22:713-737. Mombaerts P, Wang F, Dulac C, Chao SK, Nemes A, Mendelsohn M, Edmondson J, Axel R (1996) Visualizing an olfactory sensory map. Cell 87:675-686. Moon C, Yoo JY, Matarazzo V, Sung YK, Kim EJ, Ronnett GV (2002) Leukemia inhibitory factor inhibits neuronal terminal differentiation through STAT3 activation. Proc Natl Acad Sci U S A 99:9015-9020. Moretti P, Levenson JM, Battaglia F, Atkinson R, Teague R, Antalffy B, Armstrong D, Arancio O, Sweatt JD, Zoghbi HY (2006) Learning and memory and synaptic plasticity are impaired in a mouse model of Rett syndrome. J Neurosci 26:319-327. Morrison BE, Majdzadeh N, D'Mello SR (2007) Histone deacetylases: Focus on the nervous system. Cell Mol Life Sci. Morrison BE, Majdzadeh N, Zhang X, Lyles A, Bassel-Duby R, Olson EN, D'Mello SR (2006) Neuroprotection by histone deacetylase-related protein. Mol Cell Biol 26:3550-3564. Mullaney BC, Johnston MV, Blue ME (2004) Developmental expression of methyl-CpG binding protein 2 is dynamically regulated in the rodent brain. Neuroscience 123:939-949.  161  Mullen RJ, Buck CR, Smith AM (1992) NeuN, a neuronal specific nuclear protein in vertebrates. Development 116:201-211. Mumm JS, Shou J, Calof AL (1996) Colony- forming progenitors from mouse olfactory epithelium: evidence for feedback regulation of neuron production. Proc Natl Acad Sci U S A 93:11167-11172. Najbauer J, Leon M (1995) Olfactory experience modulated apoptosis in the developing olfactory bulb. Brain Res 674:245-251. Nakashima K, Takizawa T, Ochiai W, Yanagisawa M, Hisatsune T, Nakafuku M, Miyazono K, Kishimoto T, Kageyama R, Taga T (2001) BMP2- mediated alteration in the developmental pathway of fetal mouse brain cells from neurogenesis to astrocytogenesis. Proc Natl Acad Sci U S A 98:5868-5873. Nan X, Meehan RR, Bird A (1993) Dissection of the methyl-CpG binding domain from the chromosomal protein MeCP2. Nucleic Acids Res 21:4886-4892. Nan X, Campoy FJ, Bird A (1997) MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin. Cell 88:471-481. Nan X, Cross S, Bird A (1998a) Gene silencing by methyl-CpG-binding proteins. Novartis Found Symp 214:6-16; discussion 16-21, 46-50. Nan X, Tate P, Li E, Bird A (1996) DNA methylation specifies chromosomal localization of MeCP2. Mol Cell Biol 16:414-421. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A (1998b) Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393:386-389. Nau H, Hauck RS, Ehlers K (1991) Valproic acid- induced neural tube defects in mouse and human: aspects of chirality, alternative drug development, pharmacokinetics and possible mechanisms. Pharmacol Toxicol 69:310-321. Ng HH, Jeppesen P, Bird A (2000) Active repression of methylated genes by the chromosomal protein MBD1. Mol Cell Biol 20:1394-1406. Ng HH, Zhang Y, Hendrich B, Johnson CA, Turner BM, Erdjument-Bromage H, Tempst P, Reinberg D, Bird A (1999) MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nat Genet 23:58-61. Nguyen S, Meletis K, Fu D, Jhaveri S, Jaenisch R (2007) Ablation of de novo DNA methyltransferase Dnmt3a in the nervous system leads to neuromuscular defects and shortened lifespan. Dev Dyn 236:1663-1676.  