Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Developmental consequences of imprinted transcription at the Mest locus MacIsaac, Julia Lynn 2013

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

Item Metadata

Download

Media
24-ubc_2013_fall_macisaac_julia.pdf [ 5.82MB ]
Metadata
JSON: 24-1.0073875.json
JSON-LD: 24-1.0073875-ld.json
RDF/XML (Pretty): 24-1.0073875-rdf.xml
RDF/JSON: 24-1.0073875-rdf.json
Turtle: 24-1.0073875-turtle.txt
N-Triples: 24-1.0073875-rdf-ntriples.txt
Original Record: 24-1.0073875-source.json
Full Text
24-1.0073875-fulltext.txt
Citation
24-1.0073875.ris

Full Text

DEVELOPMENTAL CONSEQUENCES OF IMPRINTED TRANSCRIPTION AT THE MEST LOCUS  by Julia Lynn MacIsaac M.Sc., University of British Columbia, 2005 B.Sc., Colorado State University, 2002  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in  THE FACULTY OF GRADUATE STUDIES (Medical Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) May 2013  © Julia Lynn MacIsaac, 2013  Abstract The Mest locus is regulated by genomic imprinting in mammals and only the paternally inherited allele is expressed. A targeted mutation at this locus revealed that it plays an important role in the regulation of embryonic growth and adult behavior. The Mest locus is located in a conserved imprinted domain on mouse chromosome 6 where it is thought to play a key role in the regulation of neighboring maternally expressed genes Copg2 and Klf14 since it contains the only potential imprinting center (IC) identified thus far in this domain, a differentially methylated region (DMR) methylated in oogenesis. Here we describe new larger isoforms of the Mest mRNA, referred to as MestXL, that are generated via alternative polyadenylation and transcribed more than 10kb into the adjacent antisense gene Copg2 exclusively in the developing central nervous system. The MestXL isoforms appear to regulate the allelic usage at Copg2, but not at Klf14, in embryonic neural tissues, as Copg2 is preferentially maternally expressed only in these tissues presumably due to transcriptional interference from MestXL on the paternal chromosome. Our results therefore establish the Mest DMR as an IC and propose a new mechanism to regulate allelic usage and imprinting at sense-antisense gene pairs in mammalian genomes, via tissue-specific alternative polyadenylation and transcriptional interference. Imprinted transcription at the Mest locus also produces a microRNA, miR335, that acts to down-regulate target genes via binding to their 3’UTRs and ultimately repressing their translation. Here we show that production of miR-335 is imprinted and that its levels are reduced from the mutant MestKO allele. Additionally, we identify several candidate target genes of miR-335 by RNA-seq analysis on primary mouse fibroblasts that under-express miR-335. Our investigation of MestXL and miR-335, two unique alternative functions of the Mest locus, demonstrates that the Mest locus is involved in two types of RNA-mediated regulation and ultimately contributes to the understanding of genomic imprinting and microRNAs in mammalian biology.  ii  Preface For Chapter 3 and Appendix A, Dr. Louis Lefebvre generated Figures 3.1, 3.2, 3.3, 3.4, and A.1, and Tables A.1 and A.2. Additionally, he performed the first initial northern blot experiment for Figure 3.3. Aaron Bogutz analyzed RNA-seq data and generated Figure A.4 and also generated Figure 3.10 from my immunofluorescence experiments. Sorana Morrissy, who used to be in Marco Marra’s laboratory, analyzed the SAGE tags in the Mest-Copg2 region. For Chapter 5 and Appendix C, Misha Bilenky from the Genome Sciences Centre in Vancouver, BC, analyzed the RNA-seq results and performed the normalization using custom scripts. He provided tables and scatter plots that were used to generate Figure 5.7 and Tables C.1-C.6. Aaron Bogutz generated the venn diagrams used to generate Figure 5.8 All animal experiments were performed under certificate A07-0160 from the UBC Animal Care Committee and complied with the Canadian Council on Animal Care guidelines on the ethical care and use of experimental animals.  iii  Table of contents Abstract ......................................................................................................................... ii Preface .......................................................................................................................... iii Table of contents ......................................................................................................... iv List of tables ................................................................................................................ vii List of figures ............................................................................................................. viii List of abbreviations ..................................................................................................... x Acknowledgements..................................................................................................... xii  Chapter 1: Introduction ................................................................................................. 1 1.1 Overview of genomic imprinting................................................................................. 1 1.1.1 Historical overview and evolutionary models ................................................ 1 1.1.2 Overview of the imprinting cycle ................................................................... 2 1.1.3 Imprinting centers and mechanisms of imprinted regulation via gametic DMRs .......................................................................................................... 3 1.1.4 Tissue-specific imprinting and allelic ratios ................................................... 6 1.1.5 Summary...................................................................................................... 7 1.2 Additional roles of non-coding RNAs in imprinted domains........................................ 7 1.3 Sub-proximal MMU6 ................................................................................................ 12 1.3.1 Mest ........................................................................................................... 12 1.3.2 Copg2......................................................................................................... 14 1.3.3 Klf14 ........................................................................................................... 16 1.3.4 Copg2AS and Mit1 RNAs ........................................................................... 17 1.3.5 miR-335 ...................................................................................................... 18 1.3.6 Human homology and disease ................................................................... 21 1.4 MestKO mutant mouse line ....................................................................................... 22 1.5 Thesis theme and objectives ................................................................................... 22 Chapter 2: Materials and methods ............................................................................. 25 2.1 Mouse strains .......................................................................................................... 25 2.2 Genotyping .............................................................................................................. 25 2.3 Embryo dissection ................................................................................................... 26 2.4 SAGE analysis ........................................................................................................ 26 2.5 RNA purification ...................................................................................................... 27 iv  2.6 RT-PCR/qRT-PCR analysis..................................................................................... 27 2.7 DIG-labeled RNA probe preparation ........................................................................ 28 2.8 Northern blot analysis .............................................................................................. 28 2.9 In Situ hybridization ................................................................................................. 29 2.10 Allele specific analysis ........................................................................................... 30 2.11 5’ rapid amplification of cDNA end ......................................................................... 30 2.12 Immunofluorescence and β-galactosidase staining ............................................... 31 2.13 Analysis of miR-335 expression ............................................................................ 32 2.14 pCmir335E and pC∆mir335E plasmid construction ............................................... 32 2.15 Mouse primary embryonic fibroblast isolation, genotyping, and cell line conditions ............................................................................................................. 33 2.16 Electroporation and flow cytometry ........................................................................ 34 2.17 Whole transcriptome sequencing (RNA-seq) libraries............................................ 34 Chapter 3: The Mest locus produces larger isoforms, MestXL, through alternative polyadenylation ........................................................................................................... 39 3.1 Introduction ............................................................................................................. 39 3.2 Results .................................................................................................................... 41 3.2.1 Longer variants of Mest mRNA: the MestXL isoforms ............................... 41 3.2.2 Loss of MestXL transcripts from the mutant MestKO allele ......................... 47 3.2.3 MestXL transcripts are produced in the developing nervous system ........ 54 3.3 Discussion ............................................................................................................... 58 Chapter 4: Allelic usage at Copg2, but not at Klf14, is regulated by MestXL in the developing central nervous system ........................................................................... 63 4.1 Introduction ............................................................................................................. 63 4.2 Results .................................................................................................................... 65 4.2.1 Allelic usage at Copg2 is regulated by MestXL in the developing CNS ..... 65 4.2.2 Loss of MestXL in the developing mouse CNS leads to an overall increase in Copg2 expression levels ...................................................................... 68 4.2.3 MestXL does not regulate Klf14 imprinting ................................................ 71 4.2.4 MestXL transcripts are not retained in the nucleus .................................... 75 4.3 Discussion ............................................................................................................... 77 Chapter 5: Candidate developmental targets of miR-335 ......................................... 82 5.1 Introduction ............................................................................................................. 82 v  5.2 Results .................................................................................................................... 83 5.2.1 MiR-335 expression is decreased in Mest+/KO embryos and shares imprinted expression with Mest................................................................................. 83 5.2.2 Using primary mouse fibroblasts as a model system to study over- and under-expression of miR-335 .................................................................... 85 5.2.3 RNA-seq data from Mest+/KO PEFs reveals candidate target genes of miR335............................................................................................................ 89 5.2.4 Validating target genes by qRT-PCR .......................................................... 97 5.3 Discussion ............................................................................................................... 99 Chapter 6: Discussion .............................................................................................. 106 6.1 Summary of results and conclusions ..................................................................... 106 6.2 General discussion ................................................................................................ 107 6.2.1 Mest imprinting contrasts with high vertebrate conservation of gene ....... 107 6.2.2 MestXL: A coding or non-coding RNA ..................................................... 109 6.2.3 The Mest DMR: an IC that acts as both a promoter and an insulator? .... 110 6.2.4 Important roles of miR-335...................................................................... 113 6.2.5 Significance of the MEST locus............................................................... 114 6.3 Future directions.................................................................................................... 115 References ................................................................................................................. 119 Appendix A: Supplementary material for chapter 3................................................ 135 Appendix B: Supplementary material for chapter 4................................................ 144 Appendix C: Supplementary material for chapter 5................................................ 145  vi  List of tables Table 1.1 Table 1.2 Table 2.1 Table 2.2  Table 5.1 Table A.1 Table A.2 Table C.1 Table C.2 Table C.3 Table C.4 Table C.5 Table C.6  A survey of large and small non-coding RNAs located in mouse imprinted domains ................................................................................................. 10 Identified target genes of human, rat, and mouse miR-335 reported in the literature ................................................................................................. 19 Primer sequences for genotyping, RT-PCR, qRT-PCR, probe construction, and 5'RACE ...................................................................... 35 Strand-specific RNA probes used for Northern and ISH analyses and the restriction enzymes and RNA polymerases used to generate the probes from their respective plasmids ................................................................ 38 Candidate target genes of miR-335 chosen for further investigation by qRT-PCR ............................................................................................... 98 (+) strand ESTs from 5’ cluster ............................................................. 140 (+) strand ESTs from 3’ cluster ............................................................. 141 Significantly down-regulated genes in the Mest+/KO PEFs compared to the WT PEFs.............................................................................................. 145 Significantly up-regulated genes in the Mest+/KO PEFs compared to the WT PEFs.............................................................................................. 150 Significantly down-regulated genes in the WT+miR PEFs compared to the WT PEFs.............................................................................................. 156 Significantly up-regulated genes in the WT+miR PEFs compared to the WT PEFs.............................................................................................. 159 Significantly down-regulated genes in the Mest+/KO PEFs compared to the WT+miR PEFs ..................................................................................... 162 Significantly up-regulated genes in the Mest+/KO PEFs compared to the WT+miR PEFs ..................................................................................... 166  vii  List of figures Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 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 3.10 Figure 3.11 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  Key imprinting mechanisms that regulate the monoallelic expression of genes in cis .............................................................................................. 4 Schematic representation of the genomic organization and imprinting in the MMU6 sub-proximal domain............................................................. 13 Diagrams of the Mest and MestKO loci .................................................... 15 Products of imprinted transcription from the Mest locus .......................... 24 EST evidence for longer isoforms of Mest .............................................. 40 Main RNA isoforms at the Mest locus ..................................................... 42 Northern blot analysis of Mest expression .............................................. 43 Analysis of SAGE tags for the Mest and MestXL transcripts ................... 45 qRT-PCR analysis of Mest, the MestKO allele, and MestXL (Mit1) RNA levels in various embryonic tissues at E14.5 .......................................... 48 Loss of MestXL expression from the mutant MestKO allele ..................... 50 Strand-specific northern blot analyses of Mest, MestKO, and MestXL expression in adult brain tissue .............................................................. 51 5’ structure of the MestKO mRNA analyzed by 5’ RACE .......................... 53 MestXL is expressed in the developing central nervous system of WT but not Mest+/KO embryos at E14.5 ............................................................... 55 MEST and NCAM1 expression in the developing spinal cord ................. 57 PCR amplification of cDNA fragments from MestXL in wild-type CNS .... 59 Allelic usage at Copg2 is regulated by MestXL in the developing CNS... 67 In situ hybridization analysis of Copg2 mRNA expression at E14.5 ....... 69 Copg2 expression levels in various E14.5 tissues .................................. 70 Klf14 is not regulated by MestXL (part 1) ................................................ 72 In situ hybridization analysis of Klf14 mRNA expression at E14.5 ......... 73 Klf14 is not regulated by MestXL (part 2) ................................................ 74 MestXL transcripts do not localize to the nucleus ................................... 76 Model for regulation of allelic bias at Copg2 by MestXL .......................... 78 Expression analyses of miR-335, Mest, and MestKO RNA levels in E14.5 embryos ................................................................................................. 84 Expression analysis of miR-335 in E14.5 embryos from Mest+/KO intercrosses............................................................................................ 86 Structure of the pCmiR335E and pC∆miR335E plasmids ....................... 88 Flowchart for miR-335 under- and over-expression experiments ............ 90 Over- and under-expression analysis of miR-335 in GFP+-sorted PEFs 91 Transcription from pC∆miR335E and pCmiR335E maps to the Mest/miR335 locus................................................................................................ 93 Distributions of expression levels for annotated transcripts in the Mest+/KO, WT, and WT+miR PEFs ......................................................................... 94 Comparison of significantly differentially expressed genes in WT, WT+miR, and Mest+/KO PEFs.................................................................. 95 Expression levels of candidate target genes of miR-335 ...................... 100 viii  Figure 6.1 Figure A.1  Figure A.2 Figure A.3 Figure A.4 Figure B.1 Figure C.1 Figure C.2 Figure C.3 Figure C.4  A proposed model for the regulation of Klf14 via insulator activity at the Mest DMR ............................................................................................ 112 Tissue distribution of expressed sequence tags (ESTs) from the (+) strand mapping to the Mest locus and to the EST clusters 5’ and 3’ of the gap within intron 20 of Copg2 ............................................................... 135 RNA gels for Northern blot analyses .................................................... 136 In situ hybridization analysis of Copg2 and primary Copg2 mRNA expression at E14.5 ............................................................................. 137 Analysis of RNA-seq data in the Mest-Copg2 region ............................ 138 Effect of loss of maternal methylation of Mest, MestXL, and Copg2 Expression ........................................................................................... 144 DAVID analysis for significantly up-regulated genes in the Mest+/KO PEFs versus the WT PEFs ............................................................................ 169 DAVID analysis for significantly down-regulated genes in the WT+miR PEFs versus the WT PEFs ................................................................... 170 DAVID analysis for up-regulated genes in the WT+miR PEFs versus the WT PEFs.............................................................................................. 171 DAVID analysis for significantly down-regulated genes in the Mest+/KO PEFs versus the WT PEFs ................................................................... 172  ix  List of abbreviations 3’UTR 5mC A-to-I AS Βgeo bp Chr Chr6 CNS CpG DAVID DMR DNA DNMT ds E EMT ES FACS GFP GFP+ GSP H3K4me2 H3K4me3 H3K9me3 H3K27me3 IC ICE ICR IF IRES ISH IUGR Kb KO lncNRA miRNA Mb MHB MMU6 ncRNA ORF PBS PCR PEF pA piRNA poly(A)  3’untranslated region 5-methyl-cytosine adenosine to inosine antisense lacZ-neo fusion base pairs chromosome mouse chromosome 6 central nervous system cytosine-phosphate-guanidine dinucleotide Database for Annotation, Visualization, and Integrated Discovery differentially methylated region deoxyribonucleic acid DNA methyltransferase double-stranded embryonic day epithelial to mesenchymal transition embryonic stem (cell) fluorescence-activated cell sorting green fluorescent protein green fluorescent protein-positive phenotype gene specific primer dimethylation of histone H3 at lysine 4 trimethylation of histone H3 at lysine 4 trimethylation of histone H3 at lysine 9 trimethylation of histone H3 at lysine 27 imprinting center imprinting control element imprinting control region immunofluorescence Internal ribosome entry site in situ hybridization intrauterine growth restriction kilobase pairs knockout large non-coding RNA microRNA megabase pairs E14.5 mid/hind brain mouse chromosome 6 non-coding RNA open reading frame phosphate-buffered saline polymerase chain reaction primary embryonic fibroblast polyadenylation Piwi-interacting RNA polyadenylated x  qRT-PCR quantitative reverse transcription polymerase chain reaction RACE rapid amplification of cDNA end RISC RNA-induced silencing complex RNAi RNA interference RT-PCR reverse transcriptase-polymerase chain reaction S sense SC E14.5 spinal cord snoRNA C/D-box small nucleolar RNAs SNP single nucleotide polymorphism SRS Silver-Russell syndrome TI transcriptional interference uaRNA 3’UTR associated RNA UPD uniparental disomy WT wild type  xi  Acknowledgements First and foremost, I would like to thank my supervisor, Dr. Louis Lefebvre, for his time, support, guidance, expertise, advice, humor, and for letting me pursue my degree in his lab. Overall I had a great experience and would not have wanted to do it anywhere else. Thank you Louis! I would like to thank former and current lab members, especially Rosemary, Meaghan, and Aaron. Thank you for your help, conversations, friendship, and camaraderie, as it made every day enjoyable. Thank you Aaron for your invaluable help and always being there to answer a question. Thank you Rose for helping me learn ISH. I would like to thank my thesis advisory committee, Dr. Matthew Lorincz, Dr. Pamela Hoodless, and Dr. Michael Underhill, for their advice, comments, and taking the time to attend my meetings and for reviewing my thesis. I really appreciate it. I would like to thank my family, my in-law family, and friends for their love and support. You are the best! I feel very lucky. I would like to give a special thanks to Joanne, my mother-in-law, for hanging out with Alexandra so that I could finally finish this thesis. Thank you especially to my husband Mark for always supporting me and being my rock. I love you so much and could not have done this without you. And thank you Alexandra for making our lives more full and bringing so much joy to every day.  xii  Chapter 1: Introduction 1.1 Overview of genomic imprinting  1.1.1 Historical overview and evolutionary models  Genomic imprinting initially became apparent when early studies demonstrated that the mammalian maternal and paternal genomes were not functionally equivalent. Reconstitution of fertilized eggs to generate androgenetic (two paternal genomes) and gynogenetic (two maternal genomes) embryos revealed that they were indeed biologically different and failed to develop to term (1-3). At the time, it was postulated that the functional differences between the parental genomes were attributable to germline-specific heritable epigenetic modifications that drive the phenomenon known as genomic imprinting (1; 4). At the transcriptional level, the result is an epigenetic regulatory mechanism that directs expression of a single copy of a gene based on its parental origin. To date approximately 80-100 imprinted genes have been identified and confirmed in the mammalian genome (5). Depending on the approach and tissue analyzed, together with bioinformatic predictions, the number of estimated imprinted genes varies and has reached upwards to 2,000 genes (6; 7). Even though imprinted genes represent a small fraction of the transcribed genome, many have essential functions in embryonic development and are highly expressed in embryonic and extraembryonic tissues. For this reason, a popular explanation for the origin of imprinting, referred to as the parental conflict hypothesis, is that imprinting evolved in mammals when they acquired the ability to bear live young and therefore the maternal and paternal genes fell into conflict with regard to the transfer of nutrients from the mother to her offspring (8-10). In brief terms, the concept is that maternally expressed genes suppress fetal growth allowing nutrient supply to be available for future pregnancies, whereas paternally expressed genes enhance fetal growth to ensure survival of their genetic offspring. In addition to the parental conflict hypothesis, alternative theories for how imprinting evolved exist, which include the ovarian time bomb hypothesis and the host defense model. Briefly, the ovarian time bomb hypothesis suggests paternally expressed genes are necessary for development of extraembryonic tissues to protect female mammals from activation of eggs that could lead to subsequent cancerous invasion of 1  extraembryonic-like tissue (11). The host defense model proposes that genomic imprinting evolved to protect the organism from invasion of foreign DNA by silencing it through epigenetic mechanisms (12). In support of this model there are at least 9 imprinted retrotransposed genes identified thus far in the mammalian genome (13). Some genes located in clusters likely became imprinted by the bystander effect, simply due to their close proximity to an imprinted gene or center (11). Though these theories differ from each other a common theme emerges which indicates imprinting evolved as a mechanism to control gene dosage.  1.1.2 Overview of the imprinting cycle  Based on the androgenic and gynogenic experiments mentioned above, it was predicted that germ-line-specific heritable epigenetic modifications were the force behind instituting parental-specific monoallelic expression and that this imprinting was established and maintained via an imprinting cycle that originates in the germ-line (1; 4). Many aspects of this model have now been confirmed. Today it is well-established that the key element that controls genomic imprinting is DNA methylation that is acquired in oocytes  and  sperm  (14-16).  This  DNA  methylation  is  laid  down  by  the  DNMT3A/DNMT3L de novo DNA methyltransferase complex and passed to the zygote upon fertilization establishing differentially methylated regions (DMRs) (17-20). The mechanism whereby these specific DMRs are only methylated in the male or female germ-line in the presence of all the sequences in the mammalian genome is still not fully understood. However, components thought to be involved in this process include optimal CpG spacing and histone modifications (20-22). Additionally, to explain the acquisition of maternal imprints, Chotalia et al. further suggested that transcription across the regions that acquire this methylation in oocytes is required to create or maintain open chromatin domains to allow the methylation complex access to its targets (23). These maternally and paternally methylated DMRs, referred to as gametic DMRs, are thought to regulate all imprinted expression by various mechanisms discussed below. Continuing the cycle, the imprints are further maintained and lastly erased before they are reestablished again for transmission to the next generation (reviewed in ref. 24). For maintenance, DNA-binding protein PGC7, zinc finger protein ZFP57, and DNA methyltransferase DNMT1 are involved to protect the methylation at gametic DMRs from the genome-wide demethylation that occurs in the paternally- and maternally-inherited 2  genomes in early embryos and to maintain the allele-specific methylation in somatic and extraembryonic tissues, respectively (25-27). The erasure phase subsequently occurs in the germ-line of this said generation, involving epigenetic reprogramming and erasure of the DNA methylation imprints at some point during migration of their primordial germ cells into the genital ridge (28-30). The imprints are then reestablished appropriately during gametogenesis, according to the sex of the individual, to continue the cycle for the next generation (14-16; 24).  1.1.3 Imprinting centers and mechanisms of imprinted regulation via gametic DMRs  An important characteristic of most imprinted genes is their frequent clustering in imprinted chromosomal domains (A genomic imprinting map can be found at http://www.mgu.har.mrc.ac.uk/research/imprinting/index.html). These domains contain between 3 and 11 imprinted genes including protein-coding genes and at least one noncoding (large and/or small) gene (31; 32). The clusters are regulated by cis–acting elements called imprinting centers (IC), which are also referred to as imprinting control elements (ICE) or imprinting control regions (ICR) in the literature. Typically they are gametic DMRs that are responsible for controlling the imprinting of genes via mechanisms of imprinted regulation (see below). In addition to differential methylation, ICs also exhibit allele-specific histone modifications associated with both active and repressed chromatin states (22; 33-36). Whereas the active allele is marked with trimethylation of histone H3 at lysine 4 (H3K4me3), the repressed allele is marked with trimethylation of histone H3 at lysine 9 (H3K9me3) and/or lysine 27 (H3K27me3). To date, three key conserved imprinting mechanisms have been described that regulate the monoallelic expression of genes in cis within the domains. Firstly, the most recognized mechanism involves direct promoter methylation, via 5-methyl-cytosine (5mC) at CpG nucleotides, which controls the imprinting of the associated gene (Figure 1.1A). For imprinted genes, allele-specific methylation occurs resulting in a DMR. The allele associated with the hypermethylated promoter remains silent, whereas the allele associated with the hypomethlyated promoter is transcribed and expressed. These genes are considered primary imprinted genes because epigenetic marks at the locus facilitate the imprinting of the gene. Many paternally expressed genes fall under this category and there are several (at least 16) imprinted loci associated with maternalspecific methylation at associated promoters (24). For some of these genes, the 3  Figure 1.1 Key imprinting mechanisms that regulate the monoallelic expression of genes in cis. (A) Allele-specific direct promoter methylation, via 5-methyl-cytosine at CpG nucleotides. The allele associated with the hypermethylated promoter remains silent, whereas the allele associated with the hypomethlyated promoter is transcribed and expressed. The numbered boxes indicate exons and the line above exon 1 indicates the DMR, which in this example overlaps 5’ flanking sequences, exon 1, and part of intron 1, such as the one found at the Mest locus. (B) 5mC-sensitive insulator function. Using the H19/Igf2 domain as an example, IC1 is an intergenic DMR flanked by Igf2 and H19, which either binds the CTCF protein or becomes methylated, on the maternal or paternal allele, respectively. This figure was adapted from (40). (C) LncRNA-mediated epigenetic silencing. Using Kcnq1ot1 as an example, transcription of the lncRNA is associated with the bi-directional silencing of genes in cis on the paternal chromosome so that consequently only the maternal alleles are expressed. Kcnq1ot1 is a polyadenylated lncRNA, as indicated by the “AAAA” sequence. This figure was adapted from (32). For A, B, and C, the active genes are represented as white boxes, whereas the silent genes are represented by grey boxes. The white and black lollipops represent unmethylated and methylated CpG-rich sequences, respectively. These diagrams are not drawn to scale. M=maternal allele, P=paternal allele. 4  maternally methylated DMR also acts as an IC for the domain which comprises a large number of the identified ICs. In contrast, paternally methylated DMRs/ICs, of which there are only 4 described to date, are usually located in intergenic regions that regulate multiple imprinted genes some distance away from the DMR (24). A second recognized imprinting mechanism involves 5mC-sensitive insulator function. The best characterized example for this model of regulation is the Igf2/H19 cluster on distal mouse chromosome (Chr) 7 (Figure 1.1B) (37-39; reviewed in ref. 40). The IC that exhibits insulator activity is an intergenic DMR flanked by Igf2 and H19, which either binds the CTCF protein or becomes methylated, on the maternal or paternal allele, respectively. In vertebrates, CTCF is considered to be a versatile transcription regulator  whose  functions  include,  but  are  not  limited  to,  transcription  activation/repression, chromatin insulation, and enhancer-blocking activity (41). When CTCF binds the unmethylated maternal allele a chromatin insulation boundary between Igf2 and the enhancers that lie 3’ of H19 is created, therefore silencing Igf2 and allowing expression of H19. On the paternal chromosome, methylation is inherited in mature sperm at the IC1 insulator sequence which prevents CTCF binding and allows the interaction between the enhancers and the Igf2 promoter, so that Igf2 is expressed and H19 is silenced. More recently, the paternally expressed gene Rasgrf1 was also shown to be regulated in this manner (42). A methylation-sensitive tandem repeat sequence (DMR) was discovered 30kb upstream of the Rasgrf1 promoter that either binds CTCF or acquires methylation on the maternal and paternal alleles, respectively. Similar to H19/Igf2 regulation, the enhancer located on the other side of the DMR is either blocked from Rasgrf1 on the maternal allele or is able to facilitate expression of Rasgrf1 on the paternal allele (42). The third major imprinting mechanism involves the imprinted transcription of long non-coding RNA (lncRNA)-mediated epigenetic silencing (see Figure 1.1C) (reviewed in refs 32 and 40). Simply put, transcription of the lncRNA is associated with the bidirectional silencing of genes in cis. There are many ideas by which lncRNAs are thought to mediate this repression. RNA-based models consist of the RNA spreading on the chromatin to recruit repressive chromatin modifications or to create a transcription free sub-nuclear compartment (32). Alternative possibilities involve transcriptional models, which entail the act of transcription affecting the topological nature of the chromatin of the locus and triggering repressive marks (32). There are two well-known imprinting domains where this type of regulation has been described. Both Airn and 5  Kcnq1ot1 on mouse Chr 17 and distal Chr 7, respectively, are paternally expressed lncRNAs that are regulated by their own promoters marked by a gametic DMR and located in introns of protein-coding genes (Igf2r and Kcnq1, respectively). They act to repress expression of protein-coding genes on the paternal chromosome so that consequently only the maternal alleles of neighboring protein-coding genes are expressed (43; 44). Truncation or loss of Airn and Kcnq1ot1 results in the loss of silencing along the paternal chromosome and therefore the activation of the normally silent paternal alleles of multiple imprinted protein-coding genes (43; 45). Misregulation of KCNQ1OT1 on human Chr 11 is implicated in a wide variety of malignancies and the Beckwith-Wiedemann syndrome (44; 46; 47).  1.1.4 Tissue-specific imprinting and allelic ratios  Imprinted genes display a wide range of expression patterns and are often expressed and imprinted in a tissue- and developmental stage-specific manner. There are genes that are highly expressed and imprinted in both embryonic and extraembryonic tissues during development such as Mest, the gene focused on in this thesis. Several, however, are found to be imprinted only in certain tissues, such as in the placenta, the yolk sac or specific embryonic tissues (48). Given that the placenta is a direct mediator of maternal resources, it is not surprising that it comprises a large majority of genes that were evolutionary selected to be only imprinted and maternally expressed in the placenta (48). For an example, many maternally expressed genes in an imprinted domain on mouse Chr7 that are regulated by the lncRNA Kcnq1ot1 (discussed above) are silenced only in the placenta (34; 49; reviewed in ref. 50). Another tissue where genes tend to be tissue-specifically imprinted is the developing and adult nervous system (48; 51). Imprinted genes also exhibit varying degrees of allelic expression. Whereas many genes are strictly transcribed from only one of the parental alleles, several exhibit a preferential allelic bias where both alleles are expressed but one is expressed more strongly than the other. In one particular study reviewing the literature for preferential bias of imprinted genes, 26 out of 50 murine imprinted genes showed a preferential expression pattern in a tissue-specific manner and preferential maternally expressed genes comprised most of this group (52). Whether these genes fulfill the criteria to truly be called imprinted, however, is debatable. 6  1.1.5 Summary  In summary, most imprinted protein-coding and non-coding genes in the mammalian system are distributed across 16 clusters. Generally, they are regulated by 5mC-methylation-sensitive gametic DMRs/ICs which control the allelic usage and gene dosage of imprinted genes via varying mechanisms. Whereas the imprinting within some domains largely revolves around a central mechanism of regulation (like those discussed above), others not mentioned, such as the Snurf-Snprn, Dlk1-Gtl2, and Gnas domains are regulated by a complicated combination of DMRs and ncRNAs and are not fully characterized and understood (discussed below). Furthermore, there are a few uncharacterized domains where the regulation of the secondary imprinted genes is not yet known, such as the domain focused on in this thesis. The way by which the ICs control imprinting varies from cluster to cluster but the end result is the same in that genes become imprinted and expressed from a single parental allele.  1.2 Additional roles of non-coding RNAs in imprinted domains  Imprinted gene clusters contain at least one ncRNA that displays reciprocal imprinted expression relative to neighboring protein-coding genes (31; 32). Some clusters contain only one ncRNA while others contain several ncRNAs. This is not surprising since ncRNAs are now known to be widespread in the mammalian genome and thought to function as transcriptional regulators (53). The mode of action for ncRNAs in imprinted domains ranges and is different from cluster to cluster. Whereas the expression of some of them play a major functional role in the regulation of genomic imprinting, like in the case of the lncRNAs Airn and Kcnq1ot1 described above, the functions of other ncRNAs play smaller yet important roles in their domains or their functions remain unknown. In addition, ncRNAs are regulated in a developmental and tissue-specific manner. Major regulators such as Airn and Kcnq1ot1 tend to be more ubiquitously expressed from early development onwards, while other ncRNAs, such as in the Snurf/Snrpn, Dlk1-Gtl2, and Gnas domains, exhibit specific expression patterns in accordance to their function (54; 55; reviewed in ref. 56). For example, the complex Snurf/Snrpn domain in mice, also known as the Prader-Willi/Angelman cluster in humans, contains an interesting tissue-specific transcriptional unit referred to as Long  7  Non-Coding Antisense Transcript (LNCAT) transcribed exclusively from the paternal chromosome in post-mitotic neurons (see below) (57; 58). Several imprinted domains have also been found to contain many small regulatory RNAs, with the majority being organized into clusters that are embedded within and processed from introns of lncRNAs (59-61). The small RNAs fall into three classes: C/D-box small nucleolar RNAs (snoRNAs), microRNAs (miRNAs), and Piwiinteracting RNAs (piRNAs). In regulatory terms, they mainly act through sequencespecific recognition of other cellular RNAs (32). Briefly, snoRNAs are ~80-300-nt-long nuclear RNAs that reside in nucleoli and function in the post-transcriptional modification of rRNA nucleotides by acting as RNA guides to modify rRNAs or spliceosomal UsnRNAs (61; 62). Of interest, however, snoRNAs in imprinted domains are considered ‘orphan C/D RNAs’ since they accumulate within the nucleolus but do not appear to function like normal C/D RNA genes and may represent a novel class of nuclearretained transcripts (58; 63). MiRNAs are ~21-nt-long ncRNAs that act at the posttranscriptional level, binding to the 3’untranslated regions (3’UTRs) of their target genes, usually through an imperfect complementary manner, and inhibiting protein synthesis or, in occasional circumstances, guiding RISC-mediated cleavage of its mRNA (64). PiRNAs are a novel class of germ-line specific small RNAs that specifically interact with members of the Piwi family to protect the mammalian genome from transposon activity through mechanisms such as germ-line DNA methylation (65-67). Since small RNAs generally are trans-regulators and regulation of imprinted expression usually involves cis-acting mechanisms it is likely that most small RNAs in imprinted domains are involved in functions unrelated to the regulation of genomic imprinting per se although there are a few exceptions. Three noteworthy domains that include several large and small ncRNAs are the Snurf/Snrpn, Dlk1-Gtl2, and Gnas domains. In the mouse Snurf/Snrpn domain, LNCAT (Long Non-Coding Antisense Transcript) spans more than 1 Mb, likely codes for complex alternatively spliced transcripts, hosts several repeated snoRNA genes, and encompasses and shares exons with the Snurf/Snrpn locus (32; 56-58; 68; 69). This large domain includes maternally and paternally expressed protein-coding genes that are under control of a complex bipartite IC located at the 5’ end of Snurf/Snrpn (70). The Dlk1-Gtl2 cluster also contains a complex population of ncRNA genes that encode long non-coding, snoRNA, and miRNA genes, are exclusively expressed from the maternal allele, regulated by a DMR methylated in the male germ-line, transcribed in the same 8  orientation, and apparently share a similar expression profile (71-75). Given these characteristics, it is considered that these ncRNAs could belong to a single large transcriptional unit extending > 200 kb, like LNCAT, but this question remains open as the molecular characterization is incomplete (32). The other complex cluster, the Gnas cluster, contains two described paternally expressed ncRNAs Nespas and Exon 1A that are each transcribed from their own promoters and are each associated with an IC (76). Table 1.1 provides a survey of ncRNAs in mouse imprinted domains and their reported or proposed functional roles.  