UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of imprinted genes in mouse models of IUGR Lee, Kang Yun Connie 2011

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

Item Metadata


24-ubc_2011_spring_lee_kangyun.pdf [ 1.81MB ]
JSON: 24-1.0071656.json
JSON-LD: 24-1.0071656-ld.json
RDF/XML (Pretty): 24-1.0071656-rdf.xml
RDF/JSON: 24-1.0071656-rdf.json
Turtle: 24-1.0071656-turtle.txt
N-Triples: 24-1.0071656-rdf-ntriples.txt
Original Record: 24-1.0071656-source.json
Full Text

Full Text

THE ROLE OF IMPRINTED GENES IN MOUSE MODELS OF IUGR  by  Kang Yun Connie Lee  B.Sc., University of British Columbia, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE STUDIES  (Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  May 2011  © Kang Yun Connie Lee, 2011  Abstract Each year in Canada, 5% of ongoing pregnancies are affected by intrauterine growth restriction (IUGR), a condition diagnosed when a baby's birth weight is less than the bottom 5th percentile. Placental dysfunction is thought to be the main contributor to IUGR and many genetic aberrations can lead to problems in the placenta. The epigenetic phenomenon, genomic imprinting, has evolved with placentation and gene knockout studies of several imprinted genes in the mouse result in IUGR. The main goal of this thesis is to examine how gene expression of all imprinted genes is affected in mouse models of IUGR (Mmp2-/-, Mest+/- and Surgical). The first step is to find suitable IUGR mouse models by comparing the embryonic weights of potential models to normal mouse embryos. Next, I assessed gene expression using genomewide assays and looked at how expression of imprinted genes is altered in the IUGR mouse model. Amongst the three models, only the surgical model was identified as having IUGR and RNA samples from this model were used in genome-wide expression assays. We found 68 candidate IUGR genes, 42 genes had a 2-fold difference in IUGR embryos or placentae, with 26 genes altered in both tissue. Genes that function in the transport of substances are the most altered in both tissue. The genes that are involved in the development of anatomical structures were affected more in the IUGR embryos whilst stress response genes were more affected in the IUGR placentae. For imprinted genes, only 4 genes in the embryo (H19, Igf2, Slc38a4, and Dlk1) and 6 genes in the placenta (Slc38a4, Sfmbt2, Slc22a3, Phlda2, Cdkn1c, and Dlk1) exhibited significant difference in gene expression between wild-type and IUGR. The majority of these imprinted candidates have been linked to IUGR in either mouse and/or human studies. Overall, imprinted genes as a whole are not more affected in IUGR samples than would be expected by chance based on the chi-square test. These results illustrate that though individual ii  imprinted genes may be important regulator of IUGR, genes regulated by genomic imprinting as a whole are not more affected in this pregnancy complication.  iii  Preface All animal experiments were performed under certificate A08-0454 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.  iv  Table of Contents  Abstract ......................................................................................................................................... ii Preface ......................................................................................................................................... iv Table of Contents ..........................................................................................................................v List of Tables ............................................................................................................................. viii List of Figures ............................................................................................................................. ix List of Abbreviations ....................................................................................................................x Acknowledgements ..................................................................................................................... xi  Chapter 1: Introduction ...............................................................................................................1 1.1 Overview ...................................................................................................................................1 1.2 Intrauterine growth restriction ..................................................................................................2 1.3 Placental development ..............................................................................................................3 1.4 Oxygen and nutrient exchange in IUGR ..................................................................................5 1.5 Genomic imprinting ..................................................................................................................7 1.6 Genomic imprinting and the placenta .......................................................................................9 1.7 Imprinted genes and fetal growth ...........................................................................................10 1.8 Imprinted genes and IUGR .....................................................................................................13 1.9 Imprinted gene network ..........................................................................................................14 1.10 Rationale and experimental approach ...................................................................................15 1.10.1 Research hypothesis ....................................................................................................16 1.10.2 Research objectives ....................................................................................................16  Chapter 2: Materials and methods ............................................................................................18  2.1 Animal work ...........................................................................................................................18 2.2 Genotyping .............................................................................................................................19 2.3 RNA extraction .......................................................................................................................21 2.4 Hematoxylin and Eosin staining .............................................................................................21 2.5 In situ Hybridization ...............................................................................................................22 2.5.1 Probe preparation .........................................................................................................22 2.5.2 Tissue preparation and crysectioning ...........................................................................23 2.5.3 Prehybridization and hybridization ..............................................................................24 2.5.4 Antibody reaction and mounting ..................................................................................24 2.6 Quantitative real-time PCR .....................................................................................................26 2.7 Unilateral ovariectomy ...........................................................................................................27 2.8 Illumina microarray ................................................................................................................27 2.9 Statistical analysis ...................................................................................................................28  Chapter 3: Exploration and characterizations of IUGR mouse models ................................29 3.1 Introduction .............................................................................................................................29 3.2 Results .....................................................................................................................................31 3.2.1 Mmp2 expression in the developing placenta ..............................................................31 3.2.2 Mmp2-deficient placentae exhibit normal morphology ...............................................33 3.2.3 Mmp2-/- conceptuses do not exhibit IUGR ..................................................................33 3.2.4 Imprinted gene expression in Mest mutants .................................................................36 3.2.5 Variability in gene expression by qRT-PCR ................................................................38 3.2.6 Crowded uterine horn ...................................................................................................40 3.3 Discussion ...............................................................................................................................43 vi  3.3.1 First IUGR model: Mmp2 -/- conceptuses do not exhibit IUGR .................................43 3.3.2 Second IUGR model: Mest/Peg1 model has high variation in gene expression ..........44 3.3.3 Placenta has greater variation in gene expression than the embryo .............................46 3.3.4 Third IUGR model: crowded uterine horn can lead to development of IUGR ............47  Chapter 4: Effects of surgically-induced IUGR on whole-genome expression .....................50 4.1 Introduction .............................................................................................................................50 4.2 Results .....................................................................................................................................51 