162  O'Hare MJ, Hou ST, Morris EJ, Cregan SP, Xu Q, Slack RS, Park DS (2000) Induction and modulation of cerebellar granule neuron death by E2F-1. J Biol Chem 275:25358-25364. Okano M, Xie S, Li E (1998) Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nat Genet 19:219-220. Okano M, Bell DW, Haber DA, Li E (1999) DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99:247-257. Ono T, Uehara Y, Kurishita A, Tawa R, Sakurai H (1993) Biological significance of DNA methylation in the ageing process. Age Ageing 22:S34-43. Orrico A, Lam C, Galli L, Dotti MT, Hayek G, Tong SF, Poon PM, Zappella M, Federico A, Sorrentino V (2000) MECP2 mutation in male patients with non-specific X- linked mental retardation. FEBS Lett 481:285-288. Oswald J, Engemann S, Lane N, Mayer W, Olek A, Fundele R, Dean W, Reik W, Walter J (2000) Active demethylation of the paternal genome in the mouse zygote. Curr Biol 10:475-478. Paquette AJ, Perez SE, Anderson DJ (2000) Constitutive expression of the neuron-restrictive silencer factor (NRSF)/REST in differentiating neurons disrupts neuronal gene expression and causes axon pathfinding errors in vivo. Proc Natl Acad Sci U S A 97:12318-12323. Pasterkamp RJ, De Winter F, Holtmaat AJ, Verhaagen J (1998) Evidence for a role of the chemorepellent semaphorin III and its receptor neuropilin-1 in the regeneration of primary olfactory axons. J Neurosci 18:9962-9976. Patnaik A, Rowinsky EK, Villalona MA, Hammond LA, Britten CD, Siu LL, Goetz A, Felton SA, Burton S, Valone FH, Eckhardt SG (2002) A phase I study of pivaloyloxymethyl butyrate, a prodrug of the differentiating agent butyric acid, in patients with advanced solid malignancies. Clin Cancer Res 8:2142-2148. Pauer LR, Olivares J, Cunningham C, Williams A, Grove W, Kraker A, Olson S, Nemunaitis J (2004) Phase I study of oral CI-994 in combination with carboplatin and paclitaxel in the treatment of patients with advanced solid tumors. Cancer Invest 22:886-896. Paulsen M, Ferguson-Smith AC (2001) DNA methylation in genomic imprinting, development, and disease. J Pathol 195:97-110. Peddada S, Yasui DH, LaSalle JM (2006) Inhibitors of differentiation (ID1, ID2, ID3 and ID4) genes are neuronal targets of MeCP2 that are elevated in Rett syndrome. Hum Mol Genet 15:2003-2014.  163  Persengiev SP, Kilpatrick DL (1997) The DNA methyltransferase inhibitor 5-azacytidine specifically alters the expression of helix- loop-helix proteins Id1, Id2 and Id3 during neuronal differentiation. Neuroreport 8:2091-2095. Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, Klein PS (2001) Histone deacetylase is a direct target of valproic acid, a potent anticonvulsant, mood stabilizer, and teratogen. J Biol Chem 276:36734-36741. Pili R, Kruszewski MP, Hager BW, Lantz J, Carducci MA (2001) Combination of phenylbutyrate and 13-cis retinoic acid inhibits prostate tumor growth and angiogenesis. Cancer Res 61:1477-1485. Pradhan S, Bacolla A, Wells RD, Roberts RJ (1999) Recombinant human DNA (cytosine-5) methyltransferase. I. Expression, purification, and comparison of de novo and maintenance methylation. J Biol Chem 274:33002-33010. Prokhortchouk A, Hendrich B, Jorgensen H, Ruzov A, Wilm M, Georgiev G, Bird A, Prokhortchouk E (2001) The p120 catenin partner Kaiso is a DNA methylationdependent transcriptional repressor. Genes Dev 15:1613-1618. Puche AC, Poirier F, Hair M, Bartlett PF, Key B (1996) Role of galectin-1 in the developing mouse olfactory system. Dev Biol 179:274-287. Ramsahoye BH, Biniszkiewicz D, Lyko F, Clark V, Bird AP, Jaenisch R (2000) Non-CpG methylation is prevalent in embryonic stem cells and may be mediated by DNA methyltransferase 3a. Proc Natl Acad Sci U S A 97:5237-5242. Razin A, Riggs AD (1980) DNA methylation and gene function. Science 210:604-610. Reed RR (1992) Signaling pathways in odorant detection. Neuron 8:205-209. Ressler KJ, Sullivan SL, Buck LB (1993) A zonal organization