9  Table 1.1 A survey of large and small ncRNAs located in mouse imprinted domains. Imprinted domain  ncRNA  Encodes  Kcnq1  Kcnq1ot1  itself  P  Igf2r  Airn  itself  P  Igf2/H19  H19  miR-675  M  Snurf/Snrpn  Long Noncoding Antisense Transcript (LNCAT)  See below  P  Ube3aATS  P  Represses antisense gene, Ube3a, on paternal allele  (68)  Ipw  P  In humans, misregulation/deletion of IPW may contribute to the PWS phenotype directly. Alternatively, the mRNA product may be involved in the imprinting process of genes in the region.  (78,79)  5' U exons (Upstream from Snurf /Snprn locus)  P  Thought to be involved in the neuronalspecific expression of Ube3a-ATS  (69)  MBII-85 sno-RNA cluster  P  Misregulation/deletion of HBII-85 in humans majorly contributes to Prader-Willi syndrome. In mice, its deletion is the likely cause for the neonatal lethality in PWS model mice  (58,62,63,80, 81)  MBII-52 sno-RNA cluster  P  HBII-52 in humans regulates alternative splicing of the serotonin receptor 2C  (58,62,63,82, 83)  P  Might silence overlapping gene C15orf2 which might have evolved by retroviral integration. 3 of the piRNAs match the long terminal repeat sequence at the 3’ end of C15orf2  (32,84,85)  M  A possible cis-repression of the reciprocally imprinted Dlk1 gene has been investigated but they are generally not co-expressed in tissues where they both exhibit imprinting. The function may only be to encode miR770.  (86)  (58,86,87)  (88)  Cl.132 piRNA cluster in humans (n=17)  itself  M/P  Dlk1-Gtl2  Function or predicted function Represses protein-coding genes on paternal Chr Represses protein-coding genes on paternal Chr May regulate mRNAs in development and/or in oncogenesis Codes for complex alternatively spliced transcripts  Gtl2/Meg3  miR-770  Rian/Irm/ Meg8  snoRNA cluster  M  snoRNAs are mainly retained in the nucleus but do not appear to act as RNA guides and thus they might represent a novel class of nuclear-retained transcripts  miR-370  M  Implicated in cancer and predicted to target oncogene MAP3K8  References (44,45) (43) (77) (57)  10  Imprinted domain Dlk1-Gtl2  ncRNA  Rian/Irm/ Meg8  Mico1 /Mico1os  Mirg (MicroRNA containing gene)  AntiRtl1  Gnas  Plag1/Zac1  Nespas  Encodes  M/P  Function or predicted function  References  A possible piRNA gene cluster, Cl.124, located at the snoRNA cluster (humans, n=11)  M  The putative piRNAs are homologous with several human C/D RNA sequences but this remains to be elucidated  (32)  itself  M  Unknown function but they are exact complements of each other  (86)  more than 40 miRNAs  M  Target mRNAs; many are homologous with each other and likely exhibit functional redundancy  (73,86)  miR-376  M  Proposed to regulate PRPS1, an important enzyme of the purine metabolism  (89)  miR-134  M  Regulates the localized translation of the Limk1 mRNA which encodes a regulator of the size of dendritic spines in rats  (90)  miR-431, miR-433, miR-434, miR-136  M  In sheep, this miRNA cluster acts in trans to regulate the expression of the perfect complementary antisense gene Rtl1 by RISC-mediated cleavage of its mRNA  (64)  miR-127  M  Implicated in tumor biology and thought to target proto-oncogene BCL6  (91)  P  Associated with silencing of the antisense maternally expressed protein-coding gene Nesp, and associated with the expression of protein-coding gene Gnasxl  (76)  (76)  (92)  itself  Exon 1A  itself  P  Represses expression of protein-coding gene Gnas in specific tissues in an unknown sense cis mechanism as it is transcribed upstream of Gnas on the same sense strand in the same direction.  Hymai  itself  P  Retained in nucleus and might act to keep the paternal allele unmethylated and in a transcriptionally permissive state.  Retained in nucleus and might act to keep the paternal allele unmethylated and in a transcriptionally permissive state. M=maternally expressed, P= paternally expressed Plagl1it  itself  P  (92)  11  1.3 Sub-proximal MMU6  The proximal end of mouse Chr 6 (MMU6) contains two imprinted domains associated with abnormal phenotypes in maternal uniparental disomy (mUPD) mice. Mice carrying a mUPD for the sub-proximal region, defined from cytogenetics G-bands 6A3.2 to 6C1, are growth retarded at birth and thus show an intrauterine growth restriction (IUGR) phenotype (93; 94). This domain contains a genomic gap in the current mouse genome sequence that is set arbitrarily at 50 kb and is not supported by physical data. As for imprinted genes, this small domain contains the well-known paternally expressed gene Mest/Peg1 (Mesoderm specific transcript/Paternally expressed gene 1), the maternally expressed genes Copg2 (Coatomer protein complex subunit gamma 2) and Klf14 (Kruppel-like factor 14), and paternally expressed RNAs Copg2AS and Mit1 (Mest-linked transcript 1) (Figure 1.2) (13; 95; 96; 97). Additionally, this domain contains a miRNA, referred to as miR-335, located in intron 2 of Mest (98). Tsga13 is a testisspecific gene whose imprinting status is inconclusive in mice and humans (99; 100). To date, the Mest locus contains the only potential imprinting center identified thus far in this domain, a DMR methylated in oogenesis, which may also be involved in regulating the neighboring protein-coding genes (16; 101).  1.3.1 Mest Mest codes for a highly conserved enzyme of the α/β-hydrolase fold superfamily, although the substrate of MEST is still unknown (95; 102). This evolutionary conservation contrasts with the imprinting of the locus, which has only been observed in marsupials and placental mammals (103; 104). It is plausible that selective pressures for imprinting at the locus followed acquisition of novel mammalian-specific functions of the locus. In accordance with this proposition, Mest is highly expressed in extraembryonic mesoderm where it may play important roles in the establishment or function of the placental vascular network (105). Previous work has also reported that the expression profile of Mest includes embryonic mesoderm (and the mesodermal derivatives), in areas of the developing and adult brain, and in adult mouse tissues at low levels (95; 106; 107). As for the function of MEST, some reports have implicated it in having an important role in adipogenic differentiation and establishing patterned gene expression in 12  Figure 1.2 Schematic representation of the imprinting and genomic organization in the MMU6 sub-proximal domain. Genomic organization of the Mest-Klf14 region to scale showing the four protein-coding genes, Mest, Copg2, Tsga13, and Klf14, the two uncharacterized RNAs Copg2AS and Mit1 from the EST clusters, and the microRNAs miR-335-5p and miR-335-3p in the Mest locus. Mest, Copg2AS, and Mit1 are transcribed from the forward (+) strand, whereas Klf14, Tsga13, and Copg2 are transcribed from the reverse (-) strand in the opposite direction. Mest, Copg2AS, and Mit1 are paternally expressed genes indicated by the blue boxes, whereas Klf14 is maternally expressed indicated by the red box. MiR-335, indicated by the small yellow box in the Mest locus is presumably paternally expressed as well. Copg2, indicated by the pink boxes on both alleles, is not strictly imprinted but has been shown to be preferentially maternally expressed in certain tissues. Tsga13, indicated in black, is a testis-specific gene whose imprinting status is inconclusive. The map also shows the location of the arbitrarily sized gap, represented by the grey box, within intron 20 of Copg2. The white and black lollipops represent unmethylated and methylated promoter sequences, respectively. Notably, the Mest locus contains the only potential imprinting center (DMR) identified thus far in this domain. M=maternal allele, P=paternal allele.  13  the developing neocortex through the Wnt/β-caterin and Fgf8-regulated signaling pathways, respectively (102; 108). Additionally, MEST has been shown to localize to the endoplasmic reticulum/Golgi apparatus in adipocytes, perhaps to enable it to function in lipid accumulation and facilitate storage of fat, and to interact with LRP6 (lipoprotein receptor-related protein) to control glycosylation of LRP6 (108; 109). Mest, also known as Peg1, was first identified as a paternally expressed imprinted gene in a systematic screen using subtraction hybridization between cDNAs from normal and parthenogenetic (maternal contribution only) mouse embryos (95). Mest is imprinted in all marsupials and placental mammals analyzed to date (93; 95; 104; 110). The genomic structure of Mest consists of 12 coding exons and the promoter sequences are embedded within a CpG island that spans across exon 1 (101) (Figure 1.3A). This CpG island is methylated specifically in oogenesis and the methylation is maintained on the maternal chromosome throughout embryonic development (16; 101). Additionally, the Mest locus exhibits both H3K4me3 and H3K9me3 histone marks, which represent active and repressed transcriptional states, respectively (111). Together, the epigenetic marks presented at the Mest locus facilitate the imprinting of Mest and thus Mest fulfills the criteria to be considered a primary imprinted gene, directly regulated by a germ line-derived imprint.  1.3.2 Copg2  Copg2 is located immediately downstream of Mest and shares a tail-to-tail ~50 bp overlap with Mest which is conserved in humans and other mammals (96; 112) (see Figure 1.2). Additionally, the orthologues of both genes reside next to each other further down in the evolutionary ladder in the zebrafish and pufferfish genomes suggesting the genomic organization had been established in the early stage of vertebrate evolution and preserved ever since (103; 113; 114). Copg2 is comprised of 24 coding exons and, interestingly, there is a genomic gap within intron 20 in the latest build of the mouse genome. Regarding the protein function, COPG2 acts to facilitate intracellular trafficking of proteins through budding from the Golgi membrane (115; 116). Copg2 does not have a strict imprinting pattern like Mest and the imprinting status of Copg2 appears to differ across mammals. In humans, one disputable study showed that COPG2 was paternally expressed in several fetal tissues, but biallelically expressed in fetal brain (113). Other groups reported that COPG2 was biallelically 14  Figure 1.3 Diagrams of the Mest and MestKO loci. A) Schematic diagram of the Mest allele. The Mest ORF is indicated by the black boxes representing the coding exons and the 5’ and 3’ untranslated regions are indicated in light grey. B) Schematic diagram of the MestKO allele. The Mest exons are represented by black rectangles. The MestKO fragment consists of a splice acceptor-IRES-lacZ/neo fusion-poly(A) cassette, represented by the white rectangle, replacing Mest exons 3-8, and part of exon 9. The 5’ and 3’ untranslated regions are indicated in light gray. For A and B, the intronic miRNA, miR-335, is indicated by the small black under the second intron and the location of the CpG island is indicated by the grey line under exon 1.  15  expressed in fetal tissues and adult blood lymphocytes (100; 112). In sheep and cattle, Copg2 was found to be biallelically expressed, yet maternally expressed in swine placenta (110; 117; 118). Even within the same species, such as in mouse, conflicting results have been published. Copg2 was shown to be preferentially expressed from the maternal allele only in specific intra-subspecific hybrids but not in interspecific hybrids involving Mus spretus (96; 119). For the former study, Copg2 was shown to be imprinted in a tissue-specific manner where it was preferentially maternally expressed in adult brain tissue and at varying developmental ages from E11.5-17.5. A more relaxed biallelic pattern was observed in adult heart, lung, and muscle tissues (96). A recent study on imprinted genes in the developing and adult mouse brain, using intra-subspecific hybrids as well, consistently showed that Copg2 was preferentially maternally expressed (~7085%) in E15 brain, adult cortex and hypothalamus tissues (120). Taking all the results together, it appears the imprinting at Copg2 is a dynamic process given that the imprinting is developmental and tissue-specific and possibly species-specific. In tissues that exhibit Copg2 imprinting, it is not clear what drives the allelic usage. A study investigating the methylation at the promoter of Copg2 in these tissues arguably revealed minor differential methylation between the parental alleles suggesting that Copg2 is not regulated by a DMR at the locus (119). Hence, the mechanism that regulates the maternal bias at the Copg2 locus is still unknown and likely due to influencing factors outside of the 5’ region of the locus.  1.3.3 Klf14  Klf14 is a maternally expressed gene located approximately 60 kb downstream from the Copg2 promoter and transcribed from the (-) strand, like Copg2 (see Figure 1.2). It is a member of the Sp/KLF family of transcription factors and thought to have arisen through the retrotransposition of Klf16 (13). In humans, the KLF14 protein has been shown to be a regulatory protein of the Transforming Growth Factor β pathway that acts to silence expression of TGFβRII (Transforming Growth Factor-β Receptor II) by forming a complex with other silencing factors such as HDAC2 at the promoter of TGFβRII (121). In addition, one report in a high-profile journal also suggested that KLF14 is a major regulator of gene expression in adipose tissue (122). Regarding imprinting, Klf14 was reported to be strictly imprinted in all extraembryonic, embryonic, and adult mouse tissues tested (13). The mechanism whereby 16  Klf14 becomes maternally expressed, however, is still unknown. Klf14 was found to not contain any of the known imprinting hallmarks such as a DMR or differential histone modifications that would mediate its imprinting; yet maternal expression of Klf14 was found to rely on maternally inherited DNA methylation marks (13). This finding came from the observation that offspring from Dnmt3a conditional knockout females do not express Klf14 because germ-line methylation on the maternal allele was not properly inherited from their mothers. Given this important outcome, it was suggested that Klf14 expression is dependent upon a maternally methylated region and the DMR at the Mest locus was suggested to be a strong candidate that regulates Klf14 expression in cis (13). Whether or not the DMR at Mest indeed acts as an imprinting center also regulating imprinted expression at Klf14 is currently unknown.  1.3.4 Copg2AS and Mit1 RNAs Copg2AS overlaps the 3’UTR of Mest and the 3’UTR of Copg2 and is transcribed on the same strand as Mest (see Figure 1.2) (96). Lee et al. demonstrated that Copg2AS is imprinted and expressed only from the paternal allele in all tested tissues in intra-subspecific F1 hybrid mice. These include the developing embryo at different stages and in adult tissues such as muscle, brain, heart and lung tissues (96). Specifically in the mouse brain, Copg2AS is expressed broadly throughout the brain, but also exhibits a restricted pattern of expression within the thalamus and hippocampus and in some layers of the main olfactory bulb and cerebral cortex (123). Given the attributes of Copg2AS, it was suggested it could be part of an alternative Mest isoform with a larger 3’UTR (96). Mit1 is considered to be a ncRNA located in a cluster of expressed sequence tags (ESTs) that are also transcribed downstream and from the same strand as Mest (see Figure 1.2). They are, however, located on the other side of the genomic gap in intron 20 of Copg2. The transcription products of Mit1 are a heterogeneous population of RNAs ranging in sizes with the major transcript being ~7 kb in length and analysis of the sequence composition revealed no considerable open reading frame (96). They are mainly expressed in the embryo and adult brain and were also shown to be imprinted and expressed from the paternal allele in these tissues (96; 120).  17  1.3.5 miR-335  MiR-335, located in intron 2 of Mest, is the only miRNA identified thus far in the Mest-Copg2-Klf14 imprinting domain (see Figures 1.2 and 1.3). Mature miRNAs are short ncRNAs that are generally 21-25 nucleotides long that regulate gene expression of target genes by mRNA degradation or translational inhibition. The processing of miRNAs involves RNA polymerase II-mediated transcription into longer molecules that are capped, polyadenylated, and processed into hairpin RNAs. The hairpin RNA is then transported to the cytoplasm and processed by Dicer to generate the mature miRNA (124-128). At this point, miRNAs are incorporated into an RNA-induced silencing complex (RISC) which binds to the 3’UTR or elsewhere on a coding mRNA transcript to induce translation suppression or mRNA degradation depending on the degree of complementarity with the target mRNA (129-131). Studies have shown that miRNAs play significant roles in fine-tuning complex processes including, but not limited to, mammalian development, morphogenesis, organogenesis and cellular differentiation, by providing a means of spatial and temporal control of gene expression (132). MiR-335 was first identified in a study that sought to discover miRNAs that regulate translation in mammalian neurons (98). This result was not surprising since Mest is transcribed and expressed at high levels in the adult brain. Since miR-335 is processed from the intron 2 of Mest, it shares coordinated expression with Mest (133136). From an evolutionary standpoint, miR-335 seems to be conserved in placental mammals, including humans, suggesting a function is to fine-tune physiological processes occurring in extraembryonic tissues and in the developing embryo, in tissues where Mest is expressed. With the discovery of miRNAs, much research has focused on determining their biological functions and the mRNAs that they target post-transcriptionally to downregulate protein levels. MiRNA target prediction software such as TargetScan, miRanda, and PicTar, predicts that MiR-335 targets 1838 genes (134). Much of the research on miR-335 has focused on its role in cancer and one study in particular put miR-335 into the spotlight. This report by Tavazoie et al. showed that miR-335 inhibited metastatic cell invasion and thus identified it as a metastasis suppressor in human breast cancer (137). A subsequent study found that genetic deletion of human miR-335 or Mest promoter hypermethylation were mechanisms by which silencing of miR-335 occurred in breast cancer patients with malignant cell populations (135). Another study reported that miR18  335 plays a major role in the BRCA1 regulatory cascade by targeting upstream genes such as ERα, IGF1R, SP1, and ID4, which controls the maintenance of homeostasis in breast tissue (138). In addition to human breast cancer, deregulation of miR-335, both up- and down-regulation, has been implicated in several other human cancers (133; 136; 139-152). Given that miR-335 is predicted to target hundreds of genes and deregulation of miR-335 has been associated in several cancers, it is not surprising that normal miR-335 function has been implicated in several physiological processes. According to the literature, the scope and diversity of miR-335 functions seems widespread and reveals the importance of miR-335-mediated regulation in numerous biological processes. These include, but are not limited to, mesenchymal proliferation, migration, and differentiation, skeletal muscle regeneration, osteogenic differentiation, lipid metabolism, epididymal development, renal aging, and hepatic stellate cell migration and activation (134; 153-158). Target genes determined from these studies and among others are summarized in Table 1.2.  Table 1.2 Identified target genes of human, rat, and mouse miR-335 reported in the literature. Target gene  Function of gene  References  SOX4  SOX family members are transcription factors that regulate progenitor cell development and migration.  (134,137,156)  TNC  An extracellular matrix glycoprotein involved in cell migration  (137,158)  ERα IGF1R  Known regulator of the breast cancer susceptibility gene BRCA1 Known regulator of the breast cancer susceptibility gene BRCA1  (138) (138)  SP1  Known regulator of the breast cancer susceptibility gene BRCA1, Sequence-specific DNA-binding protein that is an extremely versatile protein involved in the expression of many different genes  (138,147)  ID4  BRCA1 suppression factor  (138)  Bcl-w  An anti-apoptotic member of the Bcl-2 protein family that suppresses apoptosis by interacting directly with proapoptotic members to block their apoptotic activities  (147)  19  Target gene  Function of gene  References  Rb1  Part of the Retinablastoma protein family that controls cell cycle genes through interaction with the E2F family of transcription factors, as well as by direct recruitment of chromatin regulators to promoters. Associated with the p53 tumor suppressor pathway that limits cell proliferation and neoplastic cell transformation.  (144,146)  Daam1  A member of the formin protein family acting downstream of WNT signaling that plays an important role in regulating the actin cytoskeleton in promoting proper cell polarization, migration, proliferation and tissue morphogenesis during embryonic development.  (148)  Rock1  In the TGF-β non-canonical pathway, phosphorylates myosin light chain by phosphorylating at serine 19  (151)  Mapk1  In the TGF-β non-canonical pathway, phosphorylates myosin light chain kinase, which in turn also phosphorylates serine 19 of mysoin light chain  (151)  Lrg1  Suspected to be involved with TGF-β signaling  (151)  Rasa1  Contributes to the epididymal development through involvement of cell proliferation and anti-apoptosis.  (155)  Runx2  A key transcription factor involved in osteogenesis  (134)  Dkk1  A protein that is essential to maintain skeletal homeostasis as an inhibitor of Wnt signaling and osteogenic differentiation  (156)  Sod2  An antioxidative enzyme located in the mitochondria that may play a key role in modulating cellular aging by detoxifying reactive oxygen species generated in the mitochondria.  (157)  20  1.3.6 Human homology and disease  The sub-proximal MMU6 domain is tightly conserved between mice and humans. All the genes described above are present and most are similarly imprinted with the exception of COPG2 (13; 112; 93; 159). The imprinting at COPG2 is disputed in the literature with a report of paternal expression and another of biallelic expression and no imprinting (112; 113). Similar to mice, the human MEST locus contains the only DMR identified in the domain and was also shown to display a “double hit,” which means MEST is marked with the activating mark, H3K4Me2, and repressive DNA methylation (36; 159). The mouse and human sequences for intron 20 of Copg2/COPG2 reveal they are at least 80% identical although both genomic sequences are incomplete and contain a gap, signifying the complexity of the region. A longer 3’UTR was identified for human MEST and an antisense COPG2 transcript was discovered in intron 20, referred to as COPG2IT1, which are homologous to mouse Copg2AS and Mit1, respectively (112). Of interest, an additional paternally expressed 4.2 kb ncRNA in the MEST locus was identified in humans, referred to as MESTIT1 (Mest intronic transcript 1), that is transcribed upstream of the CpG island on the opposite strand and is speculated to possibly be involved in the regulation of MEST expression during development (160). The mouse domain shares syntenic homology with a region on human Chr 7q32 that is associated with growth retardation and Silver-Russell syndrome (SRS) in maternal UPDs, where both copies of Chr 7 are inherited from the mother (161-163). Phenotypes that manifest from UPDs are generally indicative of imprinting defects and in this case, the observed phenotype from maternal UPD 7 is thought to be due to a deficiency of paternally expressed genes. As a consequence, MEST is designated as a key candidate locus for the aetiology of SRS in these cases. Many studies, however, have failed to provide any molecular evidence that the SRS aetiology is associated with UPD at 7q32 except for one recent study where a patient with SRS features was shown to have a deletion of the paternal allele of the MEST region (164-166). These results indicate that the mechanism by which SRS and growth retardation arises in mUPD7 remains largely unknown and more research is needed to uncover this increasingly complex region.  21  1.4 MestKO mutant mouse line A targeted mutation at Mest, the MestKO allele (official name Mesttm1Lef) was established more than a decade ago (106). The targeting construct inserts a splice acceptor-IRES-βgeo (lacZ-neo fusion) and deletes exons 3-8 and part of exon 9 (see Figure 1.3B), those most likely important for catalytic activity. The βgeo is immediately followed by an introduced polyadenylation (pA) signal to terminate the transcript at that point. LacZ staining experiments on reciprocal MestKO/+ and Mest+/KO embryos, those that express the MestKO allele either maternally or paternally, respectively, showed that Mest imprinting remains faithfully recapitulated, as expression of lacZ was only observed in embryos that paternally inherit the MestKO allele (Mest+/KO) (106). The MestKO allele was generated when miR-335 had not yet been discovered and the important sequences of miR-335 (e.g. primary miRNA) remain intact in intron 2 of the mutant allele. Disrupting the Mest allele and therefore creating the null mouse knockout showed that Mest deficiency was not embryonic lethal, although it is normally highly expressed during embryonic development. Heterozygous mice expressing the paternally inherited MestKO allele (Mest+/KO) exhibited two main phenotypes. Firstly, Mest deficiency caused embryonic growth retardation, seemingly a general growth retardation of embryonic and extra-embryonic structures that are otherwise morphologically normal (106). The second involved Mest+/KO mutant females exhibiting abnormal maternal behavior towards their pups. They did not ingest their newborn pups’ extra-embryonic tissues at birth and were generally not attentive to their pups. This fascinating observation suggested that Mest has a role in postnatal behavior, the first evidence for such a role for an imprinted gene. Although originally generated to determine the phenotypic outcome of loss of the MEST protein, the MestKO allele has serendipitously become an invaluable tool to study alternative functions of the Mest locus and the mutant mouse line was the main tool used in my studies.  1.5 Thesis theme and objectives  Imprinted transcription at the Mest locus appears to be involved in alternative functions independent of its coding potential. Firstly, the fact that Copg2AS and Mit1 are uncharacterized RNAs transcribed downstream of Mest on the same (+) strand and are exclusively expressed from the paternal allele, like Mest, suggests they constitute larger 22  isoforms produced from the Mest locus (96). Intriguingly, in tissues that express Mit1 such as the developing embryo and adult brain, Copg2 was reported to be preferentially maternally expressed, leading to the notion that this alternative function, production of larger RNAs, may have a role in regulating Copg2 and the other maternally expressed gene Klf14 in the sub-proximal MMU6 domain. Secondly, the Mest locus was found to encompass a miRNA which has already been recognized to have a role in the progression of metastatic cancer in humans, indicating miR-335 could have important roles in development (98; 137). Since the Mest locus is predicted to have additional functions, the overarching hypothesis of this thesis is that imprinted transcription at Mest was evolutionary selected for three distinct products of the Mest locus (Figure 1.4): the MEST protein, longer Mest isoforms, referred to as MestXL, and the miRNA located in intron 2, miR-335. More specifically, the work presented in this thesis is based on the following hypotheses: (i) The Mest locus produces longer regulatory transcripts which control allelic usage at the neighboring Copg2 locus; (ii) The intronic miR-335 is paternally expressed and regulates important regulators of embryonic growth during development. The specific objectives are:  1) To determine whether the Mest locus produces larger isoforms inclusive of paternally expressed RNAs Copg2AS and Mit1 (Chapter 3). 2) To investigate whether larger Mest isoforms have a role in regulating the imprinting of downstream maternally expressed genes Copg2 and Klf14 (Chapter 4). 3) To explore the developmental role of miR-335 by determining additional target genes (Chapter 5).  23  Figure 1.4 Products of imprinted transcription from the Mest locus. Schematic diagram of the Mest locus and the RNAs/products generated from imprinted transcription. A) The primary Mest mRNA is transcribed and processed to produce the coding mRNA that is translated to produce the MEST protein. MiR-335 is also produced simultaneously in the processing of the primary transcript. B) The Mest locus is predicted to produce larger RNAs, referred to as MestXL, which may have a role in regulating the imprinting of downstream genes, Copg2 and Klf14 (not pictured). Whether these variants would be processed the same way as the Mest mRNA to code for MEST and produce miR-335 is currently unknown. The exons are represented by black rectangles and the untranslated regions are indicated in grey. The location of the CpG island is indicated by the grey line under Mest exon 1 and miR-335 is indicated by the black star in intron 2.  24  Chapter 2: Materials and methods 2.1 Mouse strains The MestKO line (MGI: 2181803, official name Mesttm1Lef) was derived by targeting a promoter-trap construct to the Mest locus in ES cells, followed by germ line transmission from chimeric mice and subsequent breeding as previously described (106). The MestKO targeted allele replaces important coding exons of Mest with a spliceacceptor-IRES-βgeo-pA cassette. The MestKO mouse line is maintained on a CD-1 outbred background in a heterozygous state where the MestKO allele is always paternally transmitted. Generally Mest+/KO males were mated to wild-type (Mest+/+) female littermates and male pups genotyped to contain the MestKO allele were bred to continue the line. Note that for all genotypes the maternally inherited allele is always presented first. For some experiments outbred CD-1 females were ordered from the UBC Animal Care Centre. The mouse line with Chr 6 M. m. castaneus SNPs on the 129S1 background (129S1cCAST6/Lef, hereafter referred to as CAST6) was derived by me in our laboratory. Firstly, a female from strain 129S1/SvImJ (M. mus. musculus) was mated to a M. mus. castaneus (CAST) male to generate F1 animals. The F1 female was then backcrossed to the CAST male to produce the N2 generation. A N2 male and female genotyped to be heterozygous for two markers, D6Mit269 and D6Mit180, which flank the Mest locus, were then interbred to generate homozygous CAST6 animals. The CAST sequences on proximal Chr 6 cover at least 5 Mb as measured by the distance between the two markers. The CAST line was subsequently brought to homozygosity and maintained by intercrosses between mice homozygous for the CAST variant on proximal 6. All animal experimentation was carried out following the guidelines from the Canadian Council on Animal Care (CCAC) under UBC animal care license numbers A03-0289 and A03-0292.  2.2 Genotyping  Animals were genotyped at weaning by ear punch lysis and PCR, as described (167). When E14.5 embryos were collected, a small sample of yolk sac was retained and used for genotyping as well. Briefly, ear punches or yolk sacs were lysed overnight 25  at 50°C in proteinase K (proK) buffer (50mM KCL, 10mM Tris HCL, 2mM MgCl2, 0.1 mg/ml gelatin, 0.45% NP-40 substitute, and 0.45% Tween 20) with 2.5 mg/ml ProK (Roche). The resulting lysate was used in a PCR reaction as follows: 10 minutes at 95°C (to inactivate the ProK), followed by a pause at 85°C to add the PCR master mix (1x PCR buffer, 2mM MgCl2 200µM dNTPs, 2.5mM each primer, and 0.1U Tsg polymerase) then 35 cycles of 95°C for 30 s, 59°C for 30s, and 72°C for 30s and final analysis by agarose gel electrophoresis. Genotyping of the MestKO allele was performed with primers 702A and IRES1. Genotyping of the wild-type (WT) Mest allele was performed with primers 702A and 702C. Yolk sacs were additionally scored for lacZ expression from the MestKO allele by X-gal staining (1 mg/ml X-gal, 4 mM K4Fe(CN)6, 4 mM K3Fe(CN)6, 2 mM MgCl2, in PBS) as described (168). To derive the CAST6 mouse line, markers D6Mit180 and D6Mit269 that flank the Mest locus were used to genotype the mice. D6Mit180 is 1.5 Mb proximal to Mest and D6Mit269 is 3.6 Mb distal to Mest. Primer pairs 702A -IRES1 and Zfy1a-b were used to determine the genotype and sex of the primary embryonic fibroblasts, respectively, generated for the miRNA experiments. Primer sequences for genotyping are given in Table 2.1. 2.3 Embryo dissection Heterozygous Mest+/KO males were mated to CD-1 or 129S1cCAST6/Lef (CAST6) female mice. For timed mating, the day of the vaginal plug is E0.5. Females were sacrificed at E14.5 and yolk sac samples were taken for PCR genotyping and scored for lacZ expression for the MestKO allele by X-gal staining as described (168). Embryos collected were frozen at -80°C for subsequent RNA and/or DNA extraction. For in situ hybridization they were immediately put in 4% PFA for fixing. For tissue-specific analysis they were further dissected to collect the liver, heart, mid/hind brain, spinal cord, front brain, tongue, and lungs which were frozen at -80°C for subsequent RNA extraction.  2.4 SAGE analysis  Publicly available LongSAGE and SAGELite libraries were generated as part of the Mouse Atlas Project and the Mammalian Organogenesis – Regulation by Gene Expression Networks (MORGEN) Project  (www.mouseatlas.org; 169-171). Tag 26  sequences mapping to the sense and antisense strand of the Mest/Copg2 region were identified using in-house Perl scripts by Sorana Morrissy from Marco Marra’s laboratory. 2.5 RNA purification  Conceptuses from crosses between female CAST6 or CD-1 mice and male Mest  +/KO  heterozygotes were obtained at E14.5 and frozen on dry ice until needed for  RNA isolation. For additional experiments, certain tissues such as the developing mid and hind brain (MHB), spinal cord (SC), heart, and liver, were further dissected. For PEF lines, cells were expanded and harvested at confluency and the cell pellets were frozen down until needed. Total RNA was purified using Trizol (Invitrogen) according to manufacturer’s directions. For reverse transcription, RNA was further DNase-treated (RQ1, Promega, 1 unit) for at least 1 hour to remove contaminating gDNA.  2.6 RT-PCR/qRT-PCR analysis  Anywhere from 500 ng to 1 µg of DNase-treated RNA was reverse-transcribed with SSII (Invitrogen), using random (N15) primers, according to SSII protocol. For PCR, 1 µl of cDNA was used in a 25-µl reaction as follows: 95° C for 30 s, 56°-62°C for 30 s, and 72°C for 30 s. QRT-PCR was performed on a Step-One Plus Real time PCR system (Applied Biosystems) using Eva Green (Biotium) as the dye.1 µl of cDNA was used in a 25 µl reaction as follows: 35-40 cycles of 95° C for 30 s, 56°-62°C for 30 s, 72°C for 30 s, and 79°-89° C 10 s with a plate read depending on the primer reaction, followed by a melt curve. Ct values of three biological replicates (or more), obtained by the LinReg PCR program, were averaged and used to calculate relative amounts of transcripts, normalized to Ppia (172; www.linregpcr.nl). A two-tailed t-test was used to determine significance (p-value ≤ 0.05) between WT and Mest+/KO groups. For long range RT-PCR, RNA was reverse-transcribed with SSIII (Invitrogen), using random (N15) primers, according to SSIII protocol. The cDNA was RNase-H (New England Biolabs) treated before PCR by adding 5 units to 20 μl of cDNA and incubating for 30 minutes at 37°. For PCR, 1 μl of cDNA was used in a 25-μl reaction as follows for 26-30 cycles: 95° C for 30 s, 56°-60°C for 30 s, and 72°C for 30 s to 1 minute depending on length of PCR product. Primer sequences are given in Table 2.1.  27  2.7 DIG-labeled RNA probe preparation Selected mouse cDNA PCR fragments for Mest, the MestKO allele, Copg2, and MestXL were TA-cloned into the pGEM-T vector (Promega). The exception was the Mest ex12 probe, which was sub-cloned from the pKOPEG1.1a plasmid by excising out Mest exon 9 to intron 11 with BamHI and re-ligating back together. pKOPEG1.1a was originally generated by Louis Lefebvre by TA-cloning a PCR fragment from Mest exon 9 to exon 12 into the pCRII vector (Invitrogen). Approximately 25 µg of plasmid DNA was linearized with the respective restriction enzymes and purified by phenol-chloroform extraction. Approximately 1 µg was transcribed by T7 or SP6 RNA polymerase (40 units, Roche) in the presence of digoxigenin (DIG)-labeled UTP (Roche) to generate sense and antisense DIG-labeled probes. The in vitro transcription was carried out at 37° for two hours, incubated with DNase (RQ1, Promega, 1 unit) for 15 minutes to remove the plasmid DNA, and purified by precipitation with 4 M LiCl. Primer reactions used to generate the cDNA fragments and primer sequences are given in Table 2.1. The enzymes and RNA polymerases used to generate each probe are given in Table 2.2. 2.8 Northern blot analysis  For the analysis of WT embryonic and adult tissues, northern hybridization was performed on commercial blots (Clontech) containing 2 μg of poly(A)+ RNA from mouse tissues (Cat. # 7762-1) and mouse embryos (Cat. # 7763-1) using a  32  P-labeled Mest  ORF clone (exons 1 to 12) as a probe. The blots were hybridized at 68° overnight using Clontech quickhyb buffer and washed at maximum recommended stringency (2XSSC,0.1% SDS, 30 mins; 55° 0.1XSSC, 0.1% SDS 30 mins). Dynabeads (Invitrogen) were used to obtain polyadenylated RNA from total RNA from E14.5 conceptuses from crosses between female CD-1 mice and Mest+/KO heterozygous males. The RNA was separated by electrophoresis on a 1% agarose, 2.2 M formaldehyde denaturing gel. Neutral transfer (20X SSC) was performed onto nylon membranes, followed by UV-light crosslinking. Northern blot hybridization was carried out with DIG-UTP labeled RNA probes, with an overnight incubation at 65°C in DIG hybridization buffer (50% formamide, 5XSSC, 2% blocking reagent (Roche), 0.1% sarkosyl, 0.5% SDS, and 200 µg/ml denatured salmon sperm DNA). Membranes were washed twice with 1XSSC, 0.5% SDS, 0.1% sarkosyl for 5-10 minutes each, twice with 28  0.1XSSC, 0.5% SDS, 0.1% sarkosyl for 30 minutes each, and then twice with DIG-buffer (0.1M Maleic acid pH 7.5, 0.15 M NaCl, 0.1% Tween-20) for 5 minutes each to prepare for the DIG antibody. The membranes were then incubated at room temperature for 4560 minutes in blocking solution (DIG-buffer with 1% blocking reagent (Roche)), and 4560 minutes with DIG antibody (Roche) (1:10,000 dilution) in blocking solution, then followed by two washes with DIG-buffer for 15 minutes. After equilibrating in 100 mM Tris pH9.5, 100 mM NaCl, chemiluminescent detection with CDP-star reagent (Amersham Biosciences) was performed for 5 minutes before exposing to film. Exposure times ranged from 15 minutes to overnight. For one blot we used an alpha-32P -CTP labeled DNA probe (Mest exon 12) generated through linear amplification (20 cycles of 95° C for 30 s, 60°C for 30 s, and 72°C for 1.5 min) from a reverse primer, P1 E12 NO R1, in Mest exon 12 to generate antisense transcripts.  Hybridization was carried out in Church’s buffer (350 mM  Na2HPO4, 150 mM NaH2PO4, 10 mM EDTA, 7% SDS) overnight at 65°C, washed the following morning twice with low stringency wash (1XSSC, 0.1% SDS) for 10 minutes at 37°C, and twice with high stringency wash (0.1XSSC, 0.1% SDS) at 65°C for 10 minutes each. The membrane was then exposed to a phosphor screen overnight and imaged with the Storm Phospho-imager autoradiography system.  2.9 In situ hybridization (ISH) Conceptuses from crosses between female CD-1 mice and male Mest+/KO heterozygotes were obtained at E14.5. Embryos were fixed in 4% paraformaldehyde (PFA) overnight, treated in 30% sucrose the following night, and embedded and frozen in OCT (TissueTek). They were sectioned at 12 microns on a Leica cryostat (model CM3050 S). To prepare the sections for hybridization, they were fixed with 4% PFA, washed in 1 X PBS, treated with proteinase K, fixed again in 4% PFA, and put through an acetylation step (TEA/AA). The slides were then hybridized (50% formamide, 5XSSC, 5XDenhardts, 0.25 mg/ml yeast tRNA, and 0.5 mg/ml single stranded DNA) overnight at 65°C with their respective probe. The following day the slides were washed in different concentrations of SSC to remove non-specific probe, put through an RNase A step, the DIG antibody (1:2,000 dilution) reaction, and finally the color reaction step (usually overnight) using NBT/BCIP (Roche) as the substrate. To mount, the slides were prepared by washes with PBS, fixed for two hours (3.7% formaldehyde, 1XMEM buffer), counterstained with 29  nuclear fast red, dehydrated with varying amounts of ethanol concentrations, and mounted with Entellan. Experiments were performed at least twice for each probe and genotype (at separate times). The number of times each experiment was performed for each probe is as follows: Mest ex1-2 = 3, Mest ex3-8 = 6; Mest in11 = 2, Mest ex12 = 4, neo = 4; Copg2 3’UTR = 3; Copg2 in23 = 2; Copg2 in18-ex20 = 2; Copg2 ex14-20 = 4; BB EST = 2; Mit1 = 5; Klf14 = 3; and Cdkn1c = 2. 2.10 Allele specific analysis  The allele-specific Copg2 hot-stop PCR analysis was based on a polymorphic BstUI site present in the M. m. castaneus RT-PCR product (exon 3 SNP rs36827909) but absent in the CD-1 product. The 250-bp PCR products digested with BstUI yield bands of 200 and 50 bp for the CAST allele. Random-primed cDNA was generated from dissected E14.5 embryonic organ RNA and cDNA amplification was carried out with primer reaction Copg2 ex1-5 for 35 cycles (Tm=58°C). 5 µl of the PCR product was transferred into another master mix containing gamma-32P-CTP to go through one final PCR cycle. The hot PCR products were digested with BstUI (New England Biolabs) and analyzed on an 8% polyacrylamide gel. The bands were detected using the Storm Phospho-imager autoradiography system and the band densities (maternal and paternal alleles) were analyzed with the Image J software (173). As well, Copg2 ex1-5 and Copg2 ex17-19 RT-PCR products from F1 cDNA were directly sequenced. The Copg2 ex17-19 reaction contains 2 additional SNPs (exon 19; rs49583711 and rs47417386). The phred program was used to read the sequence trace data and assign quality values to the bases, normalized to F1 gDNA (174). For the allele-specific analysis of Klf14 expression, an expressed SNP was identified by sequencing amplified CAST genomic DNA for the single exon of Klf14. The SNP used in our analysis is at position 1245 of the Klf14-001 Ensembl transcript (ENSMUST00000101589; CD-1, A: CAST, G). Klf14 imprinting was assessed by sequencing Klf14 F4–R4 RT-PCR products on the antisense strand. Klf14 expression levels were assessed by using primers Klf14 F6-R6. Ct values of three biological replicates, obtained by the LinReg PCR program, were averaged and used to calculate relative amounts of transcripts, normalized to Ppia (172). All primer sequences are given in Table 2.1.  30  2.11 5’ rapid amplification of cDNA end (RACE) Total RNA purified from E14.5 wild-type Mest+/+ and Mest+/KO  heads were  reverse transcribed with SSII (Invitrogen) as described, using the gene-specific primer 1 (GSP 1) IRES R2, which is complementary to the IRES of the mutant allele. After first strand synthesis, the cDNA was treated with RNase-H (New England Biolabs) for 30 minutes at 37° and subsequently purified with the Wizard DNA Clean-Up System (Promega). The cDNA was artificially polyadenylated with terminal transferase (TdT) as follows: 5 μl of melted cDNA (95° for 10 minutes) was added to a solution consisting of 10 mM dATP, 2 μl 10X supplied buffer, 2 μl 10X CoCl2, 20 units Terminal Transferase (New England Biolabs) in a total volume of 20 μl. The reaction was incubated at 37° for 30 minutes, followed by a brief 10 minute 70° heating step to terminate the reaction. A first round PCR reaction was set up using primers RACE anchor primer + p(dT) and IRES1 (GSP 2), followed by a second nested PCR reaction using primers RACE anchor primer and Mest ex2 Ra (GSP 3). Following gel electrophoresis, the PCR bands were gel purified (Qiagen QIAEX II gel extraction kit) and sequenced. Primer sequences are given in Table 2.1. 