4.2.1 Gene expression in wild type C57BL/6J embryos and placentae ................................51 4.2.2 Variability in gene expression in the placenta vs. in the embryo .................................53 4.2.3 Most variable genes in the placenta and embryo .........................................................55 4.2.4 Imprinted gene expression and variability in C57BL/6J embryos and placentae ........58 4.2.5 Differential expression between wild-type and IUGR .................................................60 4.2.6 Imprinted gene expression in IUGR samples ...............................................................63 4.3 Discussion ...............................................................................................................................65 4.3.1 Genome-wide expression analysis using Illumina MouseRef8.0 Beadchip ................65 4.3.2 Variability of gene expression is more significant in the placenta than in the embryo 66 4.3.3 Comparison of whole-genome expression between normal C57BL/6J and IUGR .....69 4.3.4 Imprinted candidates of IUGR .....................................................................................74  Chapter 5: General discussion ...................................................................................................81 5.1 Summary of results .................................................................................................................83 5.2 Future directions .....................................................................................................................86 References ....................................................................................................................................88 Appendix A: Supplementary information ..............................................................................103 vii  List of Tables Table 2.1 Primers for genotyping, qRT-PCR, and Mmp2 ISH probe construction .....................20 Table 2.2 ISH Solutions ...............................................................................................................25 Table 4.1 Genes with the most variability in expression in E15.5 Bl6 embryo and placenta .....56 Table 4.2 Differentially expressed imprinted genes in IUGR samples ........................................64 Table 4.3 Chi-square test of known imprinted genes and imprinting candidates ........................64 Suppl Table 1. Genes differentially expressed in embryo versus placenta at E15.5 ..................103 Suppl Table 2. Differentially expressed genes in E15.5 IUGR samples ....................................109  viii  List of Figures Fig. 1.1. Comparison of mouse and human placenta .....................................................................5 Fig. 3.1 Basis of the "Crowded Uterine Horn" mouse model of IUGR .......................................31 Fig. 3.2 Expression of Mmp2 mRNA in E14.5 Mmp2+/+ placentae by ISH ..............................32 Fig. 3.3 Placental morphology of E14.5 Mmp2 placentae by H&E .............................................33 Fig. 3.4 Weight comparisons between Mmp2+/+, Mmp2+/-, and Mmp2-/- ................................35 Fig. 3.5 Comparison of candidate imprinted gene expression between E14.5 WT (Mest+/+) and IUGR (Mest+/-) placentae by qRT-PCR ......................................................................................37 Fig. 3.6 Variability in gene expression .........................................................................................39 Fig. 3.7 Growth phenotype of surgical model ..............................................................................42 Fig. 4.1 Gene expression in E15.5 WT C57BL/6J embryo and placenta by Illumina expression profiling system ............................................................................................................................52 Fig. 4.2. Comparison of gene expression variation between E15.5 WT embryos and placentae .54 Fig. 4.3 Whole-genome comparison of variance in gene expression between E15.5 C57BL/6J embryo and placenta .....................................................................................................................57 Fig. 4.4 Imprinted genes' expression and variability in E15.5 WT embryo and placenta ............59 Fig. 4.5 Clustering of WT and IUGR samples .............................................................................61 Fig. 4.6 Differentially expressed genes with > 2-fold different in IUGR ....................................62 Suppl Fig. 1 Weight and size measurements of Surgical IUGR model ......................................103 Suppl Fig. 2 Grouping of WT and IUGR samples .....................................................................104 Suppl Fig. 3 Plasmid map of pGEM-T vector with