of odorant receptor gene expression in the olfactory epithelium. Cell 73:597-609. Ressler KJ, Sullivan SL, Buck LB (1994) Information coding in the olfactory system: evidence for a stereotyped and highly organized epitope map in the olfactory bulb. Cell 79:12451255. Rett A (1966) [On a unusual brain atrophy syndrome in hyperammonemia in childhood]. Wien Med Wochenschr 116:723-726. Riechmann V, van Cruchten I, Sablitzky F (1994) The expression pattern of Id4, a novel dominant negative helix- loop- helix protein, is distinct from Id1, Id2 and Id3. Nucleic Acids Res 22:749-755.  164  Robert E, Guibaud P (1982) Maternal valproic acid and congenital neural tube defects. Lancet 2:937. Roopra A, Qazi R, Schoenike B, Daley TJ, Morrison JF (2004) Localized domains of G9amediated histone methylation are required for silencing of neuronal genes. Mol Cell 14:727-738. Roskams AJ, Cai X, Ronnett GV (1998) Expression of neuron-specific beta-III tubulin during olfactory neurogenesis in the embryonic and adult rat. Neuroscience 83:191-200. Roskams AJ, Bethel MA, Hurt KJ, Ronnett GV (1996) Sequential expression of Trks A, B, and C in the regenerating olfactory neuroepithelium. J Neurosci 16:1294-1307. Rouaux C, Jokic N, Mbebi C, Boutillier S, Loeffler JP, Boutillier AL (2003) Critical loss of CBP/p300 histone acetylase activity by caspase-6 during neurodegeneration. Embo J 22:6537-6549. Russo VEA, Martienssen RA, Riggs AD (1996) Epigenetic mechanisms of gene regulation. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Saha RN, Pahan K (2006) HATs and HDACs in neurodegeneration: a tale of disconcerted acetylation homeostasis. Cell Death Differ 13:539-550. Saito A, Yamashita T, Mariko Y, Nosaka Y, Tsuchiya K, Ando T, Suzuki T, Tsuruo T, Nakanishi O (1999) A synthetic inhibitor of histone deacetylase, MS-27-275, with marked in vivo antitumor activity against human tumors. Proc Natl Acad Sci U S A 96:4592-4597. Saito M, 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. Salminen A, Tapiola T, Korhonen P, Suuronen T (1998) Neuronal apoptosis induced by histone deacetylase inhibitors. Brain Res Mol Brain Res 61:203-206. Sammeta N, Yu TT, Bose SC, McClintock TS (2007) Mouse olfactory sensory neurons express 10,000 genes. J Comp Neurol 502:1138-1156. Sansom OJ, Berger J, Bishop SM, Hendrich B, Bird A, Clarke AR (2003) Deficiency of Mbd2 suppresses intestinal tumorigenesis. Nat Genet 34:145-147. Santos F, Hendrich B, Reik W, Dean W (2002) Dynamic reprogramming of DNA methylation in the early mouse embryo. Dev Biol 241:172-182.  165  Schanen C, Houwink EJ, Dorrani N, Lane J, Everett R, Feng A, Cantor RM, Percy A (2004) Phenotypic manifestations of MECP2 mutations in classical and atypical Rett syndrome. Am J Med Genet A 126:129-140. Schoenherr CJ, Anderson DJ (1995) The neuron-restrictive silencer factor (NRSF): a coordinate repressor of multiple neuron-specific genes. Science 267:1360-1363. Schwartz Levey M, Chikaraishi DM, Kauer JS (1991) Characterization of potential precursor populations in the mo use olfactory epithelium using immunocytochemistry and autoradiography. J Neurosci 11:3556-3564. Sealy L, Chalkley R (1978) The effect of sodium butyrate on histone modification. Cell 14:115121. Sekimata M, Homma Y (2004a) Sequence-specific transcriptiona l repression by an MBD2interacting zinc finger protein MIZF. Nucleic Acids Res 32:590-597. Sekimata M, Homma Y (2004b) Regulation of Rb gene expression by an MBD2- interacting zinc finger protein MIZF during myogenic differentiation. Biochem