2.12 Immunofluorescence (IF) and β-galactosidase staining  E14.5 embryos were fixed in 4% paraformaldehyde (PFA, Sigma) for 15 minutes, embedded, and transversely sectioned as described in materials and methods. For IF, cryosections were briefly fixed for 15 minutes in 4% paraformaldehyde, permeabilized for 30 minutes in 0.1% Triton-X 100 in PBS and blocked in 4% donkey serum, 0.5% BSA for 1 hour at room temperature, as previously described (175). For NCAM1 IF, 2.5% mouse-on-mouse reagent (Vector Labs) was added at the blocking step. Detection of epitopes was performed with rabbit anti-MEST (1:500, Proteintech #ag1617) or mouse anti-NCAM supernatant (5B8 at 1:10, Developmental Studies Hybridoma Bank) primary antibodies for 1 hour at room temperature and a secondary antibody incubation step for 45 minutes at room temperature with A594 anti-rabbit and A594 anti-mouse (1:200, Molecular Probes) antibodies, respectively. The sections were counterstained with DAPI (1:5000 in PBS, Sigma), mounted with Vectashield (Vector Labs) and imaged on a Leica DMI6000B inverted fluorescent microscope at 100X. Images were captured with a Qimaging Retiga 4000R monochrome camera and processed with Openlab (Improvision). 31  For LacZ staining, frozen cryosections were rinsed in PBS and stained overnight with X-gal (1 mg/ml X-gal, 4 mM K4Fe(CN)6, 4 mM K3Fe(CN)6, 2 mM MgCl2, 0.01% NP40 substitute in PBS) at room temperature. Images were captured with a Q-imaging MicroPublisher 5.0 RTV color camera and processed with Openlab (Improvision).  2.13 Analysis of miR-335 expression  E14.5 head or PEF total RNA was purified using Trizol (Invitrogen) according to manufacturer’s directions and DNase-treated (RQ1, Promega) for at least 1 hour to remove contaminating gDNA. For the first step of the TaqMan assay, anywhere from 1050 ng of RNA was reverse transcribed using the miR-335-5p and RNU6B (U6 small nuclear RNA) stem-loop RT-primers according to manufacturer’s instructions to obtain both miR-335-5p and RNU6B RT samples (Applied Biosystems). The stem-loop provides specificity for only the mature miRNA target and forms a RT primer/mature miRNA-chimera that extends the 3’ end of the miRNA. The second step, the stem-loop qRT-PCR, involved quantifying miR-335-5p and RNU6B resulting RT products using the miR-335-5p and RNU6B gene-specific probes, respectively, by the standard TaqMan qRT-PCR method (Applied Biosystems). The qRT-PCR was performed using the StepOne Plus Real time PCR system (Applied Biosystems) and Ct values of three technical triplicates for each sample, obtained by the LinReg PCR program, were averaged and used to calculate relative amounts of transcripts, normalized to RNU6B (172; 176). 2.14 pCmir335E and pC∆mir335E plasmid construction  The pCmir335E plasmid was derived from the original pCCALL2-miR335-IRESEGFP plasmid constructed by myself. pCCALL2-miR335-IRES-EGFP was generated by inserting the “miR-335 transgene” into a conditional expression vector named pCCALL2IRES-EGFP which was gifted to us from Dr. Corrine Lobe. The “miR-335 transgene” was generated through PCR amplification with primers Mest ex2F and Mest ex3R, which amplified a genomic fragment that consists of Mest intron 2 and miR-335 and also parts of the flanking Mest exons 2 and 3. This genomic fragment was inserted into the pCCALL2-IRES-EGFP plasmid with BamHI and XhoI and sequencing was performed to confirm  the  correct  genomic  sequence.  Briefly,  pCCALL2-miR335-IRES-EGFP  contained a βgeo (lacZ-neo fusion) and three pA signals (3XpA) located between loxP 32  sites followed by the “miR-335 transgene” and an IRES-EGFP. To activate the “miR-335 transgene,” the fragment between the loxP sites containing the βgeo-3XpA was excised by Cre recombinase. After realizing the IRES was faulty, the IRES-EGFP was removed by cutting the plasmid with XhoI and NotI and replaced with just EGFP which was amplified by PCR with primers pCmir335E EGFP-F and pCmir335E EGFP-R and fused in frame to the transgene with XhoI and NotI. The plasmid, now referred to as pCmiR335E, was confirmed by sequencing and positive GFP expression. The control plasmid, pC∆miR335E, was constructed by excising out the miR-335 sequences with NsiI and BglII, and re-ligating back together. Figure 5.3 shows the final vector structures and the primers used in construction of these plasmids are given in Table 2.1.  2.15 Mouse primary embryonic fibroblast (PEF) isolation, genotyping, and cell line conditions Two E14.5 litters were isolated from crosses between a CD-1 female and a Mest  +/KO  male to yield WT and Mest+/KO E14.5 embryos. Each embryo was placed in its  own dish and separated from its placenta and surrounding membranes. The head/brain and internal organs were cut away and the remainder was washed with fresh PBS to remove as much blood as possible. The remaining tissue was then finely minced with razor blades until they became “pipettable.” The suspension was then placed in several mls of trypsin-EDTA in a falcon tube and incubated with 5 mm beads with shaking at 37°C for 15 mins to1 hr until the resulting cell suspension appeared not viscous. DNase was also added to the suspension (100 units per ml – Sigma). Each mixture was then subject to low-speed centrifugation for 5 minutes and the supernatant carefully taken off. Approximately 10 mls of PEF medium (DMEM- 10% FCS, 1/100 (v/v) L-glutamine, 1/100 pen/strep) was added to each resulting cell pellet and the mixtures were plated out per gelatin-coated 10 cm dish and incubated in a humidified atmosphere of 5% CO 2 at 37°C. The medium was changed the following day and the cells from each embryo were passaged to an extra-large gelatin-coated 15 cm plate upon confluency. The cells were then frozen down into 3 vials when they reached confluency. Twenty-four fibroblast cultures were generated from the 2 litters. Genotyping of the MestKO allele was performed with primers 702A and IRES1, the same used above for genotyping embryos. For sexing, primers ZFy1a-b were used that detects a gene on the Y chromosome. Please see Table 2.1 for their sequences. After genotyping for the MestKO allele and Y  33  chromosome, the 23 cultures were found to consist of 5 WT XX, 3 WT XY, 11 Mest+/KO XX, and 4 Mest+/KO XY (data not shown). WT PEFs 1.8 and Mest+/KO PEFs 1.10 were used for the RNA seq experiments and additional WT PEFs 1.2 and 1.6 and Mest+/KO PEFs 1.1 and 1.7 were used in the qRT-PCR experiments.  2.16 Electroporation and flow cytometry Frozen PEFs (passage 2 or 3) from either the 1.8 WT PEFs or 1.10 Mest+/KO line were thawed onto an extra-large gelatin-coated plate and cultured in PEF medium (DMEM- 10% FCS, 1/100 (v/v) L-glutamine, 1/100 pen/strep) and incubated in a humidified atmosphere of 5% CO2 at 37°C. After 6 days and two extra passages, the resulting ~5-6x106 million PEFs from 9 extra-large plates were harvested between 7090% confluency and resuspended in PBS and divided into 6 cuvettes. 80-100 µg of pCmir335E or pC∆mir335E was added to each cuvette containing the divided PEFs. The 4mm cuvette containing the PEFs and plasmid DNA was electroporated using a Bio-Rad Gene Pulser Xcell set at a 15 millisecond time constant and 250 volts (177). Each electroporated sample from each cuvette was put back onto a large gelatin-coated plates containing PEF medium. The following day the electroporated PEFs were examined for GFP expression with a Leica DMI6000B inverted fluorescent microscope. To prepare the PEFs for flow cytometry, they were harvested 24 hours postelectroporation and resuspended in FACs buffer (PBS with 2mM EDTA, 2% FCS (Gibco), and 2mg/ml propidium iodide (PI, Gibco)), and passed through a cell sieve (Fisher). The resulting 5x105-2x106 surviving cells were sorted with a BDFACS Aria IIu by Justin Wong or Andy Johnson from the UBC Flow Cytometry Facility. The electroporation frequency ranged from 1-9.5%, as measured by the frequency of GFP+ cells. Three samples were obtained; 1.8 WT + pCmiR-335E (over-expression), 1.10 Mest+/KO + pC∆miR-335E (under-expression), and WT + pC∆miR-335E (comparative control).  2.17 Whole transcriptome sequencing (RNA-seq) libraries  RNA was isolated from GFP+-sorted PEFs using Trizol reagent according to the manufacturer’s instructions (Invitrogen). The WT, Mest+/KO (miR-335 under-expression), and WT+miR (miR-335 over-expression) strand-specific libraries were constructed from 34  mRNA as described in Morin et al. from approximately 2 µg of DNaseI-treated total RNA (178). Paired-end sequencing was performed on an Illumina Genome Analyzeriix, according to the recommended protocol (Illumina Inc., Hayward, CA). Briefly, BWA 0.5.7 (Burrows-Wheeler Aligner) was used to map reads to a whole-transcriptome shotgun sequencing resource that included genome plus junctions obtained from the v65 Ensembl annotations for a total of 22, 318 genes (179; 180). RPKM normalization was carried out by custom scripts by Misha Bilenky from the Genome Sciences Centre in Vancouver, BC. For determination of significant differentially expressed genes, the selection criteria was stringent and a significant fold change was defined as at least a 4fold change with a small error estimate on the fold change, plus genes were required to be expressed at a certain significant level (0.025 RPKM). Fold changes were determined by dividing the RPKM values for each gene in each sample over each other. The error estimate on the fold change was performed by using number of reads mapping to each gene and then using an assumption of Poisson distribution.  Table 2.1 Primer sequences for genotyping, RT-PCR, qRT-PCR, probe construction, and 5'RACE. Primer name  Sequence (5' - 3')  Reaction  Reference  GTTATAGCCAGTCGAGTAAGTATGC TCCCAGTGGATCACCTGAGC AGACCGCGAAGAGTTTGTCCTC GAATGTATTATGTGAGTGTGTACATGA AGGTGCCTGCACATATACAGG TGCTTTGAAGATGGCAGATG ACCCCACGCAATTCCTAAG GACTAGACATGTCTTAACATCTGTCC CCTATTGCATGGACAGCAGCTTATG  Mest WT or KO Mest WT Mest KO D6Mit269 D6Mit269 D6Mit180 D6Mit180 ZFy1 ZFy1  (105) (105) (105) (178) (178) (178) (178) Louis Lefebvre's lab Louis Lefebvre's lab  Ppia Ppia Mest ex1-2 Mest ex1-2 Mest ex10-12 Mest ex10-12 (+) strand RT (-) strand RT MestXL/Mit1  (180) (180) this study this study this study this study this study this study this study (Fig.3.10, #15)  Genotyping 702A 702C IRES1 D6Mit269A D6Mit269B D6Mit180A D6Mit180B ZFy1a ZFy1b  RT-PCR and qRT-PCR Ppia F Ppia R Mest 5'U-1F Mest 2Ra Mest e10/11-12 F Mest E12 R5 Mit1 RT F5 Mit1 RT R5 Mit1 F6  CGCGTCTCCTTCGAGCTGTTTG TGTAAAGTCACCACCCTGGCACAT TGAGAGAGTGGTGGGTCCAAGTAG CTTCCATGAGTGCAGAGCAGG TGTCCATCCCCATTCATTTT GAGTTCCAGCTGCCTGATTC TCGTTGTTTTGTGTATTTTGGT TCCAGCCTTAATGCAGCTCT GTGTGGGTCCCCTGAGTTATCTA  35  Primer name  Sequence (5' - 3')  Reaction  Reference  Mit1 R6 XIST qF1 XIST qR1 mXist qF1 mXist qR1 Lit1r4F Lit1r4R BB qF2 BB qR2 BY qF2 BY qR2 Copg2 E5-F2 Copg2 E1-R1 Copg2 E17-F2 Copg2 E19-R2 Klf14 F4 Klf14 R4 Klf14 F6 Klf14 R6 Mest ex12 3’ F2 BU9 R1 BU9 F1 BU6 R1 5’ gap R1 BY F3 No EST R1 CF F1 AF R1 AF F1 AK38 R2 AK95 F2 Acan qF1 Acan qR1 Cd34 qF1 Cd34 qR1 Cd200 qF1 Cd200 qR1 Ednrb qF1 Ednrb qR1 Krt18 qF1 Krt18 qR1 Prg4 qF1 Prg4 qR1 Thbs4 qF1 Thbs4 qR1 Cmah qF1 Cmah qR1 Derl3 qF1 Derl3 qR1 Rb1 F1 Rb1 R1  CTTTCATTGCTTCAGAATTCTTCCA TTCAGATTGTGGAGGAAAAGTG AGGTTTAGCTCATGCAATGCA GCTTCTGCGTGATACGGCTAT AGCTAGCGCAGCGCAATT CTCAGTTCCACGATACCCTTCC CTTACAGAAGCAGGGGTGGTCT CACACTCCCATCATTGCTGAGA CAGTATCCTCTTCTGTGCACTGTTC TGTCAGCTCAGCAGCAATTC AATCATGGCTCCCTGGAAAT TCCTTGATGGTGAGGTAGCA TCCAATCCTTTCCAGCATTT CATTGGAGCCTTCAGAAAAACC GCTTCACACAGCGAACGAAAT AAGCGACATCAGTGCTCCTT CTCGGTCTAATGACCCTGGA ACGTACCGAAGGAGGCAGATTA GTCGAGCCAATCACAGGAGAA GCAACTTTGAAACTGGAATGAA GGTTTTGGTTTTGCTTCTGC GCAGAAGCAAAACCAAAACC GCTGTCTGCCATTCATTTTG AAGTTCTGCTCTTGCCCTGA AGGGCAGAAGACAAAAAGCA CTGGCCTTTGTTCCCCTATT GCAAGCAAGAGGATGAAAGG CACAGCTCTCACCCTTCCTC GACAGTGAGCAGCAACTGGA GCTGGGTTACTTACTGACTTGC GGGTTGGGTAAGGGGAGTAA AGCTGCCCTTCACGTGTAAAA TGCAGGTGATTCGAGGCTCTT AAACTGCCTTCTGCACATTCAAC CAACTTTTTTGGCAGCTCCTCTAG CCCAGGTCCTCGGTGTTTG CATGGCACTGCATTGCTCTAC TCTGGCTCTGGGAGACCTACTG CTTACACATCTCAGCTCCAAATGG CATCCGCGCCCAGTATG GGCAGACTTGGTGGTGACAAC CAATCCCATGCTTTCAGATGAG ATGACCTCGAAATGCAACTAATGT AGTGCAAATACCATCCCTGCTATC ACCATGGGCCCTGTGAAAC TGGAATATGGCAACAGGTAGACAA CTGATGCACCTCCTGCGAAA CCAGGTCTGGAGGCTCATCA AAAACAGGCTGCCCAGGAAT GCAACCCCCCCAAACCA CTTGCTCTCATTCATTCTCTGCTTT  MestXL/Mit1 XIST XIST Xist Xist Kcnq1ot1 Kcnq1ot1 BB BB BY BY Copg2 ex1-5 Copg2 ex1-5 Copg2 ex17-19 Copg2 ex17-19 Klf14 F4-R4 Klf14 F4-R4 Klf14 qPCR Klf14 qPCR Mest ex12 BU922167 EST BU922167 EST BU614826 EST MestXL RT BY247745 EST  this study (Fig.3.10, #14) this study this study Carolyn Brown's lab Carolyn Brown's lab (49) (49) this study (Fig.3.10, #5) this study this study this study this study this study this study this study this study this study this study this study this study (Fig-3.10, #1) this study (Fig-3.10, #2) this study (Fig.3.10, #3) this study (Fig.3.10, #4) this study (Fig.3.10, #6) this study (Fig.3.10, #7) this study (Fig.3.10, #8) this study (Fig.3.10, #9) this study (Fig.3.10, #10) this study (Fig.3.10, #11) this study (Fig.3.10, #12) this study (Fig.3.10, #13) this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study  CF745870 EST AF217545 EST AF217545 EST  Acan Acan Cd34 Cd34 Cd200 Cd200 Ednrb Ednrb Krt18 Krt18 Prg4 Prg4 Thbs4 Thbs4 Cmah Cmah Derl3 Derl3 Rb1 Rb1  36  Primer name  Sequence (5' - 3')  Reaction  Reference  ISH and Northern probes Mest E1-F1 Mest 2Ra Mest 3Fa 0495K R Mest in11-F2 Mest in11-R1 Neo-f Neo-r Cop E24-F Cop 3 NO R2 Cop in23-F Cop in23-R Cop E14-F2 Cop E20-R2 Cop in18-F Cop E20 ISH R BB692936 F BB692936 R BY247745 F BY247745 R Mit1 F2 Mit1 R2 P1 E12 NO R1 GSP1-IRES R2 GSP2-IRES1 GSP3-Mest 2Ra RACE1 RACE2  Mest 5' UTR ACATCCCGGTGCTTCTTCT Mest 5' UTR CTTCCATGAGTGCAGAGCAGG Mest ex3-8 CTGTCGGTGTGGTCGGAAGC Mest ex3-8 GCCGTCATTGTTGCGAAT Mest intron 11 CCCTGACACTTCTTGGCACT Mest intron 11 CAGCATGGAGGTGAACTTGA neo ORF GGGTGGAGAGGCTATTCGGCTAT GAAGAACTCGTCAAGAAGGCGATAGAA neo ORF Copg2 3'UTR ACCGAGGCTGAGTGAGTGAG Copg2 3'UTR ACCATGCAGGTGACTGTGAG Copg2 in23 CTGAGAGAAAAAGCGAGGAAAA Copg2 in23 TGCTGCTTCTGTCATCCATT Copg2 E14-20 GGCCCTAGAACTCCTGTTCC Copg2 E14-20 TTCATCTGGCAATCGAACAA Copg2 in18-ex20 CATCCACTGCTTTTATTCATTTTG Copg2 in18-ex20 TTCATCTGGCAATCGAACAA AGTGGACAGAGCATGCACAC BB692936 EST TGGATCCACTCAGTCCACAA BB692936 EST AGCAGTGACCTGCTCACAAA BY247745 EST GCTCCTATCTTTGCCACCAA BY247745 EST Mit1 AGGAGGATGGGCTCCATTAC Mit1 CTGAACCCATTGAAAACTCCA TCTTCTTAGCAAGGGCCAAC linear amplification KO Mest cDNA ACCGGCCTTATTCCAAGCG KO Mest , 5’ RACE AGACCGCGAAGAGTTTGTCCTC Mest ex2 CTTCCATGAGTGCAGAGCAGG GGCCACGCGTCGACTAGTACTTTTTTT anchor + p(dT) TTTTTTTTTTVN GGCCACGCGTCGACTAGTAC anchor  this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study this study (105) this study Invitrogen Invitrogen  Plasmid construction CCGGGATCCTGCTCTGCACTCATGGAA Mest intron 2 G CGCCTCGAGCTTCCGACCACACCGAC Mest intron 2 Mest ex3 R AG pCmir335E EGFP-GCGCTCGAGAACCATGGTGAGCAAGG pCmir335E EGFP F/XhoI GC pCmir335E EGFP-GGAAGCGGCCGCTTTACTTGTACAGCT pCmir335E EGFP R/NotI CGT Mest ex2 F  this study this study this study this study  37  Table 2.2 Strand-specific RNA probes used for Northern and ISH analyses and the restriction enzymes and RNA polymerases used to generate the probes from their respective plasmids.  Probe Mest Ex1-2 Mest Ex3-8 Mest Ex12 Mest intron 11 Copg2 E14-20 Copg2 intron 23 Copg2 3'UTR Mit1 BB692936 EST BY247745 EST Copg2 in18-e20 Klf14 Neo ORF  AS S AS S AS S AS S AS S AS S S AS AS S AS S AS S AS S AS S AS S  Linearization  RNA polymerase  NotI SphI NcoI SacI BamHI XbaI NotI NcoI NcoI NotI NcoI SacI NotI NcoI SacI NcoI NotI NcoI NcoI NotI NcoI NotI NotI SacII SacII NotI  T7 SP6 SP6 T7 T7 SP6 T7 SP6 SP6 T7 SP6 T7 T7 SP6 T7 SP6 T7 SP6 SP6 T7 SP6 T7 T7 SP6 SP6 T7  Probe length (bp) 370 550 1300 720 750 370 640 620 430 430 1300 660 700  38  Chapter 3: The Mest locus produces larger isoforms, MestXL, through alternative polyadenylation1 3.1 Introduction  The present annotation of the Mest locus on the Ensembl browser (NCBIM37) depicts Mest as a protein-coding gene whose transcription terminates within a 1.4-kb exon. Intriguingly, it also presents evidence for two clusters of expressed sequence tags (ESTs) that are transcribed downstream of this terminal exon and within close proximity to Mest from the (+) strand, like Mest, that extend from the end of exon 12 of Mest all the way into intron 20 of the neighboring antisense gene Copg2 (Figure 3.1). The two clusters flank a genomic gap in the current mouse genome sequence that is set arbitrarily to 50 kb and not supported by physical data. Most of the ESTs originate from embryonic or neural tissue suggesting they are likely tissue-specific. Interestingly, imprinted transcription from the paternal allele has previously been described from each cluster (96). The transcripts closest to Mest, referred to as Copg2AS, overlaps the 3’UTR of Mest and the last terminal exon of Copg2 suggesting they could be part of a larger uncharacterized Mest transcript. The transcripts located on the other side of the genomic gap and in intron 20 of Copg2 are referred to as Mit1 (Mest-linked transcript 1). They encompass a heterogeneous RNA population covering >5 kb of intron 20 sequences with no considerable open reading frame (96). Given that Copg2AS and Mit1 transcripts are paternally expressed, “promoterless”, and transcribed downstream of Mest, we hypothesized that they and other ESTs in the region constitute parts of larger Mest transcriptional units. This hypothesis was primarily tested by investigating the transcription at the 3’ end of Mest and of downstream transcripts (i.e. Mit1), or lack thereof, in Mest+/KO mice by several methods. Since the MestKO allele contains an inserted cassette that introduces a new polyadenylation signal, it was predicted that synthesis of transcripts downstream of the mutant locus on the paternal allele would be abolished if they were indeed generated from the same locus. Further experiments investigated the tissue-specific expression and regulation of the larger Mest isoforms, referred to as MestXL. It is predicted that the 1  Parts and/or versions of Figures 3.1-3.11 and A.1, A.3, and A.4 have been published in MacIsaac et al. Tissue-specific alternative polyadenylation at the imprinted gene Mest regulates allelic usage at Copg2. Nucleic Acids Research 40(4):1523-1535 (2012).  39  Figure 3.1 EST evidence for longer isoforms of Mest. Genomic organization of the Mest-Copg2 region extracted from Ensembl as a 100 kb window. Note that the full-length Mest transcript (isoform 1A) is not annotated; the Ensembl transcript Mest-201 corresponds to isoform 1C (Fig. 3.2), with a final exon of only 393 bp, 1 kb shorter than the exon 12 isoform 1A. The data shows the location of the arbitrarily sized gap within intron 20 of Copg2, flanked by clusters of ESTs transcribed from the (+) strand, including the Copg2AS and Mit1 transcripts. Spliced Mest transcripts with 3’UTRs extending 5’ (MestXLa) and 3’ (MestXLb) of the gap are shown below the map. The size of the MestXLb variant assumes a negligible size for the gap in the mouse genome. 40  synthesis of MestXL is regulated by tissue-specific alternative polyadenylation in neural tissues. 3.2 Results  3.2.1 Longer variants of Mest mRNA: the MestXL isoforms  The current annotation of the Mest locus presents a complex production of different isoforms of the mRNA (see Figure 3.1). First, the available EST data suggest variability in promoter usage, with three minor isoforms (#2-4) transcribed from upstream promoters, relative to the main isoform 1, which initiates at the differentially methylated CpG island defining exon 1 (Figure 3.2A and B). Second, by mapping SAGE tags from the Mouse Atlas of Gene Expression for the Mest mRNA, variability in the size of the 3‘UTR of exon 12 leading to the production of three isoforms (A to C; Figure 3.2C) was identified (171). This analysis showed that most Mest transcripts are processed by transcriptional termination and polyadenylation at the most downstream termination signal (AATAAA), which yield a terminal exon of 1406 bp. The production of these isoforms in adult tissues and during development was assessed by northern blotting using a random-primed dsDNA probe against the Mest coding region (Figure 3.3). The results show that the main 2.5 kb isoform of Mest (isoform 1A, band IV) is broadly expressed in the samples analyzed (see Figure 3.2D for molecular weights of isoforms). Although this analysis cannot resolve the isoforms 1 to 4 generated by alternative promoter usage, most ESTs support formation of isoform 1 in adult and embryonic tissues (96.6% of all ESTs). The northern blot analysis also identified lower molecular weight RNA species consistent with the structures of the isoforms B (band V) and C (band VI) produced by alternative polyadenylation within exon 12 and supported by the LongSAGE data (Figure 3.2C). Importantly, we also detected polyadenylated transcripts larger than the fully spliced Mest mRNA, in both adult tissues and in the embryos. In adult tissues, the mature Mest mRNA of 2.5 kb is detected at high levels in heart, brain and lung. The larger molecular weight species (bands I to III) were mostly detected in the brain where they represent close to 30% of the transcripts (Figure 3.3A). These larger transcripts represent variants greater than 10.5 kb (band I) and ~6.5 kb of transcribed sequences (bands II and III). During development, the main Mest transcript of 2.5 kb is very 41  Figure 3.2 Main RNA isoforms at the Mest locus. A) Genomic structure of the Mest gene presenting a scaled diagram of the intron-exon arrangement of the main isoform 1A. B) Isoforms 1 to 4 are produced by alternative promoter usage. The alternative exons 1a, 1b2, and 1c are respectively, 4, 395 bp, 5, 311 bp and ~14 kb upstream of exon 1. The structure and production of each isoform is supported by EST data. The number of annotated ESTs (UCSC Browser, Aug 2010) is given for each isoform (right). C) Isoforms A, B, and C are produced by alternative polyadenylation within exon 12 of Mest, which contains three transcriptional termination signals (AAUAAA, asterisks). The figure shows the structure of each alternative exon 12, its predicted size assuming cleavage and polyadenylation 15 nt downstream of the termination signal, and its overall representation in SAGE tags libraries (Mouse Atlas; http://www.mouseatlas.org/). D) Expected size in nucleotides of the 12 possible isoforms of Mest (mature transcripts). The main isoform of Mest (1A) is in bold. 42  Figure 3.3 Northern blot analysis of Mest expression. A dsDNA probe of the entire Mest open reading frame (spanning exons 1–12) was hybridized on mouse northern blots containing 2 µg of poly(A)+ mRNA per lane. (A) RNA samples from adult tissues. (B) RNA samples from whole mouse embryos at the indicated developmental stages. The different Mest isoforms detected were labeled I to VI. Band IV corresponds to the main isoform A of Mest (see Figure 3.2).  43  abundant at all stages analyzed (see Figure 3.3B, band IV). In addition to the shorter isoforms with alternative polyadenylation within exon 12 (band VI, ~1.7 kb; band V, ~2.3 kb) embryos also show the two higher molecular forms at ~ 6.5 kb (bands II and III) at all stages analyzed. Although the larger variants could represent partially spliced forms of Mest, the fact that they are polyadenylated suggested that they might in fact represent longer forms of Mest, with an extended terminal exon. Taking into account the current annotation of the genomic browser (Build37/mm9) (see Figure 3.1), spliced Mest transcripts with exon 12 variants extending up to the genomic gap in intron 20 of Copg2 would make transcripts of a predicted size of ~6.8 kb. Those extending beyond the gap, assuming a negligible size for this gap, and up to exon 20 of Copg2 would span ~13 kb (see Figure 3.1). Notably, these predicted sizes correlate nicely with the sizes shown by northern blot (see Figure 3.3) if in fact the gap is insignificant. To obtain further experimental evidence for these longer variants, referred to as MestXL, data from SAGE tags libraries from the Mouse Atlas of Gene Expression Project were analyzed (171). They provide the location and abundance of expressed sequence tags located close to terminal or internal pA stretches. Six main sequence tags expressed from the (+) strand of the genome in the region of overlap between Mest and Copg2 were identified (Figure 3.4). The first four tags (tags 1 to 4) map within the exon 12 of Mest, as described above, and represent expression of the short isoforms of Mest. They are expressed at low levels in adult tissues (notably in brain, pituitary and adrenal glands) but show high levels of expressions during development (Figure 3.4B). Their expression patterns in 136 different adult and embryonic LongSAGE libraries show high levels of correlation (Figure 3.4D) (169). The SAGE tags 5 and 6 both map within intron 20 of Copg2, beyond the gap in the mouse genome sequence, relative to exon 12 of Mest (Figure 3.4A), and within the Mit1 cluster. These tags are representative of longer isoforms of Mest, the MestXL variants. Expression of tags 5 and 6 show the highest degree of correlation amongst the tags analyzed in the 136 LongSAGE libraries (Figure 3.4D). They show moderate to low levels of expression in adult and embryonic tissues expressing the short Mest isoform, but they are also sometimes expressed at higher levels than tags 1 to 4 (for instance adult visual cortex and lung, Figure 3.4B). The analysis of amplified libraries (SAGELite) showed that tags 5 and 6 are often expressed at equal or greater levels than the short isoforms in several regions of the adult brain (Figure 3.4C) (170). During embryogenesis 44  Figure 3.4 Analysis of SAGE tags for the Mest and MestXL transcripts. Please see the following page for figure legend.  45  Figure 3.4 Analysis of SAGE tags for the Mest and MestXL transcripts. (A) Schematic representation of the Mest-Copg2 region, drawn to scale. The position of the gap in the genome sequence is indicated (below) but omitted in the estimation of the size of intron 20 of Copg2. Above the (+) strand, on which terminal Mest exons are represented by black rectangles, six SAGE tags are positioned (arrows). Tags 1 to 4 map to exon 12 and represent short isoforms of Mest. Tags 5 and 6, located between the gap and exon 20 of Copg2 represent the larger MestXL isoforms. Probes used in northern and ISH analyses are labeled d to k (gray lines). (B) Gene expression profiles for SAGE tags 1 to 6 in adult and embryonic LongSAGE (un-amplified) libraries from the given tissues, presenting the tag counts according to the shading scale defined below. Statistics for each tag is from 136 different LongSAGE libraries. Tags counts are normalized to 100,000 tags. (C) Expression levels of tags 1-6 in adult and embryonic SAGELite (amplified) libraries from the given adult tissues. For the embryos, whole head RNA was analyzed from the E8.5 to E11.5, and whole brain from E12.5 to 15.5. Statistics for each is from 77 different libraries. Tags counts are normalized to 100,000 tags. (D) Pearson’s correlation for the six SAGE tags in the 136 LongSAGE libraries.  46  there appears to be a gradual shift from Mest to MestXL expression in the developing brain, the latter of which is first detected at developmental stage ~E12 (see Figure 3.4C). Together with the northern blot analysis, these results suggest the existence of longer isoforms of Mest, and that the production of MestXL transcripts is under developmental and tissue-specific regulation. Lastly, ESTs from the (+) strand at Mest (from UniGene entry Mm. 335639) and within the two EST clusters on either side of the gap (Table A.1 and A.2) were analyzed. Whereas only 35% of Mest ESTs are from libraries generated from the central nervous system (CNS) tissues, most ESTs from the 5’ and 3’ clusters are from the CNS (77 and 84%, respectively; Figure A.1). Together with the northern blot analysis, these results suggest the existence of longer isoforms of Mest, and that the production of MestXL transcripts is under developmental and tissue-specific regulation. 3.2.2 Loss of MestXL transcripts from the mutant MestKO allele The Mest targeted mutation Mesttm1Lef (or MestKO) is based on a promoter-trap construct in which exons 3 to 8 of Mest were replaced by a splice-acceptor-IRES-βgeopA cassette (see Figure 1.3B in Introduction) (106). Transcription from this mutant allele is imprinted and expression of the βgeo reporter faithfully recapitulates Mest expression in paternal heterozygous Mest+/KO embryos (106). Based on the structure of the mutant allele, transcriptional termination was predicted to occur upstream of the exon 9 sequences and thus to produce truncated Mest transcripts. Whether MestKO transcripts are indeed truncated was first assessed in paternal heterozygous Mest+/KO embryos using quantitative RT-PCR (qRT-PCR) (Figure 3.5). The PCR reactions encompassed Mest exons upstream (1 to 2) and downstream (10 to 12) of the targeted mutation and further downstream into intron 20 of Copg2 (Mit1). Since the SAGE data suggest that the formation of MestXL is under tissue-specific regulation, the analysis was performed on RNA samples purified from four different embryonic tissues dissected at E14.5 (Figure 3.5). From the analysis upstream of the insertion in the MestKO allele (exons 1-2 PCR), the MestKO allele was found to be produced at approximately 50% of the wild-type levels (Figure 3.5, left panel). The analysis of transcripts extending downstream of the insertional mutation (exons 10-12 PCR) essentially assessed the amount of readthrough beyond the inserted polyadenylation signal of the MestKO allele. Very little to no 47  Figure 3.5 qRT-PCR analysis of Mest, the MestKO allele, and MestXL (Mit1) RNA levels in various embryonic tissues at E14.5. The white and gray bars represent wild-type (Mest+/+) and Mest+/KO genotypes, respectively. A significant difference between genotypes was detected in all tissues for the Mest ex1-2 and MestXL reactions, indicated by the asterisks (p<0.005; two-tailed unpaired t-test). The Mest ex10-12 reaction did not detect expression from the MestKO allele in Mest+/KO tissues. mRNA levels were normalized to Ppia and the error bars indicate the standard deviation from three biological replicates, assayed in triplicates, from the tissues analyzed. The tissues analyzed are: MHB, mid/hind brain; SC, spinal cord; H, heart; L, liver.  48  transcription was detected downstream of the MestKO pA in the mutants. This suggested that there was efficient termination within the insertion of the mutant allele (see Figure 3.5, middle panel). Since the MestKO allele is not transcribed beyond exon 8, we were able to assess directly the effect of this truncation on the production of MestXL transcripts. As described for the SAGE tags 5 and 6, the MestXL transcripts are detected in wild-type embryos at higher levels in the neuronal tissues analyzed (mid/hind brain (MHB) and spinal cord (SC), see Figure 3.5, right panel). In the MHB and SC, we observed a significant 6-11-fold decrease in MestXL levels from the mutant paternal allele (p=0.0005 and p=0.001, respectively). The significant decrease we observed for MestXL (Mit1) by qRT-PCR in mutant samples provides evidence that transcription mostly terminates at the introduced pA at the MestKO allele and that MestXL isoforms are also affected by this mutation. Most importantly, this shows that Mit1 constitutes part of the longer MestXL transcripts. To provide further support that these transcripts are produced from the Mest promoter, poly(A)+ RNA from wild-type and Mest+/KO E14.5 embryos and adult brain tissue was analyzed by Northern blot (Figures 3.6B and 3.7, respectively). Since only the paternal allele of Mest is active for transcription, all Mest transcripts are from the mutant MestKO allele in Mest+/KO heterozygotes. Several strand-specific probes were used that hybridize to different regions of Mest, to the mutant allele (neo probe), and to three different locations along the MestXL isoforms (BB692936 EST, BY247745 EST, and Mit1). Probe locations are presented in Figures 3.4A and 3.6A. Within the Mest locus, probes against exons 1-2 (probe a), exons 3-8 (probe b) and exon 12 (probe d) detected the Mest isoforms in WT E14.5 head poly(A)+ RNA (Figure 3.6B; see Figure A.2 for pictures of the RNA gels stained with ethidium bromide). These blots are over-exposed so that we could observe the weaker higher molecular weight signals. The MestXL isoforms were detected by probes against exons 3-8 and exon 12, but not by the exons 1-2 probe. Northern blots on poly(A)+ adult brain RNA were also performed to increase the chances of detecting higher molecular forms with the exon 1-2 probe, since the previous northern (see Figure 3.3) showed a ratio that is approximately 3:1 for Mest and MestXL. This northern on adult brain RNA, however, showed similar results with the same exons 1-2 and exon 12 probes (Figure 3.7). The discrepancy between the probes is perhaps due to the weaker signal obtained with this short probe, or to an exclusion of exons 1 and 2 from the larger isoforms, possibly due to the processing of miR-335 located in intron 2 of Mest or alternative promoter usage. 49  Figure 3.6 Loss of MestXL expression from the mutant MestKO allele. (A) Schematic diagrams of the Mest and MestKO alleles. The exons are represented by black rectangles. The MestKO allele consists of a splice acceptor-IRES-lacZ/neo fusionpoly(A) cassette (white rectangles) replacing Mest exons 3-8, and part of exon 9. The positions of the RNA probes a to e are shown above the alleles (gray lines). (B) Strandspecific northern blot analyses of Mest, MestKO, and MestXL expression. Various DIGlabeled RNA probes were hybridized to mouse northern blots containing 1.5 µg of poly(A)+ RNA purified from E14.5 WT and Mest+/KO embryo heads. See Figures 3.4A and 3.6A for probe positions, identified by the letters given below the blots. BB and BY are (+) strand ESTs (BB692936 and BY247745, probes h and i, Figure 3.4A) mapping to 5’ and 3’ of the gap, respectively. The position of molecular weight marker bands in kb (MW) is indicated on the left sides of each blot.  50  Figure 3.7 Strand-specific northern blot analyses of Mest, MestKO, and MestXL expression in adult brain tissue. Two antisense Mest DIG-labeled RNA probes were hybridized to mouse northern blots containing 2 µg of poly(A)+ RNA purified from wild-type and Mest+/KO adult brain. The molecular weights are indicated on the left sides of each blot. Mest (isoform 1A) is detected by both probes at 2.5 kb, whereas only the exon 12 probe (2nd panel) detects higher molecular weight transcripts above 5 kb, which are absent in the mutant sample.  51  Since exons 3-8 are deleted in the MestKO allele, the absence of signal with the exons 3-8 probe on RNA from Mest+/KO embryos confirms imprinting of the locus and that only the paternal allele is active for production of all the isoforms detected in the WT embryos. Premature transcriptional termination at the inserted poly(A) cassette of the MestKO allele is confirmed by the absence of signal with the exon 12 probe on RNA from Mest+/KO embryos as well as the detection of the expected 5-kb mutant mRNA using a probe against the neo marker. To note, when the blot probed with an exon 12 probe is extremely over-exposed a very faint band (> 7kb) in the Mest+/KO lane is detected which could be suggestive of a minute amount of read-thru or cryptic splicing. We did not obtain a good signal from the mutant transcript using the small exons 1-2 probe on embryonic RNA although the 5-kb mutant mRNA was detected with the same probe on adult brain RNA (see Figure 3.7), as well as with the neo probe on embryonic RNA (see Figure 3.6). The difference between the MestKO signals with these two probes on embryonic RNA suggests that the Mest promoter may not only be producing the expected mutant allele, but other isoforms as well, perhaps through alternative and cryptic splicing in embryonic RNA. For confirmation that the mutant transcript indeed initiates at exon 1 of Mest and is part of the MestKO allele, in the developing embryo, rapid amplification of cDNA end (5’-RACE) experiments were performed and exons 1 and 2 were established to be present at the 5’ end of the mutant transcripts (Figure 3.8). Therefore it is possible that the formation of stable mutant transcripts including exons 1 and 2 is affected by abnormal processing of the miRNA in intron 2 during development. The possibility that the higher molecular weight isoforms might represent the premRNA (10.5 kb) or partially spliced forms of Mest was also considered. To address this, we used a genomic probe against intron 11 (probe c). The overexposed signal presents a smear that ranges from 2.8 to 7 kb with no prominent bands detected (see Figure 3.6B). The absence of the higher molecular weight bands detected with Mest cDNA probes suggests that they represent processed Mest isoforms and not unspliced primary transcripts. The absence of signal on the mutant RNA sample provides further evidence for the absence of read-through from the inserted polyA cassette, a result which is also supported with qRT-PCR analysis (see Figure 3.5). To analyze the formation of the MestXL isoforms, we used three probes (BB692936 EST {h}, AS/BY247745 EST {i}, and AS/Mit1 AS {j}) from intron 20 of Copg2 (see Figure 3.4A for probe locations). These probes, which span >7.5 kb of genomic 52  Figure 3.8 5’ structure of the MestKO mRNA analyzed by 5’ RACE. (A) Predicted structure of the spliced MestKO transcript (truncated) presenting exons 1 and 2 of Mest, spliced onto the inserted IRES-βgeo cassette. The positions of 3 genespecific primers (GSP) used are shown above the mRNA. Two different 5’RACE products were obtained and sequenced, indicated by the white boxes. (B) Agarose gel electrophoresis of 5’RACE products, which are only detected in the mutant sample (Mest+/KO).  53  sequences, detect RNA bands of ~4.8 kb, ~6.0 kb, as well as higher molecular weight forms (see Figure 3.6B). Notably, the probe against the EST BB692936, located at the end of Copg2 intron 20 and 5’ of the gap (probe h), detects the higher molecular weight isoforms corresponding to bands I-III seen with the Mest cDNA probes. The probes against the ESTs BY247745 EST and Mit1, located 3’ of the gap, also detect bands of similar sizes. Both probes show distinct bands at ~4.8 kb and ~6.0 kb, although interestingly in opposite intensities, suggesting that these two probes probably detect the same isoforms. The composition of these transcripts remains unknown and whether they are part of processed isoforms or stable 3’ products on a quickly processed very large MestXL remains unknown. The complex results demonstrated by the northern blot analyses suggest that there is processing and probably splicing going on throughout this region, especially on the distal side of the gap where there appears to be numerous isoforms. Taken together, the northern and qRT-PCR analyses demonstrates that the Mest and MestXL transcripts are imprinted and expressed only from the paternal allele during development. In Mest+/KO embryos, Mest transcription terminates at the inserted polyA cassette of the mutant allele and production of the MestXL isoforms is concomitantly eliminated. Notably, these findings show that Mest+/KO mice are not only deficient for the MEST protein, but also for MestXL, the loss of which might contribute to the phenotypes observed in these mice.  3.2.3 MestXL transcripts are produced in the developing nervous system  The analysis of SAGE tags for Mest and MestXL suggests that formation of the longer transcripts is under tissue-specific regulation during development. To look at the developmental expression pattern of MestXL transcripts in more detail, in situ hybridization (ISH) experiments were performed on sagittal sections of E14.5 embryos, at a stage when both Mest and MestXL SAGE tags are detected in the developing brain (see Figure 3.4C). First, antisense Mest probes were used to look at expression of Mest in WT and Mest  +/KO  heterozygous embryos (Figure 3.9A-B). The expression of the neo marker  present in the β-geo cassette of the MestKO allele was also analyzed (Figure 3.9B, probe e). The results confirm the widespread mesodermal expression of Mest during  54  Figure 3.9 MestXL is expressed in the developing central nervous system of WT but not Mest+/KO embryos at E14.5. ISH was performed on 12 mm sagittal cryosections of WT (Mest+/+) (A and C) and Mest+/KO (B and D) E14.5 embryos with various DIG-labeled RNA probes. The probe locations are indicated in Figures 3.4A and 3.6A using the letter code presented below each panel. All probes hybridize to the (+) strand. Blue staining indicates gene expression and nuclear fast red was used as a counterstain. An antisense probe against the imprinted gene Cdkn1c was used as a control.  55  development as well as the previously described expression in neuronal tissues (102; 106). In the E14.5 spinal cord, immunofluorescence (IF) experiments showed that both Mest and the MestKO allele are broadly expressed in NCAM1-positive neurons (Figure 3.10). The ISH signal is seen for all the antisense Mest cDNA probes assayed, but not for the intron 11 probe (see Figure 3.6A for position of probes located in the Mest locus). For the exon 1-2 probe (probe a), we detected weaker expression for the mutant allele than for the WT transcript, which is consistent with our qRT-PCR data (see Figure 3.5). Together these results show that mRNA levels from the promoter-trap MestKO allele are lower than those detected in WT embryos as a consequence of the insertion of the βgeo cassette. Our results also show absence of transcription from the WT maternal allele in Mest+/KO heterozygotes, using the exon 3 to 8 probe, which hybridizes to coding sequences deleted in the MestKO allele. Instead, the paternal mutant allele was found to express the neo marker Mest-expressing tissues (Figure 3.9B, probe e). We evaluated the amount of read-through beyond the inserted polyadenylation signal of the MestKO allele by using an exon 12 probe. In agreement with our qRT-PCR and northern blot results, very little transcription was detected downstream of the MestKO pA. Given that our probes generally give signal in only one or the other background, with the exception of Mest ex1-2 AS, we also used an additional control probe, antisense to the Cdkn1c gene, to show similar RNA levels between the wild-type and Mest+/KO embryo sections. The expression pattern of MestXL was studied using RNA ISH probes complementary to the (+) or (-) strand that hybridizes downstream of the Mest isoforms 1A to ESTs annotated in the genome browsers. In addition one probe was generated for which there is no supporting EST data, the Copg2 intron 23 probe (probe g, see Figure 3.4A for probe positions). The BB692936 EST and Mit1 are located in intron 20 of Copg2 and transcribed from the (+) strand. The four probes designed to detect MestXL show a localized expression signal in the developing nervous system in WT embryos (see Figure 3.9C). More specifically, MestXL appears to be expressed in most parts of the developing mid/hind brain and in the spinal cord, but not in the developing forebrain. This CNS-specific transcription is absent in Mest+/KO embryos, confirming imprinting of MestXL and suggesting that MestXL represents larger Mest transcripts produced by tissue-specific alternative polyadenylation (see Figure 3.9D). A probe designed to detect transcription downstream of Mit1 (Copg2 in18-ex20, probe k) shows that transcription of MestXL terminates in intron 20, which is supported by SAGE tag 6 (see Figures 3.4A and 3.9A). 56  Figure 3.10 MEST and NCAM1 expression in the developing spinal cord. Frozen transverse sections of E14.5 wild-type embryos were analyzed for MEST (A,B) and NCAM1 (C,D) expression with anti-MEST and anti-NCAM1 antibodies in the developing spinal cord. Panels A and C show MEST and NCAM expression (red), respectively. The arrows in panel C show background primary-independent staining of the epithelial layer. Panels B and D add the DAPI counterstain (blue). Panel E presents lacZ staining on an E14.5 Mest+/KO transverse section, showing where the MestKO allele is expressed.  