Mmp2 ISH probe as insert .........................105  ix  List of Abbreviations AS  Angelman Syndrome  BWS  Beckwith-Wiedemann Syndrome  EB  Empirical Bayes  ER  endoplasmic reticulum  EtOH  ethanol  DMR  differentially-methylated region  DNA  deoxyribonucleic acid  ICR  imprinting control region  IGN  imprinted gene network  ISH  in situ hybridization  IUGR  intrauterine growth restriction  KO  knockout  LOI  loss of imprinting  MeAIB  methyl-alpha aminoisobutyric acid  MeV  MultiExperiment Viewer  PCA  Principal Component Analysis  PCR  polymerase chain reaction  PWS  Prader-Willi Syndrome  RNA  ribonucleic acid  SAM  Significant Analysis of Microarray  SD  standard deviation  SGA  small for gestational age  SRS  Silver-Russell Syndrome  UPD  uniparental disomy  VST  Variance-stabilized Transformation  WT  wild-type  x  Acknowledgements First I would like to thank my supervisor Dr. Louis Lefebvre for giving me the opportunity to pursue a M.Sc. with him in the field of epigenetics. Without his ingenuity and guidance, I would not be able to accomplish all the goals we have set forth in this project. I would also like to acknowledge Dr. Wendy Robinson for the funding as well as support for this project. To my other committee member, Dr. Catherine Van Raamsdonk, thank you for your invaluable input to the direction of my project, and for letting me borrow equipment from your lab. To all my lab mates in the Lefebvre lab, you have helped me immensely on almost every aspect of my project I could not imagine where I would be without you. I will miss the lab parties and the trips that we have taken together. To my friends from the wing, thank you for your company during the daily coffee runs. I will miss them too. To my parents, thank you for the financial support and for moving out to Vancouver so I could live closer to campus while I pursue my degree. Last but not least, I would like to thank my boyfriend Andrew Haack for everything, especially for helping me to edit my thesis.  xi  Chapter 1: Introduction 1.1 Overview Pregnancy complications occur in approximately 20% of pregnancies every year in Canada (BC Vital Statistics, 2004). These can lead to spontaneous abortions as well as other harmful effects to the mother. Preeclampsia, a maternal hypertensive disorder, and intrauterine growth restriction (IUGR) account for most adverse perinatal outcomes. IUGR is defined as having fetal weight of <5 percentile for the gestational age, though having weight of <10 percentile has been found to be clinically relevant. IUGR babies have varying degrees of perinatal pathology such as decreased organ size, brain hypoxia, and hypoglycemia (Cox and Marton, 2009). Long-term effects including diabetes and abnormal psychological profile have also been associated with IUGR (Nicholls et al., 1998; Sebire and Fisher, 2005; van der Smagt et al., 2006). The placenta is the tissue that lies at the fetal-maternal interface and is the site of gas and nutrient exchange between the mother and the fetus during development. Therefore, problems in the placenta, such as abnormal vasculature, can directly impact the growth of the developing embryo(s) and possibly lead to IUGR. In fact, it is suggested that the majority of IUGR-only pregnancies have underlying problems in the placenta (Cox and Marton, 2009). A major epigenetic contributor to IUGR may be genomic imprinting. Many mouse knockout studies of imprinted genes have documented IUGR in null mutant mice, indicating that proper expression of imprinted genes is necessary for normal development of the embryo. Several of these genes have high expression in the placenta. Interestingly, there are more genes that are only imprinted in the placenta than in the embryo. The link between imprinting in the placenta and how 1  imprinted genes contribute to IUGR is the main premise of this thesis. We wish to assess if imprinted genes as a group are affected more than other groups of genes in IUGR.  1.2 Intrauterine growth restriction There are many different causes for IUGR. Chromosome and genetic abnormalities often lead to syndromes in which IUGR is one of the phenotypes observed. For instance, SilverRussell Syndrome (SRS) is characterized by IUGR, postnatal growth restriction, and craniofacial features. The known causes of the syndrome are maternal uniparental disomy (UPD) of chromosome 7 as well as aberrations to the epigenetic modification of H19 and IGF2 on chromosome 11p15, though these causes only account for <50% of SRS cases (Bliek et al., 2006; Eggerding et al., 1994; Gabory et al., 2009; Kozak et al., 1997). Fetuses with trisomies also exhibit IUGR as a pathology (Cox and Marton, 2009). Development of IUGR has been linked to fetal infections by cytomegalovirus or rubella viruses (Choong et al., 2000; Schuster et al., 1993). However, the majority of causes