Biophys Res Commun 325:653-659. Sekimata M, Takahashi A, Murakami-Sekimata A, Homma Y (2001) Involvement of a novel zinc finger protein, MIZF, in transcriptional repression by interacting with a methylCpG-binding protein, MBD2. J Biol Chem 276:42632-42638. Serizawa S, Miyamichi K, Takeuchi H, Yamagishi Y, Suzuki M, Sakano H (2006) A neuronal identity code for the odorant receptor-specific and activity-dependent axon sorting. Cell 127:1057-1069. Shahbazian M, Young J, Yuva-Paylor L, Spencer C, Antalffy B, Noebels J, Armstrong D, Paylor R, Zoghbi H (2002a) Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron 35:243-254. Shahbazian MD, Zoghbi HY (2001) Molecular genetics of Rett syndrome and clinical spectrum of MECP2 mutations. Curr Opin Neurol 14:171-176. Shahbazian MD, Antalffy B, Armstrong DL, Zoghbi HY (2002b) Insight into Rett syndrome: MeCP2 levels display tissue- and cell-specific differences and correlate with neuronal maturation. Hum Mol Genet 11:115-124. Shen S, Li J, Casaccia-Bonnefil P (2005) Histone modifications affect timing of oligodendrocyte progenitor differentiation in the developing rat brain. J Cell Biol 169:577-589. Shetty RS, Bose SC, Nickell MD, McIntyre JC, Hardin DH, Harris AM, McClintock TS (2005) Transcriptional changes during neuronal death and replacement in the olfactory epithelium. Mol Cell Neurosci 30:90-107.  166  Shou J, Rim PC, Calof AL (1999) BMPs inhibit neurogenesis by a mechanism involving degradation of a transcription factor [see comments]. Nat Neurosci 2:339-345. Shou J, Murray RC, Rim PC, Calof AL (2000) Opposing effects of bone morphogenetic proteins on neuron production and survival in the olfactory receptor neuron lineage [In Process Citation]. Development 127:5403-5413. Siebzehnrubl FA, Buslei R, Eyupoglu IY, Seufert S, Hahnen E, Blumcke I (2007) Histone deacetylase inhibitors increase neuronal differentiation in adult forebrain precursor cells. Exp Brain Res 176:672-678. Silberstein SD (1996) Divalproex sodium in headache : literature review and clinical guidelines. Headache 36:547-555. Slater GE, Johnston D (1978) Sodium valproate increases potassium conductance in Aplysia neurons. Epilepsia 19:379-384. Smrt RD, Eaves-Egenes J, Barkho BZ, Santistevan NJ, Zhao C, Aimone JB, Gage FH, Zhao X (2007) Mecp2 deficiency leads to delayed maturation and altered gene expression in hippocampal neurons. Neurobiol Dis 27:77-89. Soppet D, Escandon E, Maragos J, Middlemas DS, Reid SW, Blair J, Burton LE, Stanton BR, Kaplan DR, Hunter T, et al. (1991) The neurotrophic factors brain-derived neurotrophic factor and neurotrophin-3 are ligands for the trkB tyrosine kinase receptor. Cell 65:895903. Stancheva I, Collins AL, Van den Veyver IB, Zoghbi H, Meehan RR (2003) A mutant form of MeCP2 protein associated with human Rett syndrome cannot be displaced from methylated DNA by notch in Xenopus embryos. Mol Cell 12:425-435. Stockhausen MT, Sjolund J, Manetopoulos C, Axelson H (2005) Effects of the histone deacetylase inhibitor valproic acid on Notch signalling in human neuroblastoma cells. Br J Cancer 92:751-759. Strahl BD, Allis CD (2000) The language of covalent histone modifications. Nature 403:41-45. Strotmann J, Wanner I, Helfrich T, Beck A, Breer H (1994a) Rostro-caudal patterning of receptor-expressing olfactory neurones in the rat nasal cavity. Cell Tissue Res 278:11-20. Strotmann J, Levai O, Fleischer J, Schwarzenbacher K, Breer H (2004) Olfactory receptor proteins in axonal processes of chemosensory neurons. J Neurosci 24:7754-7761. Strotmann J, Wanner I, Helfrich T, Beck A, Meinken C, Kubick S, Breer H (1994b) Olfactory neurones expressing distinct odorant receptor subtypes are spatially segregated in the nasal neuroepithelium. Cell Tissue Res 276:429-438.  