57  The reciprocal probe partners detecting transcription from the (-) strand, either the Copg2 mRNA or primary mRNA, were analyzed as well. For all probes assayed, minimal expression was observed for Copg2 mRNA or primary mRNA transcripts in both wildtype and MestKO heterozygous embryos with the exception of the Copg2 exons 1420 cDNA probe (Figure A.3). As observed for MestXL, Copg2 is mainly expressed in the developing neural tissues. Together with our northern data, the expression analysis by ISH suggests that MestXL is imprinted, produced mainly in the developing and adult CNS, and lost in Mest+/KO mutants. These conclusions are also supported by extensive RT-PCR experiments detecting brain-specific transcripts in the Mest-Copg2 region only in WT E14.5 brain samples (Figure 3.11). The analysis of expression profiling by highthroughput sequencing (RNA-seq) also revealed transcription extending from exon 12 of Mest to exon 20 of Copg2, specifically in developing and adult neuronal tissues (Figure A.4).  3.3 Discussion  The work presented in this chapter demonstrates that the Mest locus is under the regulation of tissue-specific alternative polyadenylation which results in the production of longer variants, the MestXL transcripts. Together with the fact that the Mest locus also consists of a variety of alternatively spliced transcripts coupled with alternative promoter usage, the Mest locus appears to be more complex than previously thought. The presence of the physical sequencing gap within the region of MestXL transcription (Intron 20 of Copg2) makes this locus even more multifaceted. As for the composition of the MestXL isoforms, the results from the northern blots with the Mest cDNA probes (see Figure 3.6, probes b and d) suggest MestXL consists of the mature Mest mRNA (or part of it) with an extended terminal exon. This conclusion is supported by the results using downstream probes proximal to the genomic gap (probe h). Using additional probes detecting transcripts distal to the gap (probes i and j), however, yielded similar results showing prominent bands at ~6 and also ~4.8 kb, where one would expect to observe only higher molecular bands that are >10 kb assuming the gap size is negligible. One possible explanation is that these transcripts are differentially processed isoforms, similar to the immunoglobulin heavy chain genes (181). Alternatively, they could be cleavage products due to instability of the MestXL 58  Figure 3.11 PCR amplification of cDNA fragments from MestXL in wild-type CNS. (A) Genomic structure of the Mest-Copg2 region presenting a scaled diagram of the intron-exon arrangement of the region, with the (+) strand shown at the top. The positions of ESTs from the (+) strand are shown by gray boxes and SAGE tags 4-6 representing pA signals are indicated by arrowheads. Primer pairs were designed to amplify across and within MestXL ESTs indicated by the white boxes under the MestCopg2 diagram. The numbers represent the primers used and the sequences can be found in Table 2.1 (B) RT-PCR was performed on wild-type (+/+) and Mest+/KO (+/KO) mid/hind brain (MHB) and liver (liv) cDNAs to provide support for MestXL transcription throughout this entire region. MestXL is only expressed in the wild-type MHB sample. So far, we have not been able to amplify across the gap on genomic DNA or cDNA.  59  messages. Given that a 4.8 kb band is not detected from the Mest probes, it is likely that this transcript does not include the Mest mRNA transcript and is comprised of sequences mostly on the distal side of the gap, perhaps as a cleavage product. The ~6 kb band detected with the distal probes is also probably comprised of sequences mostly on the distal side of the gap and not the same 6 kb band detected with probes on the proximal side of the gap. The highest molecular weight transcripts themselves that we detect could also be processed transcripts if the gap is very large. The gap might complicate and impede our understanding of how the MestXL transcripts are regulated and processed throughout this region, but our consistent results provide evidence that their expression relies on transcription from the Mest locus and are under the regulation of alternative polyadenylation. The mechanism of this CNS-specific alternative polyadenylation at Mest remains to be elucidated but is certainly not unique to Mest. Over half of mammalian genes contain multiple polyadenylation sites that lead to variants with different 3’UTRs or coding regions (182). Using SAGE data, several mRNAs expressed in neuronal tissues were shown to have progressively longer 3’UTRs throughout embryonic and postnatal developmental stages, a tissue-specific characteristic shared by MestXL (183). Examples of neuronal-specific larger variants of imprinted genes have also been documented. Dlk1, located in the Dlk1-Dio3 domain on mouse Chr 12, was found to undergo alternative polyadenylation and produce a transcript in brain that is 2.7 kb larger than the main transcript, referred to as “DAT” (Dlk1 associated transcript) (86). To date the function of this larger neural Dlk1 transcript remains unknown. As well, there are examples of neural ncRNAs that have different size isoforms. One gene in particular, Men β/ε, produces a ~20 kb isoform (β) in addition to a 3.2 kb isoform (ε), and both localize to the nucleus where they are required for the structural integrity of nuclear paraspeckles (184). A main role of 3'UTRs is to dictate the post-transcriptional fate of mRNAs. They contain cis elements and sequences important for translation efficiency, mRNA stability, and polyadenylation (182). Specifically in the CNS, distinct physiological functions for longer 3’UTRs have been reported to be involved in cellular processes such as translational control/repression and differential localization of transcripts (185-188). For example, the Bdnf (brain-derived neurotrophic factor) locus produces two pools of mRNA transcripts from the same promoter but with two different sizes of the 3’UTR via alternative polyadenylation that encode the same BDNF protein. Interestingly, in 60  hippocampal neurons, the long 3’UTR was shown to be responsible for targeting Bdnf mRNA into neuronal dendrites, whereas the short 3’UTR mRNA was restricted to the somata (187). In another report the transcripts with short 3’UTRs were shown to facilitate active translation to maintain basal levels of protein production, whereas the transcripts with long 3’UTRs suppressed translation until neuronal activation where they are rapidly activated, at least in a reporter system (188). Whether MestXL is involved in any physiological cellular processes involving the translational repression/activation or localization of the RNA remains to be elucidated. An additional way that 3’UTRs are used for post-transcriptional regulation of gene expression is through miRNA binding. MiRNAs are small regulatory ncRNAs (2125 bps) that bind to complementary sequences of target mRNAs, mainly in the 3’UTRs, resulting in translational repression or target degradation and gene silencing of the target gene (64). There are several hundred miRNAs in the mammalian genome. Obviously a gene with a longer 3’UTR will contain more sequence and ultimately more binding sites for miRNAs, which can alter the regulation of the gene. This has been shown with Bdnf (the same gene described above), which has two isoforms defined by the length of the 3’UTR as a result of alternative polyadenylation. MiR-206, a miRNA known to regulate differentiation of myoblasts, has three binding sites in the long 3’UTR of Bdnf, but only one in the short isoform. During myoblast differentiation, which involves the suppression of Bdnf, Miura et al. showed that the long Bdnf isoform is suppressed to a greater extent than the short isoform as a consequence of the extra binding sites (189). This suggested that the greater miR-206-mediated suppression of the longer isoform of Bdnf is important for myogenic differentiation and responsible for the expression pattern of Bdnf during this process. Since MestXL is mainly expressed in the CNS, it is tempting to speculate that MestXL harbors several miRNA binding sites important for physiological processes that require fine-tuning miRNA-mediated regulation in the nervous system. Numerous genes are under the regulation of alternative polyadenylation including genes that are epigenetically regulated such as imprinted genes. Intriguingly, one study by Wood et al. has also shown that the model can be the other way around, whereby alternative polyadenylation is regulated in an epigenetic manner by genomic imprinting (190). The H13 locus produces five isoforms that are defined by different polyA sites. Whereas H13a-c isoforms exhibit preferential maternal expression, H13d and H13e exhibit preferential paternal expression. This difference in allele-specific utilization of alternative polyA sites was found to be due to a maternally methylated gametic DMR 61  that is situated between the H13c and H13d polyA sites and controls the imprinting of another gene, Mcts2 (190). On the maternal allele, H13 preferentially utilizes the three polyA sites downstream of the DMR presumably because transcription of H13a-c is permitted through the methylated and transcriptionally inactive DMR (Mcts2 promoter). On the paternal allele, H13 preferentially generates isoforms (d and e) that terminate upstream of the unmethylated region perhaps due to blocking signals from the RNA polymerase II initiation complex on the active Mcts2 promoter/DMR (190). Although this alternative polyadenylation story is different than ours, it highlights another relationship whereby genomic imprinting and alternative mRNA processing are connected. Coming back to known imprinting mechanisms, MestXL is a large paternally expressed RNA in an imprinted domain. As described in the Introduction (Chapter 1), most imprinted genes reside within imprinting domains and are regulated by various mechanisms, one of them being long noncoding-RNA-mediated epigenetic silencing. The finding that the Mest locus produces much larger transcripts mainly in the CNS raises the possibility that paternally expressed MestXL is involved in regulating the maternally expressed genes Copg2 and Klf14 in these tissues. Of particular interest is the relationship between MestXL and Copg2, given that their 3’ ends overlap and were found to overlap by several more kilobases in the CNS with our description of the MestXL isoforms. Chapter 4 aims to determine the role of the Mest locus and specifically the MestXL transcripts in regulating the maternally expressed genes within this domain.  62  Chapter 4: Allelic usage at Copg2, but not at Klf14, is regulated by MestXL in the developing central nervous system2 4.1 Introduction  The murine imprinted domain on subproximal Chr 6, which shares synteny with human 7q32, contains the paternally expressed genes Mest and the tissue-specific MestXL isoforms we identified that correspond to the previously described RNAs Copg2AS and Mit1, and the maternally expressed genes Copg2 and Klf14 (13; 96). The mechanism of imprinting of the maternally expressed genes in this region is still currently unknown. The Mest locus itself is regulated by maternal 5mC-methylation, acquired during oogenesis, which is responsible for the direct silencing of the maternal allele (16; 101). These properties, therefore, establish Mest/MestXL as a primary imprinted gene, one that is directly regulated by epigenetic signals inherited from the gametes. Whereas epigenetic marks regulate their host gene themselves, they may also be involved in regulating the imprinting of other genes within the vicinity, so that the imprinting of those genes is secondary to the main primary imprint. Intriguingly, Mest may be the only primary imprinted gene within this small domain because it hosts the only gametic DMR identified thus far in the region. Together with the findings from Chapter 3 that the Mest locus produces the MestXL isoforms, the Mest locus is predicted to be the main locus involved in regulating the maternally expressed genes in the area, Copg2 and Klf14 (see Figure 1.2 for a schematic map of region) (13; 96). The notable feature regarding the Mest and Copg2 loci is their 3’ends overlap, which is conserved in humans (96; 112). Previous mapping showed that the Mest mRNA, transcribed on the (+) strand, overlaps the 3’ end of Copg2 by 52 base pairs on the (-) strand. We now know this overlap extends even further into Copg2 in certain tissues by several kilobases, with evidence of transcription of MestXL terminating before exon 20 on the (+) strand (see Chapter 3, Figure 3.9B, probe k). Whereas Mest and MestXL are strictly paternally expressed, Copg2, in contrast, shows a looser preferential maternally expressed pattern, at least in intra-subspecific F1 hybrid crosses of mice involving C57BL/6J (mostly Mus musculus domesticus) and KJR/Msf (Mus musculus  2  Parts and/or versions of Figures 4.1-4.8 and B.1 have been published in MacIsaac et al. Tissue-specific alternative polyadenylation at the imprinted gene Mest regulates allelic usage at Copg2. Nucleic Acids Research 40(4):1523-3515 (2012).  63  molossinus) (96). Intriguingly, imprinting and maternal allele-specific expression of Copg2 was only observed in tissues where MestXL is expressed, in the developing embryo and in adult brain (96). Furthermore, DNA methylation analyses at the Copg2 promoter, the next logical step to look for differential methylation that could mediate imprinting at this locus, yielded arguable results showing very few methylation differences between the parental alleles (119). These results would suggest that both alleles of Copg2 are actually transcribed. Since Copg2 appears to be imprinted in tissues where MestXL is expressed, we alternatively hypothesized that the preferential maternal expression observed at Copg2 was due to a short-range cis effect from MestXL transcription on the paternal allele. As a consequence Copg2 would therefore be considered a secondary imprinted gene due to this interaction with MestXL. To investigate this hypothesis the allele-specific expression of Copg2 was studied in Mest+/KO heterozygous animals, in which production of MestXL is abolished. Having established that the MestXL transcripts are mainly detected in the CNS in the developing mouse embryo (Chapter 3, See Figure 3.9), the parental allele contributions and expression of Copg2 were examined in a tissue-specific manner. Klf14 is a maternally expressed gene located approximately 60 kb downstream from the Copg2 promoter and transcribed from the (-) strand, like Copg2. A previous report demonstrated that the Klf14 promoter itself does not harbor any of the known imprinting hallmarks such as a DMR or differential histone modifications at the locus that would mediate its imprinting; yet maternal expression was reliant on maternally inherited DNA methylation (13). This conclusion came from the observation that offspring from Dnmt3a germ line conditional knockout females fail to express Klf14 (13). The interpretation of those results is that a maternally-inherited DNA methylation mark is required for Klf14 expression. Since the Mest DMR is the closest in proximity, it was hypothesized to be the DMR that controls Klf14 imprinting. There are two possible main mechanisms for which the maternally methylated Mest DMR might regulate this imprinting: via the production of large regulatory RNAs, MestXL, exerting a long-range cis effect or the establishment of an insulator boundary on the paternal allele. The results on embryos from Dnmt3a-null females can be explained by both mechanisms. If Mest becomes biallelically expressed due to the maternal allele not being methylated in oogenesis then consequently MestXL could become biallelically expressed as well, which could silence the maternal Klf14 allele. In fact, microarray data from E8.5 Dnmt3l-/+ embryos, also lacking maternal DNA methylation, show over-expression of Mest and of 64  MestXL consistent with biallelic expression (191) (2008; GEO series GSE8756, Figure B.1). On the other hand, the lack of DNA methylation might be grounds to set up an additional insulator boundary on the maternal allele that could lead to the silencing of maternal Klf14. We investigated the first option, again taking advantage of our Mest+/KO line that contains a truncated MestXL. This chapter aims to establish the relationship between production of MestXL and the two downstream maternally expressed imprinted genes transcribed on the opposite strand. If MestXL has a role in regulating Copg2 and Klf14 imprinting the prediction is that there will be a change in the imprinting pattern and an increase in overall expression levels in Mest+/KO animals since MestXL transcription is abolished on the paternal chromosome. This investigation will determine whether transcription of MestXL is a mechanism by which secondary imprinting of Copg2 and Klf14 occurs through short- and long -range regulation, respectively.  4.2 Results  4.2.1 Allelic usage at Copg2 is regulated by MestXL in the developing CNS  The newly identified MestXL isoforms of Mest overlap Copg2 on the paternal chromosome by several kilobases in the developing CNS, thus creating a very large putative transcriptional overlap which is hypothesized to inhibit paternal expression of Copg2. To investigate this hypothesis, allele-specific experiments were conducted in a tissue-specific manner to explore the imprinting pattern of Copg2. Experiments were performed on intra-subspecific F1 (M. m. castaneus X CD-1) E14.5 wild-type (MestC/+) and MestC/KO tissues: two where MestXL is expressed, the mid/hind brain (MHB) and spinal cord (SC), and two where MestXL is not expressed; the heart and liver. Initially, a normal RT-PCR/RFLP analysis was used to determine the parental allele contributions but the results indicated that Copg2 was biallelically expressed in all analyzed tissues, even in MestC/+ tissues (data not shown), making it hard to differentiate and quantify the results. For a more accurate approach, a hot-stop PCR/RFLP analysis was executed. Performing one final PCR cycle in the presence of a labeled nucleotide is advantageous because it circumvents the problem that can occur when heteroduplexes form between mixed genetic backgrounds. This in turn can lead to the RFLP becoming non-digestible and presenting inaccurate results (192). Our analysis relied on a SNP we identified in 65  exon 3 of Copg2 (rs36827909) that generates a BstUI site unique for the M. m. castaneus allele. By measuring the band densities of the maternal (CAST) and paternal (CD1) alleles we obtained quantitative measurements and calculated maternal to paternal (M/P) allele ratios for MestC/+ and MestC/KO samples purified from F1 hybrid embryos (Figure 4.1A and B). The MHB shows the highest Copg2 M/P ratio as the maternal allele is expressed 2.3 times more than the paternal allele in MestC/+ RNA. When MestXL expression is lost in the MestC/KO MHB this allelic bias is lost and the M/P ratio plummets to 0.86. A similar situation occurs in the SC but it is not as striking; the MestC/+ and MestC/KO M/P ratios are 1.40 and 0.92, respectively. In tissues where MestXL is not expressed, the heart and liver, the M/P ratios basically stay the same between MestC/+ and MestC/KO samples. Here the wild-type and MestC/KO ratios are 1.38 and 1.49 for the heart and 0.78 and 0.71 for the liver (see Figure 4.1B). To confirm the hot-stop PCR/RFLP analysis, the exon 3 SNP analyzed above and two additional SNPs we identified in exon 19 (rs49583711 and rs47417386) were also directly sequenced from amplified cDNA fragment. The quantification for this method involved using the phred software to calculate the area under the called base (SNP) tracing (174). After normalizing the read-outs to the F1 genomic DNA tracing, the M/P ratios were calculated by averaging the reads for each SNP in 4 MestC/+ and 6 MestC/KO samples from each E14.5 tissue (see Figure 4.1C). The results showed similar patterns to the hot-stop PCR/RFLP analysis which also suggested that allelic usage at Copg2 is regulated by MestXL, although the M/P ratios were consistently higher. In MestC/+ MHB the maternal Copg2 allele is expressed 3.09 times more than the paternal allele. When MestXL expression is lost in the MestC/KO MHB, the M/P ratio decreases to 1.87 (p<0.0001). In MestC/+ and MestC/KO SC tissues, the M/P ratios are 2.97 and 2.06, respectively (p<0.0005). In tissues where MestXL is not expressed, the heart and liver, the M/P ratios are quite similar between MestC/+ and MestC/KO samples, as is seen in the hot-stop PCR analysis. Here the MestC/+ and MestC/KO ratios are 2.72 and 2.41 for the heart and 1.55 and 1.55 for the liver (p=0.25 and 0.99, respectively). These results strongly suggest that paternal transcription of MestXL hinders paternal expression of Copg2 such that Copg2 becomes preferentially maternally expressed, in tissues that express MestXL. When transcription of MestXL is perturbed, such as in MestC/KO embryonic neural tissues, Copg2 expression from the paternal allele is no longer hindered and presumably increases which consequently evens out the allelic differences  66  Figure 4.1 Allelic usage at Copg2 is regulated by MestXL in the developing CNS. (A) Hot-stop PCR/RFLP analysis on various E14.5 F1 MestC/+ (1) and MestC/KO (2) tissues reveals allelic parental differences for Copg2 in tissues that express MestXL. 32Plabeled Copg2 RT-PCR products were digested with BstUI, which cuts only the maternal CAST (C) allele, and separated by electrophoresis on an 8% polyacrylamide gel. Controls on paternal (CD-1) and maternal (C) RNA show complete digestion of the maternal CAST allele with BstUI. P:paternal allele; M:maternal allele. (B) Maternal and paternal band densities were calculated using Image J analysis software and an M/P ratio was generated for each sample (173). (C) Sequencing of three Copg2 SNPs in 4 MestC/+ and 6 MestC/KO E14.5 F1 tissues revealed similar M/P ratio results. The phred program was used to read the sequence trace data and assign quality values to the bases and the numbers were generated after normalizing to F1 genomic DNA (172). Differences between the genotypes for each E14.5 tissue were evaluated using the twotailed unpaired t- test (p-values) and those differences that were significant are marked with an asterisk (p<0.0001 in MHB and 0.0004 in SC). Error bars indicate the standard deviation from four MestC/+ and six MestC/KO biological replicates. 67  of Copg2. The results, taken together with the hot-stop PCR analysis, suggest that allelic usage at Copg2 is regulated by MestXL in the developing CNS.  4.2.2 Loss of MestXL in the developing mouse CNS leads to an overall increase in Copg2 expression levels Considering that the paternal allele of Copg2 is aberrantly expressed in MestC/KO embryonic neural tissues, the next question was whether this leads to an overall increase in steady-state Copg2 expression levels, as suggested by the allele-specific data. This question was important because an increase in Copg2 levels could be implicated in the phenotype observed in Mest+/KO female mice, those that fail to take care of their newborn pups (106). To address this ISH and qRT-PCR experiments were performed. The ISH analysis showed Copg2 to be ubiquitously expressed from low to moderate levels during development. Importantly, Copg2 expression was detected in the E14.5 brain and spinal cord (Figure 4.2, probe Copg2 ex14-20, (-) strand), where MestXL is expressed, but a notable difference between the genotypes was not observed. For a more quantitative approach, qRT-PCR analysis was performed on cDNA from F1 E14.5 wild-type (MestC/+) and MestC/KO tissues: two where MestXL is expressed, the MHB and SC, and two where MestXL is not expressed; the heart and liver. This experiment, on the other hand, showed significant differences between MestC/+ and MestC/KO samples only in tissues where MestXL is expressed (Figure 4.3). In the mutant MHB, Copg2 was expressed 1.4X higher than in the MestC/+ MHB (P=0.012), a significant difference between the two genotypes. In the MestC/KO SC, Copg2 was expressed 1.2X higher than in the WT, a difference that approached significance (P=0.09). Conversely, in the heart and liver, Copg2 expression levels remained similar between the MestC/+ and MestC/KO samples (P=0.81 and 0.62, respectively). According to these results, loss of MestXL in MestC/KO neural tissues leads to an overall increase in Copg2 expression, presumably through the aberrant expression of stable Copg2 mRNA transcripts from the paternal allele. Whether this detected increase is biologically significant and contributes to the phenotype observed in mutant Mest+/KO females remains unknown.  68  Figure 4.2 In situ hybridization analysis of Copg2 mRNA expression at E14.5. ISH was performed on 12 µm sagittal cryosections of Mest+/+ and Mest+/KO E14.5 embryos with a sense and antisense DIG-labeled Copg2 exons 14-20 RNA probe, hybridizing to the (+) and (-) strand, respectively. Blue staining indicates gene expression and nuclear fast red was used as a counterstain. For both genotypes, expression of Copg2 is ubiquitously detected throughout the embryo only with the AS probe (detecting the (-) strand, bottom panels).  69  Figure 4.3 Copg2 expression levels in various E14.5 tissues. qRT-PCR analysis was performed on cDNA from F1 E14.5 wild-type (MestC/+) and MestC/KO tissues (MHB - mid/hind brain, SC - spinal cord, heart, liver) to determine whether steady-state Copg2 levels increase as a result of losing expression of MestXL, specifically in MestC/KO MHB and SC. A significant difference between genotypes was only detected in MHB tissue (p=0.012; two-tailed unpaired t-test). Copg2 mRNA levels were normalized to Ppia and the error bars indicate the standard deviation from four MestC/+ and six MestC/KO biological replicates, assayed in triplicates, for each tissue analyzed.  70  4.2.3 MestXL does not regulate Klf14 imprinting  Maternally expressed gene Klf14 is located 60 kilobases downstream from Copg2. The possibility that MestXL regulation propagates downstream to Klf14 was investigated as well. Again, the F1 MestC/KO mouse line that contains a truncated MestXL was used to address this question. Firstly, an expressed SNP was identified at position 1245 of the Klf14-001 Ensembl transcript (ENSMUST00000101589; CD-1, T: CAST, C) by sequencing amplified CAST genomic DNA for the single exon of Klf14. The allelespecific analysis was then performed in F1 MestC/+ and MestC/KO E14.5 tissues by direct sequencing of RT-PCR products containing the described SNP. The results showed that strict maternal expression of Klf14 is normally maintained in the MestC/KO tissues analyzed, and importantly in MestC/KO mid/hind brain where MestXL is truncated (Figure 4.4). Since the paternal allele of Klf14 remains silenced in MestC/KO tissues, MestXL does not appear to be involved in establishing Klf14 imprinting. For additional support, ISH and qRT-PCR analyses were performed to evaluate overall Klf14 expression levels (Figures 4.5 and 4.6). The ISH analysis showed Klf14 mRNA to be expressed at low to moderate levels in the mesodermal derivatives in both Mest+/+ and Mest+/KO E14.5 embryos (Figure 4.5). Interestingly, this expression pattern resembles that of Mest (see Figure 3.9, Chapter 3) perhaps suggesting they share mesodermal enhancers in the region. Additionally, no significant differences in Klf14 expression levels between wild-type (MestC/+) and MestC/KO tissues were found with the qRT-PCR analysis in the same F1 E14.5 tissues used for the allelic analysis (Figure 4.6). Again, Klf14 was expressed the highest in E14.5 heart tissue, like Mest. Together with the allele-specific analysis, these results suggest that MestXL is not involved in regulating Klf14, as loss of MestXL does not affect the imprinting or expression levels of Klf14 in the developing CNS.  71  Figure 4.4 Klf14 is not regulated by MestXL (part 1). Klf14 imprinting was assessed by directly sequencing RT-PCR products from F1 E14.5 MestC/+ and MestC/KO heart and mid/hind brain tissues. Shown are the sequencing traces for the cDNA strands. The maternal CAST allele (C) contains a C, whereas the paternal CD-1 allele contains a T, as shown by the F1 gDNA control. Strict maternal expression of Klf14 was maintained in MestC/KO mid/hind brain tissue, where expression of MestXL is abolished, suggesting MestXL does not have a role in regulating imprinting of Klf14.  72  Figure 4.5 In situ hybridization analysis of Klf14 mRNA expression at E14.5. ISH was performed on 12 µm sagittal cryosections of Mest+/+ and Mest+/KO E14.5 embryos with an antisense DIG-labeled Klf14 RNA probe, hybridizing to the (-) strand. Blue staining indicates gene expression and nuclear fast red was used as a counterstain. Expression of Klf14 is detected mainly in mesodermal tissues and mesodermal derivatives.  73  Figure 4.6 Klf14 is not regulated by MestXL (part 2). qRT-PCR analysis of Klf14 mRNA levels in various F1 E14.5 MestC/+ and MestC/KO tissues, represented by the white and grey bars, respectively, shows little difference between genotypes. Klf14 mRNA levels were normalized to Ppia and the error bars indicate the standard deviation from three biological replicates, assayed in triplicates, for the tissues analyzed. MHB=mid/hind brain, SC=spinal cord, H=heart, L=liver.  74  4.2.4 MestXL transcripts are not retained in the nucleus  One major mechanism by which imprinted genes are regulated involves imprinted long non-coding RNAs (lncRNAs) acting in cis to silence genes along their respective chromosome. Both Air and Kcnq1ot1 are very large (>100 kb) ncRNAs that are paternally expressed and silence genes along the paternal chromosome such that the genes they regulate are consequently maternally expressed (193). Additionally, both lncRNAs have been shown to be retained in the nucleus, consistent with a function in transcriptional silencing (194). Given that MestXL is paternally expressed and located in a small imprinting domain that includes maternally expressed genes, the question was asked whether MestXL transcripts also shared this feature with these ncRNAs to assist in determining its function. To address this question nuclear and cytoplasmic RNA fractions from 3 adult female brain samples were separated, where MestXL and Xist (another nuclear ncRNA) are expressed. qRT-PCR was performed on a panel of genes and on 3 ESTs that represent MestXL (Figure 4.7; BB692936, BY247745, and Mit1). Additionally, an exogenous RNA control was introduced; a human Xist transcript generated by in vitro transcription, to the nuclear and cytoplasmic RNA preparations before the reversetranscription to control for external variables. The nuclear controls, Xist and Kcnq1ot1, showed that the nuclear separation worked well as they are 96% and 93% enriched in the nucleus (Figure 4.7). The cytoplasmic control, Ppia mRNA (used as our comparative control in our qRT-PCR experiments), showed only an 11% nuclear enrichment. The other protein-coding genes analyzed in this experiment, Mest (ex10-12) and Copg2, showed a 13% and 20% nuclear enrichment, respectively. Of interest, the Mest exon 1-2 reaction showed a 40% nuclear enrichment, which is statistically different from the Mest exon 10-12 reaction (13%). This difference could potentially be due to the fact that the miRNA miR-335 resides in intron 2 which requires extra processing. The three ESTs analyzed for MestXL nuclear localization exhibited nuclear enrichments of 18%, 24%, and 20%, for BB692936, BY247745, and Mit1, respectively, suggesting that MestXL is not enriched in the nucleus and likely to not have a major silencing role like that of Kcnq1ot1. This finding makes sense in light of our results that Klf14 imprinting does not involve long-range regulation from MestXL.  75  Figure 4.7 MestXL transcripts do not localize to the nucleus. Nuclear and cytoplasmic RNA fractions from female adult brain tissue were separated using a sodium citrate/Triton X-100 protocol and analyzed by qRT-PCR. Positive controls Xist and Kcnq1ot1 show nuclear enrichment, whereas Mest, Copg2, and Ppia show cytoplasmic enrichment. Results for MestXL, represented by BB692936, BY247745, and Mit1, imply that MestXL is not enriched in the nucleus. mRNA levels were normalized to an exogenous XIST control that was generated by in vitro transcription and added to the nuclear and cytoplasmic RNA samples at the RT step. The error bars indicate the standard deviation from three biological triplicates, assayed in triplicates. A value of 1 indicates 100% nuclear enrichment and the values were calculated by dividing the nuclear E^dCt values by the total (nuclear E^dCT + cytoplasmic E^dCt) values.  76  4.3 Discussion  The production of MestXL in the developing CNS, through alternative polyadenylation at the Mest locus, regulates the preferential maternal allelic bias at Copg2 but is not involved in regulating Klf14. The present study reports for the first time a new model where the allelic usage at Copg2 is regulated by the production of larger variants at the Mest locus, via tissue-specific transcriptional interference (Figure 4.8). Consequently, allelic usage at Copg2 is regulated secondarily to imprinting at the Mest/MestXL locus. In tissues that express MestXL, Copg2 was found to be preferentially maternally expressed due to the “transcriptional interference” on the paternal chromosome. This imprinting is lost in Mest+/KO tissues where MestXL is truncated due to the introduced pA signal in the MestKO allele. The effect is an overall increase in Copg2 expression levels which may in fact contribute to the phenotype described for Mest+/KO mice. In tissues where Mest is basally expressed such as in the developing liver, Copg2 is biallelically expressed because there is no competing transcriptional interference. Interestingly, a similar story was reported by another group that showed that Murr1 exhibited preferential maternal expression in adult brain tissue which was also hypothesized to be due to transcriptional interference by the antisenseoriented paternally expressed gene U2af1-rs1 (195). Evidence for our model is further supported by previous microarray expression data from Dnmt3l-/+ E8.5 embryos, which do not inherit proper methylation marks from the maternal germ line (GEO series GSE8756, Figure B.1) (191). Since the maternal Mest promoter is not methylated in these embryos, Mest is up-regulated 1.5-fold presumably due to the aberrant transcription from the unmethylated maternal allele. A probe located downstream of Mest that maps to BB692936 (see Chapter 3, Figure 3.4A; probe h) shows a 1.8-fold up-regulation, as well, which is likely due to the maternal allele being transcribed. Most importantly, in support of our proposed model, Copg2 is downregulated and only expressed at half the level of an E8.5 wild-type embryo. We propose that it is the biallelic expression of MestXL that interferes with the transcription of Copg2 on both parental alleles in these mutant embryos to cause the down-regulation observed at Copg2. The mechanism of this transcriptional interference (TI) is unknown but current models propose a few possibilities. The TI “collision” model suggests that physical collisions between converging sense and antisense elongation complexes in cis can 77  Figure 4.8 Model for regulation of allelic bias at Copg2 by MestXL. On the basis of the results presented in this study, we propose the following model for the regulation of allelic bias at Copg2. Transcription from the Mest promoter is directly regulated by a germ-line DNA methylation imprint inherited from oocytes (black lollipop on the maternal allele, (M), such that only the paternal allele (P) is active for transcription. (A) In tissues where Mest transcription terminates within exon 12, such as embryonic liver, Copg2 is biallelically expressed (M/P=1). (B) In the CNS, alternative polyadenylation leads to the formation of the MestXL transcripts which extend all the way to the beginning of intron 20 of Copg2, in addition to the normal Mest transcripts. We propose that MestXL exerts a cis negative effect on the production of the Copg2 mRNA from the paternal allele (reduced black arrow), via transcriptional interference. As a consequence, stable Copg2 transcripts are preferentially derived from the maternal allele (M/P=2.3). (C) In Mest+/KO embryos, MestXL transcripts are truncated by the inserted polyadenylation sequence of the mutant allele and Copg2 is biallelically expressed in the CNS.  78  lead to premature termination of the transcriptional progress of one or both complexes in yeast (196). This concept could equally apply in mammalian systems. Elongation of the regulatory MestXL Pol II complex could dislodge the target Copg2 Pol II complex terminating its transcription. Conceivably, the Copg2 Pol II complex could also terminate MestXL prematurely, but the affect would be much smaller since MestXL is expressed at higher levels and would likely over-power transcription from Copg2 on the paternal allele. Besides physical transcriptional interference models, another way MestXL could obstruct Copg2 on the paternal allele is via the termination complex of MestXL interfering with changes in chromatin structure needed for transcription of Copg2, an idea previously proposed by others (197, 198). Alternatively, if the overlap between MestXL and Copg2 generates a doublestranded RNA (dsRNA), the dsRNA could be a target for RNA-editing, possibly leading to the subsequent degradation of the paternal allele of Copg2. Selective RNA editing by deamination of adenosine to inosine (A-to-I) is an important posttranscriptional mechanism that modifies pre-mRNA transcripts, which in turn can alter individual codons in open reading frames and affect the splicing or untranslated regions of genes (199, 200). One recent report analyzing RNA editing sites in multiple mouse tissues found that the majority of sites occurred in 3’UTR of genes and were interestingly biased at miRNA target sequences (201). Hyper-editing, where at least 20% A-to-I editing has occurred in sequence greater than 50 nucleotides, is an antiviral mechanism that targets viral dsRNA for degradation (202, 203). Perhaps the overlapping double-stranded 3’UTRs of MestXL and Copg2 lead to selective or hyper A-to-I editing of the paternal Copg2 3’UTR tagging the mRNA for degradation, either by inducing a conformational change of the mRNA thus making it unstable or from the mere presence of inosine (203, 204). The speculative MestXL/Copg2 dsRNA could also be a target for the RNA interference (RNAi) pathway. This would involve the dsRNA being processed into smaller fragments by the Dicer protein, one of the mature strands being loaded into the RISC complex enabling target RNA recognition through complementary base pairing, and finally target mRNA degradation (reviewed in ref. 205). This mechanism, however, would presumably work in trans and target both Copg2 parental mRNAs (if the Copg2 strand is loaded into RISC) since the RISC complex would not be able to differentiate between the two in the cytoplasm. Additionally, from the perspective of Copg2, the majority of the composition of the dsRNA would be Copg2 intronic sequences which  79  would decrease the probability of targeting the mature mRNA. For these reasons, RNAi is not likely to be the mechanism that regulates the allelic usage at Copg2. The imprinting at the Copg2 locus is peculiar in that it appears to show a lot of variability in different genetic backgrounds of F1 hybrid mouse strains. There have been three different kinds of strain hybrids that have been used to study the imprinting at Copg2 and all have provided different results. In an intra-specific F1 hybrid between C57BL/6 and M. m. molossinus, Copg2 was found to be more exclusively maternally expressed in the developing embryo and adult brain tissue (96). However, in an interspecific F1 hybrid cross between C57BL/6 and M. spretus, Copg2 was biallelically expressed in all adult tissues analyzed including in the brain where we would expect it to be maternally expressed (119). For our studies, we used a different intra-specific F1 hybrid cross between CD-1 and M. m. castaneus and found that Copg2 is preferentially maternally expressed in a tissue specific manner, but only in the range of 1.5 to 3 times greater than the paternal allele. The major difference between our study and the others mentioned is that we focused on the question of whether the Mest locus had a role in the regulation of Copg2 given their vicinity to each other. Our results suggested to us that Copg2 is not primarily imprinted. Although not studied in great detail Copg2 does not appear to have a differentially methylated region or other imprinting features that would facilitate its imprinting. Instead we propose that it is secondarily imprinted due to the production of larger Mest variants, MestXL, that are under the regulation of alternative polyadenylation. As a consequence these variants somehow interfere and regulate the allelic usage at Copg2. The discrepancy between these three kinds of hybrid strains suggests that whatever mechanism is controlling the imprinting at Copg2 it is rapidly evolving and, considering our results, perhaps the variation is at the level of alternative polyadenylation from the Mest locus. Additionally, from our allele-specific experiments, it is not clear if the effect on Copg2 is an all-or-nothing interference. The ISH results clearly show that MestXL is not expressed in all the cells in the mid/hind brain and it appears to be expressed in a ventral to dorsal gradient in the developing spinal cord (see Chapter 3, Figure 3.9). Whether a stricter maternal Copg2 expression pattern would be observed in a homogenous cell population, compared to the crude developing brain samples composed of heterogeneous cell types used in our analysis, is unknown and requires further experiments.  80  The type of regulation described above is unique to Copg2 as MestXL does not regulate allelic usage at the downstream Klf14 gene. Since Klf14 is exclusively expressed from the maternal allele, does not carry a germ line imprint, and is silenced in absence of maternal methylation imprints, it was previously proposed that Klf14 might be under long-range regulation via the Mest promoter DMR (13). The possibility that this regulation could be due to MestXL transcription was tested here and we found it not to be involved in the imprinting at Klf14. This result is in agreement with our observation that MestXL transcription terminates before exon 20 of Copg2 (see Chapter 3, Figure 3.9, probe k), therefore supporting a local cis effect. In support of this model, we additionally found that MestXL is not localized to the nucleus (see Figure 4.7), a characteristic feature of regulatory ncRNAs such as Airn, Kcnq1ot1, and Xist that mediate long-range regulation of epigenetic silencing. Thus, based on our results, we conclude that Klf14 is under the regulation of an imprinting mechanism that involves methylation on the maternal allele but which does not implicate MestXL. Since loss of MestXL does not abrogate imprinting at Klf14, our results indicate that the regulation of imprinting within this domain involves a combination of different mechanisms.  81  Chapter 5: Candidate developmental targets of miR-335 5.1 Introduction  Imprinted transcription at Mest not only produces the Mest mRNA and the MestXL isoforms described in the previous chapters, but also a miRNA located in intron 2 referred to as miR-335 (98). MiRNAs are key gene regulators that down-regulate target genes by binding to the 3’UTRs of mRNA messages to ultimately inhibit translation of the protein. During this process, target mRNAs can also become destabilized leading to a detectable down-regulation in target mRNA levels. For this reason, mRNA destabilization is thought to be the predominant reason for reduced protein output (206). The mouse miR-335 locus produces two miRNAs; the prominent miR-335-5p miRNA studied in the literature and the mirror antisense miRNA from the other side of the processing RNA hairpin, referred to as miR-335-3p. Notably, miR-3355p has been shown to be expressed approximately twenty times higher than miR-335-3p in several deep-sequencing experiments (This summarized data can be found at http://www.mirbase.org/cgi-in/mirna_entry.pl?acc=MIMAT0000766). In this thesis, miR335 refers to miR-335-5p, unless otherwise stated. Since miR-335 is an intronic miRNA it shares coordinated expression with Mest (133-136). This would also suggest that production of miR-335 is regulated by imprinting, although this has not been shown experimentally. A high profile study in 2008 implicating human miR-335 in the suppression of human breast cancer metastasis put it into the spotlight and since then much more research has focused on its roles in various other human cancers and conditions and physiological processes (described in Chapter 1) (137). Necessary questions regarding miR-335-5p that are relevant to our laboratory are examined in this chapter. These include investigating the expression level of miR335 in Mest+/KO mice and determining whether miR-335 is paternally expressed, like Mest. Since important miR-335 sequences are still present in the MestKO allele, miR-335 in Mest+/KO mice was predicted to share coordinated expression with the MestKO allele. Since Mest is strictly paternally expressed it was hypothesized that miR-335 shares this same pattern and is imprinted. To answer these questions, miR-335 expression levels were analyzed with a TaqMan assay (Applied Biosystems) specific to the mature miRNA.  