of IUGR are due to infarctions in the placenta, the tissue that lies at the fetal-maternal interface (Cox and Marton, 2009). Though the clinical diagnosis for IUGR is based on the birth weight or fetal weight, assessment using biometric parameters of the developing fetus are also utilized for research purposes. Ultrasound typically is used to estimate fetal weight parameters such as abdominal circumference, femur length, and head size (Hadlock et al., 1985). Gestational age of the pregnancy is also estimated by ultrasound during the first trimester. The two estimates together are compared to the normal fetal growth percentiles to determine if growth restriction has occurred (Neilson, 1984; Platz and Newman, 2008). Another method to diagnose fetal growth restriction is to use Doppler ultrasound to measure blood flow of major vessels in and out of the 2  placenta (Rigano et al., 2001). This diagnostic is suggested to differentiate IUGR due to placental insufficiency from other potential causes in the fetus; this delineation is important since placental-based IUGR is more manageable (Miller et al., 2008).  1.3 Placental development In order to understand how placental insufficiency can lead to IUGR, we must first understand the connection between the uterus, the placenta, and the fetus. Insights into human placental development are largely based on mouse mutants. Comprehensive comparison between the human and mouse placenta has been reviewed (Georgiades et al., 2002). After fertilization and the first several rounds of cleavage, the zygote develops into a blastocyst where the trophectoderm cells overlie the inner cell mass, which will become the embryo proper. This occurs at around three to five days into mammalian gestation. Trophoblast cells develop from trophectoderm (TE) to mediate the invasion of the uterus and initiate the necessary uterine responses for implantation (Cross et al., 1994). The trophectoderm also gives rise to structures that will fuse with the embryonic mesoderm-derived allantois. This area of fusion eventually becomes the site where fetal and maternal vasculature interact to mediate exchange of gases, nutrients and waste (Rossant and Cross, 2001). There are differences between the structure of mouse and human placenta, but key developmental events are similar. Initial invasion of the uterus by trophoblast cells must occur, resulting in implantation. Secondary invasion of maternal and fetal blood vessels into the placenta results in the formation of the fetal-maternal interface, where the exchanges of gases, nutrients and waste between the mother and the embryo occur in the chorioallantoic placentae (Rossant and Cross, 2001). 3  The layer where fetal-maternal exchange occurs is known as the labyrinth in the mouse. Fetal as well as maternal circulation can be found in this layer. The fetal capillaries in the placenta are lined by trophoblast cells and are connected to the embryo by the umbilical vessels. The exact organization of the fetal vessels differs between human and the mouse. Fetal capillaries in the mouse placenta have a maze-like pattern whereas in the human the fetal capillaries form villous trees. There are also slight differences in the arrangement of trophoblast cells but the result is the same; the maternal and fetal circulation come into close contact with only intervening trophoblast cells for the exchange of gases, nutrients, and waste (Georgiades et al., 2002). Effective circulation is not established in the human placenta until approximately 12 weeks gestation, perhaps to prevent the fetus from being exposed to high levels oxygen and blood pressure from maternal blood flow before the formation of a functional placenta (Jauniaux et al., 1995). Similarly, maternal blood is not observed in the labyrinth until 10.5 days (E10.5) into mouse gestation (Georgiades et al., 2002; Muntener and Hsu, 1977). In humans, the villous trees of the monochorial placenta are covered by a single multinucleated trophoblast layer known as the syncytiotrophoblast (Gaunt and Ockleford, 1986; Huppertz et al., 1998). The mouse has three trophoblast layers (trichorial placenta) in between the maternal blood sinus and the fetal endothelial cells that encompass fetal circulation (Fig. 1.1). Though the anatomy of the mouse “syncytium” is different, these layers behave like the syncytiotrophoblast in the human placenta. These cells have structures to increase the surface area to allow for more absorption to occur (Enders, 1965).  4  Fig. 1.1 Comparison of mouse and human placenta. Three-layer mouse placenta with depiction of fetal-maternal interaction. Villous tree-shaped human placenta with cross section of the chorionic villi.  