167  Swank MW, Sweatt JD (2001) Increased histone acetyltransferase and lysine acetyltransferase activity and biphasic activation of the ERK/RSK cascade in insular cortex during novel taste learning. J Neurosci 21:3383-3391. Takizawa T, Nakashima K, Namihira M, Ochiai W, Uemura A, Yanagisawa M, Fujita N, Nakao M, Taga T (2001) DNA methylation is a critical cell- intrinsic determinant of astrocyte differentiation in the fetal brain. Dev Cell 1:749-758. Tawa R, Ono T, Kurishita A, Okada S, Hirose S (1990) Changes of DNA methylation level during pre- and postnatal periods in mice. Differentiation 45:44-48. Teter B, Rozovsky I, Krohn K, Anderson C, Osterburg H, Finch C (1996) Methylation of the glial fibrillary acidic protein gene shows novel biphasic changes during brain development. Glia 17:195-205. Traynor J, Agarwal P, Lazzeroni L, Francke U (2002) Gene expression patterns vary in clonal cell cultures from Rett syndrome females with eight different MECP2 mutations. BMC Med Genet 3:12. Treloar H, Tomasiewicz H, Magnuson T, Key B (1997) The central pathway of primary olfactory axons is abnormal in mice lacking the N-CAM-180 isoform. J Neurobiol 32:643-658. Treloar HB, Purcell AL, Greer CA (1999) Glomerular formation in the developing rat olfactory bulb. Journal of Comparative Neurology 413:289-304. Treloar HB, Feinstein P, Mombaerts P, Greer CA (2002) Specificity of glomerular targeting by olfactory sensory axons. J Neurosci 22:2469-2477. Tudor M, Akbarian S, Chen RZ, Jaenisch R (2002) Transcriptional profiling of a mouse model for Rett syndrome reveals subtle transcriptional changes in the brain. Proc Natl Acad Sci U S A 99:15536-15541. Tunnicliff G (1999) Actions of sodium valproate on the central nervous system. J Physiol Pharmacol 50:347-365. Vakoc CR, Mandat SA, Olenchock BA, Blobel GA (2005) Histone H3 lysine 9 methylation and HP1gamma are associated with transcription elongation through mammalian chromatin. Mol Cell 19:381-391. Van den Berg RJ, Kok P, Voskuyl RA (1993) Valproate and sodium currents in cultured hippocampal neurons. Exp Brain Res 93:279-287. van der Laan JW, de Boer T, Bruinvels J (1979) Di-n-propylacetate and GABA degradation. Preferential inhibition of succinic semialdehyde dehydrogenase and indirect inhibition of GABA-transaminase. J Neurochem 32:1769-1780.  168  Van Esch H, Bauters M, Ignatius J, Jansen M, Raynaud M, Hollanders K, Lugtenberg D, Bienvenu T, Jensen LR, Gecz J, Moraine C, Marynen P, Fryns JP, Froyen G (2005) Duplication of the MECP2 region is a frequent cause of severe mental retardation and progressive neurological symptoms in males. Am J Hum Genet 77:442-453. van Gent DC, Hoeijmakers JH, Kanaar R (2001) Chromosomal stability and the DNA doublestranded break connection. Nat Rev Genet 2:196-206. Vassar R, Ngai J, Axel R (1993) Spatial segregation of odorant receptor expression in the mammalian olfactory epithelium. Cell 74:309-318. Vassar R, Chao SK, Sitcheran R, Nunez JM, Vosshall LB, Axel R (1994) Topographic organization of sensory projections to the olfactory bulb. Cell 79:981-991. Vega CJ, Peterson DA (2005) Stem cell proliferative history in tissue revealed by temporal halogenated thymidine analog discrimination. Nat Methods 2:167-169. Veldic M, Guidotti A, Maloku E, Davis JM, Costa E (2005) In psychosis, cortical interneurons overexpress DNA- methyltransferase 1. Proc Natl Acad Sci U S A 102:2152-2157. Veldic M, Caruncho HJ, Liu WS, Davis J, Satta R, Grayson DR, Guidotti A, Costa E (2004) DNA-methyltransferase 1 mRNA is selectively overexpressed in telencephalic GABAergic interneurons of schizophrenia brains. Proc Natl Acad Sci U S A 101:348353. Verhaagen J, Oestreicher AB, Gispen WH, Margolis FL (1989) The expression of the growth associated protein B50/GAP43 in the olfactory system of neonatal and adult rats. J Neurosci 9:683-691. Vignali M, Hassan AH, Neely KE, Workman JL (2000) ATP-dependent chromatin-remodeling complexes. Mol Cell Biol 20:1899-1910. Villar-Garea A, Esteller M (2004) Histone deacetylase inhibitors: understanding a new wave of anticancer agents. Int J Cancer 112:171-178. Waddington C (1942) The epigenotype. Endeavour 1:18-20. Wade PA (2001) Transcriptional control at regulatory checkpoints by histone deacetylases: molecular connections between cancer and chromatin. Hum Mol Genet 10:693-698. Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP (1999) Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat Genet 23:6266. Walz A, Rodriguez I, Mombaerts P (2002) Aberrant sensory innervation of the olfactory bulb in neuropilin-2 mutant mice. J Neurosci 22:4025-4035.  169  Wang F, Nemes A, Mendelsohn M, Axel R (1998) Odorant receptors govern the formation of a precise topographic map. Cell 93:47-60. Wang H, Chan SA, Ogier M, Hellard D, Wang Q, Smith C, Katz DM (2006) Dysregulation of brain-derived neurotrophic factor expression and neurosecretory function in Mecp2 null mice. J Neurosci 26:10911-10915. Watson P, Black G, Ramsden S, Barrow M, Super M, Kerr B, Clayton-Smith J (2001) Angelman syndrome phenotype associated with mutations in MECP2, a gene encoding a methyl CpG binding protein. J Med Genet 38:224-228. Weiler E, Farbman AI (1997) Proliferation in the rat olfactory epithelium: age-dependent changes. J Neurosci 17:3610-3622. Whittle SR, Turner AJ (1978) Effects of the anticonvulsant sodium valproate on gammaaminobutyrate and aldehyde metabolism in ox brain. J Neurochem 31:1453-1459. Wikinski SI, Acosta GB, Rubio MC (1996) Valproic acid differs in its in vitro effect on glutamic acid decarboxylase activity in neonatal and adult rat brain. Gen Pharmacol 27:635-638. Wilson VL, Smith RA, Ma S, Cutler RG (1987) Genomic 5- methyldeoxycytidine decreases with age. J Biol Chem 262:9948-9951. Winner B, Cooper-Kuhn CM, Aigner R, Winkler J, Kuhn HG (2002) Long-term survival and cell death of newly generated neurons in the adult rat olfactory bulb. Eur J Neurosci 16:1681-1689. Wood MA, Kaplan MP, Park A, Blanchard EJ, Oliveira AM, Lombardi TL, Abel T (2005) Transgenic mice expressing a truncated form of CREB-binding protein (CBP) exhibit deficits in hippocampal synaptic plasticity and memory storage. Learn Mem 12:111-119. Wu HH, Ivkovic S, Murray RC, Jaramillo S, Lyons KM, Johnson JE, Calof AL (2003) Autoregulation of neurogenesis by GDF11. Neuron 37:197-207. Yamaguchi M, Tonou-Fujimori N, Komori A, Maeda R, Nojima Y, Li H, Okamoto H, Masai I (2005) Histone deacetylase 1 regulates retinal neurogenesis in zebrafish by suppressing Wnt and Notch signaling pathways. Development 132:3027-3043. Yamamoto R, Yanagita T, Kobayashi H, Yokoo H, Wada A (1997) Up-regulation of sodium channel sub unit mRNAs and their cell surface expression by antiepileptic valproic acid: activation of calcium channel and catecholamine secretion in adrenal chromaffin cells. J Neurochem 68:1655-1662. Yoder JA, Soman NS, Verdine GL, Bestor TH (1997) DNA (cytosine-5)-methyltransferases in mouse cells and tissues. Studies with a mechanism-based probe. J Mol Biol 270:385-395.  