82  MiRNA target prediction software such as TargetScan, miRanda, and PicTar, predicts that miR-335 (5p and 3p) targets hundreds to thousands of genes (134). Since different tissues display diverse expression gene signatures, miR-335 is expected to have several functions and roles depending on what targets are expressed in any given tissue. This chapter mainly presents an initial and preliminary genome-wide experiment that aimed to determine candidate target genes of miR-335 in the developing embryo, specifically in primary mouse embryonic fibroblasts, using both over-expression and under-expression model systems. One broad hypothesis coming into this experiment was that miR-335 targets genes implicated in embryonic development since Mest is highly expressed in mesodermal derivatives. This question was addressed and putative target genes identified by examining gene expression fold differences in the over- and under-expression model systems determined by RNA-seq, a whole transcriptome shotgun deep-sequencing technique. Follow-up qRT-PCR experiments were then performed to validate some of the candidate miR-335 target genes. Ultimately the preliminary experiments described in this chapter aimed to provide a foundation on which to build future experiments to decipher the developmental role of miR-335. 5.2 Results 5.2.1 MiR-335 expression is decreased in Mest+/KO embryos and shares imprinted expression with Mest When the MestKO allele was generated miR-335 had not yet been discovered. By chance, all the sequences important for miR-335 and for the processing of miR-335 were retained, even though the MestKO cassette initiates within intron 2 (See Chapter 1, Figure 1.3B). One of the very first questions addressed regarding miR-335 was whether the level of expression in Mest+/KO E14.5 embryos was different from wild-type (WT) (Mest+/+) E14.5 embryos due to the introduced mutant cassette in intron 2, where they are both located. Mest and the MestKO alleles are both highly expressed in the developing E14.5 embryo so it was predicted that miR-335 shared this same expression domain. By performing a TaqMan assay for miR-335 expression in WT and Mest+/KO E14.5 embryos, which was normalized to a housekeeping U6 small nuclear RNA (RNU6B) involved in splicing, miR-335 expression was found to be ~50% lower in Mest+/KO compared to the WT level (Figure 5.1A). There are two main possibilities for 83  Figure 5.1 Expression analyses of miR-335, Mest, and MestKO RNA levels in E14.5 embryos. A) MiR-335 expression was assessed by a TaqMan assay on RNA from the top half of a WT and Mest+/KO E14.5 embryo. The miRNA levels were normalized to U6 small nuclear RNA (RNU6B) and the error bars indicate the standard deviation from technical triplicate samples. B) Mest and MestKO allele expression was measured by qRT-PCR on RNA from the top half of a WT and Mest+/KO E14.5 embryo. The mRNA levels were normalized to Ppia and the error bars indicate the standard deviation from technical triplicate samples. For A and B, WT levels were set to 1 to for simple comparison.  84  why this could occur; either the processing of miR-335 is perturbed in Mest+/KO embryos or the Mest+/KO mRNA levels are lower in general. The latter was found to be true when Mest and MestKO mRNA levels were analyzed by qRT-PCR (see Figure 5.1B), thereby suggesting that the processing of miR-335 is normal in Mest+/KO embryos, but that the mutant mRNA is produced at reduced levels from the MestKO allele. On a side note, miR335 levels in Mest+/KO adult brain tissue were found to be considerably higher than WT, similar to Mest+/KO versus WT mRNA levels in the same tissues (data not shown), again suggesting that miR-335 expression correlates well with the respective mRNA level and that processing of the mature miRNA is not perturbed in the mutant allele. Of interest, the finding that miR-335 levels are reduced in Mest+/KO embryos raises the possibility that this decrease could also be associated with the growth retardation phenotype observed in Mest+/KO mice that are also deficient for the MEST protein. Since miR-335 shares coordinated expression with Mest and the MestKO allele, it is also presumably paternally expressed like Mest. One way to confirm this was to also look at the expression levels in embryos from intercrosses, to generate the MestKO/+ and MestKO/KO genotypes and determine whether the same differences observed between WT and Mest+/KO embryos persisted in these additional mutant embryos. If only the paternal allele of miR-335 is expressed, WT (Mest+/+) and MestKO/+ embryos should show similar wild-type levels of miR-335 whereas Mest+/KO and MestKO/KO embryos should show similar reduced levels of miR-335. Results from the TaqMan analysis performed on the intercrosses indeed showed the expected results, suggesting that miR-335 is only expressed from the paternal allele (Figure 5.2).  5.2.2 Using primary mouse fibroblasts as a model system to study over- and underexpression of miR-335  The main objective of the current study was to identify candidate target genes of miR-335 in a developmental context through RNA-seq analysis. This labor intensive process required several steps including constructing the vectors for controls and overexpression of the miRNA, obtaining primary cells in which to perform the experiments, electroporating the vectors into the cells, sorting the green fluorescent positive (GFP+) cells that acquired the vectors 24 hours post-electroporation, isolating RNA from the samples, analyzing the levels of miR-335 in each sample, and finally sending the samples to Genome Sciences Centre in Vancouver, BC, who performed the RNA-seq 85  Figure 5.2 Expression analysis of miR-335 in E14.5 embryos from Mest+/KO intercrosses. MiR-335 expression was measured by a Taqman assay on RNA from a top half of a WT, Mest+/KO, MestKO/+, and MestKO/KO E14.5 embryo. The miRNA levels were normalized to U6 small nuclear RNA (RNU6B) and the error bars indicate the standard deviation from technical triplicate samples.  86  analysis. One of the main goals of this experiment was to establish both a miR-335 overand under-expression system, thus hopefully strengthening the results from RNA-seq and guiding us to direct targets of miR-335. With the finding that miR-335 levels were at least 50% lower in Mest+/KO embryos, the MestKO line serendipitously became a system in which to study under-expression of miR-335. For over-expression, a vector was constructed, described below, to exogenously express miR-335. Primary embryonic mouse fibroblasts (PEFs) were chosen as the cells to use because firstly, they express Mest/miR-335 and therefore developmental target genes of miR-335 and secondly, they were fast and easy to derive from mouse embryos which allowed for ready comparison of WT and Mest+/KO cells derived from the same litter of animals. Twenty three PEFs cultures were established from individual E14.5 embryos (2 litters) from a cross between a WT female and a Mest+/KO male. After genotyping for the MestKO allele and Y chromosome, the 23 cultures were found to consist of 5 WT XX, 3 WT XY, 11 Mest+/KO XX, and 4 Mest+/KO XY (data not shown). One male WT (1.8) and one male Mest+/KO (1.10) PEF cultures were selected to use for the RNA-seq experiments. Whereas the 1.10 Mest+/KO PEFs were used for the under-expression system, the 1.8 WT PEFs were used for both the over-expression experiment and for the WT control. Two vectors were constructed, one for the over-expression of miR-335 and one that was used as a control vector where the exogenous miRNA sequence was deleted from the same transgene (Figure 5.3). The over-expression plasmid, pCmiR335E, consists of the Mest genomic fragment containing miR-335 sequences (specifically intron 2 of Mest with flanking parts of Mest exons 2 and 3), hence requiring the natural splicing process to occur in order to mediate the over-expression of the miRNA (Figure 5.3A). This genomic fragment was fused, in frame, to a green fluorescent reporter protein (GFP) to follow expression of the exogenous miR-335. pC∆miR335E, the control plasmid, was also constructed to use in this experiment, for the WT comparative control (in WT PEFs) and for the under-expression of miR-335 (in Mest+/KO PEFs), to control for external variables introduced by the electroporation protocol into the PEFs. This control plasmid, pC∆miR335E, is identical to pCmiR-335E except that the miR-335 sequences are deleted from the Mest intron 2 sequences of the transgene (Figure 5.3B). Thus, in the end, there were 3 samples that went through the same sequential process; WT PEFs + pCmiR-335E (over-expression), Mest+/KO PEFs + pC∆miR-335E (under-expression), and WT PEFs + pC∆miR-335E (comparative WT control). For our purposes, they are referred in this chapter as WT, WT+miR, and Mest+/KO, respectively. 87  Figure 5.3 Structure of the pCmiR335E and pC∆miR335E plasmids. Linear diagrams of the pCmir335E and pC∆miR335E plasmids used in the miR-335 under- and over-expression experiments. The structure of the plasmids consists of a typical strong enhancer/promoter (CMV-IE) followed by the vector’s exon1 and exon3’, the inserted “transgene,” an EGFP reporter, and a strong pA signal. The “miR-335 transgene” consists of Mest intron 2 that contains miR-335 and parts of the two flanking Mest exons. The control “∆miR-335 transgene,” shown in B, is basically the same except the miR-335 sequences have been deleted. The circular forms were electroporated into either WT or Mest+/KO PEFs to exogenously express miR-335 or the control transgene. Note that the annotated exon 3' from the vector and part of Mest exon 2 are fused into a single exon, such that the transgene consists of three functional exons (1, 3'-Mest ex2, and 3-EGFP) and the entire intron 2 of Mest in the pCmiR335E vector.  88  To complete the experiments the circular plasmids were electroporated into the respective PEFs, transfected GFP-positive fibroblasts were sorted 24 hours postelectroporation, the RNA was immediately isolated from each sorted sample, and finally miR-335-5p levels were assessed by a TaqMan assay (Figure 5.4). Importantly, the TaqMan analysis determining miR-335 levels in the samples showed that miR-335 was over-expressed 14-fold in the WT+miR sample and under-expressed to a level of 15% of WT in the Mest+/KO sample (Figure 5.5). After this confirmation, the RNA was sequenced using strand-specific RNA-seq analysis. To note, the Mest+/KO and WT+miR overexpression PEFs showed no noticeable phenotypes in culture. 5.2.3 RNA-seq data from Mest+/KO PEFs reveals candidate target genes of miR-335  RNA-seq is a deep-sequencing technique that surveys the entire transcriptome in a very high-throughput and quantitative manner (207). In brief terms, RNA is converted to a library of cDNA fragments which are then sequenced and aligned to a reference genome that produces a genome-scale transcription map consisting of both the transcriptional structure and/or level of expression for each gene (207). For our purposes, RNA-seq offered a way to detect potential miR-335 targets by comparing expression levels of genes between the three samples, the WT+miR (over-expression), Mest+/KO (under-expression), and the WT (comparative control) samples. The expectation was that direct target genes of miR-335 would be up-regulated in the Mest+/KO sample and down-regulated in the WT+miR sample compared to WT, and that there would be some overlap between the two lists, strengthening the results for positive direct targets. Additionally, for those genes that overlapped both lists, the expectation was that the fold-change between the WT+miR and Mest+/KO samples would be greater than both samples compared to the WT sample so that fold-change levels would fall into the pattern of Mest+/KO > WT > WT+miR if they were indeed direct targets. Misha Bilenky from Genome Sciences provided the RNA-seq results for all coding genes from Ensembl v65, for a total of 22, 318 genes. The results were normalized with RPKM measures (Reads Per Kilobase of gene per Million reads) to reflect the molar concentration of a transcript in the starting sample by normalizing for RNA length and for the total read number in the measurement. This in turn facilitated comparison of transcript levels both within and between samples (208). Firstly, a snapshot of the UCSC genome browser displaying the custom track of the RNA-seq 89  Figure 5.4 Flowchart for miR-335 under- and over-expression experiments. Many steps were involved in this experiment to ultimately determine candidate target genes of miR-335. Firstly, plasmids pCmir335E and pC∆mir335E were constructed and WT and Mest+/KO PEFs were isolated from E14.5 embryos. The plasmids were electroporated into the PEFs and the GFP+ cells were sorted 24 hours postelectroporation to obtain the cells that expressed the transgenes. RNA was immediately isolated from the three samples and miR-335 levels were assessed by a TaqMan assay. The three samples, WT, WT+miR (over-expression), and Mest+/KO (under-expression), were then taken to Genome Sciences to perform the strand-specific RNA-seq analysis.  90  Figure 5.5 Over- and under-expression analysis of miR-335 in GFP+-sorted PEFs. MiR-335 expression was measured by a TaqMan assay on RNA from GFP+-sorted PEFs transfected with either pC∆miR335E, the control plasmid, or pCmir335E, the overexpression plasmid. These samples are known as WT, WT+miR, and Mest+/KO. The miRNA levels were normalized to U6 small nuclear RNA (RNU6B), the error bars indicate the standard deviation from technical triplicate samples, and WT was set to 1.  91  data revealed that the control and over-expression transgenes properly aligned to the miR-335 locus and were transcribed correctly in all three samples (Figure 5.6). Taken together with the analysis of miR-335 levels in the 3 samples (See Figure 5.5), this provided strong evidence that the control electroporations worked in the WT and Mest+/KO PEFs and that specifically, the over-expression experiment worked in the WT PEFs. The genome-wide average (±Standard Deviation) RPKM values of the 22, 318 genes for each sample were comparable as they were 18.5±89.23, 20.2±100.84, and 20.2±104.18, for the Mest+/KO, WT+miR, and WT samples, respectively. Misha Bilenky from Genome Sciences also provided lists of genes that were significantly up- and down-regulated between 3 conditions, Mest+/KO versus WT, WT+miR versus WT, and Mest+/KO versus WT+miR, with 2D scatter plots showing the selection process for the genes (Figure 5.7. Tables C.1-6). The correlation between expression in any two conditions was very high, as the Pearson’s correlation coefficients equaled (R=) 0.95, 0.96, and 0.96 for each pairwise comparison (Figure 5.7). The selection criteria for potential miR-335 targets were stringent and a significant fold change was defined as at least a 4-fold change, plus genes were required to be expressed at a minimal expression level (0.025 RPKM). Fold changes were determined by dividing the RPKM values for each gene in each sample over each other. As for the number of significant genes determined in these comparisons, 228 genes were found to be up-regulated and 223 down-regulated in the Mest+/KO versus the WT samples, 113 genes were found to be up-regulated and 89 genes down-regulated in the WT+miR versus WT samples, and 49 genes were found to be up-regulated and 170 genes downregulated in the Mest+/KO versus WT+miR samples (Figure 5.7D, Tables C.1-6). Of most interest were those genes that were significantly up-regulated in the Mest+/KO sample versus the WT+miR and WT samples (n=228 and 49, respectively) and down-regulated in the WT+miR versus WT sample (n=89), since they would likely represent direct candidate target genes of miR-335. A Venn diagram was generated from these 6 significant gene lists from the 3 comparisons using the VENNY program to look for overlaps with the focus on the WT+miR<WT, Mest+/KO>WT, and Mest+/KO> WT+miR comparisons, those that would most likely contain direct miR-335 targets (Figure 5.8) (209). Unfortunately, there was hardly any overlap between the up-regulated genes in Mest+/KO and down-regulated genes in WT+miR with the high stringency rules (Figure 5.8). No genes overlapped the WT+miR<WT and Mest+/KO>WT  comparison, and only 2 genes overlapped the 92  Figure 5.6 Transcription from pC∆miR335E and pCmiR335E maps to the Mest/miR335 locus. Strand-specific RNA-seq analysis from the sorted GFP+ PEFs confirms that transcription from the transfected pC∆miR335E and pCmiR335E plasmids maps appropriately to the Mest/miR-335 locus. The first and third rows show that the control plasmid, pC∆miR335E, indeed does not contain miR-335 sequences, as they are not transcribed. The third row displaying Mest+/KO confirms that Mest transcription is abolished after exon 3, at the beginning of the mutant cassette. This UCSC snapshot was downloaded from the website after our custom track was added. The numbers at the bottom indicate the Mest exons and the miR-335 position is noted. The strand polarities are indicated by (+) and (-).  93  Figure 5.7 Distributions of expression levels for annotated transcripts in the Mest+/KO, WT, and WT+miR PEFs. The distributions of the expression levels (log2 RPKMs) are shown by 2D scatter plots for the 22,318 annotated transcripts in the Mest+/KO, WT, and WT+miR samples compared in 3 conditions; A) Mest+/KO versus WT, B) WT+miR versus WT, and C) Mest+/KO versus WT+miR. Those genes/transcripts that are significantly differentially expressed (DE) between each of the 2 compared conditions are plotted in either red, indicating up-regulated expression, or blue, indicating down-regulated expression. Pearson’s correlation coefficients (r) between conditions are also presented. D) A summary table displaying the number of significantly differentially expressed genes for each comparison.  94  Figure 5.8 Comparison of significantly differentially expressed genes in WT, WT+miR, and Mest+/KO PEFs. A Venn diagram was generated by comparing the lists of significantly differentially expressed genes for each relationship (A-F) to show overlaps between the lists. The overlaps are indicated by the intersection of the circles. The 3 circles on top (A-C) and bottom (D-F), separated by the line, represent genes that are down- and up-regulated, respectively, in the noted relationships on the left. The Venny program was used to help generate this Venn diagram (209).  95  WT+miR<WT and Mest+/KO>WT+miR comparison. These 2 genes, Derl3 and 1700023e05rik, showed an up-regulation of 4.25- and 33.73-fold, respectively, and a down-regulation of 4.04- and 28.37-fold, respectively. Whereas an overlap did not occur for the genes in WT+miR<WT and Mest+/KO >WT, several other overlaps between the comparative groups did occur as shown in Figure 5.8. These overlaps perhaps demonstrated secondary targets of those targeted by miR-335, such as the genes that overlap Mest+/KO<WT and Mest+/KO<WT+miR, or genes associated with the protein MEST, since it is not produced in Mest+/KO. Alternatively, these overlaps could be meaningless since many of these genes are expressed at quite low levels or represent genes with variable expression. The results from this experiment were unclear in that they did not give the expected results. Firstly, as described above, there was no overlap between the significantly down-regulated genes in the WT+miR<WT group and up-regulated genes in the Mest+/KO>WT group (see Figure 5.8, B and D). However, more than half of the significantly down-regulated genes in WT+miR<WT (n=48) overlapped with the Mest+/KO<WT list, a puzzling result (Figure 5.8, A and B). Additionally, for the rest of the WT+miR<WT list, no genes displayed even a small up-regulation in Mest+/KO versus WT, and most of them showed a substantial down-regulation, in agreement with those significant overlapping genes (Table C.3). Secondly, more than half of the significantly up-regulated genes in WT+miR>WT (n=54), not down-regulated, overlapped with the Mest+/KO>WT list, again demonstrating perplexing results (Figure 5.8, D and E). Therefore, the differentially expressed genes in the WT+miR group compared to WT (Tables C.3 and C.4) were not considered to be candidate targets in the subsequent experiments, although it is possible this list may contain a few actual target genes. Since the WT+miR over-expression experiment did not give the expected results the decision was to focus on those genes significantly up-regulated in Mest+/KO versus WT (see Figure 5.8D, Table C.2, n=228). For the Mest+/KO sample, the decreased level of miR-335 inherently existed without requiring extra manipulation from an exogenous transgene and therefore target genes of miR-335 were already affected as well, at least for genes whose mRNA levels are affected in miRNA-mediated regulation. Out of the 228 genes up-regulated in Mest+/KO versus WT, 62% are predicted to be targets of either miR-335-5p or miR-335-3p, according to www.microRNA.org. This percentage was determined by looking up each individual gene to see whether miR-335-5p or miR-3353p was predicted to target their respective mRNA transcripts. It may seem like this list is 96  enriched, and perhaps it is, but half the genome is predicted to be targets of either miR335 mature miRNA, according to this particular prediction website. Interestingly, though, when the overlapping genes between Mest+/KO>WT and Mest+/KO>WT+miR were investigated in the same way described above (n=33), 85% were predicted to be targets of miR-335-5p, miR-335-3p, or both, suggesting this list was enriched for miR-335 target genes. To determine if the genes in the Mest+/KO >WT list were enriched in any significant biological themes, they were run though a DAVID (Database for Annotation, Visualization, and Integrated Discovery) analysis on the high stringency setting (210; 211). A few moderately significant small annotated clusters were identified with the top one being enriched for extracellular matrix proteins (Figure C.1; n=15). This enrichment suggested that miR-335 may help regulate the extracellular matrix in PEFs, a finding that is in line with the study that reported miR-335 to be a metastatic suppressor of cancer that could remodel the extracellular matrix of cancer cells (137). The other 2 significant terms were EGF-like domain (n=12) and homophilic cell adhesion (n=7).  5.2.4 Validating target genes by qRT-PCR  Several candidate target genes of miR-335 (5p and 3p) were determined through RNA-seq analysis by comparing genes significantly up-regulated in the Mest+/KO PEFs versus WT PEFs. As a starting point, 6 genes from the Mest+/KO>WT list (Table C.2) were chosen for further validation by qRT-PCR (Table 5.1). Considering expression levels of WT RPKM ≥ 1.0 and significant fold changes, genes Cd34, Cd200, Ednrb, Krt18, Prg4, and Thbs4 were assessed. Additionally, two genes that were not significantly up-regulated in the Mest+/KO PEF line were examined out of interest. Rb1 has already been verified as a target in mouse NIH 3T3 embryonic fibroblasts, mouse C2C12 myoblast cells, and human osteosarcoma (U2OS) and meningioma cells, and Derl3 was differentially expressed in the WT+miR<WT and Mest+/KO>WT+miR lists, a possible informative overlap (144; 146). To further investigate them as target genes, expression levels of these 8 candidates were measured by qRT-PCR on new independent RNA samples. The new PEFs were derived from E14.5 embryos within the same litter as those used for the RNA-seq and they were referred to as 1.2 WT XX, 1.6 WT XY, 1.1 Mest+/KO XY, and 1.7  97  Table 5.1 Candidate target genes of miR-335 chosen for further investigation by qRTPCR.  RPKM  miR-335 predicted target3  Fold-change  Gene ID  Mest+/KO  WT+miR  WT  Cd200*  5.696  1.173  1.234  4.857  Cd34  29.638  15.166  5.029  Ednrb  12.646  8.811  Krt18*  6.058  Prg4*  Mest+/KO Mest+/KO WT+miR WT  Wt+miR WT  5p  3p  4.618  0.951  0  2  1.954  5.893  3.016  2  2  2.977  1.435  4.247  2.959  0  2  8.345  1.429  0.726  4.239  5.839  1  0  32.809  2.791  1.963  11.757  16.718  1.422  0  1  Thbs4*  25.621  11.871  2.149  2.158  11.921  5.523  0  0  Derl3  0.357  0.084  0.339  4.25  1.05  0.25  _  _  Rb1  10.170  6.997  7.772  1.45  1.31  0.90  1  3  *Overlaps with another relationship 3  Predicted target according to www.microRNA.org  98  Mest+/KO XX PEFs. Together with PEFs 1.8 and 1.10 used for RNA-seq, these 6 PEFs comprised the three WT and Mest+/KO biological replicates used in the qRT-PCR experiment. One notable difference between these PEFs and those used in the RNAseq was that these new PEFs did not go through an electroporation protocol. According to the qRT-PCR results, all tested candidate genes showed variable expression between and within genotypes (Figure 5.9). Genes Cd200 and Krt18 seemed to show an up-regulation in Mest+/KO PEFs, but these results were not statistically significant as shown by the large error bars. Rb1, a gene that is a target of miR-335, was not affected in our Mest+/KO PEFs, but the effect could be hidden if only the protein levels are affected and not the mRNA levels as well. Alternatively, Rb1 might not be regulated by miR-335 in fibroblasts. Overall, this analysis showed that these specific candidate genes are variably expressed and likely not targets of miR-335. Importantly, this analysis also showed that it is necessary to assess more than one biological replicate when performing genome wide experiments like RNA-seq to rule out genes that are highly variably expressed. Of course, one reason for doing the miR-335 over-expression experiment alongside the under-expression experiment was to help rule out false positives such as these genes. 5.3 Discussion  The present study mainly reports an initial attempt to determine candidate target genes of miR-335 in a developmental context and is a project that is currently ongoing. The strategy to identify putative target genes revolved around comparing RNA-seq libraries from PEFs with opposing levels of miR-335 to screen for significant differentially expressed genes with the expectation of obtaining correlating data between samples. On one hand, RNA-seq was advantageous because it provided precise gene expression levels of the whole transcriptome but, on the other hand, it generated large volumes of data that was and is challenging to determine what is relevant and what is noise and artifacts. Additionally, our strategy assumed two things; firstly, that miR-335 targets genes in PEFs and secondly, miR-335-mediated regulation not only affects target proteins, since one of the functions of miRNAs is to induce translational repression of proteins, but also affects target mRNAs, as shown by numerous examples in the literature. There are genes, however, for which only the protein levels are affected and not both the protein and mRNA transcript levels. For example, Zhang et al. found that 99  Figure 5.9 Expression levels of candidate target genes of miR-335. qRT-PCR analysis was performed on cDNA from WT (Mest+/+) and Mest+/KO PEFs to further investigate candidate target genes of miR-335 that were determined through RNA-seq analysis. mRNA levels were normalized to Ppia and the error bars indicate the standard deviation from three WT (1.2, 1.6, and 1.8; white bars) and three Mest+/KO (1.1, 1.7, and 1.10; grey bars) biological replicates, assayed in triplicates. All genes analyzed revealed variable expression among and within genotypes and therefore no genes were significantly up-regulated in Mest+/KO PEFs to implicate true targets, at least as measured by mRNA levels.  100  miR-335-5p did not have an effect on Dkk1 mRNA levels, but significantly decreased DKK1 protein levels (156). Therefore, as a strategy, RNA-seq only provided potential valuable information regarding candidate targets of miR-335 whose mRNAs levels are also affected by miR-335-mediated regulation. For the purposes of determining target genes, the overall RNA-seq results from all three samples were generally perplexing, as described in the results section. The expectation was that if the miR-335 over-expression transfection/electroporation worked as hypothesized, the target genes that would be affected by up-regulation of miR-335 would correspond to the same genes in the Mest+/KO sample where miR-335 is downregulated. However, there was no overlap between these lists (see Tables C.2, C.3, Figure 5.8). Additionally, for those 228 genes that were significantly up-regulated in Mest+/KO>WT, only 7 showed a slight inverse insignificant down-regulation in WT+miR<WT (Table C.2), and for those 89 genes that are significantly down-regulated in WT+miR<WT, only 2 showed an inverse insignificant slight up-regulation in Mest+/KO >WT (Table C.3), therefore further suggesting that there were basically no genes in these two lists that were close to overlapping. Additionally, a DAVID analysis on these two groups presented different significant biological terms. Whereas the top term for the Mest+/KO>WT list was extracellular matrix, as described in the results, the most significant top biological term for the WT+miR<WT list was nucleosome genes (see Figures C.1, C.2). One possible reason for the non-overlap of lists was that the stringency was set too high to detect subtle changes brought on by mis-regulation of miR-335, if mis-regulation of miR-335 does not affect target mRNAs in PEFs to a 4-fold extent. Indeed if the stringency is lowered to a 2-fold change, some overlap does occur (n=72, data not shown), however, the results become less significant so they must be looked upon with some degree of uncertainty and additionally, many of these genes are expressed at low levels. Due to the non-overlap of these experiments, the Mest+/KO sample was deemed the most likely to contain true target genes of miR-335 from which to continue the study. The expression levels of some candidates that were up-regulated in the Mest+/KO PEFs (compared to the WT PEFs) were further analyzed by qRT-PCR in independent WT and Mest+/KO PEFs to investigate their potential authenticity (see Table 5.1). Many of them appeared to have variable gene expression, not only between the genotypes, but also within the biological replicates (see Figure 5.9), therefore suggesting these genes are probably not targets of miR-335. As this was just a starting point, the quest to find 101  target genes of miR-335 will hopefully continue with more analysis of genes on the list. Other genes worth investigating in the future include the extracellular matrix genes in the top enriched group in the DAVID analysis, as miR-335 has already been shown to be involved in remodeling the extracellular matrix of cancer cells and genes within smaller annotated groups related to development (see Figure C.1) (137). Although we demonstrated that miR-335 was indeed over-expressed (see Figure 5.5), the RNA-seq analysis did not provide the results we had anticipated for the WT+miR PEFs. One reason might have been that the 24 hour post-electroporation period did not allow enough time for target mRNA levels to be affected. In our system the processing of exogenous miR-335 relied on the natural splicing of the transgenic intron to occur in the plasmid. The idea was that waiting 24 hours to harvest the PEFs post-electroporation decreased the chance of off-targets and indirect targets of miR-335. One of the more popular methods of studying over-expression of miRNAs is buying the double-stranded (ds) miRNA precursor of choice from a company such as Ambion and directly transfecting it using the manufacturer’s instructions. This step bypasses several processing steps as the ds-miRNA precursor is directly integrated into the RISC complex in the cytoplasm. For these types of experiments and related ones reported in the literature, time-frames for harvesting cells post-transfection of miRNAs range from 24 to 72 hours. In hindsight, it probably would have been better to wait 48-72 hours with our system since it likely takes more time for target mRNAs to be affected. Future similar over-expression experiments harvesting PEFs 48-72 hours post-electroporation will determine whether this was an issue in our over-expression experiment. Given that miR-335 was over-expressed after electroporation of pCmir335E, as measured by the TaqMan assay (see Figure 5.5), there were a number of important factors that could have skewed the results when assessing miRNA targeting effects (212). Firstly, transfected (exogenous) miRNAs compete with endogenous miRNAs for the RISC complex needed for miRNA-mediated regulation (213). As a consequence, genes targeted by endogenous miRNAs can also be up-regulated since endogenous miRNAs are not able to function to their full potential. Perhaps those genes that were found to be significantly up-regulated in the WT+miR group versus the WT group fall into this category (Table C.4). Interestingly, a DAVID analysis on this group revealed only one marginally significant enrichment indicating that this group is more random than the others (Figure C.3). Secondly, the effect of miRNA regulation, specifically miR-335 regulation, can be diluted by target abundance meaning that down-regulation of 102  individual genes varies with the total concentration of available target transcripts (214). If there are more available target transcripts expressed in PEFs, it is predicted each individual target gene is down-regulated to a lesser extent than a cell line with a lower number of targets. Thirdly, it has been reported that genes with very long 3’UTRs are poor targets for ectopically expressed miRNAs since they have a higher chance of also being regulated by endogenous miRNAs (212; 215). Given these factors could be involved in skewing the miR-335 over-expression screen, one would still expect for there to be some overlap with the up-regulated genes from the Mest+/KO>WT list, which we did not encounter. The finding that miR-335 was decreased in Mest+/KO embryos raised the possibility that the growth retardation phenotype observed in Mest+/KO pups might not only be caused from loss of the MEST protein (106). The extent to which this reduction may contribute to the phenotype, however, is difficult to assess since it is basically impossible to dissect out the effects of each product (protein and miRNA) in the Mest+/KO mouse line and much is still unknown regarding the functions of both miR-335 and MEST. The only way to investigate this matter further would be to find a way to isolate either loss of miR-335 or MEST, such as generating a new mouse line where only miR335 is reduced or deleted while the protein levels stay intact. In accordance with the Mest+/KO line, loss of MEST in the Mest+/KO PEFs also complicates our RNA-seq analysis since loss of MEST might also affect certain genes in PEFs, supposing that MEST is part of a signaling or metabolic pathway. However, in addition to identifying targets of miR-335, the RNA-seq analysis might also serendipitously help identify physiological processes and pathways that involve MEST. Again, a better way to determine the effect of either MEST or miR-335 would be to separate them, such as restoring the MEST protein or miR-335 back to Mest+/KO PEFs and reanalyzing with RNA-seq. PEFs represent a heterogeneous population of primary mesenchymal adherent cells that have multipotential capacity to differentiate into mesenchymal lineages (216; 217). For one interesting study, over-expression of miR-335 in human mesenchymal stem cells inhibited their proliferation and migration as well as their osteogenic and adipogenic potential (134). If mice mesenchymal cells are similar, perhaps overexpression of miR-335 in PEFs also inhibits the same processes. Unfortunately, this question was never addressed since the GFP+-sorted PEFs that over-expressed miR335 were never replated given that the plasmid was only transiently expressed. Future research will determine whether miR-335 over-expressing PEFs, or any other future cell 103  line over-expressing miR-335, exhibits defects in proliferation, migration, and differentiation. On the flipside, reduction of miR-335 did not seem to affect the PEFs in a major way as the Mest+/KO PEFs were easily derived and appeared to proliferate at the same pace as the WT PEFs, although no experiments were performed to formally address this question. Most miRNAs are predicted to down-regulate large numbers of target mRNAs and numerous mammalian genes have more than one miRNA target site in their 3’UTRs (132). Target identification, however, is hindered by the fact that miRNAs do not have to be perfectly complementary to their target mRNA. If one searches for potential targets for their favorite mouse miRNA via the Sanger miRBase Targets database (www.microRNA.org) the list is likely to contain thousands of target genes. Indeed, miR335 is predicted to target approximately half the mouse genome as miR-335-5p and miR-335-3p is predicted to target 7,409 and 8,698 genes, respectively, with several of these genes overlapping. When the significantly up-regulated genes in Mest+/KO>WT were examined in www.microRNA.org (see Table C.2), 62% were predicted targets, which is not surprising given how many are predicted to be targets in the first place. Other sites such as Targetscan and Mirdb are more stringent with their target predictions and predict that miR-335-5p and miR-335-3p target 119 and 345 genes and 237 and 1138  genes,  respectively  (http://www.targetscan.org/  mmu_50/,  http://mirdb.org/miRDB/). Four genes from the Mest+/KO>WT list overlapped with these lists for predicted targets of miR-335-5p. Fat3, is a predicted target on the TargetScan list, and Naip6, Cmah, Gata5 are predicted targets on the Mirdb list. Of interest, out of the genes that are identified as miR-335 targets in the literature (see Table 1.2), none of them were significantly differentially expressed in our experiments, although Sp1 and Rb1 showed a slight up-regulation in Mest+/KO>WT, a 1.5 and 1.3 fold-change, respectively. PEFs were used as the first cells for this type of analysis because they were easy to derive from WT and Mest+/KO E.14.5 embryo littermates which allowed for ready comparison. Due to time constraints and circumstances, this study was basically meant to provide a foundation on which to build future experiments as many questions and follow-up research remain unfortunately not in the scope of this thesis. As the analysis of candidate genes is still only preliminary, the hope is that research continues to explore several more significantly up-regulated genes in the Mest+/KO PEFs as the list likely contains some true targets. For those that appear to be targets of miR-335, validation 104  will continue through techniques such as western blotting and 3’UTR luciferase reporter assays since miRNA-mediated regulation ultimately represses translation through binding to 3’UTRs of target mRNAs. In addition, repeating this experiment for different cell types, such as embryonic neuronal cell lines from WT and Mest+/KO embryos or in cells in a less differentiated state, would be of great interest as miR-335 might target genes specific to neuronal function or be important in differentiation. Expression of all intronic miRNAs was once thought to be dependent on transcription from their host genes. Recent studies, however, have since shown that a percentage of intronic miRNAs appear to be regulated independently as their expression correlates poorly with the expression of their host genes (218). Furthermore, one study reported that approximately 35% of intronic miRNAs have predicted upstream promoters that might be responsible for their activity and presented some examples whereby cloning intronic regions encompassing miRNAs and upstream sequences into a promoter-less plasmid resulted in the miRNA being expressed (219). As for miR-335, a number of reports have suggested that miR-335 fits into the majority group of intronic miRNAs that are co-regulated with their transgenes (133-136). Here we present supporting results, firstly showing that miR-335 is reduced in Mest+/KO E14.5 embryos that have a reduced level of the MestKO allele and secondly demonstrating, for the first time, that miR-335 is imprinted and paternally expressed like Mest. Taken together, these results suggest miR-335 is processed from transcripts initiating at the unmethylated paternal promoter and not from other potentially biallelic and intronic promoters.  105  Chapter 6: Discussion 6.1 Summary of results and conclusions  Though Mest and its mechanism of imprinting have been well-characterized in mammals, more questions arose regarding the Mest locus when reports revealed there were uncharacterized paternally expressed RNAs transcribed from the same strand in immediate proximity downstream of Mest (95; 96; 101; 159). We hypothesized they constituted larger isoforms of Mest, referred to as MestXL. Since the mutant MestKO allele consists of a truncated mutant Mest transcript which abolished transcription in the 3’ end, the Mest+/KO mouse line provided a good system to study the regulation of these transcripts downstream of Mest. Shown by several methods, Mit1, Copg2AS, and two other analyzed ESTs in the region were found to be expressed in WT E14.5 embryos but not in Mest+/KO E14.5 embryos, suggesting they constituted larger isoforms of Mest. Additionally, the ISH analysis established that MestXL is produced mainly in the developing central nervous system in mice. In support of our results, SAGE analysis on this region showed that the Mest locus is indeed regulated by alternative polyadenylation, using at least 2 extra polyadenylation sites ~7-10 kb downstream of Mest (in intron 20 of Copg2). The work presented in Chapter 3 demonstrated for the first time that the Mest locus produces larger transcripts, MestXL, in the developing central nervous system via alternative polyadenylation. The Mest locus contains the only potential imprinting center identified thus far in the domain, a gametic DMR methylated in oogenesis (13; 16; 101; 119). Given this, it was questioned whether the Mest DMR might be involved in regulating the allelic usage at neighboring maternally expressed protein-coding genes Copg2 and Klf14 via the production of MestXL. Allele-specific experiments were performed in WT and Mest+/KO neural and non-neural embryonic tissues, in tissues that express and do not express MestXL, respectively. Copg2 was shown to be preferentially maternally expressed in WT neural tissues which revealed that Copg2 is regulated secondarily to imprinting at Mest via production of MestXL in a tissue-specific manner. Additionally the lack of interference from MestXL on the paternal allele in Mest+/KO neural tissues led to an overall increase in Copg2 expression levels. For Klf14, loss of MestXL did not change the imprinting or expression levels in Mest+/KO neural tissues, suggesting that MestXL does not regulate allelic usage at Klf14 and that this type of regulation, likely transcriptional interference on 106  the paternal allele, is unique to Copg2. In conclusion, the work in Chapter 4 revealed that MestXL has a role in regulating the allelic usage at Copg2 in neural tissues and establishes a model for allelic interactions between neighboring transcripts as a new mechanism for the regulation of imprinted expression in mammals. Of notable interest, the Mest locus also encodes a miRNA in its intron 2, miR335, that is co-expressed with Mest (133-136). MiRNAs act as post-transcriptional gene regulators by binding to 3’UTRs of target genes to repress translation and ultimately down-regulate them. As a way to attempt to determine candidate target genes of miR335, PEFs were generated from WT and Mest+/KO littermates to use as the system for these experiments, to study the effects of over- and under-expression of miR-335. We established that miR-335 is expressed at a considerably lower level in Mest+/KO E14.5 embryos providing the opportunity to study under-expression of miR-335 in Mest+/KO PEFs. A plasmid was constructed to generate the over-expression system that exogenously expressed miR-335 when transfected into WT PEFs. The transcriptomes of both systems were analyzed by RNA-seq, as well as a WT comparative control, to obtain a list of genes that were significantly differentially expressed to determine target genes. The expectation was that positive target genes would show reciprocal differential expression in the under-expression and over-expression systems. Since there were no genes that overlapped between the systems with the high stringency rules, those genes found to be up-regulated in the Mest+/KO PEFs were deemed the most likely to be targets of miR-335. Several genes were further investigated by qRT-PCR from this list and found to be quite variably expressed suggesting they were probably not targets. At this point the work presented in Chapter 5 has not determined any targets of miR-335 but has laid down the groundwork for further exploration. 