1.4 Oxygen and nutrient exchange in IUGR The availability of oxygen governs the proliferation or differentiation of trophoblast cells. Prior to 10 weeks in human gestation, there is little secondary invasion, resulting in minimal amounts of maternal blood flow reaching the placenta (Genbacev et al., 1997). Cytotrophoblasts initially invade and plug the uterine spiral arterioles at 8 weeks gestation, resulting in low oxygen tension in the interstitial space (2.5% oxygen) (Jauniaux et al., 2000; Rodesch et al., 1992). Therefore, the placenta is placed in a relatively hypoxic environment at this point (Jauniaux et al., 2003). In mice no uterine arteriole plug is observed, but the murine placenta during early embryogenesis is also left in a hypoxic environment (Pringle et al., 2007). This hypoxia promotes the proliferation of trophoblast cells and consequently, the placenta increases rapidly in size in comparison to the embryo (Genbacev et al., 1997). It also promotes 5  angiogenesis by increasing the expression of vascular-endothelial growth factor (VEGF) , a wellstudied molecule that promotes the proliferation of endothelial cells (Phillips et al., 1995). The regulation of angiogenic factors depend on hypoxia-inducible factors (HIF). Mouse knockouts (KO) of Hif1a die around midgestation (normal gestation in the mouse is approximately 20 days) due to a vascularisation defect in the labyrinth and spongiotrophoblast or failure in chorioallantoic fusion (Kozak et al., 1997). With the onset of secondary invasion and subsequent angiogenesis, the supply of oxygen also increases after 10 weeks of gestation, resulting in trophoblast cells that differentiates into a more invasive form (Genbacev et al., 1997). Low oxygen culture condition (1% oxygen) has been found to reduce the extent of outgrowths of ectoplacental cultures and decrease expression of Hif1a (Pringle et al., 2007). Hif1a is regulated by pro-inflammatory cofactors in addition to hypoxia. Cytokines such as TNF-α and IL-1 were discovered to be elevated in placental villous explants exposed to low oxygen environment (Benyo et al., 1997). Restricting blood flow to IUGR placentae results in a larger increase in TNF-α as compared to controls (Holcberg et al., 2001). This reduction of TNF-α in IUGR may be placental-specific since it is not observed in fetal lymphocytes derived from IUGR pregnancies (Iruloh et al., 2009). Moreover, it has been demonstrated that trophoblast cultures exposed to hypoxic condition demonstrate a marked decrease in system A amino acid transport (Nelson et al., 2003). System A amino acid transport is amongst one of the most important nutrient exchange system in the developing placenta. These experiments looking at hypoxia in the placenta have outlined how this condition can lead to pregnancy complications such as preeclampsia and/or IUGR (Genbacev et al., 1996; Genbacev et al., 1997; Gerretsen et al., 1981). Nutrient transfer in the placenta involves the transfer of glucose, amino acids and fatty acids from maternal circulation to fetal circulation. Different transport proteins and receptors are 6  present on the maternal-facing plasma membrane versus those on the fetal-facing basal membrane (Jones et al., 2007). Amino acid transport appear to be the most affected in IUGR pregnancies as concentrations of many amino acids are decreased in fetal circulation with corresponding elevation of maternal concentrations of these amino acids (Cetin et al., 1996). The sharpest decrease in the fetomaternal ratio was observed for leucine, taurine, cationic amino acids, and system A amino acids (serine and glycine). Factors that regulate amino acid transport are also affected in IUGR. Additionally, there is a decrease in the levels of fetal insulin and lower expression of placental insulin receptor (Chellakooty et al., 2004; Economides et al., 1989; Potau et al., 1981). Maternal serum levels of insulin-growth factors (IGF1 and IGF2) are also lower in IUGR pregnancies (Holmes et al., 1997). Both insulin and IGF1 have been demonstrated to affect system A transport (Fang et al., 2006; Karl, 1995; Karl et al., 1992; Masuyama et al., 1996; Sferruzzi-Perri et al., 2006; Sferruzzi-Perri et al., 2007). System A transport is the major sodium-dependent amino acid transport system. In the human placenta it regulates transport of alanine, serine, and glycine on the microvillous membrane and directs the uptake of methyl-alpha aminoisobutyric acid (MeAIB) on the basal membrane (Hoeltzli and Smith, 1989; Johnson and Smith, 1988). The mTOR signalling system, which is made up of serine/threonine kinases important for sensing placental oxygenation and directing transport of leucine, has also been found to be down-regulated in IUGR placentae (Roos et al., 2007). This is indicative that proper nutrient transfer is important in determining fetal growth.  