170  Yoon HG, Chan DW, Reynolds AB, Qin J, Wong J (2003) N-CoR mediates DNA methylationdependent repression through a methyl CpG binding protein Kaiso. Mol Cell 12:723-734. Yoshida M, Kijima M, Akita M, Beppu T (1990) Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A. J Biol Chem 265:1717417179. Young JI, Zoghbi HY (2004) X-chromosome inactivatio n patterns are unbalanced and affect the phenotypic outcome in a mouse model of rett syndrome. Am J Hum Genet 74:511-520. Young JI, Hong EP, Castle JC, Crespo-Barreto J, Bowman AB, Rose MF, Kang D, Richman R, Johnson JM, Berget S, Zoghbi HY (2005) Regulation of RNA splicing by the methylation-dependent transcriptional repressor methyl-CpG binding protein 2. Proc Natl Acad Sci U S A 102:17551-17558. Young JM, Friedman C, Williams EM, Ross JA, Tonnes-Priddy L, Trask BJ (2002) Different evolutionary processes shaped the mouse and human olfactory receptor gene families. Hum Mol Genet 11:535-546. Yu CR, Power J, Barnea G, O'Donnell S, Brown HE, Osborne J, Axel R, Gogos JA (2004) Spontaneous neural activity is required for the establishment and maintenance of the olfactory sensory map. Neuron 42:553-566. Yuan PX, Huang LD, Jiang YM, Gutkind JS, Manji HK, Chen G (2001) The mood stabilizer valproic acid activates mitogen-activated protein kinases and promotes neurite growth. J Biol Chem 276:31674-31683. Zeise ML, Kasparow S, Zieglgansberger W (1991) Valproate suppresses N-methyl-D-aspartateevoked, transient depolarizations in the rat neocortex in vitro. Brain Res 544:345-348. Zhang HS, Dean DC (2001) Rb- mediated chromatin structure regulation and transcriptional repression. Oncogene 20:3134-3138. Zhang X, Firestein S (2002) The olfactory receptor gene superfamily of the mouse. Nat Neurosci 5:124-133. Zhang X, Rogers M, Tian H, Zou DJ, Liu J, Ma M, Shepherd GM, Firestein SJ (2004) Highthroughput microarray detection of olfactory receptor gene expression in the mouse. Proc Natl Acad Sci U S A 101:14168-14173. Zhang Y, Ng HH, Erdjument-Bromage H, Tempst P, Bird A, 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 X, Ueba T, Christie BR, Barkho B, McConnell MJ, Nakashima K, Lein ES, Eadie BD, Willhoite AR, Muotri AR, Summers RG, Chun J, Lee KF, Gage FH (2003) Mice lacking  171  methyl-CpG binding protein 1 have deficits in adult neurogenesis and hippocampal function. Proc Natl Acad Sci U S A 100:6777-6782. Zhou DC, Kim SH, Ding W, Schultz C, Warrell RP, Jr., Gallagher RE (2002) Frequent mutations in the ligand-binding domain of PML-RARalpha after multiple relapses of acute promyelocytic leukemia: analysis for functional relationship to response to all- trans retinoic acid and histone deacetylase inhibitors in vitro and in vivo. Blood 99:1356-1363. Zhou Z, Hong EJ, Cohen S, Zhao WN, Ho HY, Schmidt L, Chen WG, Lin Y, Savner E, Griffith EC, Hu L, Steen JA, Weitz CJ, Greenberg ME (2006) Brain-specific phosphorylation of MeCP2 regulates activity-dependent Bdnf transcription, dendritic growth, and spine maturation. Neuron 52:255-269. Zou DJ, Feinstein P, Rivers AL, Mathews GA, Kim A, Greer CA, Mombaerts P, Firestein S (2004) Postnatal refinement of peripheral olfactory projections. Science 304:1976-1979.  172  Appendix  173  174  175  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066189/manifest

Comment

Related Items