6.2 General discussion  6.2.1 Mest imprinting contrasts with high vertebrate conservation of gene  Mest is highly conserved in vertebrates throughout evolution yet is imprinted only in placental mammals and marsupials, in eutherian and metatherian lineages, respectively (95; 102-104; 220). Since this evolutionary conservation contrasts with the imprinting at the locus, it is conceivable that Mest acquired imprinting for reasons associated with novel mammalian-specific functions specific to this locus. One possibility 107  is that the MEST protein acquired new functions that generated a situation which allowed for imprinting. For example, Mest might have been selected to be imprinted upon being expressed in the extraembryonic mesoderm of the placenta if it has important roles in the establishment or function of the placental vascular network (105). In accordance with the parental conflict theory, if the Mest locus has a role in regulating the exchange between mother and her offspring in the placenta, this might have created a window of conflict between the parental alleles whereby the maternal side would want to suppress fetal growth and the paternal side would want to enhance fetal growth and eventually the paternal side dominated (8; 10). Alternatively, perhaps MEST acquired a new substrate in the metatherian and eutherian lineages if it developed a mutation which generated a situation to control the dosage of MEST, a situation that might have occurred for imprinted genes Igf2 and Igf2r. Briefly, Igfr2 does not bind Igf2 in Xenopus and chickens where they are both not imprinted, yet in metatherian and eutherian animals, maternally expressed Igf2r binds and modulate levels of the paternally expressed growth factor Igf2 (and other glycoproteins) suggesting they became imprinted after Igf2r acquired Igf2 as a substrate (221-224). On the other hand, perhaps the Mest locus was selected to become imprinted for reasons not regarding the protein, but when it acquired alternative functions, like miR335 and MestXL. In light of the results presented in Chapter 3, we now know the Mest locus produces MestXL in mice and there is evidence for production of MESTXL in humans (112). Although it is unknown whether MestXL exists in other mammals and marsupials, perhaps the Mest locus was selected to become imprinted when these longer isoforms of Mest were starting to be produced and regulated by alternative polyadenylation. Then again, perhaps the acquisition of miR-335 was the driving force behind the Mest locus becoming imprinted to control the dosage of this miRNA. MiR-335 seems to be conserved in placental animals, but its status in marsupial animals is not clear. In the wallaby, where Mest exhibits imprinting, the genome sequence is incomplete and the 5’ Mest sequences inclusive of intron 2 are not yet annotated so it is unresolved whether miR-335 is present in wallaby (104). In the opossum, miR-335 is not present in the Mest locus (according to the UCSC genome browser) and was recently reported to display monoallelic expression presumed to be from the paternal allele, presenting evidence that miR-335 is absent in marsupial animals that exhibit Mest imprinting (220). This would therefore argue against a role for miR-335 in the selection for imprinting since both marsupial and placental animals exhibit imprinting of Mest. 108  Alternatively, maybe the Mest locus was selected to become imprinted to solely control genes in the region if the Mest promoter was somehow poised to become a gametic DMR and IC. Whatever the reasons for the Mest locus becoming imprinted, it is becoming clear that the imprinting at the Mest locus impacted not only the locus but the region as a whole, whether purposely or consequently.  6.2.2 MestXL: A coding or non-coding RNA?  Known imprinted gene clusters were observed to contain at least one lncRNA, defined as greater than 200 bp, that displays reciprocal imprinted expression relative to neighboring protein-coding genes (31; 32). At one point, perhaps Copg2AS and Mit1 were considered these ncRNAs for the Mest-Copg2-Klf14 cluster, but we now know they constitute larger isoforms of Mest (123; 225). One question that remains unanswered is whether MestXL still codes for the MEST protein or is non-coding due to novel functions in neural tissues. From our Northern analyses, probes against the Mest mRNA transcript revealed larger isoforms in addition to the main Mest transcript indicating MestXL comprises the Mest mRNA and maintains the proper processing of the Mest primary transcript (see Figure 3.6). This suggests MEST is likely translated from the MestXL transcript and therefore not considered a ncRNA. This is also supported by our finding that MestXL is predominantly cytoplasmic. If MestXL is not a lncRNA, then this may be the first imprinted domain that does not contain a lncRNA, unless there are other unidentified ncRNAs in the region. Of interest, a large 68 kb ncRNA is predicted 500 MB downstream from Klf14 (225). If MestXL is a tissue-specific ncRNA then perhaps its only function is to regulate the allelic usage of Copg2, similar to other regulatory ncRNAs in imprinted domains that regulate protein-coding genes. Although questionable, this mechanism might have evolved if there was a need to control the dosage of Copg2 in the murine nervous system in smaller ways, by reducing the Copg2 level slightly, yet not totally silencing one of the alleles. Another way MestXL might be considered non-coding is if a function of the long 3’UTR is to inhibit or stall translation of MEST in neural tissues, as mentioned in the Discussion section (3.3) of Chapter 3, which would not be mutually exclusive from the transcription interfering with Copg2. Other similarities that MestXL shares with other ncRNAs includes antisense regulation of a protein-coding gene and being specifically  109  expressed in the nervous system, a tissue system that overly represents ncRNAs (123; 226; 227). In light of recent studies, perhaps MestXL comprises both a coding and noncoding RNA. A study by Furuno et al. surveying ~100,000 FANTOM mouse cDNA clones reported that transcription may also initiate within 3’UTR sequences and therefore act as a source of independent transcripts that may exhibit expression patterns different  from  their  upstream  corresponding  protein-coding  sequences  (225).  Additionally, Mercer et al. recently provided further support for this concept by mapping CAGE tags, those that indicate 5’- capped ends, to hundreds of 3’UTRs in mouse and human genes implying novel functions of these 3’UTRs and representing a novel class of RNAs produced from them, which they refer to as 3’UTR associated RNAs (uaRNAs) (228). Of great interest, out of the ~4000 mouse genes listed that had at least one high confidence CAGE mapping site in their 3’UTR, Mest made the list with 9 CAGE tags being mapped to the Mest 3’UTR perhaps implicating it in having alternative functions. Furthermore, they show that some uaRNAs probably arise as a consequence of posttranscriptional cleavage instead of the conventional transcription initiation, by being transcriptionally processed from the mRNA (228). Interestingly these proposed uaRNAs can also be developmentally regulated, showing different expression patterns than the corresponding protein-coding sequence mRNA, like we observe for the CNS-specific pattern for MestXL. This is an intriguing concept and might explain the northern blot results in Figure 3.6 where all MestXL probes (BB EST AS {h}, BY EST AS {i}, and Mit1 AS {j}), detecting transcription on both sides of the genomic gap, share a~6 kb band, perhaps suggestive of a MestXL uaRNA.  6.2.3 The Mest DMR: an IC that acts as both a promoter and an insulator?  Imprinted genes in mammals mainly reside in numerous clusters spanning across the genome. Attributed to this clustering, genes acquire allele-specific expression through varying mechanisms regulated by ICs, as described in the Introduction (Chapter 1). For the MMU6 domain, the DMR at the Mest locus is the only potential IC identified to date and therefore predicted to be the major regulator of the region. In Chapter 4, we investigated this role of the DMR as an IC and showed it controls the paternal expression of MestXL to regulate the allelic usage of the immediate adjacent gene Copg2 via tissue-specific transcriptional interference. Our results also showed that 110  MestXL does not regulate imprinting at Klf14 and revealed that the Mest IC might regulate Klf14 through a different mechanism. This is supported by the observation that maternal expression at Klf14 was reported to rely on inherited maternal DNA methylation (13). Since conducting these studies, high-throughput ChIP-seq experiments in the mouse has provided intriguing evidence that the Mest DMR may also act as an insulator, as CTCF binds at that precise location (Figure 6.1A). CTCF binding sites are considered a mark for potential insulator elements and have been shown to be required to maintain imprinting in other imprinted regions, the most well-known being the H19/Igf2 locus described in the Introduction (37; 38; 41). Usually in a model such as this, shared enhancers are also involved that reciprocally enhance expression of maternally and paternally expressed genes,  depending on whichever one has access to them, for  example through chromatin looping (229; 230). In agreement with this proposition, Mest and Klf14 appear to have strikingly similar expression profiles (see Figures 3.9 and 4.5) suggesting they indeed share mesodermal enhancers somewhere in the region. Although none have been experimentally validated, several enhancers are predicted throughout the region including upstream of Mest, between Mest and Klf14, and downstream of Klf14, at least in human CD4+ T cells (231). In support, there are several conserved non-coding elements around the region according to the PEDB: Mammalian Promoter/Enhancer DataBase and areas marked with co-activator protein p300, a mark indicative for enhancers (232-234) (http://promoter.cdb.riken.jp/). According to our proposed model for the regulation of Klf14, CTCF does not bind the maternal chromosome due to CpG methylation at the Mest DMR (and perhaps the presence of other methyl-binding proteins) allowing the maternal Klf14 allele access to enhancers, for example located upstream of Mest (Figure 6.1B). Conversely, on the paternal chromosome, CTCF binds adjacent to the Mest DMR creating an insulator to Klf14 and therefore cutting off access to enhancers, allowing Mest to somehow access the enhancers for expression (Figure 6.1B). It is also conceivable that Mest and Klf14 do not share enhancers and that only Klf14 needs enhancers for expression, which still fits with our model of the Mest DMR acting as an insulator.  111  Figure 6.1 A proposed model for the regulation of Klf14 via insulator activity at the Mest DMR. A) High-throughput CHiP-seq experiments showed that CTCF binds adjacent to the Mest DMR. This data was obtained from the UCSC ENCODE genome browser. All tissues, primary cell, and cell lines analyzed showed CTCF binding at this location but only two are displayed here, in mouse cerebellum and heart tissue, for the sake of keeping the figure small. This snapshot was downloaded from the UCSC genome browser in January 2013. B) A proposed model for the regulation of imprinting at the Klf14 locus via the Mest DMR acting as an insulator if enhancers are located upstream of Mest. On the maternal allele, putative enhancers can access Klf14 since CTCF cannot bind the methylated DMR. On the paternal allele, CTCF binds the unmethylated Mest DMR creating an insulator so that Klf14 cannot access the enhancers and therefore is not expressed. As there are many predicted enhancers throughout the region, it is possible they are located elsewhere. The white boxes represent the genes, the grey boxes represent the Mest DMR, the black triangles indicate the putative enhancer, and the white and black lollipops represent unmethylated and methylated CpG-rich sequences, respectively.  112  6.2.4 Important roles of miR-335  Although our studies to date on miR-335 have not identified new target genes and roles in development, specifically in mesenchymal PEFs, many other studies have provided valuable information regarding functional roles of miR-335. The major finding that firstly placed miR-335 in the spotlight was that human miR-335 acts as an antimetastatic regulator in breast cancer (137). Initially, this was believed to be through targeting genes involved in epithelial to mesenchymal transition (EMT), a major player in the metatstatic process, as shown for other miRNAs such as the miR-200 family (137; 235; 236). Instead, however, miR-335 is believed to target SOX4, TNC, and PTPRN2 in breast tissue to remodel the extracellular matrix of cancer cells and inhibit cell migration (137). MiR-335 was also identified to be a metastatic suppressor in gastric cancer, but interestingly, through a different mechanism involving targeting SP1 directly and indirectly through the Bcl-w-induced phosphoinositide 3-kinase-Akt-Sp1 pathway (147). In neuroblastomas, miR-335 was implicated in the TGF-β non-canonical pathways by targeting ROCK1, MAPK1, and LRG1 resulting in reduced phosphorylation of downstream members, such as myosin light chain, leading to an inhibition of the invasive and migratory potential of neuroblastoma cells (151). Additional cancer progression pathways involving miR-335 include the Rb1 signaling pathway in meningiomas and the cAMP/PKA pathway in malignant glioma cells (146; 149). Importantly, since miR-335 seems to be significantly under- or over-expressed in certain types of cancers, it is a good candidate to become a significant biomarker of metastatic tumors and cancer proliferation and differentiation (150; 237). Regarding development, miR-335 has also been implicated in numerous biological processes. For example, in humans, it has been proposed to have a role in tissue homeostasis and shown to be a negative regulator of human mesenchymal stem cell migration by targeting RUNX2, a key transcription factor involved in osteogenesis (134). In support of this proposition, miR-335 was also shown to have a key role in regulating bone development in mouse cell lines by activating Wnt signaling and promoting osteogenic differentiation by down-regulating Dkk1, a protein essential to maintaining skeletal homeostasis (156). Together with the cancer studies, it is becoming increasingly clear that miR-335 has key roles in proliferation, migration, and differentiation through various mechanisms and pathways in a cell-dependent manner.  113  6.2.5 Significance of the MEST locus  Imprinted transcription at the murine Mest locus produces three notable products which seem to have mutually exclusive roles. Importantly, this locus is conserved in humans. Human miR-335 has already been shown to have important roles in cancer, as indicated by several reports showing it regulates metastasis and other cancerprogression pathways (135-137; 146; 147; 149; 151). Although the existence of MESTXL is not completely established in humans, supporting data comes from a study that identified an alternative 3’UTR for MEST that is at least 2.7 kb larger than the main 3’UTR and demonstrated that COPG2 intronic transcript 1 (CIT1), an antisense transcript located in intron 20 of COPG2, like Mit1 in mouse, was paternally expressed (112). Additionally, the disputed study in humans by Yamasaki et al. that reported COPG2 to be paternally expressed had analyzed a SNP located within intron 22 in RTPCR products that were not strand-specific, most likely detecting the overlapping predicted paternally expressed MESTXL transcripts that are presumably expressed at higher levels than the primary COPG2 transcript (113). Since COPG2 was reported to be biallelically expressed in all human fetal tissues examined, including those that were shown to express CIT1 (putative MESTXL transcripts), such as in fetal brain, lung, and liver tissue, it remains unclear whether MESTXL would have a similar role in regulating the allelic usage at COPG2 (112). If MESTXL is not involved in this type of regulation in humans, perhaps there is another undiscovered role for this longer 3’UTR as conservation of genes is usually indicative of important functions. Regarding the role of the MEST locus in the imprinted domain, it might also have the only DMR in the region, like in mouse, which might act to regulate imprinted genes in the region. The MMU6 imprinting domain shares homology with human chromosome 7q32 that has been associated with growth retardation and Silver-Russell syndrome (SRS) in maternal uniparental disomies for this chromosomal segment, where both copies are inherited from the mother. Nearly 10% of SRS cases are associated with mUPD for this region presumably due to the loss of paternally expressed genes (162; 163). As a consequence, MEST was designated as a key candidate locus for the aetiology of SRS in these cases yet extensive sequencing and methylation studies have failed to provide any evidence and the SRS aetiology associated with UPD at 7q32 remains ambiguous and unknown (164; 165). Since MestXL is paternally expressed and its loss affects  114  Copg2 expression, this raises the possibility that loss of MESTXL and/or increased COPG2 levels might be implicated in the aetiology of SRS. 6.3 Future directions  Here, some key features of the imprinted products from the Mest locus were investigated. The one that was not examined, the MEST protein, required work beyond the scope of this thesis and my field of study. MEST is predicted to be a member of the α/β hydrolase superfamily which include, but are not limited to, hydrolases, acyltransfereases, esterases, and lipases (238). MEST has been shown to localize to the endoplasmic reticulum membrane, at least in adipocytes, and possibly interact with LRP6 (lipoprotein receptor-related protein 6) to control its glycosylation (108; 109). The structure and substrates of the MEST protein remain unknown and uncovering them will undoubtedly provide invaluable information regarding the role of MEST specifically in development and in the adult brain. As discussed above, it is unclear whether MestXL codes for MEST and if MestXL has roles outside of the one demonstrated in Chapter 4. One way which might help determine whether MestXL codes for MEST would be to derive a neuronal Mest+/KO cell line, such as embryonic primary cortical neurons, that does not produce MEST and MestXL. A next step would involve in vitro transcribing MestXL, and introducing it into the Mest+/KO neuronal cell line. If this is managed, then antibodies can be used to detect MEST translated from the exogenous MestXL transcript, either by immunohistochemistry or western blotting. Along this line, the MestXL 3’UTR could be cloned downstream from a reporter gene such as EGFP to investigate if the protein is produced, again using antibodies against it by immunohistochemistry or looking for EGFP expression with a fluorescent microscope, when transfected into a neuronal cell line. Simultaneously we would also observe whether the MestXL 3’UTR changes the subcellular location of the reporter gene in comparison with a control that contains the normal Mest 3’UTR, a technique used to observe the subcellular location of two isoforms of Bdnf (187). Alternatively, we could try to knock down MestXL in a neuronal cell line by using siRNAs against different regions of MestXL to see if there is a reduction in MEST protein levels via RNAi, as there would never be a full knock-down of MEST since Mest mRNA is still producing MEST.  115  Our results suggest that MestXL does not localize to the nucleus. However, this question was assessed only through qRT-PCR on MestXL ESTs in nuclear and cytoplasmic fractions (See Figure 4.7). If MestXL is quickly processed and cleaved, perhaps we were assessing the localization of the fragments of MestXL post-processing which are cytoplasmic. One technique that could be utilized to actually visualize the localization of MestXL is fluorescent in situ hybridization. Although cytoplasmic localization would be observed with RNA probes against MestXL transcripts, perhaps we would also catch MestXL in the nucleus, if it indeed has some nuclear localization. This technique would also examine reveal if MestXL RNA localizes to any particular region of neural cells. A study on the two isoforms of Bdnf with different 3’UTRs interestingly showed that, in hippocampal neurons, the long 3’UTR was shown to be responsible for targeting Bdnf mRNA into neuronal dendrites, whereas the short 3’UTR mRNA was restricted to the somata (187). The Mest locus was shown to be under the regulation of alternative polyadenylation but the mechanism of this “read-through” remains unknown. As a starting point to investigate this question, plasmid vectors have been constructed in the laboratory by Aaron Bogutz to attempt to recapitulate this scenario in neural cells. In brief terms, the main construct consists of 1.4 kb of the MestXL 3’UTR (Mest 3’UTR plus the sequence that overlaps the Copg2 3’UTR) flanked by two reporter genes, a tomato reporter and a IRES-EGFP reporter, which are upstream and downstream from the MestXL 3’UTR segment, respectively. A control plasmid was also constructed where the MestXL segment is flipped between the reporter genes. Experiments are currently being carried out in the laboratory by Aaron Bogutz and Ting Gu to stably transfect these constructs into ES cell lines. If this is successful, then the plan is to differentiate them down the neural lineage with retinoic acid and use FACs to analyze expression from the reporter genes. The expectation is that only the tomato reporter will be expressed in the ES cells, whereas both will be expressed in the differentiated neural cells due to the “read-through” of the Mest 3’UTR. If this situation is indeed recapitulated in this system, further experiments include verifying by 3’RACE that the “read-through” is occurring and searching for involved important sequences and proteins that bind to the Mest 3’UTR to facilitate this process. Our proposed model that the MestXL RNAs interfere with Copg2 on the paternal chromosome can be strengthened and further explored. Firstly, performing experiments to show that Mest and Copg2 colocalize, either through double ISH or IHC, would 116  provide ultimate proof they are expressed together. Since Copg2 has been shown to be preferentially maternally expressed in mouse adult cortex and hypothalamus tissues, similar experiments utilized in Chapters 3 and 4 can be performed in Mest+/KO adult brain tissue to demonstrate the expected role for MestXL in regulating Copg2 in these neural tissues (120). It is intriguing that the Mest DMR might also act as an insulator to control the imprinting at the Klf14 locus. As discussed above and in Figure 6.1, CTCF binding and enhancers are thought to be involved in this mechanism of imprinting. A strategy that would prove useful in determining this would be to generate a mouse line where the Mest DMR is deleted and/or replaced with an alternative promoter that is not methylated to see if Klf14 expression still exists. If Klf14 expression is abolished due to absence of maternal DNA methylation, this would suggest that the Mest DMR is indeed responsible for Klf14 imprinting. Alternatively, if the CTCF binding site slightly upstream from the DMR (see Figure 6.1A) could be deleted without disturbing the gametic DMR and expression and imprinting of Mest, this would assess whether Klf14 imprinting also relies on CTCF-mediated regulation which would create an insulator on the unmethylated paternal allele. The expectation would be loss of imprinting of Klf14 as the paternal allele would not bind CTCF and therefore not insulate paternal Klf14 from enhancers. Additional experiments include discovering the involved enhancers with functional enhancer assays and validating them by performing in vivo experiments such as microinjecting the putative enhancer-lacZ cassettes into mouse eggs and assessing their expression at points during embryonic development through whole mount lacZ staining. Regarding miR-335, the initial main experiment to determine target genes was to generate conditional transgenic embryonic stem (ES) cell lines that could be further differentiated into certain lineages to study over-expression of miR-335 in different environments in a controlled manner. Generating this conditional transgenic ES cell line was indeed accomplished after much time and effort but, unfortunately, whether the conditional over-expression plasmid that had randomly integrated into the mouse genome operated to a considerable extent, if any, was questioned since no expected EGFP signal was observed upon activation of the “miR-335 transgene.” This was ultimately found to be due to a faulty IRES in the IRES-EGFP fragment. However, once the Taqman assay for miR-335 was obtained, the expression levels were assessed in some of these transgenic ES cell lines and were found to be over-expressed 2.5-4-fold compared to the parental ES cell line with the inactivated transgene (data not shown). In 117  light of this, it might be worth pursuing these transgenic ES cell lines if this overexpression is enough to significantly change expression of target genes. After differentiating into a lineage of choice, RNA-seq could be used again to identify candidate target genes. Conversely, different cell lines could also be generated from Mest+/KO embryos, such as a primary cortical neuronal cell line, to explore the underexpression of miR-335 in different environments. Additional experiments include making a conditional knock out of miR-335 to determine the ultimate biological phenotype of loss of miR-335. In undertaking these studies, the mutant Mest+/KO line was the common tool utilized and obviously proved very valuable. Once thought to only have loss of the MEST protein, we now know the Mest+/KO line additionally does not express MestXL and has considerably lower levels of miR-335 (106). Conceivably, these losses might also contribute to the cause of the observed growth retardation and abnormal postnatal behavior phenotypes in Mest+/KO mice which complicates the interpretation (106). For example, MestXL appears to be neural-specific in mice, at least in our intra-specific F1 hybrid cross, and it is speculated whether the behavior exhibited by Mest+/KO mothers that do not nurture their newborn pups might solely or partly be due from the loss of MestXL and over-expression of Copg2. It would therefore be interesting to isolate the loss of MestXL, miR-335, and MEST, by making new mouse lines to dissect out the contribution of each, if any, to the observed phenotypes in Mest+/KO mice.  118  References 1.  2. 3. 4. 5. 6.  7.  8. 9. 10. 11. 12. 13.  14.  15.  16.  17. 18.  Surani, M., Barton, S., and Norris, M. (1984) Development of reconstituted mouse eggs suggests imprinting of the genome during gametogeneisis, Nature 308, 548550. McGrath, J., and Solter, D. (1984) Completion of mouse embryogenesis requires both the maternal and paternal genomes, Cell 37, 179-183. Barton, S. C., Surani, M. A., and Norris, M. L. (1984) Role of paternal and maternal genomes in mouse development, Nature 311, 374-376. Surani, M. A. (1994) Genomic imprinting: control of gene expression by epigenetic inheritance, Current Opinion in Cell Biology 6, 390-395. Morison, I. M., Ramsay, J. P., and Spencer, H. G. (2005) A census of mammalian imprinting, Trends in genetics 21, 457-465. Nikaido, I., Saito, C., Mizuno, Y., Meguro, M., Bono, H., Kadomura, M., Kono, T., Morris, G. A., Lyons, P. A., Oshimura, M., Hayashizaki, Y., and Okazaki, Y. (2003) Discovery of imprinted transcripts in the mouse transcriptome using large-scale expression profiling, Genome Research 13, 1402-1409. Luedi, P. P., Dietrich, F. S., Weidman, J. R., Bosko, J. M., Jirtle, R. L., and Hartemink, A. J. (2007) Computational and experimental identification of novel human imprinted genes, Genome Research 17, 1723-1730. Moore, T., and Haig, D. (1991) Genomic imprinting in mammalian development: a parental tug-of-war, Trends in genetics 7, 45-49. Haig, D. (1996) Altercation of generations: genetic conflicts of pregnancy, American journal of reproductive immunology 35, 226-232. Moore, T., and Reik, W. (1996) Genetic conflict in early development: parental imprinting in normal and abnormal growth, Reviews of reproduction 1, 73-77. Varmuza, S., and Mann, M. (1994) Genomic imprinting--defusing the ovarian time bomb, Trends in Genetics 10, 118-123. Barlow, D. (1993) Methylation and imprinting: from host defense to gene regulation?, Science 260, 309-310. Parker-Katiraee, L., Carson, A. R., Yamada, T., Arnaud, P., Feil, R., Abu-Amero, S. N., Moore, G. E., Kaneda, M., Perry, G. H., Stone, A. C., Lee, C., Meguro-Horike, M., Sasaki, H., Kobayashi, K., Nakabayashi, K., and Scherer, S. W. (2007) Identification of the Imprinted KLF14 Transcription Factor Undergoing HumanSpecific Accelerated Evolution, PLoS Genetics 3, e65. Davis, T. L., Trasler, J. M., Moss, S. B., Yang, G. J., and Bartolomei, M. S. (1999) Acquisition of the H19 methylation imprint occurs differentially on the parental alleles during spermatogenesis, Genomics 58, 18-28. Li, J., Lees-Murdock, D. J., Xu, G., and Walsh, C. P. (2004) Timing of establishment of paternal methylation imprints in the mouse, Genomics 84, 952960. Lucifero, D., Mertineit, C., Clarke, H. J., Bestor, T. H., and Trasler, J. M. (2002) Methylation dynamics of imprinted genes in mouse germ cells, Genomics 79, 530538. Bourc'his, D., Xu, G. L., Lin, C. S., Bollman, B., and Bestor, T. H. (2001) Dnmt3L and the establishment of maternal genomic imprints, Science 294, 2536-2539. Hata, K., Okano, M., Lei, H., and Li, E. (2002) Dnmt3L cooperates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal imprints in mice, Development 129, 1983-1993.  119  19.  20.  21. 22.  23.  24. 25. 26.  27.  28.  29.  30.  31. 32. 33.  34.  35.  Kaneda, M., Okano, M., Hata, K., Sado, T., Tsujimoto, N., Li, E., and Sasaki, H. (2004) Essential role for de novo DNA methyltransferase Dnmt3a in paternal and maternal imprinting, Nature 429, 900-903. Jia, D., Jurkowska, R. Z., Zhang, X., Jeltsch, A., and Cheng, X. (2007) Structure of Dnmt3a bound to Dnmt3L suggests a model for de novo DNA methylation, Nature 449, 248-251. Ferguson-Smith, A. C., and Greally, J. M. (2007) Epigenetics: perceptive enzymes, Nature 449, 148-149. Ooi, S. K., Qiu, C., Bernstein, E., Li, K., Da Jia, Yang, Z., Erdjument-Bromage, H., Tempst, P., Lin, S., Allis, C. D., Cheng, X., and Bestor, T. H. (2007) DNMT3L connects unmethylated lysine 4 of histone H3 to de novo methylation of DNA, Nature 448, 714-717. Chotalia, M., Smallwood, S. A., Ruf, N., Dawson, C., Lucifero, D., Frontera, M., James, K., Dean, W., and Kelsey, G. (2009) Transcription is required for establishment of germline methylation marks at imprinted genes, Genes & Development 23, 105-117. Bartolomei, M. S., and Ferguson-Smith, A. C. (2011) Mammalian Genomic Imprinting, Cold Spring Harbor Perspectives in Biology 3, a002592-a002592. Li, E., Beard, C., and Jaenisch, R. (1993) Role for DNA methylation in genomic imprinting, Nature 366, 362-365. Nakamura, T., Arai, Y., Umehara, H., Masuhara, M., Kimura, T., Taniguchi, H., Sekimoto, T., Ikawa, M., Yoneda, Y., Okabe, M., Tanaka, S., Shiota, K., and Nakano, T. (2006) PGC7/Stella protects against DNA demethylation in early embryogenesis, Nature Cell Biology 9, 64-71. Li, X., Ito, M., Zhou, F., Youngson, N., Zuo, X., Leder, P., and Ferguson-Smith, A. C. (2008) A Maternal-Zygotic Effect Gene, Zfp57, Maintains Both Maternal and Paternal Imprints, Developmental Cell 15, 547-557. Hajkova, P., Erhardt, S., Lane, N., Haaf, T., El-Maarri, O., Reik, W., Walter, J., and Surani, M. A. (2002) Epigenetic reprogramming in mouse primordial germ cells, Mechanisms of Development 117, 15-23. Hajkova, P., Ancelin, K., Waldmann, T., Lacoste, N., Lange, U. C., Cesari, F., Lee, C., Almouzni, G., Schneider, R., and Surani, M. A. (2008) Chromatin dynamics during epigenetic reprogramming in the mouse germ line, Nature 452, 877-881. Sasaki, H., and Matsui, Y. (2008) Epigenetic events in mammalian germ-cell development: reprogramming and beyond, Nature Reviews Genetics 2008, 129140. Pauler, F. M., and Barlow, D. P. (2006) Imprinting mechanisms--it only takes two, Genes & Development 20, 1203-1206. Royo, H., and Cavaillé, J. (2008) Non-coding RNAs in imprinted gene clusters, Biology of the Cell 100, 149-166. Yang, Y., Li, T., Vu, T. H., Ulaner, G. A., Hu, J., and Hoffman, A. R. (2003) The histone code regulating expression of the imprinted mouse Igf2r gene, Endocrinology 144, 5658-5670. Lewis, A., Mitsuya, K., Umlauf, D., Smith, P., Dean, W., Walter, J., Higgins, M., Feil, R., and Reik, W. (2004) Imprinting on distal chromosome 7 in the placenta involves repressive histone methylation independent of DNA methylation, Nature Genetics 36, 1291-1295. Regha, K., Sloane, M. A., Huang, R., Pauler, F. M., Warczok, K. E., Melikant, B., Radolf, M., Martens, J. H., Schotta, G., Jenuwein, T., and Barlow, D. P. (2007) Active and Repressive Chromatin Are Interspersed without Spreading in an Imprinted Gene Cluster in the Mammalian Genome, Molecular Cell 27, 353-366. 120  36.  37.  38.  39. 40. 41.  42.  43. 44.  45.  46.  47. 48. 49.  50.  51.  Wen, B., Wu, H., Bjornsson, H., Green, R. D., Irizarry, R., and Feinberg, A. P. (2008) Overlapping euchromatin/heterochromatin- associated marks are enriched in imprinted gene regions and predict allele-specific modification, Genome Research 18, 1806-1813. Thorvaldsen, J. L., Duran, K. L., and Bartolomei, M. S. (1998) Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2, Genes & Development 12, 3693-3702. Hark, A. T., Schoenherr, C. J., Katz, D. J., Ingram, R. S., Levorse, J. M., and Tilghman, S. M. (2000) CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus, Nature 405, 486-489. Bell, A. C., and Felsenfeld, G. (2000) Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene, Nature 405, 482-485. Ferguson-Smith, A. C. (2011) Genomic imprinting: the emergence of an epigenetic paradigm, Nature Reviews Genetics 12, 565-575. Ohlsson, R., Renkawitz, R., and Lobanenkov, V. (2001) CTCF is a uniquely versatile transcription regulator linked to epigenetics and disease, Trends in genetics 17, 520-527. Yoon, B., Herman, H., Hu, B., Park, Y. J., Lindroth, A., Bell, A., West, A. G., Chang, Y., Stablewski, A., Piel, J. C., Loukinov, D. I., Lobanenkov, V. V., and Soloway, P. D. (2005) Rasgrf1 imprinting is regulated by a CTCF-dependent methylation-sensitive enhancer blocker, Molecular and Cellular Biology 25, 1118411190. Sleutels, F., Zwart, R., and Barlow, D. P. (2002) The non-coding Air RNA is required for silencing autosomal imprinted genes, Nature 415, 810-813. Horike, S., Mitsuya, K., Meguro, M., Kotobuki, N., Kashiwagi, A., Notsu, T., Schulz, T. C., Shirayoshi, Y., and Oshimura, M. (2000) Targeted disruption of the human LIT1 locus defines a putative imprinting control element playing an essential role in Beckwith-Wiedemann syndrome, Human Molecular Genetics 9, 2075-2083. Mancini-DiNardo, D., Steele, S. J., Levorse, J. M., Ingram, R. S., and Tilghman, S. M. (2006) Elongation of the Kcnq1ot1 transcript is required for genomic imprinting of neighboring genes, Genes & Development 20, 1268-1282. Lee, M. P., DeBaun, M. R., Mitsuya, K., Galonek, H. L., Brandenburg, S., Oshimura, M., and Feinberg, A. P. (1999) Loss of imprinting of a paternally expressed transcript, with antisense orientation to KVLQT1, occurs frequently in Beckwith-Wiedemann syndrome and is independent of insulin-like growth factor II imprinting, Proceedings of the National Academy of Sciences of the United States of America 96, 5203-5208. Enklaar, T., Zabel, B. U., and Prawitt, D. (2006) Beckwith-Wiedemann syndrome: multiple molecular mechanisms, Expert reviews in molecular medicine 8, 1-19. Prickett, A. R., and Oakey, R. J. (2012) A survey of tissue-specific genomic imprinting in mammals, Molecular Genetics and Genomics 287, 621-630. Umlauf, D., Goto, Y., Cao, R., Cerqueira, F., Wagschal, A., Zhang, Y., and Feil, R. (2004) Imprinting along the Kcnq1 domain on mouse chromosome 7 involves repressive histone methylation and recruitment of Polycomb group complexes, Nature Genetics 36, 1296-1300. Hudson, Q. J., Kulinski, T. M., Huetter, S. P., and Barlow, D. P. (2010) Genomic imprinting mechanisms in embryonic and extraembryonic mouse tissues, Heredity 105, 45-56. Wilkinson, L. S., Davies, W., and Isles, A. R. (2007) Genomic imprinting effects on brain development and function, Nature reviews. Neuroscience 8, 832-843. 121  52. 53. 54.  55.  56.  57.  58.  59.  60.  61.  62.  63.  64.  65.  66.  Khatib, H. (2007) Is it genomic imprinting or preferential expression?, BioEssays : news and reviews in molecular, cellular and developmental biology 29, 1022-1028. Mercer, T. R., Dinger, M. E., and Mattick, J. S. (2009) Long non-coding RNAs: insights into functions, Nature Reviews Genetics 10, 155-159. Mancini-DiNardo, D. (2003) A differentially methylated region within the gene Kcnq1 functions as an imprinted promoter and silencer, Human Molecular Genetics 12, 283-294. Yamasaki, Y. (2005) Neuron-specific relaxation of Igf2r imprinting is associated with neuron-specific histone modifications and lack of its antisense transcript Air, Human Molecular Genetics 14, 2511-2520. Santoro, F., and Barlow, D. P. (2011) Developmental control of imprinted expression by macro non-coding RNAs, Seminars in Cell & Developmental Biology 22, 328-335. Le Meur, E., Watrin, F., Landers, M., Sturny, R., Lalande, M., and Muscatelli, F. (2005) Dynamic developmental regulation of the large non-coding RNA associated with the mouse 7C imprinted chromosomal region, Developmental Biology 286, 587-600. Vitali, P., Royo, H., Marty, V., Bortolin-Cavaillé, M., and Cavaillé, J. (2009) Long nuclear-retained non-coding RNAs and allele-specific higher-order chromatin organization at imprinted snoRNA gene arrays, Journal of Cell Science 123, 70-83. Cavaillé, J., Vitali, P., Basyuk, E., Hüttenhofer, A., and Bachellerie, J. P. (2001) A novel brain-specific box C/D small nucleolar RNA processed from tandemly repeated introns of a noncoding RNA gene in rats, The Journal of biological chemistry 276, 26374-26383. Cavaillé, J., Seitz, H., Paulsen, M., Ferguson-Smith, A. C., and Bachellerie, J. (2002) Identification of tandemly-repeated C/D snoRNA genes at the imprinted human 14q32 domain reminiscent of those at the Prader-Willi/Angelman syndrome region, Human Molecular Genetics 11, 1527-1538. Runte, M., Hüttenhofer, A., Gross, S., Kiefmann, M., Horsthemke, B., and Buiting, K. (2001) The IC-SNURF-SNRPN transcript serves as a host for multiple small nucleolar RNA species and as an antisense RNA for UBE3A, Human Molecular Genetics 10, 2687-2700. Cavaillé, J., Buiting, K., Kiefmann, M., Lalande, M., Brannan, C. I., Horsthemke, B., Bachellerie, J. P., Brosius, J., and Hüttenhofer, A. (2000) Identification of brainspecific and imprinted small nucleolar RNA genes exhibiting an unusual genomic organization, Proceedings of the National Academy of Sciences of the United States of America 97, 14311-14316. Vitali, P., Basyuk, E., Le Meur, E., Bertrand, E., Muscatelli, F., Cavaillé, J., and Huttenhofer, A., (2005) ADAR2-mediated editing of RNA substrates in the nucleolus is inhibited by C/D small nucleolar RNAs, The Journal of Cell Biology 169, 745-753. Davis, E., Caiment, F., Tordoir, X., Cavaillé, J., Ferguson-Smith, A., Cockett, N., Georges, M., and Charlier, C. (2005) RNAi-Mediated Allelic trans-Interaction at the Imprinted Rtl1/Peg11 Locus, Current Biology 15, 743-749. Brennecke, J., Aravin, A. A., Stark, A., Dus, M., Kellis, M., Sachidanandam, R., and Hannon, G. J. (2007) Discrete Small RNA-Generating Loci as Master Regulators of Transposon Activity in Drosophila, Cell 128, 1089-1103. Carmell, M. A., Girard, A., van de Kant, H. J., Bourc'his, D., Bestor, T. H., de Rooij, D. G., and Hannon, G. J. (2007) MIWI2 Is Essential for Spermatogenesis and Repression of Transposons in the Mouse Male Germline, Developmental Cell 12, 503-514. 122  67.  68.  69.  70.  71.  72.  73.  74.  75.  76. 77. 78.  79.  80.  81.  Kuramochi-Miyagawa, S., Watanabe, T., Gotoh, K., Totoki, Y., Toyoda, A., Ikawa, M., Asada, N., Kojima, K., Yamaguchi, Y., Ijiri, T. W., Hata, K., Li, E., Matsuda, Y., Kimura, T., Okabe, M., Sakaki, Y., Sasaki, H., and Nakano, T. (2008) DNA methylation of retrotransposon genes is regulated by Piwi family members MILI and MIWI2 in murine fetal testes, Genes Dev 22, 908-917. Meng, L., Person, R. E., and Beaudet, A. L. (2012) Ube3a-ATS is an atypical RNA polymerase II transcript that represses the paternal expression of Ube3a, Human Molecular Genetics 21, 3001-3012. Landers, M., Bancescu, D. L., Le Meur, E., Rougeulle, C., Glatt-Deeley, H., Brannan, C., Muscatelli, F., and Lalande, M. (2004) Regulation of the large (approximately 1000 kb) imprinted murine Ube3a antisense transcript by alternative exons upstream of Snurf/Snrpn, Nucleic Acids Research 32, 34803492. Johnstone, K. A., Dubose, A. J., Futtner, C. R., Elmore, M. D, Brannan, C. L., and Resnick, J.L. (2005) A human imprinting centre demonstrates conserved acquisition but diverged maintenance of imprinting in a mouse model for Angelman syndrome imprinting defects, Human Molecular Genetics 15, 393-404. Lin, S., Youngson, N., Takada, S., Seitz, H., Reik, W., Paulsen, M., Cavaillé, J., and Ferguson-Smith, A. C. (2003) Asymmetric regulation of imprinting on the maternal and paternal chromosomes at the Dlk1-Gtl2 imprinted cluster on mouse chromosome 12, Nature Genetics 35, 97-102. Lin, S., Coan, P., da Rocha, S. T., Seitz, H., Cavaillé, J., Teng, P., Takada, S., and Ferguson-Smith, A. C. (2007) Differential regulation of imprinting in the murine embryo and placenta by the Dlk1-Dio3 imprinting control region, Development 134, 417-426. Seitz, H., Royo, H., Bortolin, M., Lin, S., Ferguson-Smith, A. C., and Cavaillé, J. (2004) A large imprinted microRNA gene cluster at the mouse Dlk1-Gtl2 domain, Genome Research 14, 1741-1748. Tierling, S., Dalbert, S., Schoppenhorst, S., Tsai, C., Oliger, S., Ferguson-Smith, A. C., Paulsen, M., and Walter, J. (2006) High-resolution map and imprinting analysis of the Gtl2–Dnchc1 domain on mouse chromosome 12, Genomics 87, 225-235. Da Rocha, S. T., Edwards, C. A., Ito, M., Ogata, T., and Ferguson-Smith, A. C. (2008) Genomic imprinting at the mammalian Dlk1-Dio3 domain, Trends in Genetics 24, 306-316. Peters, J., and Williamson, C. M. (2007) Control of imprinting at the Gnas cluster, Epigenetics 2, 207-13. Cai, X., and Cullen, B. R. (2007) The imprinted H19 noncoding RNA is a primary microRNA precursor, RNA 13, 313-316. Wevrick, R., Kerns, J. A., and Francke, U. (1994) Identification of a novel paternally expressed gene in the Prader-Willi syndrome region, Human Molecular Genetics 3, 1877-1882. Wevrick, R., Kerns, J., and Francke, U. (1996) The IPW gene is imprinted and is not expressed in the Prader-Willi syndrome, Acta Genet Gemellol (Roma) 45, 191197. Ding, F., Prints, Y., Dhar, M. S., Johnson, D. K., Garnacho–Montero, C., Nicholls, R. D., and Francke, U. (2005) Lack of Pwcr1/MBII-85 snoRNA is critical for neonatal lethality in Prader–Willi syndrome mouse models, Mammalian Genome 16, 424-431. Duker, A. L., Ballif, B. C., Bawle, E. V., Person, R. E., Mahadevan, S., Alliman, S., Thompson, R., Traylor, R., Bejjani, B. A., Shaffer, L. G., Rosenfeld, J. A., Lamb, A. N., and Sahoo, T. (2010) Paternally inherited microdeletion at 15q11.2 confirms a 123  82. 83.  84.  85.  86.  87.  88.  89.  90.  91.  92.  93.  94.  95.  significant role for the SNORD116 C/D box snoRNA cluster in Prader–Willi syndrome, European Journal of Human Genetics 18, 1196-1201. Kishore, S., and Stamm, S. (2006) The snoRNA HBII-52 Regulates Alternative Splicing of the Serotonin Receptor 2C, Science 311, 230-232. Kishore, S., Khanna, A., Zhang, Z., Hui, J., Balwierz, P. J., Stefan, M., Beach, C., Nicholls, R. D., Zavolan, M., and Stamm, S. (2010) The snoRNA MBII-52 (SNORD 115) is processed into smaller RNAs and regulates alternative splicing, Human Molecular Genetics 19, 1153-1164. Nielsen, R., Bustamante, C., Clark, A. G., Glanowski, S., Sackton, T. B., Hubisz, M. J., Fledel-Alon, A., Tanenbaum, D. M., Civello, D., White, T. J., Sninsky, J. J., Adams, M. D., and Cargill, M. (2005) A Scan for Positively Selected Genes in the Genomes of Humans and Chimpanzees, PLoS Biology 3, e170. Färber, C., Gross, S., Neesen, J., Buiting, K., and Horsthemke, B. (2000) Identification of a testis-specific gene (C15orf2) in the Prader-Willi syndrome region on chromosome 15, Genomics 65, 174-183. Hagan, J. P., O'Neill, B. L., Stewart, C. L., Kozlov, S. V., and Croce, C. M. (2009) At least ten genes define the imprinted Dlk1-Dio3 cluster on mouse chromosome 12qF1, PLoS ONE 4, e4352. Xiao, Y., Zhou, H., and Qu, L. (2006) Characterization of three novel imprinted snoRNAs from mouse Irm gene, Biochemical and Biophysical Research Communications 340, 1217-1223. Meng, F., Wehbe-Janek, H., Henson, R., Smith, H., and Patel, T. (2008) Epigenetic regulation of microRNA-370 by interleukin-6 in malignant human cholangiocytes, Oncogene 27, 378-386. Kawahara, Y., Zinshteyn, B., Sethupathy, P., Iizasa, H., Hatzigeorgiou, A. G., and Nishikura, K. (2007) Redirection of Silencing Targets by Adenosine-to-Inosine Editing of miRNAs, Science 315, 1137-1140. Schratt, G. M., Tuebing, F., Nigh, E. A., Kane, C. G., Sabatini, M. E., Kiebler, M., and Greenberg, M. E. (2006) A brain-specific microRNA regulates dendritic spine development, Nature 439, 283-289. Saito, Y., Liang, G., Egger, G., Friedman, J. M., Chuang, J. C., Coetzee, G. A., and Jones, P. A. (2006) Specific activation of microRNA-127 with downregulation of the proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells, Cancer Cell 9, 435-443. Iglesias-Platas, I., Martin-Trujillo, A., Cirillo, D., Court, F., Guillaumet-Adkins, A., Camprubi, C., Bourc’his, D., Hata, K., Feil, R., Tartaglia, G., Arnaud, P., and Monk, D. (2012) Characterization of Novel Paternal ncRNAs at the Plagl1 Locus, Including Hymai, Predicted to Interact with Regulators of Active Chromatin, PLoS ONE 7, e38907. Beechey, C. V. (2000) Peg1/Mest locates distal to the currently defined imprinting region on mouse proximal chromosome 6 and identifies a new imprinting region affecting growth, Cytogenetics and cell genetics 90, 309-314. Beechey, C. V. (2004) A reassessment of imprinting regions and phenotypes on mouse chromosome 6: Nap1l5 locates within the currently defined sub-proximal imprinting region, Cytogenetic and Genome Research 107, 108-114. Kaneko-Ishino, T., Kuroiwa, Y., Miyoshi, N., Kohda, T., Suzuki, R., Yokoyama, M., Viville, S., Barton, S. C., Ishino, F., and Surani, M. A. (1995) Peg1/Mest imprinted gene on chromosome 6 identified by cDNA subtraction hybridization, Nature Genetics 11, 52-59.  124  96.  97.  98.  99.  100.  101.  102.  103.  104.  105.  106.  107.  108. 109.  110.  Lee, Y. J., Park, C. W., Hahn, Y., Park, J., Lee, J., Yun, J. H., Hyun, B., and Chung, J. H. (2000) Mit1/Lb9 and Copg2, new members of mouse imprinted genes closely linked to Peg1/Mest, FEBS Letters 472, 230-234. Scohy, S., Gabant, P., Van Reeth, T., Hertveldt, V., Drèze, P. L., Van Vooren, P., Rivière, M., Szpirer, J., and Szpirer, C. (2000) Identification of KLF13 and KLF14 (SP6), novel members of the SP/XKLF transcription factor family, Genomics 70, 93-101. Kim, J., Krichevsky, A., Grad, Y., Hayes, G. D., Kosik, K. S., Church, G. M., and Ruvkun, G. (2004) Identification of many microRNAs that copurify with polyribosomes in mammalian neurons, Proceedings of the National Academy of Sciences of the United States of America 101, 360-365. Jeong, S., Lee, Y. J., Jang, J. S., Park, C. W., Chung, J. H., Seong, J. K., Lee, K. K., and Yu, D. Y. (2000) A novel epigenetic control operating on Vme1+ locus leads to variegated monoallelic expression, Biochemical and Biophysical Research Communications 279, 884-890. Yamada, T., Mitsuya, K., Kayashima, T., Yamasaki, K., Ohta, T., Yoshiura, K., Matsumoto, N., Yamada, H., Minakami, H., Oshimura, M., Niikawa, N., and Kishino, T. (2004) Imprinting analysis of 10 genes and/or transcripts in a 1.5-Mb MEST-flanking region at human chromosome 7q32, Genomics 83, 402-412. Lefebvre, L., Viville, S., Barton, S. C., Ishino, F., and Surani, M. A. (1997) Genomic structure and parent-of-origin-specific methylation of Peg1, Human Molecular Genetics 6, 1907-1915. Sansom, S. N., Hébert, J. M., Thammongkol, U., Smith, J., Nisbet, G., Surani, M. A., McConnell, S. K., and Livesey, F. J. (2005) Genomic characterisation of a Fgfregulated gradient-based neocortical protomap, Development 132, 3947-3961. Hahn, Y., Yang, S. K., and Chung, J. H. (2005) Structure and expression of the zebrafish mest gene, an ortholog of mammalian imprinted gene PEG1/MEST, Biochimica et biophysica acta 1731, 125-132. Suzuki, S., Renfree, M. B., Pask, A. J., Shaw, G., Kobayashi, S., Kohda, T., Kaneko-Ishino, T., and Ishino, F. (2005) Genomic imprinting of IGF2, p57(KIP2) and PEG1/MEST in a marsupial, the tammar wallaby, Mechanisms of Development 122, 213-222. Mayer, W., Hemberger, M., Frank, H. G., Grümmer, R., Winterhager, E., Kaufmann, P., and Fundele, R. (2000) Expression of the imprinted genes MEST/Mest in human and murine placenta suggests a role in angiogenesis, Developmental dynamics 217, 1-10. Lefebvre, L., Viville, S., Barton, S. C., Ishino, F., Keverne, E. B., and Surani, M. A. (1998) Abnormal maternal behaviour and growth retardation associated with loss of the imprinted gene Mest, Nature Genetics 20, 163-169. Reule, M., Krause, R., Hemberger, M., and Fundele, R. (1998) Analysis of Peg1/Mest imprinting in the mouse, Development Genes and Evolution 208, 161163. Jung, H., Lee, S. K., and Jho, E. (2011) Mest/Peg1 inhibits Wnt signalling through regulation of LRP6 glycosylation, The Biochemical journal 436, 263-269. Nikonova, L., Koza, R. A., Mendoza, T., Chao, P., Curley, J. P., and Kozak, L. P. (2008) Mesoderm-specific transcript is associated with fat mass expansion in response to a positive energy balance, The FASEB journal 22, 3925-3937. Bischoff, S. R., Tsai, S., Hardison, N., Motsinger-Reif, A. A., Freking, B. A., Nonneman, D., Rohrer, G., and Piedrahita, J. A. (2009) Characterization of conserved and nonconserved imprinted genes in swine, Biology of reproduction 81, 906-920. 125  111. Dindot, S. V., Person, R., Strivens, M., Garcia, R., and Beaudet, A. L. (2009) Epigenetic profiling at mouse imprinted gene clusters reveals novel epigenetic and genetic features at differentially methylated regions, Genome Research 19, 13741383. 112. Yamasaki, K., Hayashida, S., Miura, K., Masuzaki, H., Ishimaru, T., Niikawa, N., and Kishino, T. (2000) The novel gene, gamma2-COP (COPG2), in the 7q32 imprinted domain escapes genomic imprinting, Genomics 68, 330-335. 113. Blagitko, N., Schulz, U., Schinzel, A. A., Ropers, H. H., and Kalscheuer, V. M. (1999) gamma2-COP, a novel imprinted gene on chromosome 7q32, defines a new imprinting cluster in the human genome, Human Molecular Genetics 8, 23872396. 114. Brunner, B., Grützner, F., Yaspo, M., Ropers, H., Haaf, T., and Kalscheuer, V. (2000) Molecular cloning and characterization of the Fugu rubripes MEST/COPG2 imprinting cluster and chromosomal localization in Fugu and Tetraodon nigroviridis., Chromosome Res 8, 465-476. 115. Stenbeck, G., Schreiner, R., Herrmann, D., Auerbach, S., Lottspeich, F., Rothman, J. E., and Wieland, F. T. (1992) Gamma-COP, a coat subunit of non-clathrincoated vesicles with homology to Sec21p, FEBS Letters 314, 195-198. 116. Lowe, M., and Kreis, T. E. (1998) Regulation of membrane traffic in animal cells by COPI, Biochimica et biophysica acta 1404, 53-66. 117. Khatib, H. (2005) The COPG2, DCN, and SDHD genes are biallelically expressed in cattle, Mammalian genome 16, 545-552. 118. Khatib, H. (2005) Characterization and analysis of the imprinting status of the ovine SDHD and COPG2 genes, Animal genetics 36, 186-188. 119. Yun, J., Park, C. W., Lee, Y. J., and Chung, J. H. (2003) Allele-specific methylation at the promoter-associated CpG island of mouse Copg2, Mammalian genome 14, 376-382. 120. Gregg, C., Zhang, J., Weissbourd, B., Luo, S., Schroth, G. P., Haig, D., and Dulac, C. (2010) High-resolution analysis of parent-of-origin allelic expression in the mouse brain, Science 329, 643-648. 121. Truty, M. J., Lomberk, G., Fernandez-Zapico, M. E., and Urrutia, R. (2008) Silencing of the Transforming Growth Factor- (TGF ) Receptor II by Kruppel-like Factor 14 Underscores the Importance of a Negative Feedback Mechanism in TGF Signaling, Journal of Biological Chemistry 284, 6291-6300. 122. Small, K. S., Hedman, A. K., Grundberg, E., Nica, A. C., Thorleifsson, G., Kong, A., Thorsteindottir, U., Shin, S., Richards, H. B., Soranzo, N., Ahmadi, K. R., Lindgren, C. M., Stefansson, K., Dermitzakis, E. T., Deloukas, P., Spector, T. D., and McCarthy, M. I. (2011) Identification of an imprinted master trans regulator at the KLF14 locus related to multiple metabolic phenotypes, Nature Genetics 43, 561564. 123. Mercer, T. R., Dinger, M. E., Sunkin, S. M., Mehler, M. F., and Mattick, J. S. (2008) Specific expression of long noncoding RNAs in the mouse brain, Proceedings of the National Academy of Sciences of the United States of America 105, 716-721. 124. Denli, A. M., Tops, B. B., Plasterk, R. H., Ketting, R. F., and Hannon, G. J. (2004) Processing of primary microRNAs by the Microprocessor complex, Nature 432, 231-235. 125. Gregory, R. I., Yan, K., Amuthan, G., Chendrimada, T., Doratotaj, B., Cooch, N., and Shiekhattar, R. (2004) The Microprocessor complex mediates the genesis of microRNAs, Nature 432, 235-240. 126  126. Han, J., Lee, Y., Yeom, K., Kim, Y., Jin, H., and Kim, V. N. (2004) The DroshaDGCR8 complex in primary microRNA processing, Genes & Development 18, 3016-3027. 127. Landthaler, M., Yalcin, A., and Tuschl, T. (2004) The human DiGeorge syndrome critical region gene 8 and Its D. melanogaster homolog are required for miRNA biogenesis, Current biology 14, 2162-2167. 128. Cullen, B. R. (2004) Transcription and processing of human microRNA precursors, Molecular Cell 16, 861-865. 129. Pillai, R. S., Bhattacharyya, S. N., Artus, C. G., Zoller, T., Cougot, N., Basyuk, E., Bertrand, E., and Filipowicz, W. (2005) Inhibition of translational initiation by Let-7 MicroRNA in human cells, Science 309, 1573-1576. 130. Pillai, R. S. (2005) MicroRNA function: multiple mechanisms for a tiny RNA?, RNA 11, 1753-1761. 131. Valencia-Sanchez, M. A., Liu, J., Hannon, G. J., and Parker, R. (2006) Control of translation and mRNA degradation by miRNAs and siRNAs, Genes & Development 20, 515-524. 132. Sevignani, C., Calin, G. A., Siracusa, L. D., and Croce, C. M. (2006) Mammalian microRNAs: a small world for fine-tuning gene expression, Mammalian genome 17, 189-202. 133. Ronchetti, D., Lionetti, M., Mosca, L., Agnelli, L., Andronache, A., Fabris, S., Deliliers, G., and Neri, A. (2008) An integrative genomic approach reveals coordinated expression of intronic miR-335, miR-342, and miR-561 with deregulated host genes in multiple myeloma, BMC Medical Genomics 1, 37-46. 134. Tomé, M., López-Romero, P., Albo, C., Sepúlveda, J. C., Fernández-Gutiérrez, B., Dopazo, A., Bernad, A., and González, M. A. (2011) miR-335 orchestrates cell proliferation, migration and differentiation in human mesenchymal stem cells, Cell Death and Differentiation 18, 985-995. 135. Png, K. J., Yoshida, M., Zhang, X. H., Shu, W., Lee, H., Rimner, A., Chan, T. A., Comen, E., Andrade, V. P., Kim, S. W., King, T. A., Hudis, C. A., Norton, L., Hicks, J., Massagué, J., and Tavazoie, S. F. (2011) MicroRNA-335 inhibits tumor reinitiation and is silenced through genetic and epigenetic mechanisms in human breast cancer, Genes & Development 25, 226-231. 136. Dohi, O., Yasui, K., Gen, Y., Takada, H., Endo, M., Tsuji, K., Konishi, C., Yamada, N., Mitsuyoshi, H., Yagi, N., Naito, Y., Tanaka, S., Arii, S., and Yoshikawa, T. (2013) Epigenetic silencing of miR-335 and its host gene MEST in hepatocellular carcinoma, International journal of oncology 42, 411-418. 137. Tavazoie, S. F., Alarcón, C., Oskarsson, T., Padua, D., Wang, Q., Bos, P. D., Gerald, W. L., and Massagué, J. (2008) Endogenous human microRNAs that suppress breast cancer metastasis, Nature 451, 147-152. 138. Heyn, H., Engelmann, M., Schreek, S., Ahrens, P., Lehmann, U., Kreipe, H., Schlegelberger, B., and Beger, C. (2011) MicroRNA miR-335 is crucial for the BRCA1 regulatory cascade in breast cancer development, International Journal of Cancer 129, 2797-2806. 139. Sorrentino, A., Liu, C., Addario, A., Peschle, C., Scambia, G., and Ferlini, C. (2008) Role of microRNAs in drug-resistant ovarian cancer cells, Gynecologic Oncology 111, 478-486. 140. Zhang, H., Luo, X., Zhang, P., Huang, L., Zheng, Y., Wu, J., Zhou, H., Qu, L., Xu, L., and Chen, Y. (2009) MicroRNA Patterns Associated with Clinical Prognostic Parameters and CNS Relapse Prediction in Pediatric Acute Leukemia, PLoS ONE 4, e7826. 127  141. Hummel, R., Hussey, D. J., and Haier, J. (2010) MicroRNAs: Predictors and modifiers of chemo- and radiotherapy in different tumour types, European Journal of Cancer 46, 298-311. 142. Soon, P. S., Tacon, L. J., Gill, A. J., Bambach, C. P., Sywak, M. S., Campbell, P. R., Yeh, M. W., Wong, S. G., Clifton-Bligh, R. J., Robinson, B. G., and Sidhu, S. B. (2009) miR-195 and miR-483-5p Identified as Predictors of Poor Prognosis in Adrenocortical Cancer, Clinical Cancer Research 15, 7684-7692. 143. Wang, Y. X., Zhang, X. Y., Zhang, B. F., Yang, C. Q., Chen, X. M., and Gao, H. J. (2010) Initial study of microRNA expression profiles of colonic cancer without lymph node metastasis, Journal of Digestive Diseases 11, 50-54. 144. Scarola, M., Schoeftner, S., Schneider, C., and Benetti, R. (2010) miR-335 directly targets Rb1 (pRb/p105) in a proximal connection to p53-dependent stress response, Cancer Research 70, 6925-6933. 145. Schmitz, K. J., Helwig, J., Bertram, S., Sheu, S. Y., Suttorp, A. C., Seggewiss, J., Willscher, E., Walz, M. K., Worm, K., and Schmid, K. W. (2011) Differential expression of microRNA-675, microRNA-139-3p and microRNA-335 in benign and malignant adrenocortical tumours, Journal of Clinical Pathology 64, 529-535. 146. Shi, L., Jiang, D., Sun, G., Wan, Y., Zhang, S., Zeng, Y., Pan, T., and Wang, Z. (2012) miR-335 promotes cell proliferation by directly targeting Rb1 in meningiomas, Journal of neuro-oncology 110, 155-62. 147. Xu, Y., Zhao, F., Wang, Z., Song, Y., Luo, Y., Zhang, X., Jiang, L., Sun, Z., Miao, Z., and Xu, H. (2011) MicroRNA-335 acts as a metastasis suppressor in gastric cancer by targeting Bcl-w and specificity protein 1, Oncogene 31, 1398-1407. 148. Shu, M., Zheng, X., Wu, S., Lu, H., Leng, T., Zhu, W., Zhou, Y., Ou, Y., Lin, X., Lin, Y., Xu, D., Zhou, Y., and Yan, G. (2011) Targeting oncogenic miR-335 inhibits growth and invasion of malignant astrocytoma cells, Molecular cancer 10, 59. 149. Shu, M., Zhou, Y., Zhu, W., Zhang, H., Wu, S., Chen, J., and Yan, G. (2012) MicroRNA 335 is required for differentiation of malignant glioma cells induced by activation of cAMP/protein kinase A pathway, Molecular pharmacology 81, 292298. 150. Vickers, M. M., Bar, J., Gorn-Hondermann, I., Yarom, N., Daneshmand, M., Hanson, J. E., Addison, C. L., Asmis, T. R., Jonker, D. J., Maroun, J., Lorimer, I. A., Goss, G. D., and Dimitroulakos, J. (2012) Stage-dependent differential expression of microRNAs in colorectal cancer: potential role as markers of metastatic disease, Clinical & experimental metastasis 29, 123-132. 151. Lynch, J., Fay, J., Meehan, M., Bryan, K., Watters, K. M., Murphy, D. M., and Stallings, R. L. (2012) MiRNA-335 suppresses neuroblastoma cell invasiveness by direct targeting of multiple genes from the non-canonical TGF-β signalling pathway, Carcinogenesis 33, 976-985. 152. Jiang, J., Sun, X., Wang, W., Jin, X., Bo, X., Li, Z., Bian, A., Jiu, J., Wang, X., Liu, D., Hui, X., Wang, Y., Wang, A., and Ding, L. (2012) Tumor microRNA-335 expression is associated with poor prognosis in human glioma, Medical oncology 29, 3472-3477. 153. Nakanishi, N., Nakagawa, Y., Tokushige, N., Aoki, N., Matsuzaka, T., Ishii, K., Yahagi, N., Kobayashi, K., Yatoh, S., Takahashi, A., Suzuki, H., Urayama, O., Yamada, N., and Shimano, H. (2009) The up-regulation of microRNA-335 is associated with lipid metabolism in liver and white adipose tissue of genetically obese mice, Biochemical and Biophysical Research Communications 385, 492496. 154. Greco, S., de Simone, M., Colussi, C., Zaccagnini, G., Fasanaro, P., Pescatori, M., Cardani, R., Perbellini, R., Isaia, E., Sale, P., Meola, G., Capogrossi, M. C., 128  155.  156.  157.  158.  159.  160.  161.  162.  163.  164.  165.  166.  167. 168.  Gaetano, C., and Martelli, F. (2009) Common micro-RNA signature in skeletal muscle damage and regeneration induced by Duchenne muscular dystrophy and acute ischemia, FASEB journal 23, 3335-3346. Wang, J., and Ruan, K. (2010) miR-335 is involved in the rat epididymal development by targeting the mRNA of RASA1, Biochemical and Biophysical Research Communications 402, 222-227. Zhang, J., Tu, Q., Bonewald, L. F., He, X., Stein, G., Lian, J., and Chen, J. (2011) Effects of miR-335-5p in modulating osteogenic differentiation by specifically downregulating Wnt antagonist DKK1, Journal of bone and mineral research 26, 1953-1963. Bai, X., Ma, Y., Ding, R., Fu, B., Shi, S., and Chen, X. (2011) miR-335 and miR34a Promote renal senescence by suppressing mitochondrial antioxidative enzymes, Journal of the American Society of Nephrolog 22, 1252-1261. Chen, C., Wu, C., Zhang, Z., Yao, D., and Zhu, L. (2011) Loss of expression of miR-335 is implicated in hepatic stellate cell migration and activation, Experimental Cell Research 317, 1714-1725. Riesewijk, A. M., Hu, L., Schulz, U., Tariverdian, G., Höglund, P., Kere, J., Ropers, H. H., and Kalscheuer, V. M. (1997) Monoallelic expression of human PEG1/MEST is paralleled by parent-specific methylation in fetuses, Genomics 42, 236-244. Nakabayashi, K., Bentley, L., Hitchins, M. P., Mitsuya, K., Meguro, M., Minagawa, S., Bamforth, J. S., Stanier, P., Preece, M., Weksberg, R., Oshimura, M., Moore, G. E., and Scherer, S. W. (2002) Identification and characterization of an imprinted antisense RNA (MESTIT1) in the human MEST locus on chromosome 7q32, Human Molecular Genetics 11, 1743-1756. Hayashida, S., Yamasaki, K., Asada, Y., Soeda, E., Niikawa, N., and Kishino, T. (2000) Construction of a physical and transcript map flanking the imprinted MEST/PEG1 region at 7q32, Genomics 66, 221-225. Kotzot, D., Schmitt, S., Bernasconi, F., Robinson, W. P., Lurie, I. W., Ilyina, H., Méhes, K., Hamel, B. C., Otten, B. J., and Hergersberg, M. (1995) Uniparental disomy 7 in Silver-Russell syndrome and primordial growth retardation, Human Molecular Genetics 4, 583-587. Preece, M. A., Price, S. M., Davies, V., Clough, L., Stanier, P., Trembath, R. C., and Moore, G. E. (1997) Maternal uniparental disomy 7 in Silver-Russell syndrome, Journal of Medical Genetics 34, 6-9. Riesewijk, A. M., Blagitko, N., Schinzel, A. A., Hu, L., Schulz, U., Hamel, B. C., Ropers, H. H., and Kalscheuer, V. M. (1998) Evidence against a major role of PEG1/MEST in Silver-Russell syndrome, European Journal of Human Genetics 6, 114-120. Schöherr, N., Jäger, S., Ranke, M. B., Wollmann, H. A., Binder, G., and Eggermann, T. (2008) No evidence for isolated imprinting mutations in the PEG1/MEST locus in Silver-Russell patients, European journal of medical genetics 51, 322-324. Eggermann, T., Spengler, S., Begemann, M., Binder, G., Buiting, K., Albrecht, B., and Spranger, S. (2012) Deletion of the paternal allele of the imprinted MEST/PEG1 region in a patient with Silver-Russell syndrome features, Clinical Genetics 81, 298-300. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. (2003) Manipulating the Mouse Embryo: A Laboratory Manual, 3rd ed. Tada, M., Tada, T., Lefebvre, L., Barton, S. C., and Surani, M. A. (1997) Embryonic germ cells induce epigenetic reprogramming of somatic nucleus in hybrid cells, The EMBO Journal 16, 6510-6520. 129  169. Saha, S., Sparks, A. B., Rago, C., Akmaev, V., Wang, C. J., Vogelstein, B., Kinzler, K. W., and Velculescu, V. E. (2002) Using the transcriptome to annotate the genome, Nature Biotechnology 20, 508-512. 170. Peters, D. G., Kassam, A. B., Yonas, H., O'Hare, E. H., Ferrell, R. E., and Brufsky, A. M. (1999) Comprehensive transcript analysis in small quantities of mRNA by SAGE-lite, Nucleic Acids Research 27, e39. 171. Siddiqui, A. S., Khattra, J., Delaney, A. D., Zhao, Y., Astell, C., Asano, J., Babakaiff, R., Barber, S., Beland, J., Bohacec, S., Brown-John, M., Chand, S., Charest, D., Charters, A. M., Cullum, R., Dhalla, N., Featherstone, R., Gerhard, D. S., Hoffman, B., Holt, R. A., Hou, J., Kuo, B. Y., Lee, L. L., Lee, S., Leung, D., Ma, K., Matsuo, C., Mayo, M., McDonald, H., Prabhu, A., Pandoh, P., Riggins, G. J., de Algara, T. R., Rupert, J. L., Smailus, D., Stott, J., Tsai, M., Varhol, R., Vrljicak, P., Wong, D., Wu, M. K., Xie, Y., Yang, G., Zhang, I., Hirst, M., Jones, S. J., Helgason, C. D., Simpson, E. M., Hoodless, P. A., and Marra, M. A. (2005) A mouse atlas of gene expression: large-scale digital gene-expression profiles from precisely defined developing C57BL/6J mouse tissues and cells, Proceedings of the National Academy of Sciences of the United States of America 102, 18485-18490. 172. Ramakers, C., Ruijter, J. M., Deprez, R. H., and Moorman, A. F. (2003) Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data, Neuroscience Letters 339, 62-66. 173. Schneider, C., Rasband, W., and Eliceiri, K. (2012) NIH Image to ImageJ: 25 years of image analysis, Nature methods 9, 671-675. 174. Ewing, B., Hillier, L., Wendl, M. C., and Green, P. (1998) Base-calling of automated sequencer traces using phred. I. Accuracy assessment, Genome Research 8, 175-185. 175. Jones, M. J., and Lefebvre, L. (2009) An imprinted GFP insertion reveals longrange epigenetic regulation in embryonic lineages, Developmental Biology 336, 4252. 176. Parsons, M. J., Grimm, C. H., Paya-Cano, J. L., Sugden, K., Nietfeld, W., Lehrach, H., and Schalkwyk, L. C. (2008) Using hippocampal microRNA expression differences between mouse inbred strains to characterise miRNA function, Mammalian Genome 19, 552-560. 177. Kalinski, T., Jaquet, K., Langen, R., Wang, L., Gabbert, H., and Gerharz, C. (1997) An optimized electroporation protocol for transfection of sensitive cell lines using basic laboratory equipment, Biotechnology Techniques 11, 717-722. 178. Morin, R., Bainbridge, M., Fejes, A., Hirst, M., Krzywinski, M., Pugh, T., McDonald, H., Varhol, R., Jones, S., and Marra, M. (2008) Profiling the HeLa S3 transcriptome using randomly primed cDNA and massively parallel short-read sequencing, BioTechniques 45, 81-94. 179. Li, H., and Durbin, R. (2010) Fast and accurate long-read alignment with BurrowsWheeler transform, Bioinformatics 26, 589-95. 180. Li, H., and Durbin, R. (2009) Fast and accurate short read alignment with BurrowsWheeler transform, Bioinformatics 25, 1754-60. 181. Edwalds-Gilbert, G., Veraldi, K. L., and Milcarek, C. (1997) Alternative poly(A) site selection in complex transcription units: means to an end?, Nucleic Acids Research 25, 2547-2561. 182. Bin Tian, Hu, J., Zhang, H., and Lutz, C. S. (2005) A large-scale analysis of mRNA polyadenylation of human and mouse genes, Nucleic Acids Research 33, 201-212. 183. Ji, Z., Lee, J. Y., Pan, Z., Jiang, B., and Bin Tian. (2009) Progressive lengthening of 3' untranslated regions of mRNAs by alternative polyadenylation during mouse 130  184.  185.  186.  187.  188.  189.  190.  191.  192.  193.  194.  195.  196.  197.  embryonic development, Proceedings of the National Academy of Sciences of the United States of America 106, 7028-7033. Sunwoo, H., Dinger, M. E., Wilusz, J. E., Amaral, P. P., Mattick, J. S., and Spector, D. L. (2009) MEN epsilon/beta nuclear-retained non-coding RNAs are up-regulated upon muscle differentiation and are essential components of paraspeckles, Genome Research 19, 347-359. Yu, M., Sha, H., Gao, Y., Zeng, H., Zhu, M., and Gao, X. (2006) Alternative 3′ UTR polyadenylation of Bzw1 transcripts display differential translation efficiency and tissue-specific expression, Biochemical and Biophysical Research Communications 345, 479-485. Irier, H. A., Shaw, R., Lau, A., Feng, Y., and Dingledine, R. (2009) Translational regulation of GluR2 mRNAs in rat hippocampus by alternative 3' untranslated regions, Journal of Neurochemistry 109, 584-594. An, J. J., Gharami, K., Liao, G., Woo, N. H., Lau, A. G., Vanevski, F., Torre, E. R., Jones, K. R., Feng, Y., Lu, B., and Xu, B. (2008) Distinct Role of Long 3′ UTR BDNF mRNA in Spine Morphology and Synaptic Plasticity in Hippocampal Neurons, Cell 134, 175-187. Lau, A. G., Irier, H. A., Gu, J., Tian, D., Ku, L., Liu, G., Xia, M., Fritsch, B., Zheng, J. Q., Dingledine, R., Xu, B., Lu, B., and Feng, Y. (2010) Distinct 3'UTRs differentially regulate activity-dependent translation of brain-derived neurotrophic factor (BDNF), Proceedings of the National Academy of Sciences of the United States of America 107, 15945-15950. Miura, P., Amirouche, A., Clow, C., Bélanger, G., and Jasmin, B. (2012) Brainderived neurotrophic factor expression is repressed during myogenic differentiation by miR-206, J Neurochem 120, 230-238. Wood, A. J., Schylz, R., Woodfine, K., Koltowska, K., Beechey, C. V., Peters, J., Bourc'his, D., and Oakey, R. J. (2009) Regulation of alternative polyadenylation by genomic imprinting, Genes and Development, 22, 1141-1146. Schulz, R., Woodfine, K., Menheniott, T. R., Bourc'his, D., Bestor, T., and Oakey, R. J. (2008) WAMIDEX: a web atlas of murine genomic imprinting and differential expression, Epigenetics 3, 89-96. Uejima, H., Lee, M. P., Cui, H., and Feinberg, A. P. (2000) Hot-stop PCR: a simple and general assay for linear quantitation of allele ratios, Nature Genetics 25, 375376. Fitzpatrick, G. V., Soloway, P. D., and Higgins, M. J. (2002) Regional loss of imprinting and growth deficiency in mice with a targeted deletion of KvDMR1, Nature Genetics 32, 426-431. Redrup, L., Branco, M. R., Perdeaux, E. R., Krueger, C., Lewis, A., Santos, F., Nagano, T., Cobb, B. S., Fraser, P., and Reik, W. (2009) The long noncoding RNA Kcnq1ot1 organises a lineage-specific nuclear domain for epigenetic gene silencing, Development 136, 525-530. Wang, Y., Joh, K., Masuko, S., Yatsuki, H., Soejima, H., Nabetani, A., Beechey, C. V., Okinami, S., and Mukai, T. (2004) The mouse Murr1 gene is imprinted in the adult brain, presumably due to transcriptional interference by the antisenseoriented U2af1-rs1 gene, Molecular and Cellular Biology 24, 270-279. Prescott, E. M., and Proudfoot, N. J. (2002) Transcriptional collision between convergent genes in budding yeast, Proceedings of the National Academy of Sciences of the United States of America 99, 8796-8801. Proudfoot, N. (2004) New perspectives on connecting messenger RNA 3' end formation to transcription, Current Opinion in Cell Biology 16, 272-278. 131  198. Mazo, A., Hodgson, J. W., Petruk, S., Sedkov, Y., and Brock, H. W. (2007) Transcriptional interference: an unexpected layer of complexity in gene regulation, Journal of Cell Science 120, 2755-2761. 199. Hogg, M., Paro, S., Keegan, L. P., and O'Connell, M. A. (2011) RNA editing by mammalian ADARs, Advances in genetics 73, 87-120. 200. Keegan, L. P., Gallo, A., and O'Connell, M. A. (2001) The many roles of an RNA editor, Nature Reviews Genetics 2, 869-878. 201. Gu, T., Buaas, F. W., Simons, A. K., Ackert-Bicknell, C. L., Braun, R. E., and Hibbs, M. A. (2012) Canonical A-to-I and C-to-U RNA Editing Is Enriched at 3′UTRs and microRNA Target Sites in Multiple Mouse Tissues, PLoS ONE 7, e33720. 202. DeCerbo, J., and Carmichael, G. G. (2005) Retention and repression: fates of hyperedited RNAs in the nucleus, Current Opinion in Cell Biology 17, 302-308. 203. Scadden, A. D., and Smith, C. W. (2001) Specific cleavage of hyper-edited dsRNAs, The EMBO Journal 20, 4243-4252. 204. Bass, B. L., and Weintraub, H. (1988) An unwinding activity that covalently modifies its double-stranded RNA substrate, Cell 55, 1089-1098. 205. Carthew, R. W., and Sontheimer, E. J. (2009) Origins and Mechanisms of miRNAs and siRNAs, Cell 136, 642-655. 206. Gou, H., Ingolia, N. T., Weissman, J. S., and Bartel, D. P. (2010) Mammalian microRNAs predominantly act to decrease target mRNA levels, Nature 466, 835-841. 207. Wang, Z., Gerstein, M., and Snyder, M. (2009) RNA-Seq: a revolutionary tool for transcriptomics, Nature Reviews Genetics 10, 57-63. 208. Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq, Nature methods 5, 621-628. 209. Oliveros, J. (2007) VENNY. An interactive tool for comparing lists with Venn Diagrams., http://bioinfogp.cnb.csic.es/tools/venny/index.html. 210. Huang, D. W., Sherman, B. T., and Lempicki, R. A. (2009) Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists, Nucleic Acids Research 37, 1-13. 211. Huang, W., Sherman, B., and Lempicki, R. (2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources, Nature protocols 4, 44-57. 212. Saito, T., Sætrom, P., and Sætrom, P. (2010) MicroRNAs – targeting and target prediction, New Biotechnology 27, 243-249. 213. Khan, A. A., Betel, D., Miller, M. L., Sander, C., Leslie, C. S., and Marks, D. S. (2009) Transfection of small RNAs globally perturbs gene regulation by endogenous microRNAs, Nature Biotechnology 27, 549-55. 214. Arvey, A., Larsson, E., S, C., er, Leslie, C. S., Marks, D. S., and Sander, C. (2010) Target mRNA abundance dilutes microRNA and siRNA activity, Molecular Systems Biology 6, 363 215. Saito, T., and Sætrom, P. (2012) Target gene expression levels and competition between transfected and endogenous microRNAs are strong confounding factors in microRNA high-throughput experiments, Silence 3, 3. 216. Molchadsky, A., Shats, I., Goldfinger, N., Pevsner-Fischer, M., Olson, M., Rinon, A., Tzahor, E., Lozano, G., Zipori, D., Sarig, R., and Rotter, V. (2008) p53 Plays a Role in Mesenchymal Differentiation Programs, in a Cell Fate Dependent Manner, PLoS ONE 3, e3707. 132  217. Quintana, L., Muiños, T. F., Genove, E., del Olmos, M. M., Borrós, S., and Semino, C. E. (2009) Early tissue patterning recreated by mouse embryonic fibroblasts in a three-dimensional environment., Tissue engineering. Part A 15, 45-54. 218. Baskerville, S., and Bartel, D. P. (2005) Microarray profiling of microRNAs reveals frequent coexpression with neighboring miRNAs and host genes, RNA 11, 241247. 219. Monteys, A. M., Spengler, R. M., Wan, J., Tecedor, L., Lennox, K. A., Xing, Y., and Davidson, B. L. (2010) Structure and activity of putative intronic miRNA promoters, RNA 16, 495-505. 220. Radhika, das, Anderson, N., Koran, M. I., Weidman, J. R., Mikkelsen, T. S., Kamal, M., Murphy, S. K., Linblad-Toh, K., Greally, J. M., and Jirtle, R. L. (2012) Convergent and divergent evolution of genomic imprinting in the marsupial Monodelphis domestica., BMC genomics 13, 394. 221. Kornfeld, S. (1992) Structure and function of the mannose 6-phosphate/insulinlike growth factor II receptors, Annual review of biochemistry 61, 307-330. 222. Dahms, N., Brzycki-Wessell, M., Ramanujam, K., and Seetharam, B. (1993) Characterization of mannose 6-phosphate receptors (MPRs) from opossum liver: opossum cation-independent MPR binds insulin-like growth factor-II, Endocrinology 133, 440-446. 223. Clairmont, K. B., and Czech, M. P. (1989) Chicken and Xenopus mannose 6phosphate receptors fail to bind insulin-like growth factor II, The Journal of biological chemistry 264, 16390-16392. 224. Killian, J. K., Byrd, J. C., Jirtle, J. V., Munday, B. L., Stoskopf, M. K., MacDonald, R. G., and Jirtle, R. L. (2000) M6P/IGF2R imprinting evolution in mammals, Molecular Cell 5, 707-716. 225. Furuno, M., Pang, K. C., Ninomiya, N., Fukuda, S., Frith, M. C., Bult, C., Kai, C., Kawai, J., Carninci, P., Hayashizaki, Y., Mattick, J. S., and Suzuki, H. (2006) Clusters of internally primed transcripts reveal novel long noncoding RNAs., PLoS Genetics 2, e37. 226. Katayama, S., Tomaru, Y., Kasukawa, T., Waki, K., Nakanishi, M., Nakamura, M., Nishida, H., Yap, C. C., Suzuki, M., Kawai, J., Suzuki, H., Carninci, P., Hayashizaki, Y., Wells, C., Frith, M., Ravasi, T., Pang, K. C., Hallinan, J., Mattick, J., Hume, D. A., Lipovich, L., Batalov, S., Engström, P. G., Mizuno, Y., Faghihi, M. A., Sandelin, A., Chalk, A. M., Mottagui-Tabar, S., Liang, Z., Lenhard, B., Wahlestedt, C., Group, R. G., Group, G. S., and Consortium, F. (2005) Antisense transcription in the mammalian transcriptome, Science 309, 1564-1566. 227. Mehler, M. F., and Mattick, J. S. (2006) Non-coding RNAs in the nervous system, The Journal of physiology 575, 333-341. 228. Mercer, T. R., Wilhelm, D., Dinger, M. E., Soldà, G., Korbie, D. J., Glazov, E. A., Truong, V., Schwenke, M., Simons, C., Matthaei, K. I., Saint Robert, Koopman, P., and Mattick, J. S. (2011) Expression of distinct RNAs from 3' untranslated regions, Nucleic Acids Research 39, 2393-2403. 229. Murrell, A., Heeson, S., and Reik, W. (2004) Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 into parent-specific chromatin loops, Nature Genetics 36, 889-893. 230. Kurukuti, S., Tiwari, V. K., Tavoosidana, G., Pugacheva, E., Murrell, A., Zhao, Z., Lobanenkov, V., Reik, W., and Ohlsson, R. (2006) CTCF binding at the H19 imprinting control region mediates maternally inherited higher-order chromatin conformation to restrict enhancer access to Igf2, Proceedings of the National Academy of Sciences of the United States of America 103, 10684-10689. 133  231. Fernández, M., and Miranda-Saavedra, D. (2012) Genome-wide enhancer prediction from epigenetic signatures using genetic algorithm-optimized support vector machines, Nucleic Acids Research 40, e77. 232. Kumaki, Y., Ukai-Tadenuma, M., Uno, K. D., Nishio, J., Masumoto, K., Nagano, M., Komori, T., Shigeyoshi, Y., Hogenesch, J. B., and Ueda, H. R. (2008) Analysis and synthesis of high-amplitude Cis-elements in the mammalian circadian clock, Proceedings of the National Academy of Sciences of the United States of America 105, 14946-14951. 233. Visel, A., Blow, M. J., Li, Z., Zhang, T., Akiyama, J. A., Holt, A., Plajzer-Frick, I., Shoukry, M., Wright, C., Chen, F., Afzal, V., Ren, B., Rubin, E. M., and Pennacchio, L. A. (2009) ChIP-seq accurately predicts tissue-specific activity of enhancers, Nature 457, 854-858. 234. Blow, M. J., McCulley, D. J., Li, Z., Zhang, T., Akiyama, J. A., Holt, A., PlajzerFrick, I., Shoukry, M., Wright, C., Chen, F., Afzal, V., Bristow, J., Ren, B., Black, B. L., Rubin, E. M., Visel, A., and Pennacchio, L. A. (2010) ChIP-Seq identification of weakly conserved heart enhancers, Nature Genetics 42, 806-810. 235. Zhang, J., and Ma, L. (2012) MicroRNA control of epithelial–mesenchymal transition and metastasis, Cancer and Metastasis Reviews 31, 653-662. 236. Korpal, M., and Kang, Y. (2008) The emerging role of miR-200 family of microRNAs in epithelial-mesenchymal transition and cancer metastasis, RNA Biology 5, 115-119. 237. White, N., Bao, T., Grigull, J., Youssef, Y., Girgis, A., Diamandis, M., Fatoohi, E., Metias, M., Honey, R., Stewart, R., Pace, K., Bjarnason, G., and Yousef, G. (2011) miRNA profiling for clear cell renal cell carcinoma: biomarker discovery and identification of potential controls and consequences of miRNA dysregulation, Journal of Urology 186, 1077-1083. 238. Holmquist, M. (2000) Alpha/Beta-hydrolase fold enzymes: structures, functions and mechanisms, Curr Protein Pept Sci. 1, 209-35. 239 Trapnell,C., Pachter,L. and Salzberg,S.L. (2009) TopHat: discovering splice junctions with RNA-Seq, Bioinformatics 25, 1105-11. 240. Robinson,J.T., Thorvaldsdóttir,H., Winckler,W., Guttman,M., Lander,E., Getz, G., and Mesirov,J. (2011) Integrative genomics viewer, Nat Biotechnol 29, 24-26.  134  Appendix A: Supplementary material for chapter 3  Figure A.1 Tissue distribution of expressed sequence tags (ESTs) from the (+) strand mapping to the Mest locus and to the EST clusters 5’ and 3’ of the gap within intron 20 of Copg2. Mest ESTs are from the 1, 460 EST sequences for the Mest UniGene entry Mm.335639. ESTs from the 5’ cluster (n=122, 19 from the (+) strand) and 3’ clusters (n=114, 73 from the (+) strand) are presented in Tables A.1 and A.2, respectively. ESTs were grouped in two different classes representing the tissue source of the EST. Expressed sequences from tumors, mixed or undefined libraries were not included in this analysis.  135  Figure A.2 RNA gels for Northern blot analyses. 2 µg of poly(A)+ RNA purified from wild-type and Mest+/KO E14.5 embryos and adult brain (per sample) was separated by electrophoresis on a 1% agarose, 2.2 M formaldehyde denaturing gel and stained with ethidium bromide. The gel scans show equivalent amounts of RNA in each lane for wild-type and Mest+/KO samples.  136  Figure A.3 In situ hybridization analysis of Copg2 and primary Copg2 mRNA expression at E14.5. ISH was performed on 12 μm sagittal cryosections of wild-type (Mest+/+) (A) and Mest+/KO (B) E14.5 embryos with various DIG-labeled RNA probes. The antisense probe locations are indicated in Figures 3.4A and 3.6A in Chapter 3 using the letter code presented below and hybridize to the (-) strand. Blue staining indicates gene expression and nuclear fast red was used as a counterstain.  137  Figure A.4 Analysis of RNA-seq data in the Mest-Copg2 region. Please see the following page for figure legend.  138  Figure A.4 Analysis of RNA-seq data in the Mest-Copg2 region. TopHat analysis of high-throughput RNA sequencing from a CD-1 E9.5 placenta, a C57Bl/6 E15 head, an adult C57Bl/6 medial prefrontal cortex (PFC) and an adult C57Bl/6 preoptic area (POA). Exonic reads for Mest, and to a lesser extent Copg2, are seen in the placental sample while reads located in intron 20 of Copg2 are seen in the neural samples. Log-scale analysis of the reads shows alignments of MestXL reads in all 4 samples but supports their preponderance in the neural samples. Repetitive regions (such as the middle of Mest ex12) show no alignments as the sequence was masked before alignment was performed. The four different libraries of expression profiling by high-throughput sequencing of cDNA (RNA-seq) analyzed were from E15 whole brain (GEO GSM563764), adult preoptic area (GEO GSM563768), and adult medial prefrontal cortex (GEO GSM563770), and from a mesodermal tissue, the extraembryonic mesoderm from a E9.5 placenta (L. Lefebvre, unpublished data). Sequence reads were aligned to a repeat-masked version of the genomic sequence containing Mest, Copg2 and 50 kb surrounding the two genes (mm9) using TopHat (239). TopHat output was analyzed in the Integrative Genomics Viewer (240).  139  Table A.1 (+) strand ESTs from 5’ cluster #  GenBank  size (bp)  library  1  CB234278.1  782  brain  2  BU922167.1  600  E14.5 neural retina  3  CX213951.1  588  adult brain, lateral wall of lateral ventricle, neurospheres  4  BG086321.2  572  mixed tissues  5  BG064605.2  563  mixed tissues  6  AW227625.1  574  tumor, metastatic to mammary (Wnt-1 tg)  7  BB692936.1  539  2 days neonate sympathetic ganglion  8  BM220201.2  514  E12.5 male genital ridge/mesonephros  9  AW545338.1  512  E7.5 Extraembryonic Portion  10  AV443337.1  507  Abe mouse ES cell  11  W66960.1  362  E13.5-14.5 total fetus  12  BU614826.1  653  E15.5 whole brain  13  BE945786.1  409  27-32 days hippocampus  14  BY661711.1  383  14.5 days embryo RP+/+ Rathke-s pouches  15  AI851468.1  328  27-32 days, ten regions of the mouse brain  16  CO424394.1  268  newborn, whole eye  17  BX631827.1  278  unknown  18  BF319175.1  402  tumor, metastatic to mammary (Wnt-1 tg)  19  CB233684.1  936  brain  140  Table A.2 (+) strand ESTs from 3’ cluster GenBank  size (bp)  library  1  BY718675.1  960  adult male medulla oblongata  2  CF617792.1  888  E16, placenta  3  CB181671.1  892  E10.5-11.5 pooled mouse embryonic limb, maxilla and mandible  4  BY718330.1  978  adult male, medulla oblongata  5  CD348666.1  812  E13.5-17.5 embryo, whole brain  6  BQ952052.1  898  retina  7  CD578842.1  766  1, 5 and 15 days newborn, whole brain  8  BU707629.1  749  E13.5-17.5 embryo, whole brain  9  CB234355.1  770  brain  10  AU080676.1  756  adult female, brain  11  CX242332.1  775  C57BL/6 adult, lateral ventricle wall  12  BQ444669.1  714  E15.5 whole brain  13  CD348658.1  697  E13.5-17.5 embryo, whole brain  14  CA317223.1  737  E13.5-17.5 embryo, whole brain  15  CB057654.1  690  E9-12, pituitary gland  16  CF746092.1  710  1, 5 and 15 days newborn, whole brain  17  CB059106.1  660  adult, pituitary gland  18  CB058980.1  665  adult, pituitary gland  19  BU058909.1  642  E13.5-17.5 embryo, whole brain  20  CX731783.1  634  adult, whole eye  21  CB524143.1  632  1, 5 and 15 days newborn, whole brain  22  CD804425.1  630  1, 5 and 15 days newborn, whole brain  23  DT919282.1  821  hematopoietic stem cells  24  BY718746.1  627  adult male, medulla oblongata  25  BG294462.1  639  retina  26  BU705429.1  689  E12.5, whole brain 141  GenBank  size (bp)  library  27  CN664257.1  603  E13.5, whole embryo  28  BQ555002.1  604  mixed  29  BM936491.1  582  27-32 days, brain regions  30  BB622187.1  659  adult male, medulla oblongata  31  CB055817.1  571  juvenile, 13-15 days, pituitary gland  32  BX523722.1  541  irradiated colon  33  BB644886.1  544  adult male, corpora quadrigemina  34  BQ555026.1  533  mixed  35  BB633186.1  578  adult male, spinal cord  36  BB316578.2  567  adult male, corpora quadrigemina  37  BI157391.1  878  mammary tumor, gross tissue  38  CX222337.1  537  adult brain, lateral wall of lateral ventricle, neurospheres  39  BB604503.2  631  0 day neonate, lung  40  W57281.1  482  E13.5-14.5, total fetus  41  BE860082.1  487  27-32 days, hippocampus  42  AA611551.1  511  irradiated colon  43  BB664248.1  508  0 day neonate, lung  44  BY247745.1  483  visual cortex  45  DN173937.1  482  C57BL/6 adult, lateral ventricle wall  46  AI430073.1  464  E13.5-14.5, total fetus  47  AI664413.1  450  lactating female, mammary gland  48  AI428992.1  463  irradiated colon  49  CX730814.1  565  adult, whole eye (retinal degeneration)  50  BB184889.2  442  adult male, spinal cord  51  BQ747804.1  516  E12.5 whole brain  52  CX213741.1  450  adult brain, lateral wall of lateral ventricle, neurospheres  53  BY271818.1  443  visual cortex  142  GenBank  size (bp)  library  54  CD355286.1  424  E13.5-17.5 embryo, whole brain  55  BY250425.1  447  visual cortex  56  CX734371.1  440  adult, whole eye  57  CB722626.1  695  1, 5 and 15 days newborn, whole brain  58  CA316586.1  599  E13.5-17.5 embryo, whole brain  59  BQ042233.1  493  E15.5, whole brain  60  BY288214.1  418  visual cortex  61  BY632357.1  397  visual cortex  62  BY608282.1  369  visual cortex  63  BY239946.1  411  visual cortex  64  CD352457.1  499  E13.5-17.5 embryo, whole brain  65  BY610290.1  394  visual cortex  66  CB321435.1  1318  embryonic limb, maxilla and mandible  67  BY245849.1  429  visual cortex  68  CA320171.1  321  E13.5-17.5 embryo, whole brain  69  BE859443.1  327  27-32 days, striatum  70  BP427243.1  283  E18-P56, cerebellum  71  BB663781.1  673  0 day neonate, lung  72  AI846076.1  291  27-32 days, basal ganglia  73  BI738357.1  783  retina  143  Appendix B: Supplementary material for chapter 4  Figure B.1 Effect of loss of maternal methylation of Mest, MestXL, and Copg2 expression. Microarray expression summary from data compiled from the Wamidex website (https://atlas.genetics.kcl.ac.uk/) analyzing expression differences between E8.5 wildtype and Dnmt3l-/+ embryos (191). Two replicates were analyzed for each probe and the data was summarized by averaging the fold-changes obtained through three methods which include PLIER, GC-RMA, and RMA. The shown values are log2 of the fold change.  144  Appendix C: Supplementary material for chapter 5 Table C.1 Significantly down-regulated genes in the Mest+/KO PEFs compared to the WT PEFs. This list corresponds to Group A in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. Note: A fold-change of 1000 indicates that the denominator RPKM value used to determine the fold-change was zero. RPKM Gene ID 1500009C09Rik 1600029D21Rik 1700009N14Rik 1700015F17Rik 1700024N20Rik 1700024P16Rik 1700057K13Rik 2010005H15Rik 2310002L13Rik 2310005G13Rik 3110039M20Rik 8030474K03Rik 8430432A02Rik 9230109A22Rik Acp1 Acta1 Ahcy Akr1c18 Akr1c19 Alox8 Als2cr12 Ang2 Ankrd1 Arg1 Asb5 AU018091 BC061194 Btn1a1 C130079G13Rik C3ar1 Car8 Cd200r2 Cd200r3 Cd200r4 Chrnb3 Cldn4  Mest+/KO WT+miR 0.007 0.050 0.109 0.111 0.089 0.162 0.085 0.000 4.025 0.006 0.067 0.026 0.018 0.019 2.871 4.837 0.828 0.246 0.056 0.027 0.170 1.080 6.067 0.395 1.446 0.078 0.030 0.051 0.002 0.351 0.314 0.007 0.000 0.008 0.007 0.305  0.017 0.069 0.351 0.371 0.225 1.460 0.121 0.683 13.401 0.029 0.305 0.202 0.091 0.028 13.263 14.170 10.347 0.686 0.119 0.025 0.377 4.712 11.497 0.886 3.390 0.089 0.089 0.155 0.000 1.125 3.117 0.065 0.018 0.122 0.025 1.395  Fold-change WT 0.064 0.266 0.625 0.629 0.425 3.394 0.359 0.528 16.694 0.044 1.157 0.118 0.166 0.099 12.615 30.463 9.192 1.403 0.722 0.115 0.707 5.287 27.741 1.716 7.364 0.328 0.149 0.388 0.034 2.812 2.732 0.090 0.033 0.148 0.041 1.264  +/KO  Mest Mest+/KO WT+miR WT+miR WT WT 0.11 0.41 0.26 0.19 0.71 0.26 0.17 0.31 0.56 0.18 0.30 0.59 0.21 0.40 0.53 0.05 0.11 0.43 0.24 0.70 0.34 0.00 0.00 1.29 0.24 0.30 0.80 0.13 0.19 0.67 0.06 0.22 0.26 0.22 0.13 1.71 0.11 0.19 0.55 0.19 0.69 0.28 0.23 0.22 1.05 0.16 0.34 0.47 0.09 0.08 1.13 0.18 0.36 0.49 0.08 0.47 0.16 0.23 1.08 0.21 0.24 0.45 0.53 0.20 0.23 0.89 0.22 0.53 0.41 0.23 0.45 0.52 0.20 0.43 0.46 0.24 0.88 0.27 0.20 0.34 0.59 0.13 0.33 0.40 0.05 1000.00 0.00 0.12 0.31 0.40 0.11 0.10 1.14 0.08 0.11 0.72 0.01 0.01 0.56 0.05 0.06 0.82 0.18 0.29 0.61 0.24 0.22 1.10  x x x x x x x x x x  x x x x x 145  RPKM Gene ID  Mest+/KO WT+miR  Clec2h Cma2 Col6a4 Cpne4 Crct1 Csf2 Ctgf Cx3cr1 Cxcl2 Cxcl3 Cyp1a1 Cyp24a1 Dbh Dntt Ear11 Eif3j Enthd1 Fam110c Fam71e2 Fasl Fkbp6 Fnd3c2 Folh1 Foxa1 Foxg1 Foxn1 Gal Galnt13 Gdf3 Gjb2 Gm10008 Gm10923 Gm11127 Gm13283 Gm13288 Gm14399 Gm14446 Gm17680 Gm266 Gm4759 Gm4858 Gm5152  0.057 0.515 0.019 0.008 0.130 0.216 144.429 0.022 3.573 5.757 0.099 0.014 0.004 0.030 0.050 0.776 0.041 4.893 0.030 0.058 0.001 0.000 0.003 0.055 0.037 0.015 0.247 0.652 0.012 6.823 0.014 0.028 0.103 0.003 0.003 0.065 20.938 0.000 0.197 0.022 0.000 0.003  0.080 0.804 0.089 0.030 1.048 0.375 196.824 0.076 3.375 8.160 0.181 0.046 0.030 0.053 0.424 3.776 0.228 10.607 0.104 0.215 0.046 0.008 0.034 0.316 0.080 0.021 0.453 1.703 0.050 12.762 0.075 0.139 0.993 0.064 0.034 0.269 71.647 0.795 1.871 0.828 0.012 0.013  Fold-change WT 0.390 3.758 0.087 0.043 3.123 1.239 823.556 0.291 17.667 24.184 1.697 0.091 0.065 0.222 0.804 4.367 0.233 25.140 0.139 0.301 0.064 0.043 0.044 0.292 0.424 0.235 1.120 3.228 0.095 32.806 0.174 0.304 1.411 0.110 0.037 0.295 99.566 0.605 0.821 0.388 0.045 0.099  +/KO  Mest Mest+/KO WT+miR WT+miR WT WT 0.15 0.72 0.20 0.14 0.64 0.21 0.22 0.22 1.03 0.18 0.26 0.71 0.04 0.12 0.34 0.17 0.58 0.30 0.18 0.73 0.24 0.07 0.29 0.26 0.20 1.06 0.19 0.24 0.71 0.34 0.06 0.55 0.11 0.16 0.32 0.50 0.06 0.12 0.45 0.14 0.57 0.24 0.06 0.12 0.53 0.18 0.21 0.86 0.18 0.18 0.98 0.19 0.46 0.42 0.21 0.29 0.75 0.19 0.27 0.71 0.02 0.02 0.72 0.00 0.00 0.19 0.07 0.10 0.77 0.19 0.17 1.08 0.09 0.47 0.19 0.06 0.68 0.09 0.22 0.54 0.40 0.20 0.38 0.53 0.13 0.24 0.52 0.21 0.53 0.39 0.08 0.18 0.43 0.09 0.20 0.46 0.07 0.10 0.70 0.02 0.04 0.58 0.08 0.09 0.94 0.22 0.24 0.91 0.21 0.29 0.72 0.00 0.00 1.31 0.24 0.11 2.28 0.06 0.03 2.14 0.00 0.00 0.26 0.03 0.20 0.13  x x x x x x x x x x x x  x x x x x x  x x x x x x x x x x x 146  RPKM Gene ID  Mest+/KO WT+miR  Gm5468 Gm5623 Gm6505 Gpat2 Gpr114 Gpr149 Gsx1 Guca1b Gvin1 Gzmd Gzme H2-T22 Hey2 Hist1h1a Hist1h1d Hist1h1t Hist1h2af Hist1h2ai Hist1h2ak Hist1h2an Hist1h2ba Hist1h2be Hist1h2bl Hist1h2bm Hist1h2bn Hist1h3e Hist1h3h Hist1h4c Hist1h4d Hist1h4f Hist1h4h Hist1h4i Hist1h4n Hist2h2ac Hist2h2bb Hist2h3b Hist4h4 Hmgn2 Hsd3b1 Hsd3b6 Ifna4 Ifnb1  0.109 0.119 0.045 0.003 0.355 0.647 0.091 0.143 0.002 1.014 4.623 14.297 0.619 0.229 0.190 0.009 0.050 0.366 0.271 0.144 0.000 0.564 0.117 0.138 0.195 0.398 0.087 0.664 0.729 0.247 3.486 5.543 0.965 0.111 0.078 0.175 0.259 0.753 0.014 0.026 0.149 0.522  0.281 0.115 3.564 0.020 0.710 3.251 0.203 0.663 1.481 1.363 5.096 96.404 0.930 0.596 0.767 0.087 0.042 0.811 0.612 0.130 0.128 1.473 0.556 0.272 0.397 1.176 0.224 0.647 1.689 0.366 4.726 12.418 1.776 0.265 0.188 0.375 0.773 7.092 0.059 0.093 0.598 1.240  Fold-change WT 0.518 1.232 4.196 0.050 1.567 4.946 0.566 0.715 1.460 4.359 23.644 88.282 2.948 1.079 1.248 0.210 0.447 1.542 1.823 0.643 0.678 3.420 1.433 1.195 0.788 2.368 0.492 3.317 6.036 1.676 16.375 23.018 5.781 1.162 0.670 0.860 1.959 3.907 0.120 0.203 0.687 2.547  +/KO  Mest Mest+/KO WT+miR WT+miR WT WT 0.21 0.39 0.54 0.10 1.03 0.09 0.01 0.01 0.85 0.07 0.16 0.40 0.23 0.50 0.45 0.13 0.20 0.66 0.16 0.45 0.36 0.20 0.22 0.93 0.00 0.00 1.01 0.23 0.74 0.31 0.20 0.91 0.22 0.16 0.15 1.09 0.21 0.66 0.32 0.21 0.38 0.55 0.15 0.25 0.61 0.04 0.11 0.41 0.11 1.19 0.09 0.24 0.45 0.53 0.15 0.44 0.34 0.22 1.11 0.20 0.00 0.00 0.19 0.16 0.38 0.43 0.08 0.21 0.39 0.12 0.51 0.23 0.25 0.49 0.50 0.17 0.34 0.50 0.18 0.39 0.45 0.20 1.03 0.20 0.12 0.43 0.28 0.15 0.67 0.22 0.21 0.74 0.29 0.24 0.45 0.54 0.17 0.54 0.31 0.10 0.42 0.23 0.12 0.41 0.28 0.20 0.47 0.44 0.13 0.33 0.39 0.19 0.11 1.82 0.12 0.24 0.49 0.13 0.28 0.46 0.22 0.25 0.87 0.20 0.42 0.49  x x  x x x x x  x x x  x x x x  x x  x  x x x  147  RPKM Gene ID  Mest+/KO WT+miR  Ifnz Il11 Il1a Il1f6 Inhba Isoc2b Itgam Kcnk10 Kcnk13 Kcnmb2 Kctd8 Kpna7 Kprp Krt20 Krt79 Krtap1-5 Krtap16-8 Krtap3-2 Krtap3-3 Krtap4-16 Lair1 Lars2 Lce1f Lce1g Lce1h Lce1l Lce1m Lingo2 Lipm Mcpt1 Ms4a4c Muc15 Nek11 Nkx6-2 Nlrp10 Nr1i2 Nr4a3 Obfc2a Oca2 Olfm2 Olfr1394 Olfr432  0.000 46.403 0.102 0.012 90.441 0.406 0.037 0.756 0.038 0.026 0.199 0.030 0.006 0.545 0.002 0.656 0.000 0.008 0.000 0.009 0.036 114.718 0.071 0.074 0.173 0.300 0.010 0.077 0.032 0.058 0.148 0.005 0.039 0.026 0.035 0.010 0.811 13.505 0.002 2.163 0.014 0.166  0.077 55.166 0.089 0.014 82.820 1.413 0.075 1.182 0.051 0.022 0.242 0.084 0.022 1.267 0.016 1.185 0.000 0.067 0.013 0.000 0.103 154.726 0.261 0.074 0.407 1.635 0.095 0.481 0.062 0.138 0.797 0.025 0.112 0.101 0.145 0.035 1.437 25.343 0.011 3.481 0.108 0.196  Fold-change Mest Mest+/KO WT+miR WT+miR WT WT 0.089 0.00 0.00 0.86 194.164 0.24 0.84 0.28 0.495 0.21 1.14 0.18 0.151 0.08 0.87 0.09 387.640 0.23 1.09 0.21 1.698 0.24 0.29 0.83 0.171 0.22 0.50 0.44 3.920 0.19 0.64 0.30 0.160 0.24 0.74 0.32 0.185 0.14 1.16 0.12 0.837 0.24 0.83 0.29 0.333 0.09 0.36 0.25 0.123 0.05 0.29 0.18 3.286 0.17 0.43 0.39 0.055 0.04 0.13 0.30 2.856 0.23 0.55 0.41 0.270 0.00 0.00 0.00 0.663 0.01 0.12 0.10 0.136 0.00 0.00 0.10 0.182 0.05 1000.00 0.00 0.160 0.22 0.35 0.65 1580.170 0.07 0.74 0.10 2.446 0.03 0.27 0.11 1.703 0.04 0.99 0.04 4.787 0.04 0.42 0.09 6.167 0.05 0.18 0.27 0.454 0.02 0.10 0.21 0.695 0.11 0.16 0.69 0.173 0.18 0.51 0.36 0.846 0.07 0.42 0.16 1.122 0.13 0.19 0.71 0.088 0.06 0.21 0.29 0.163 0.24 0.35 0.69 0.122 0.22 0.26 0.83 0.351 0.10 0.24 0.41 0.075 0.13 0.29 0.46 3.354 0.24 0.56 0.43 54.735 0.25 0.53 0.46 0.047 0.04 0.16 0.22 9.838 0.22 0.62 0.35 0.149 0.10 0.13 0.73 0.670 0.25 0.84 0.29 WT  +/KO  x x x x  x  x  x x x x x x x x x x x x x  x  x x  148  RPKM Gene ID  Mest+/KO WT+miR  Olfr433 Olfr456 Olfr513 Opn5 Orai2-ps P2ry12 Parm1 Pdc Pdcd1lg2 Pfpl Pgpep1l Plg Ppbp Prl3a1 Prl3d1 Prl3d2 Prl3d3 Prl5a1 Prl7a1 Prl8a6 Prl8a9 Prrxl1 Prss27 Ptgs2 Ptprc Pyhin1 Rbm24 Rbpsuh-rs3 Rgs16 Rhox9 Sdcbp2 Serpina9 Serpinb2 Serpinb9e Slc6a12 Sost Spata4 Sprr1a Sprr2h Sprr2k Tarm1 Tcl1b2  0.063 0.224 0.009 0.016 0.003 0.025 1.982 0.159 0.131 0.000 0.033 0.069 31.070 0.162 0.060 0.381 0.070 0.000 0.016 0.083 0.032 0.005 0.057 110.521 0.040 2.696 0.431 0.086 20.029 0.011 3.539 0.010 0.591 0.330 0.482 0.002 0.036 5.763 0.034 1.558 0.257 0.009  0.188 0.407 0.163 0.070 0.256 0.177 2.781 0.342 0.309 0.003 0.163 0.108 41.377 0.704 0.166 0.657 0.097 0.009 0.063 0.207 0.121 0.034 0.081 104.524 0.088 14.107 1.099 3.283 23.255 0.169 17.838 0.047 1.391 0.292 1.183 0.010 0.060 23.477 0.046 2.696 0.580 0.039  Fold-change WT 0.303 2.073 0.332 0.162 0.301 0.588 10.237 0.690 0.677 0.078 0.354 0.162 152.849 0.686 0.306 1.752 0.325 0.202 0.123 0.371 0.349 0.060 0.416 560.919 0.341 11.896 2.573 3.670 98.367 0.161 37.943 0.082 9.369 1.518 2.365 0.035 0.199 37.379 0.243 6.343 1.603 0.124  +/KO  Mest Mest+/KO WT+miR WT+miR WT WT 0.21 0.33 0.62 0.11 0.55 0.20 0.03 0.06 0.49 0.10 0.23 0.43 0.01 0.01 0.85 0.04 0.14 0.30 0.19 0.71 0.27 0.23 0.47 0.50 0.19 0.42 0.46 0.00 0.00 0.04 0.09 0.20 0.46 0.43 0.64 0.67 0.20 0.75 0.27 0.24 0.23 1.03 0.20 0.36 0.54 0.22 0.58 0.38 0.22 0.72 0.30 0.00 0.00 0.05 0.13 0.26 0.51 0.22 0.40 0.56 0.09 0.27 0.35 0.09 0.16 0.56 0.14 0.70 0.19 0.20 1.06 0.19 0.12 0.45 0.26 0.23 0.19 1.19 0.17 0.39 0.43 0.02 0.03 0.89 0.20 0.86 0.24 0.07 0.06 1.05 0.09 0.20 0.47 0.12 0.21 0.57 0.06 0.43 0.15 0.22 1.13 0.19 0.20 0.41 0.50 0.06 0.22 0.27 0.18 0.61 0.30 0.15 0.25 0.63 0.14 0.75 0.19 0.25 0.58 0.42 0.16 0.44 0.36 0.07 0.22 0.31  x x x x x  x x x x  x  x x x x x x x x x x x  x x  149  RPKM Gene ID  Mest+/KO WT+miR  Tmem171 Tmem91 Tmprss11bnl Tmprss11d Tnfsf18 Trim10 Trim12a Trim5 Trpc4 Ush1g Uts2d Vmn1r65 Vmn2r79 Vsig2 Vsx2 Xcr1 Xirp2 Zcchc16 Zfp804b Average StDev  0.337 0.046 0.018 0.006 0.272 0.117 0.002 0.176 0.056 0.043 0.000 0.006 0.000 0.217 0.013 0.008 0.006 0.007 0.001 3.170 16.004  0.784 0.147 0.022 0.085 0.649 0.313 8.853 1.867 0.187 0.119 0.024 0.016 0.058 0.858 0.080 0.004 0.011 0.019 0.142 5.139 20.976  Fold-change WT 1.907 0.192 0.078 0.110 1.657 0.483 7.709 2.263 0.296 0.177 0.300 0.054 0.386 1.784 0.073 0.042 0.092 0.132 0.114 20.770 128.144  +/KO  Mest Mest+/KO WT+miR WT+miR WT WT 0.18 0.43 0.41 0.24 0.31 0.77 0.23 0.79 0.29 0.06 0.07 0.78 0.16 0.42 0.39 0.24 0.37 0.65 0.00 0.00 1.15 0.08 0.09 0.82 0.19 0.30 0.63 0.24 0.36 0.67 0.00 0.00 0.08 0.12 0.40 0.29 0.00 0.00 0.15 0.12 0.25 0.48 0.18 0.17 1.10 0.18 1.74 0.10 0.06 0.50 0.12 0.05 0.37 0.14 0.01 0.01 1.24  x  x x  x x x x x x x  Table C.2 Significantly up-regulated genes in the Mest+/KO PEFs compared to the WT PEFs. This list corresponds to group D in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. Note: A fold-change of 1000 indicates that the denominator RPKM value used to determine the fold-change was zero. RPKM Gene ID  Mest+/KO WT+miR  1600012P17Rik 1810041L15Rik 2610044O15Rik 2810459M11Rik 3110035E14Rik 4933415A04Rik 9130219A07Rik A230065H16Rik A2m Aass Acan  0.032 0.094 2.166 0.035 0.070 0.070 0.414 0.460 0.108 0.201 37.945  0.000 0.086 0.565 0.018 0.020 0.020 0.129 0.128 0.055 0.077 7.728  Fold-change WT 0.004 0.014 0.433 0.003 0.012 0.012 0.015 0.032 0.019 0.030 5.048  +/KO  Mest WT+miR 1000.00 1.10 3.83 1.96 3.54 3.54 3.21 3.60 1.96 2.61 4.91  Mest+/KO WT 8.58 6.65 5.01 11.05 5.72 5.72 28.20 14.45 5.55 6.80 7.52  WT+miR WT 0.00 6.05 1.31 5.64 1.61 1.61 8.77 4.01 2.83 2.60 1.53  x  x x x  150  RPKM Gene ID  Mest+/KO WT+miR  Accn3 Acss3 Agt Agtr2 Aldh1a1 Aldh1a2 Amph Angptl1 AU023871 B3galt2 Barx2 Bex4 Bmp2 Bmp5 C130050O18Rik C1qtnf7 C530028O21Rik Camkv Camsap3 Car9 Ccdc3 Ccdc67 Ccr7 Ccrl1 Cd200 Cd244 Cd34 Ces2g Chdh Cib4 Clec11a Cmah Cntnap3 Col8a2 Col9a1 Col9a2 Comp Cox4i2 Cpne7 Crb2  2.015 0.114 0.192 5.627 7.109 5.297 0.175 0.246 0.026 0.035 0.316 0.123 8.149 2.007 0.229 0.138 7.959 0.200 0.287 2.084 5.778 0.038 0.212 0.015 5.696 0.095 29.638 0.677 0.174 0.304 3.470 0.109 0.196 1.276 0.060 0.423 2.147 2.290 0.209 1.002  0.794 0.044 0.135 1.669 3.008 2.230 0.071 0.111 0.017 0.026 0.141 0.094 3.962 0.987 0.199 0.106 0.361 0.106 0.195 0.730 2.207 0.030 0.087 0.004 1.173 0.026 15.166 0.358 0.099 0.091 1.534 0.004 0.028 0.415 0.065 0.139 0.595 1.431 0.100 0.421  Fold-change WT 0.281 0.025 0.047 0.441 1.504 0.998 0.041 0.039 0.002 0.008 0.063 0.000 1.745 0.280 0.051 0.028 0.259 0.024 0.072 0.342 0.988 0.006 0.032 0.004 1.234 0.018 5.029 0.132 0.033 0.061 0.784 0.005 0.025 0.181 0.015 0.041 0.382 0.565 0.045 0.136  +/KO  Mest WT+miR 2.54 2.56 1.42 3.37 2.36 2.38 2.47 2.21 1.54 1.35 2.24 1.31 2.06 2.03 1.15 1.31 22.06 1.90 1.47 2.85 2.62 1.27 2.44 4.02 4.86 3.67 1.95 1.89 1.76 3.33 2.26 26.94 6.99 3.07 0.92 3.04 3.61 1.60 2.08 2.38  Mest+/KO WT 7.18 4.51 4.09 12.75 4.73 5.31 4.30 6.25 11.61 4.31 5.03 1000.00 4.67 7.16 4.45 4.94 30.68 8.29 4.00 6.09 5.85 6.01 6.68 3.33 4.62 5.22 5.89 5.14 5.25 4.98 4.43 23.23 7.86 7.06 4.12 10.29 5.63 4.05 4.68 7.37  WT+miR WT 2.83 1.76 2.87 3.78 2.00 2.23 1.74 2.82 7.52 3.18 2.24 1000.00 2.27 3.52 3.88 3.78 1.39 4.37 2.72 2.13 2.23 4.73 2.74 0.83 0.95 1.42 3.02 2.72 2.98 1.50 1.96 0.86 1.12 2.30 4.45 3.39 1.56 2.53 2.25 3.10  x  x  x  x  151  RPKM Gene ID  Mest+/KO WT+miR  Cyp2f2 Cyp2s1 Cyp4f18 Cyp4f39 D1Pas1 D630039A03Rik Dcaf12l2 Dchs1 Ddc Ddx4 Defb1 Disp2 Dmp1 Dmrtb1 Dock10 Dpp4 E030010N08Rik Ednrb Egfl6 Elfn2 Enpep Entpd3 Epyc Esm1 Fam151a Fat3 Fbn2 Fmo2 Frmpd3 Fut9 Fxyd1 Gabra3 Gal3st3 Galnt5 Galnt9 Gas2l2 Gata5 Gdf10 Gdf5 Glt1d1  0.034 1.807 0.047 0.175 0.032 0.073 0.088 14.796 1.024 1.223 0.170 0.105 0.057 0.190 2.129 0.319 1.867 12.646 2.261 0.199 0.868 0.111 0.465 2.528 0.045 0.099 33.342 0.094 0.133 0.062 0.908 0.157 0.725 0.360 0.655 0.039 0.132 1.962 7.061 0.125  0.027 0.663 0.011 0.068 0.008 0.021 0.128 9.373 0.876 0.612 0.011 0.034 0.007 0.056 1.234 0.180 0.981 8.811 1.505 0.131 0.539 0.026 0.156 1.244 0.032 0.060 11.330 0.024 0.038 0.031 0.634 0.106 0.402 0.108 0.305 0.002 0.095 0.698 4.409 0.048  Fold-change WT 0.000 0.147 0.004 0.028 0.003 0.014 0.018 2.298 0.183 0.224 0.012 0.025 0.008 0.040 0.470 0.078 0.270 2.977 0.542 0.028 0.148 0.020 0.078 0.336 0.000 0.016 6.453 0.014 0.007 0.013 0.145 0.026 0.076 0.039 0.061 0.001 0.009 0.173 0.786 0.030  +/KO  Mest WT+miR 1.27 2.73 4.33 2.57 4.13 3.48 0.68 1.58 1.17 2.00 15.06 3.06 8.04 3.41 1.72 1.77 1.90 1.44 1.50 1.51 1.61 4.27 2.97 2.03 1.39 1.65 2.94 3.88 3.48 2.00 1.43 1.48 1.81 3.34 2.15 20.17 1.39 2.81 1.60 2.61  Mest+/KO WT 1000.00 12.32 13.00 6.30 9.71 5.06 4.83 6.44 5.59 5.46 14.15 4.13 7.56 4.70 4.53 4.09 6.91 4.25 4.17 7.08 5.86 5.61 5.98 7.51 1000.00 6.07 5.17 6.77 17.86 4.84 6.27 6.08 9.60 9.25 10.76 45.88 14.45 11.34 8.98 4.14  WT+miR WT 1000.00 4.52 3.00 2.45 2.35 1.45 7.05 4.08 4.78 2.73 0.94 1.35 0.94 1.38 2.63 2.31 3.63 2.96 2.78 4.68 3.64 1.31 2.01 3.70 1000.00 3.68 1.76 1.75 5.14 2.42 4.38 4.10 5.32 2.77 5.01 2.27 10.39 4.03 5.61 1.59  x x  x x x  x  x  x x x x x x x x  152  RPKM Gene ID  Mest+/KO WT+miR  Gm10369 Gm10749 Gm10828 Gm12824 Gm13247 Gm14403 Gm14410 Gm16223 Gm3940 Gm4737 Gm52 Gm5451 Gm5567 Gm5662 Gm8787 Gm9104 Gm9755 Gm98 Gpm6a Gpr39 Gpr68 Gpr88 Gprc6a Gprin3 Gtf2a1l Hmcn1 Hopx Il16 Il22ra1 Il28ra Il31ra Irs4 Itgb3bp Itih5 Kank4 Kcne1l Kcnk15 Krt14 Krt18 Krt8  0.082 0.372 0.038 6.111 0.111 5.999 3.087 0.027 0.247 14.480 0.289 0.247 0.025 0.052 0.040 0.106 0.041 11.571 2.779 3.272 0.375 0.346 0.034 1.066 0.099 1.118 8.280 0.829 0.488 0.137 0.364 0.031 1.071 3.852 5.434 3.128 0.253 6.747 6.058 2.079  0.004 0.000 0.025 2.359 0.044 1.332 0.743 0.004 0.096 0.086 0.076 0.125 0.010 0.013 0.021 0.091 0.023 6.343 0.999 1.064 0.084 0.157 0.008 0.221 0.032 0.329 2.879 0.595 0.086 0.083 0.208 0.010 0.410 2.803 0.190 3.113 0.073 3.029 8.345 1.394  Fold-change WT 0.000 0.000 0.003 0.988 0.027 1.209 0.602 0.000 0.024 0.104 0.032 0.057 0.002 0.005 0.000 0.000 0.003 2.392 0.266 0.740 0.085 0.030 0.004 0.064 0.021 0.189 1.659 0.179 0.015 0.033 0.043 0.007 0.246 0.948 0.108 0.373 0.061 0.840 1.429 0.199  +/KO  Mest WT+miR 19.15 1000.00 1.52 2.59 2.52 4.50 4.15 7.05 2.58 168.12 3.80 1.97 2.67 4.00 1.89 1.16 1.75 1.82 2.78 3.08 4.46 2.21 4.30 4.82 3.09 3.40 2.88 1.39 5.67 1.64 1.75 3.08 2.61 1.37 28.53 1.00 3.48 2.23 0.73 1.49  Mest+/KO WT 1000.00 1000.00 11.46 6.18 4.05 4.96 5.13 1000.00 10.19 138.85 9.14 4.35 15.05 11.27 1000.00 1000.00 14.16 4.84 10.44 4.42 4.44 11.43 9.25 16.63 4.75 5.91 4.99 4.62 33.33 4.22 8.38 4.71 4.35 4.06 50.28 8.39 4.15 8.04 4.24 10.42  WT+miR WT 1000.00 1000.00 7.52 2.39 1.61 1.10 1.23 1000.00 3.95 0.83 2.41 2.22 5.64 2.82 1000.00 1000.00 8.07 2.65 3.76 1.44 0.99 5.18 2.15 3.45 1.54 1.74 1.74 3.32 5.87 2.57 4.78 1.53 1.67 2.96 1.76 8.35 1.19 3.61 5.84 6.99  x  x  x x  x  x x 153  RPKM Gene ID  Mest+/KO WT+miR  Lbp Lcn2 Lctl Lrrn4 Lsamp Ly86 Mamstr Matn1 Mc3r Mc4r Meox1 Meox2 Mmrn1 Muc16 Mycn Myrip Naip6 Ndp Ndufa4l2 Nfam1 Nfe2 Nkx2-3 Nos1 Odz4 Olfml1 Otor Pabpc5 Padi1 Panx3 Pcdhb10 Pcdhb12 Pcdhb13 Pcdhb15 Pcdhb21 Pdzk1ip1 Pgr Pkhd1l1 Plcd4 Plekhb1 Popdc3  0.561 0.326 0.133 10.537 0.064 0.141 0.950 0.031 0.031 0.056 1.850 2.753 0.076 0.320 0.968 0.030 0.063 0.979 6.667 0.082 0.055 0.339 0.116 2.104 0.887 0.239 0.134 0.179 0.032 0.045 0.074 0.174 0.592 0.925 0.099 0.139 0.943 0.113 0.530 0.865  0.305 0.519 0.068 4.358 0.076 0.296 0.443 0.020 0.006 0.017 0.759 1.476 0.017 0.194 0.533 0.004 0.030 0.413 2.346 0.057 0.029 0.246 0.025 0.547 0.405 0.129 0.091 0.074 0.009 0.021 0.021 0.035 0.130 0.239 0.096 0.048 0.227 0.041 0.092 0.551  Fold-change WT 0.088 0.074 0.020 1.192 0.006 0.010 0.159 0.000 0.000 0.008 0.324 0.618 0.012 0.056 0.237 0.000 0.014 0.087 0.669 0.020 0.000 0.030 0.011 0.356 0.048 0.033 0.021 0.031 0.004 0.000 0.017 0.020 0.129 0.171 0.011 0.024 0.051 0.020 0.095 0.183  +/KO  Mest WT+miR 1.84 0.63 1.96 2.42 0.85 0.48 2.14 1.56 5.51 3.22 2.44 1.87 4.52 1.65 1.82 7.34 2.12 2.37 2.84 1.44 1.89 1.38 4.54 3.85 2.19 1.85 1.48 2.40 3.74 2.18 3.52 5.03 4.55 3.87 1.03 2.91 4.15 2.76 5.74 1.57  Mest+/KO WT 6.40 4.38 6.76 8.84 10.35 13.45 5.97 1000.00 1000.00 7.36 5.72 4.46 6.08 5.75 4.08 1000.00 4.43 11.30 9.96 4.05 1000.00 11.40 10.14 5.91 18.44 7.23 6.45 5.84 8.00 1000.00 4.30 8.78 4.59 5.40 8.72 5.86 18.51 5.63 5.59 4.73  WT+miR WT 3.49 6.98 3.45 3.66 12.21 28.23 2.79 1000.00 1000.00 2.28 2.35 2.39 1.34 3.49 2.25 1000.00 2.09 4.77 3.51 2.81 1000.00 8.27 2.23 1.54 8.42 3.91 4.36 2.43 2.14 1000.00 1.22 1.75 1.01 1.40 8.46 2.01 4.46 2.04 0.97 3.01  x  x x  x  x x  x x  x x  154  RPKM Gene ID  Mest+/KO WT+miR  Ppl Ppp1r3g Ppp2r2b Prg4 Prkcb Prr15 Prr15l Rassf4 Rassf9 Reps2 Rgs14 Rnf112 Rspo1 Rtn1 Rxfp1 Ryr2 Scara5 Scube3 Sez6l Sfrp2 She Slc14a1 Slc17a6 Slc27a6 Slc45a1 Slc4a10 Slurp1 Smpd3 Sned1 Sox6 Speer3 Sprn Sstr2 St8sia1 Sult5a1 Susd5 Sv2a Tagap Tcf21 Tcf23  5.278 0.334 0.485 32.809 0.140 0.373 0.072 0.327 0.292 0.766 0.028 0.064 1.601 0.142 0.379 0.033 0.649 3.413 0.069 29.784 0.215 5.233 0.552 0.593 0.098 0.057 1.144 3.059 11.951 2.301 0.247 0.027 5.840 0.096 0.036 0.569 1.287 0.192 0.466 0.032  3.422 0.296 0.172 2.791 0.059 0.228 0.028 0.102 0.114 0.404 0.016 0.021 0.914 0.061 0.120 0.021 0.353 1.100 0.010 19.508 0.078 1.187 0.171 0.425 0.054 0.041 1.054 0.906 5.093 1.074 0.033 0.012 0.980 0.051 0.016 0.215 0.834 0.077 0.221 0.024  Fold-change WT 1.227 0.000 0.046 1.963 0.013 0.030 0.006 0.013 0.067 0.181 0.002 0.003 0.399 0.031 0.046 0.007 0.086 0.516 0.004 6.923 0.043 0.807 0.131 0.131 0.015 0.014 0.227 0.661 1.311 0.564 0.039 0.003 0.771 0.012 0.000 0.114 0.260 0.023 0.026 0.005  +/KO  Mest WT+miR 1.54 1.13 2.82 11.76 2.36 1.64 2.58 3.21 2.56 1.89 1.80 2.99 1.75 2.32 3.17 1.53 1.84 3.10 7.10 1.53 2.77 4.41 3.23 1.40 1.83 1.38 1.09 3.38 2.35 2.14 7.54 2.29 5.96 1.89 2.18 2.64 1.54 2.51 2.11 1.35  Mest+/KO WT 4.30 1000.00 10.55 16.72 10.66 12.38 11.87 24.38 4.38 4.24 13.09 21.92 4.01 4.60 8.22 4.67 7.54 6.61 15.85 4.30 4.96 6.49 4.22 4.53 6.70 4.13 5.03 4.63 9.11 4.08 6.29 8.59 7.58 8.03 1000.00 4.98 4.94 8.32 18.17 6.34  WT+miR WT 2.79 1000.00 3.74 1.42 4.51 7.57 4.59 7.60 1.71 2.24 7.26 7.32 2.29 1.98 2.60 3.04 4.10 2.13 2.23 2.82 1.79 1.47 1.31 3.24 3.66 3.00 4.64 1.37 3.88 1.90 0.83 3.76 1.27 4.25 1000.00 1.89 3.20 3.32 8.61 4.70  x  x x x x  x  x  x  x  155  RPKM Gene ID  Mest+/KO WT+miR  Thbs4 Thsd4 Timp4 Tmem26 Tnfsf8 Tpsab1 Trhr2 Tspan18 Upk1b Upk3b Vat1l Wnt16 Wnt7b Wt1 Zbtb8b Zcchc12 Zfp92 Average StDev  25.621 0.017 0.111 9.435 0.100 0.089 0.795 0.346 25.311 19.287 2.126 1.972 0.145 2.288 0.099 0.220 0.042 2.420 5.842  11.871 0.000 0.083 4.721 0.048 0.021 0.085 0.208 12.222 6.644 1.315 1.448 0.082 1.408 0.059 0.275 0.005 1.000 2.498  Fold-change WT 2.149 0.014 0.019 1.918 0.011 0.009 0.116 0.078 4.680 3.134 0.518 0.415 0.020 0.310 0.013 0.044 0.000 0.375 0.947  +/KO  Mest WT+miR 2.16 1000.00 1.33 2.00 2.08 4.15 9.35 1.66 2.07 2.90 1.62 1.36 1.77 1.62 1.68 0.80 9.12  Mest+/KO WT 11.92 1.23 5.73 4.92 9.10 9.80 6.88 4.47 5.41 6.15 4.11 4.76 7.17 7.38 7.68 5.05 1000.00  WT+miR WT 5.52 0.00 4.29 2.46 4.38 2.36 0.74 2.68 2.61 2.12 2.54 3.49 4.06 4.54 4.58 6.31 1000.00  x x x  x x x x  Table C.3 Significantly down-regulated genes in the WT+miR PEFs compared to the WT PEFs. This list corresponds to group B in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. RPKM Gene ID 1700023E05Rik 1810046K07Rik 4732440D04Rik 4930500M09Rik A730020M07Rik Akr1c19 Alox8 Ankrd22 AY074887 Bcl2l15 C130079G13Rik  Mest+/KO WT+miR 0.044 0.036 0.038 0.006 0.027 0.056 0.027 0.009 0.226 0.192 0.002  0.001 0.002 0.004 0.007 0.010 0.119 0.025 0.000 0.128 0.157 0.000  Fold-change WT 0.037 0.072 0.087 0.094 0.109 0.722 0.115 0.048 0.684 0.634 0.034  Mest+/KO WT+miR 33.73 15.34 8.52 0.83 2.69 0.47 1.08 1000.00 1.76 1.22 1000.00  Mest+/KO WT 1.19 0.50 0.43 0.07 0.25 0.08 0.23 0.19 0.33 0.30 0.05  WT+miR WT 0.04 0.03 0.05 0.08 0.09 0.16 0.21 0.00 0.19 0.25 0.00  x x x x x  x 156  RPKM Gene ID  Mest+/KO WT+miR  Calr4 Ccl20 Clec2h Cma2 Col10a1 Ctgf Cxcl2 Cyp1a1 Derl3 Dntt Fnd3c2 Foxg1 Foxn1 Gldn Gm10645 Gm10718 Gm10800 Gm17509 Gm17535 Gm5152 Gm5623 Gzme Hist1h2af Hist1h2an Hist1h2ba Hist1h2bm Hist1h4c Hist1h4f Hist2h2aa2 Hist2h2ac Il12b Il1a Il1f6 Inhba Ins2 Kcnmb2 Klrg1 Kprp Krtap16-7 Krtap16-8  0.108 6.336 0.057 0.515 28.664 144.429 3.573 0.099 0.357 0.030 0.000 0.037 0.015 1.546 0.048 0.055 2.983 0.009 0.019 0.003 0.119 4.623 0.050 0.144 0.000 0.138 0.664 0.247 0.140 0.111 0.076 0.102 0.012 90.441 0.000 0.026 0.089 0.006 0.052 0.000  0.074 4.784 0.080 0.804 9.592 196.824 3.375 0.181 0.084 0.053 0.008 0.080 0.021 0.439 0.045 0.005 0.782 0.002 0.001 0.013 0.115 5.096 0.042 0.130 0.128 0.272 0.647 0.366 0.098 0.265 0.027 0.089 0.014 82.820 0.000 0.022 0.035 0.022 0.007 0.000  Fold-change WT 0.366 25.264 0.390 3.758 40.874 823.556 17.667 1.697 0.339 0.222 0.043 0.424 0.235 2.352 0.185 0.075 5.887 0.035 0.086 0.099 1.232 23.644 0.447 0.643 0.678 1.195 3.317 1.676 0.411 1.162 0.169 0.495 0.151 387.640 0.124 0.185 0.140 0.123 0.139 0.270  Mest+/KO WT+miR 1.48 1.32 0.72 0.64 2.99 0.73 1.06 0.55 4.25 0.57 0.00 0.47 0.68 3.52 1.06 10.83 3.81 4.35 15.07 0.20 1.03 0.91 1.19 1.11 0.00 0.51 1.03 0.67 1.42 0.42 2.78 1.14 0.87 1.09 0.00 1.16 2.55 0.29 7.87 0.00  Mest+/KO WT 0.30 0.25 0.15 0.14 0.70 0.18 0.20 0.06 1.05 0.14 0.00 0.09 0.06 0.66 0.26 0.73 0.51 0.26 0.22 0.03 0.10 0.20 0.11 0.22 0.00 0.12 0.20 0.15 0.34 0.10 0.45 0.21 0.08 0.23 0.00 0.14 0.63 0.05 0.37 0.00  WT+miR WT 0.20 0.19 0.20 0.21 0.23 0.24 0.19 0.11 0.25 0.24 0.19 0.19 0.09 0.19 0.24 0.07 0.13 0.06 0.01 0.13 0.09 0.22 0.09 0.20 0.19 0.23 0.20 0.22 0.24 0.23 0.16 0.18 0.09 0.21 0.00 0.12 0.25 0.18 0.05 0.00  x x x x x x x x  x x x x  x x  x x x x  x  x x x  157  RPKM Gene ID  Mest+/KO WT+miR  Krtap3-2 Krtap3-3 Krtap4-16 Lalba Lars2 Lce1f Lce1g Lce1h Lce1m Mcpt1 Myl10 Oca2 Olfr456 Olig3 Opn1mw Otol1 Pfpl Pglyrp3 Pkp1 Prl5a1 Prss27 Ptgs2 Rgs16 Scrt1 Serpinb2 Serpinb9e Serpinb9f Serpinb9g Spink8 Sprr2h Tktl1 Tmem190 Uts2d Vmn2r79 Wbscr28 Xcr1 Xirp2 Zcchc16 Average Stdev  0.008 0.000 0.009 0.012 114.718 0.071 0.074 0.173 0.010 0.058 0.052 0.002 0.224 0.015 0.004 0.018 0.000 0.035 2.324 0.000 0.057 110.521 20.029 0.026 0.591 0.330 0.080 0.155 0.022 0.034 0.109 0.357 0.000 0.000 0.000 0.008 0.006 0.007 6.031 24.424  0.067 0.013 0.000 0.008 154.726 0.261 0.074 0.407 0.095 0.138 0.012 0.011 0.407 0.000 0.000 0.000 0.003 0.006 1.369 0.009 0.081 104.524 23.255 0.008 1.391 0.292 0.019 0.120 0.019 0.046 0.075 0.084 0.024 0.058 0.000 0.004 0.011 0.019 6.691 29.615  Fold-change Mest+/KO WT+miR 0.663 0.12 0.136 0.00 0.182 1000.00 0.124 1.55 1580.170 0.74 2.446 0.27 1.703 0.99 4.787 0.42 0.454 0.10 0.846 0.42 0.115 4.45 0.047 0.16 2.073 0.55 0.069 1000.00 0.052 1000.00 0.046 1000.00 0.078 0.00 0.147 5.55 6.032 1.70 0.202 0.00 0.416 0.70 560.919 1.06 98.367 0.86 0.040 3.35 9.369 0.43 1.518 1.13 0.193 4.28 0.603 1.29 0.170 1.16 0.243 0.75 0.331 1.47 0.339 4.25 0.300 0.00 0.386 0.00 0.062 1000.00 0.042 1.74 0.092 0.50 0.132 0.37 40.718 199.591 WT  Mest+/KO WT 0.01 0.00 0.05 0.09 0.07 0.03 0.04 0.04 0.02 0.07 0.45 0.04 0.11 0.21 0.07 0.38 0.00 0.24 0.39 0.00 0.14 0.20 0.20 0.65 0.06 0.22 0.41 0.26 0.13 0.14 0.33 1.05 0.00 0.00 0.00 0.18 0.06 0.05  WT+miR WT 0.10 0.10 0.00 0.06 0.10 0.11 0.04 0.09 0.21 0.16 0.10 0.22 0.20 0.00 0.00 0.00 0.04 0.04 0.23 0.05 0.19 0.19 0.24 0.19 0.15 0.19 0.10 0.20 0.11 0.19 0.22 0.25 0.08 0.15 0.00 0.10 0.12 0.14  x x x x x x x x x x x  x x x x  x x x x  x x  x x  158  Table C.4 Significantly up-regulated genes in the WT+miR PEFs compared to the WT PEFs. This list corresponds to group E in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. Note: A fold-change of 1000 indicates that the denominator RPKM value used to determine the fold-change was zero. RPKM  Fold-change  Gene ID  Mest+/KO  WT+miR  WT  1110032F04Rik 1700011M02Rik 1810041L15Rik 4933415A04Rik 9130219A07Rik A230065H16Rik Adcyap1 AI427809 Alpi Amac1 Ankrd7 Arpm1 Barhl2 Bex4 C130026L21Rik Cabp5 Calcr Camkv Ccdc67 Ccna1 Cdh19 Ces4a Clec12b Col9a1 Csn3 Cyp11a1 Cyp2f2 Cyp2s1 Dcaf12l2 Dchs1 Ddc Dlx6 Dmrta1  0.165 0.153 0.094 0.070 0.414 0.460 0.033 0.292 0.004 0.006 0.009 0.029 0.000 0.123 0.052 0.018 0.026 0.200 0.038 0.038 0.141 0.023 0.008 0.060 0.081 0.103 0.034 1.807 0.088 14.796 1.024 0.298 0.029  0.949 0.304 0.086 0.020 0.129 0.128 0.162 0.452 0.029 0.049 0.039 0.047 0.045 0.094 0.104 0.078 0.048 0.106 0.030 0.050 0.221 0.029 0.032 0.065 0.122 0.346 0.027 0.663 0.128 9.373 0.876 0.398 0.052  0.156 0.025 0.014 0.012 0.015 0.032 0.035 0.082 0.004 0.007 0.009 0.004 0.010 0.000 0.024 0.000 0.011 0.024 0.006 0.005 0.050 0.000 0.007 0.015 0.028 0.033 0.000 0.147 0.018 2.298 0.183 0.077 0.008  Mest+/KO WT+miR 0.17 0.50 1.10 3.54 3.21 3.60 0.20 0.65 0.12 0.12 0.23 0.62 0.00 1.31 0.50 0.24 0.54 1.90 1.27 0.76 0.64 0.80 0.26 0.92 0.67 0.30 1.27 2.73 0.68 1.58 1.17 0.75 0.56  Mest+/KO WT 1.05 6.22 6.65 5.72 28.20 14.45 0.94 3.55 0.82 0.82 1.03 7.58 0.00 1000.00 2.15 1000.00 2.39 8.29 6.01 8.07 2.85 1000.00 1.23 4.12 2.95 3.11 1000.00 12.32 4.83 6.44 5.59 3.90 3.68  WT+miR WT 6.08 12.36 6.05 1.61 8.77 4.01 4.64 5.48 6.57 6.84 4.41 12.29 4.50 1000.00 4.33 1000.00 4.42 4.37 4.73 10.66 4.46 1000.00 4.70 4.45 4.42 10.43 1000.00 4.52 7.05 4.08 4.78 5.20 6.61  x x x x x x x x  x x x x x  x  x x x x x  159  RPKM Gene ID Dub3 Elfn2 En2 Fam151a Fam46d Fam69c Foxd3 Frmpd3 Fxyd1 Gabra3 Gal3st3 Galnt9 Gata5 Gdf10 Gdf5 Gfra3 Gjc3 Glb1l2 Gm17269 Gm4392 Gm5045 Gm5105 Gm5887 Gm6578 Gm6871 Gm9104 Gpr88 H2-T3 Hmgcs2 Hmx3 Il22ra1 Il31ra Kcne1l Kcnh3 Klhl34 Klk1b11 Klk6 Krt18 Krt8  Fold-change  Mest+/KO  WT+miR  WT  0.015 0.199 0.012 0.045 0.052 0.014 0.092 0.133 0.908 0.157 0.725 0.655 0.132 1.962 7.061 0.034 0.009 0.062 0.128 0.102 0.015 0.022 0.038 0.030 0.007 0.106 0.346 0.036 0.014 0.006 0.488 0.364 3.128 0.019 0.004 0.010 0.028 6.058 2.079  0.123 0.131 0.030 0.032 0.163 0.048 0.230 0.038 0.634 0.106 0.402 0.305 0.095 0.698 4.409 0.140 0.025 0.084 0.259 0.153 0.095 0.041 0.043 0.049 0.060 0.091 0.157 0.046 0.038 0.037 0.086 0.208 3.113 0.073 0.043 0.068 0.043 8.345 1.394  0.013 0.028 0.004 0.000 0.040 0.004 0.040 0.007 0.145 0.026 0.076 0.061 0.009 0.173 0.786 0.016 0.000 0.018 0.039 0.031 0.018 0.008 0.008 0.005 0.008 0.000 0.030 0.005 0.004 0.007 0.015 0.043 0.373 0.011 0.000 0.006 0.009 1.429 0.199  Mest+/KO WT+miR 0.12 1.51 0.41 1.39 0.32 0.29 0.40 3.48 1.43 1.48 1.81 2.15 1.39 2.81 1.60 0.24 0.34 0.74 0.50 0.67 0.15 0.53 0.87 0.61 0.11 1.16 2.21 0.78 0.36 0.16 5.67 1.75 1.00 0.27 0.10 0.14 0.66 0.73 1.49  Mest+/KO WT 1.16 7.08 2.86 1000.00 1.31 3.20 2.32 17.86 6.27 6.08 9.60 10.76 14.45 11.34 8.98 2.20 1000.00 3.37 3.26 3.32 0.82 2.89 4.50 5.71 0.79 1000.00 11.43 6.93 3.33 0.82 33.33 8.38 8.39 1.74 1000.00 1.60 3.27 4.24 10.42  WT+miR WT 9.54 4.68 7.05 1000.00 4.05 10.87 5.81 5.14 4.38 4.10 5.32 5.01 10.39 4.03 5.61 9.00 1000.00 4.57 6.57 4.96 5.33 5.42 5.17 9.30 7.07 1000.00 5.18 8.93 9.16 5.17 5.87 4.78 8.35 6.50 1000.00 11.28 4.94 5.84 6.99  x x x  x x x x x x x x x  x  x x x  x x x x x x x x 160  RPKM  Fold-change  Gene ID  Mest+/KO  WT+miR  WT  Lcn2 Lipn Lsamp Ly6c2 Ly86 Msc Ndp Neto1 Nfe2 Nkx2-3 Olfml1 Olfml2a Pabpc5 Pdzk1ip1 Pglyrp1 Pkhd1l1 Pon1 Pou3f2 Ppp1r3g Prkcb Prr15 Prr15l Ptch2 Rassf4 Scara5 Slurp1 St8sia1 Tceal6 Tcf15 Tcf21 Thbs4 Timp4 Tmem156 Tnfsf8 Tril Ubl4b Wnt7b Wt1 Zbtb8b  0.326 0.034 0.064 0.290 0.141 0.012 0.979 0.008 0.055 0.339 0.887 0.522 0.134 0.099 0.113 0.943 0.031 0.010 0.334 0.140 0.373 0.072 0.143 0.327 0.649 1.144 0.096 0.023 0.035 0.466 25.621 0.111 0.040 0.100 2.287 0.022 0.145 2.288 0.099  0.519 0.049 0.076 0.666 0.296 0.111 0.413 0.029 0.029 0.246 0.405 1.067 0.091 0.096 0.331 0.227 0.087 0.029 0.296 0.059 0.228 0.028 0.460 0.102 0.353 1.054 0.051 0.041 0.100 0.221 11.871 0.083 0.056 0.048 3.922 0.068 0.082 1.408 0.059  0.074 0.009 0.006 0.158 0.010 0.026 0.087 0.003 0.000 0.030 0.048 0.198 0.021 0.011 0.077 0.051 0.020 0.000 0.000 0.013 0.030 0.006 0.110 0.013 0.086 0.227 0.012 0.009 0.012 0.026 2.149 0.019 0.009 0.011 0.864 0.008 0.020 0.310 0.013  Mest+/KO WT+miR 0.63 0.70 0.85 0.44 0.48 0.11 2.37 0.30 1.89 1.38 2.19 0.49 1.48 1.03 0.34 4.15 0.36 0.35 1.13 2.36 1.64 2.58 0.31 3.21 1.84 1.09 1.89 0.57 0.35 2.11 2.16 1.33 0.71 2.08 0.58 0.32 1.77 1.62 1.68  Mest+/KO WT 4.38 3.85 10.35 1.84 13.45 0.47 11.30 2.88 1000.00 11.40 18.44 2.64 6.45 8.72 1.47 18.51 1.53 1000.00 1000.00 10.66 12.38 11.87 1.30 24.38 7.54 5.03 8.03 2.46 2.96 18.17 11.92 5.73 4.23 9.10 2.65 2.86 7.17 7.38 7.68  WT+miR WT 6.98 5.54 12.21 4.22 28.23 4.20 4.77 9.77 1000.00 8.27 8.42 5.39 4.36 8.46 4.32 4.46 4.29 1000.00 1000.00 4.51 7.57 4.59 4.18 7.60 4.10 4.64 4.25 4.32 8.55 8.61 5.52 4.29 5.98 4.38 4.54 8.93 4.06 4.54 4.58  x x x x x x x x x x x  x x x x x x x x  x x x  x x x 161  RPKM Gene ID  Mest+/KO  WT+miR  Zcchc12 Zfp536 Average StDev  0.220 0.057 0.759 2.895  0.275 0.119 0.559 1.691  Fold-change Mest+/KO WT+miR 0.044 0.80 0.024 0.47 0.105 0.336 WT  Mest+/KO WT 5.05 2.40  WT+miR WT 6.31 5.07  x  Table C.5 Significantly down-regulated genes in the Mest+/KO PEFs compared to the WT+miR PEFs. This list corresponds to group C in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. Note: A fold-change of 1000 indicates that the denominator RPKM value used to determine the fold-change was zero. RPKM Gene ID 1110032F04Rik 1700011F14Rik 1700024P16Rik 1700061I17Rik 2010001M09Rik 2010005H15Rik 2210010C17Rik 2310005G13Rik 3110039M20Rik 4930528F23Rik 4931417G12Rik 4933427I04Rik 8030474K03Rik 8430432A02Rik Acp1 Adcyap1 Ahcy Akr1c20 Alpi Amac1 Ang2 Anxa10 Aqp8 Atp6v1b1 Barhl2 BC048609  Mest+/KO  WT+miR  0.165 0.007 0.162 0.007 0.048 0.000 0.011 0.006 0.067 0.171 0.025 0.000 0.026 0.018 2.871 0.033 0.828 0.064 0.004 0.006 1.080 0.002 0.005 0.018 0.000 0.009  0.949 0.075 1.460 0.039 0.200 0.683 0.050 0.029 0.305 0.703 0.108 0.027 0.202 0.091 13.263 0.162 10.347 0.321 0.029 0.049 4.712 0.028 0.031 0.096 0.045 0.080  Fold-change +/KO  Mest WT+miR 0.17 0.156 0.09 0.055 0.11 3.394 0.17 0.045 0.24 0.131 0.00 0.528 0.22 0.014 0.19 0.044 0.22 1.157 0.24 0.472 0.23 0.048 0.00 0.029 0.13 0.118 0.19 0.166 0.22 12.615 0.20 0.035 0.08 9.192 0.20 0.199 0.12 0.004 0.12 0.007 0.23 5.287 0.09 0.026 0.17 0.013 0.19 0.033 0.00 0.010 0.12 0.051 WT  Mest+/KO WT 1.05 0.12 0.05 0.15 0.37 0.00 0.80 0.13 0.06 0.36 0.53 0.00 0.22 0.11 0.23 0.94 0.09 0.32 0.82 0.82 0.20 0.09 0.41 0.55 0.00 0.19  WT+miR WT 6.08 1.37 0.43 0.86 1.53 1.29 3.58 0.67 0.26 1.49 2.26 0.94 1.71 0.55 1.05 4.64 1.13 1.61 6.57 6.84 0.89 1.05 2.34 2.90 4.50 1.56  x x  x x x  x x x x x x x x  x  162  RPKM Gene ID BC051142 BC094916 Bpifb2 Cabp5 Calca Car8 Ccr10 Cd200r2 Cd200r4 Cd3d Cd7 Cldn4 Col6a4 Crct1 Csrp3 Ctxn3 Dbh Defb43 Dub3 Ear11 Ecel1 Eif3j Enthd1 Esx1 Etos1 Fbll1 Fbxw18 Fcgr1 Fgf6 Fgg Fkbp6 Folh1 Foxa1 Gbx2 Gdf3 Gfra3 Gip Glb1l3 Glyat Gm10008  Fold-change  Mest+/KO  WT+miR  WT  0.005 0.264 0.015 0.018 0.035 0.314 0.010 0.007 0.008 0.006 0.018 0.305 0.019 0.130 0.078 0.015 0.004 0.009 0.015 0.050 1.563 0.776 0.041 0.010 0.008 0.000 0.006 0.003 0.003 0.004 0.001 0.003 0.055 0.011 0.012 0.034 0.000 0.002 0.005 0.014  0.053 1.082 0.075 0.078 0.143 3.117 0.046 0.065 0.122 0.033 0.165 1.395 0.089 1.048 0.381 0.067 0.030 0.133 0.123 0.424 6.784 3.776 0.228 0.062 0.039 0.033 0.032 0.029 0.040 0.039 0.046 0.034 0.316 0.050 0.050 0.140 0.064 0.026 0.025 0.075  0.033 0.910 0.025 0.000 0.055 2.732 0.038 0.090 0.148 0.000 0.043 1.264 0.087 3.123 0.174 0.085 0.065 0.156 0.013 0.804 4.882 4.367 0.233 0.036 0.044 0.014 0.034 0.016 0.042 0.022 0.064 0.044 0.292 0.038 0.095 0.016 0.053 0.008 0.028 0.174  Mest+/KO WT+miR 0.10 0.24 0.20 0.24 0.25 0.10 0.22 0.11 0.06 0.19 0.11 0.22 0.22 0.12 0.21 0.22 0.12 0.07 0.12 0.12 0.23 0.21 0.18 0.16 0.22 0.00 0.20 0.12 0.09 0.11 0.02 0.10 0.17 0.22 0.24 0.24 0.00 0.06 0.18 0.18  Mest+/KO WT 0.15 0.29 0.60 1000.00 0.64 0.11 0.26 0.08 0.05 1000.00 0.42 0.24 0.22 0.04 0.45 0.17 0.06 0.06 1.16 0.06 0.32 0.18 0.18 0.27 0.19 0.00 0.19 0.20 0.08 0.20 0.02 0.07 0.19 0.29 0.13 2.20 0.00 0.19 0.17 0.08  WT+miR WT 1.59 1.19 3.02 1000.00 2.58 1.14 1.20 0.72 0.82 1000.00 3.87 1.10 1.03 0.34 2.19 0.78 0.45 0.85 9.54 0.53 1.39 0.86 0.98 1.72 0.89 2.35 0.94 1.76 0.94 1.77 0.72 0.77 1.08 1.30 0.52 9.00 1.21 3.05 0.89 0.43  x x x x  x x x  x x x x x  x x x x x  x 163  RPKM Gene ID  Mest+/KO  WT+miR  Gm10115 Gm10212 Gm10479 Gm10923 Gm11127 Gm13283 Gm13287 Gm13288 Gm14226 Gm14393 Gm14399 Gm16686 Gm1679 Gm17641 Gm17680 Gm266 Gm3591 Gm4759 Gm5045 Gm5533 Gm6505 Gm6728 Gm6871 Gm6990 Gp6 Gpr149 Guca1b Gvin1 Gykl1 H2-T22 H2-T23 Hamp2 Hbb-bh1 Hist1h1d Hist1h1t Hist1h2ba Hist1h2bl Hk3 Hmgn2 Hmx3  0.000 0.030 0.020 0.028 0.103 0.003 0.000 0.003 0.003 0.009 0.065 0.002 0.003 0.004 0.000 0.197 0.007 0.022 0.015 0.003 0.045 0.006 0.007 0.001 0.022 0.647 0.143 0.002 0.009 14.297 17.246 0.036 0.104 0.190 0.009 0.000 0.117 0.007 0.753 0.006  0.084 0.153 0.155 0.139 0.993 0.064 0.066 0.034 0.036 0.075 0.269 0.102 0.057 0.046 0.795 1.871 0.034 0.828 0.095 0.030 3.564 0.088 0.060 0.119 0.090 3.251 0.663 1.481 0.036 96.404 70.452 0.157 0.499 0.767 0.087 0.128 0.556 0.032 7.092 0.037  Fold-change Mest+/KO WT+miR 0.00 0.025 0.20 0.044 0.13 0.067 0.20 0.304 0.10 1.411 0.04 0.110 0.00 0.049 0.09 0.037 0.07 0.019 0.12 0.049 0.24 0.295 0.02 0.076 0.05 0.044 0.08 0.022 0.00 0.605 0.11 0.821 0.22 0.029 0.03 0.388 0.15 0.018 0.11 0.020 0.01 4.196 0.07 0.070 0.11 0.008 0.01 0.090 0.24 0.135 0.20 4.946 0.22 0.715 0.00 1.460 0.24 0.031 0.15 88.282 0.24 48.719 0.23 0.060 0.21 0.307 0.25 1.248 0.11 0.210 0.00 0.678 0.21 1.433 0.22 0.032 0.11 3.907 0.16 0.007 WT  Mest+/KO WT 0.00 0.69 0.30 0.09 0.07 0.02 0.00 0.08 0.14 0.18 0.22 0.03 0.07 0.16 0.00 0.24 0.26 0.06 0.82 0.16 0.01 0.09 0.79 0.01 0.16 0.13 0.20 0.00 0.27 0.16 0.35 0.60 0.34 0.15 0.04 0.00 0.08 0.22 0.19 0.82  WT+miR WT 3.30 3.48 2.32 0.46 0.70 0.58 1.36 0.94 1.88 1.52 0.91 1.35 1.31 2.07 1.31 2.28 1.16 2.14 5.33 1.50 0.85 1.25 7.07 1.31 0.67 0.66 0.93 1.01 1.14 1.09 1.45 2.63 1.63 0.61 0.41 0.19 0.39 1.01 1.82 5.17  x x x x  x  x x x x x x  x x x x  x x x x x x 164  RPKM Gene ID Hsd3b1 I830077J02Rik Icam2 Ifna2 Ifna4 Ifnz Klhl34 Klk1b11 Krtap3-2 Lce1l Lce1m Lingo2 Lrrc14b Ms4a4b Ms4a4c Ms4a6d Msc Muc15 Myog Nat8b Nlrp10 Olfr1033 Olfr1394 Olfr183 Olfr513 Olfr658 Opn5 Orai2-ps Ovol2 P2ry12 Pax1 Pax2 Pcp4 Pgpep1l Plg Ppapdc1a Prl3a1 Prl7d1 Prrxl1 Pyhin1  Mest+/KO  WT+miR  0.014 0.033 0.016 0.018 0.149 0.000 0.004 0.010 0.008 0.300 0.010 0.077 0.020 0.010 0.148 0.018 0.012 0.005 0.000 0.018 0.035 0.000 0.014 0.004 0.009 0.019 0.016 0.003 0.004 0.025 0.952 0.005 0.153 0.033 0.003 0.000 0.162 0.008 0.005 2.696  0.059 0.140 0.069 0.142 0.598 0.077 0.043 0.068 0.067 1.635 0.095 0.481 0.107 0.048 0.797 0.103 0.111 0.025 0.060 0.090 0.145 0.030 0.108 0.044 0.163 0.085 0.070 0.256 0.036 0.177 4.328 0.058 0.652 0.163 0.032 0.036 0.704 0.045 0.034 14.107  Fold-change Mest+/KO WT+miR 0.24 0.120 0.24 0.098 0.23 0.051 0.13 0.046 0.25 0.687 0.00 0.089 0.10 0.000 0.14 0.006 0.12 0.663 0.18 6.167 0.10 0.454 0.16 0.695 0.19 0.059 0.21 0.055 0.19 1.122 0.18 0.105 0.11 0.026 0.21 0.088 0.00 0.021 0.20 0.090 0.24 0.351 0.00 0.011 0.13 0.149 0.08 0.017 0.06 0.332 0.22 0.041 0.23 0.162 0.01 0.301 0.12 0.027 0.14 0.588 0.22 2.127 0.09 0.036 0.24 0.236 0.20 0.354 0.10 0.058 0.00 0.000 0.23 0.686 0.17 0.019 0.16 0.060 0.19 11.896 WT  Mest+/KO WT 0.12 0.34 0.30 0.39 0.22 0.00 1000.00 1.60 0.01 0.05 0.02 0.11 0.34 0.18 0.13 0.17 0.47 0.06 0.00 0.20 0.10 0.00 0.10 0.20 0.03 0.46 0.10 0.01 0.16 0.04 0.45 0.15 0.65 0.09 0.05 1000.00 0.24 0.41 0.09 0.23  WT+miR WT 0.49 1.42 1.34 3.07 0.87 0.86 1000.00 11.28 0.10 0.27 0.21 0.69 1.81 0.87 0.71 0.98 4.20 0.29 2.82 1.00 0.41 2.59 0.73 2.51 0.49 2.07 0.43 0.85 1.32 0.30 2.04 1.63 2.76 0.46 0.55 1000.00 1.03 2.36 0.56 1.19  x  x x x x x x x x  x x  x x x x x x  x x x x x 165  RPKM Gene ID Rasal1 Rbpsuh-rs3 Resp18 Rhox9 S1pr5 Sdcbp2 Sectm1b Serpina9 Slc45a2 Sprr1a Thap8 Tmem8c Tmprss11d Trim12a Trim30b Trim30d Trim5 Ttc6 Txndc3 Vmn2r79 Vsx2 Xlr4b Zfp42 Zfp804b Average StDev  Mest+/KO  WT+miR  0.006 0.086 0.006 0.011 0.005 3.539 0.003 0.010 0.007 5.763 0.011 0.000 0.006 0.002 0.043 6.798 0.176 0.006 0.006 0.000 0.013 0.006 0.005 0.001 0.384 1.871  0.032 3.283 0.094 0.169 0.045 17.838 0.032 0.047 0.030 23.477 0.058 0.026 0.085 8.853 0.183 33.786 1.867 0.026 0.034 0.058 0.080 0.033 0.035 0.142 2.133 9.814  Fold-change Mest+/KO WT+miR 0.17 0.027 0.03 3.670 0.07 0.084 0.06 0.161 0.11 0.022 0.20 37.943 0.10 0.012 0.21 0.082 0.24 0.016 0.25 37.379 0.20 0.043 0.00 0.000 0.07 0.110 0.00 7.709 0.23 0.136 0.20 26.063 0.09 2.263 0.24 0.024 0.16 0.015 0.00 0.386 0.17 0.073 0.19 0.021 0.13 0.017 0.01 0.114 2.115 8.973 WT  Mest+/KO WT 0.21 0.02 0.08 0.07 0.23 0.09 0.27 0.12 0.44 0.15 0.27 1000.00 0.06 0.00 0.31 0.26 0.08 0.27 0.38 0.00 0.18 0.30 0.27 0.01  WT+miR WT 1.20 0.89 1.12 1.05 2.06 0.47 2.62 0.57 1.82 0.63 1.36 1000.00 0.78 1.15 1.35 1.30 0.82 1.10 2.32 0.15 1.10 1.58 2.03 1.24  x x x x x  x x  x  x x  x  Table C.6 Significantly up-regulated genes in the Mest+/KO PEFs compared to the WT+miR PEFs. This list corresponds to group F in Figure 5.8. Those genes that overlap another relationship are indicated by an "x" in the last column. Note: A fold-change of 1000 indicates that the denominator RPKM value used to determine the fold-change was zero. RPKM Gene ID  Mest+/KO  WT+miR  1600012P17Rik 1700023E05Rik Acan  0.032 0.044 37.945  0.000 0.001 7.728  Fold-change Mest+/KO WT+miR 8.58 0.004 33.73 0.037 4.91 5.048 WT  Mest+/KO WT 8.58 1.19 7.52  WT+miR WT 0.00 0.04 1.53  x x x 166  RPKM  Fold-change  Gene ID  Mest+/KO  WT+miR  WT  BC051628 C330046G03Rik C530028O21Rik Cbfa2t3 Cd200 Cmah Cntnap3 Cyct D10Bwg1379e Defb1 Derl3 Dmp1 Entpd3 F5 Gas2l2 Gm10369 Gm10749 Gm12033 Gm14403 Gm14410 Gm1661 Gm4613 Gm4737 Gpr68 Gprin3 Hist1h4m Il22ra1 Kank4 Mmrn1 Myrip Nmbr Nos1 Padi3 Pcdhb13 Pcdhb15 Pkhd1l1 Plekhb1 Prg4 Rhd  0.077 0.041 7.959 0.041 5.696 0.109 0.196 0.071 0.063 0.170 0.357 0.057 0.111 0.924 0.039 0.082 0.372 0.339 5.999 3.087 0.075 0.208 14.480 0.375 1.066 1.032 0.488 5.434 0.076 0.030 0.034 0.116 0.040 0.174 0.592 0.943 0.530 32.809 0.213  0.009 0.003 0.361 0.010 1.173 0.004 0.028 0.002 0.013 0.011 0.084 0.007 0.026 0.212 0.002 0.004 0.000 0.051 1.332 0.743 0.015 0.000 0.086 0.084 0.221 0.242 0.086 0.190 0.017 0.004 0.001 0.025 0.003 0.035 0.130 0.227 0.092 2.791 0.048  0.021 0.013 0.259 0.028 1.234 0.005 0.025 0.025 0.033 0.012 0.339 0.008 0.020 0.565 0.001 0.000 0.000 0.104 1.209 0.602 0.060 0.151 0.104 0.085 0.064 0.559 0.015 0.108 0.012 0.000 0.025 0.011 0.009 0.020 0.129 0.051 0.095 1.963 0.122  Mest+/KO WT+miR 8.41 12.13 22.06 4.11 4.86 26.94 6.99 29.75 4.75 15.06 4.25 8.04 4.27 4.37 20.17 19.15 1000.00 6.63 4.50 4.15 4.94 1000.00 168.12 4.46 4.82 4.27 5.67 28.53 4.52 7.34 22.49 4.54 12.59 5.03 4.55 4.15 5.74 11.76 4.48  Mest+/KO WT 3.72 3.15 30.68 1.49 4.62 23.23 7.86 2.88 1.91 14.15 1.05 7.56 5.61 1.64 45.88 1000.00 1000.00 3.25 4.96 5.13 1.25 1.38 138.85 4.44 16.63 1.84 33.33 50.28 6.08 1000.00 1.36 10.14 4.73 8.78 4.59 18.51 5.59 16.72 1.75  WT+miR WT 0.44 0.26 1.39 0.36 0.95 0.86 1.12 0.10 0.40 0.94 0.25 0.94 1.31 0.37 2.27 1000.00 1000.00 0.49 1.10 1.23 0.25 0.00 0.83 0.99 3.45 0.43 5.87 1.76 1.34 1000.00 0.06 2.23 0.38 1.75 1.01 0.22 0.97 1.42 0.39  x x x x  x x x x x x x x x  x x x x x x x x x x x x x  167  RPKM  Fold-change  Gene ID  Mest+/KO  WT+miR  WT  Sez6l Slc14a1 Speer3 Sstr2 Stra6 Trhr2 Zfp92 Average StDev  0.069 5.233 0.247 5.840 0.106 0.795 0.042 2.752 7.350  0.010 1.187 0.033 0.980 0.008 0.085 0.005 0.376 1.182  0.004 0.807 0.039 0.771 0.037 0.116 0.000 0.305 0.795  Mest+/KO WT+miR 7.10 4.41 7.54 5.96 13.07 9.35 9.12  Mest+/KO WT 15.85 6.49 6.29 7.58 2.87 6.88 1000.00  WT+miR WT 2.23 1.47 0.83 1.27 0.22 0.74 1000.00  x x x x x x  168  Figure C.1 DAVID analysis for significantly up-regulated genes in the Mest+/KO PEFs versus the WT PEFs. The significantly up-regulated genes in the Mest+/KO PEFs compared to the WT PEFs were analyzed by the DAVID program to look for significant biological enrichments (Table C.2; 210). The 12 annotated clusters presented here had significant enrichment scores greater than 1.3 as a score of 1.3 is equivalent to non-log scale 0.05 (210). For each annotated cluster, the top individual term within the cluster with the greatest -log10[p-value(Benjamini)] value (most significant term) is presented. The Benjamini score is one of the multiple testing correction techniques that globally corrects enrichment p-values of individual term members (210). To note, only the top 3 annotated clusters had terms with significant Benjamini scores (p<0.05).  169  Figure C.2 DAVID analysis for significantly down-regulated genes in the WT+miR PEFs versus the WT PEFs. The significantly down-regulated genes in the WT+miR PEFs compared to the WT PEFs were analyzed by the DAVID program to look for significant biological enrichments (Table C.3; 210). The 6 annotated clusters presented here had significant enrichment scores greater than 1.3 as a score of 1.3 is equivalent to non-log scale 0.05 (210). For each annotated cluster, the top individual term within the cluster with the greatest -log10[p-value(Benjamini)] value (most significant term) is presented. The Benjamini score is one of the multiple testing correction techniques that globally corrects enrichment p-values of individual term members (210). To note, only the top 3 annotated clusters had terms with significant Benjamini scores (p<0.05).  170  Figure C.3 DAVID analysis for up-regulated genes in the WT+miR PEFs versus the WT PEFs. The significantly up-regulated genes in the WT+miR PEFs compared to the WT PEFs were analyzed by the DAVID program to look for significant biological enrichments (Table C.4; 210). The 5 annotated clusters presented here had significant enrichment scores greater than 1.3 as a score of 1.3 is equivalent to non-log scale 0.05 (210). For each annotated cluster, the top individual term within the cluster with the greatest -log10[p-value(Benjamini)] value (most significant term) is presented. The Benjamini score is one of the multiple testing correction techniques that globally corrects enrichment p-values of individual term members (210). To note, only the top annotated cluster had terms with significant Benjamini scores and one term in cluster 4 (p<0.05).  171  Figure C.4 DAVID analysis for significantly down-regulated genes in the Mest+/KO PEFs versus the WT PEFs. The significantly down-regulated genes in the Mest+/KO PEFs compared to the WT PEFs were analyzed by the DAVID program to look for significant biological enrichments (Table C.1; 210). The fifteen annotated clusters presented here had significant enrichment scores greater than 1.3 as a score of 1.3 is equivalent to non-log scale 0.05 (210). For each annotated cluster, the top individual term within the cluster with the greatest -log10[p-value(Benjamini)] value (most significant term) is presented. The Benjamini score is one of the multiple testing correction techniques that globally corrects enrichment p-values of individual term members (210). To note, only the top 7 annotated clusters had terms with significant Benjamini scores (p<0.05).  172  

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.24.1-0073875/manifest

Comment

Related Items