1.5 Genomic imprinting The first indication of the importance of equal maternal and paternal contribution for the proper development of the embryo and extraembryonic tissue comes from experiments in the 7  1980’s that succeeded in generating bi-maternal (gynogenetic) and bi-paternal (androgenetic) mouse embryos. Both types of uniparental embryos exhibit gross abnormalities and die around mid-gestation. Parthenogenetic (and gynogenetic) conceptuses exhibit growth restriction with very limited and abnormal development of the extraembryonic tissue whereas the androgenetic conceptuses have overgrown extraembryonic tissues with poor development of the embryo proper (McGrath and Solter, 1984; Surani et al., 1984). It has also been observed in human diseases that deletion of the same chromosomal region can result in different syndromes depending on the parental inheritance of the deletion (Kagami et al., 2008; Ledbetter and Engel, 1995). Based on these results, it was postulated that the mammalian genome might contain developmentally important genes expressed only from one of the two parental alleles. To explain such a mode of expression, these genes were postulated to be differentially epigeneticallymarked in the two parental germlines, a model which led to the hypothesis of genomic imprinting. Genomic imprinting is a type of gene regulation that results in monoallelic gene expression. During gametogenesis, all epigenetic modifications are erased and re-set according to the sex of the parent. Consequently, differential DNA methylated regions (DMR) and/or histone modifications exist between oocytes and sperm. These epigenetic marks, or imprints, are maintained after fertilization and govern the monoallelic expression of these genes. Imprinted genes are often found in clusters in large chromosomal domains and are regulated by noncoding RNAs, histone modifications, or both (Koerner et al., 2009; Nagano et al., 2008; Wagschal et al., 2008). Imprinted genes have been associated with embryonic development ever since their identification in the early 1980s. They are especially important to the proper growth of the 8  embryo as will be outlined in the following paragraphs. Igf2 is the earliest gene discovered to be imprinted and several other factors affected by IGF-II are also imprinted, including Slc38a4, which codes for one of the System A transporters, SNAT4 (Mackenzie and Erickson, 2004).  1.6 Genomic imprinting and the placenta The presence of a placenta distinguishes mammals from the rest of the animal kingdom with exceptions only in some reptilian species and in egg-laying mammals (monotremes). It has been implied that the emergence of genomic imprinting is closely associated with the evolutionary development of the placenta. Indeed, genomic imprinting has been suggested to have evolved as a result of a competition for resources between the mother (maternal genome only) and the fetus (both maternal and paternal genomes) (Mochizuki et al., 1996; Moore and Haig, 1991). As an extreme example, the complete paternal human conceptus manifests as a large mass of placental tissue with overgrowth of the trophoblast known as the hydatidiform mole. Since the placenta is the site of nutrient extraction for the embryo, this suggests that the paternal genome has evolved to maximize the chance of it being passed onto future generations by maximizing the chance of survival of the embryo. Conversely, bi-maternal conceptuses in mouse have a very small placenta, indicative of the role the maternal genome plays in restricting nutrient extraction of the embryo to ensure the mother's survival (McGrath and Solter, 1984; Surani et al., 1984). Several mouse knockouts have demonstrated the necessity of some imprinted genes in the placenta. The genes Peg10, Rtl1, Igf2, Phlda2, and Ascl2 all directly affect the structure of the placenta. Peg10 and Ascl2 are essential for the development of the spongiotrophoblast of the mouse placenta whereas Rtl1 is needed for development of the fetal capillaries in the labyrinth; 9  absence of these factors leads to failure to thrive by midgestation (Gabory et al., 2009; Guillemot et al., 1995; Guillemot et al., 1994; Ono et al., 2006; Sekita et al., 2008). Phlda2 is suggested to regulate glycogen storage of glycogen-containing cells of the spongiotrophoblast, which is necessary for the continual invasion of the maternal tissue during mouse gestation (Tunster et al., 2010). Igf2 P0 KO (Igf2 P0+/-) is a placental-specific knockout of Igf2 and Igf2 P0+/- conceptuses have growth-restricted placenta beginning E14, accompanied by later reduction in fetal growth (Constancia et al., 2002). Recent studies have demonstrated the importance of Igf2 in regulating nutrient exchange in the placenta (Coan et al., 2010; Constancia et al., 2005; Constancia et al., 2002). It was shown that nutrient diffusion capacity is reduced in the Igf2 P0+/- placenta, which eventually affects nutrient uptake by the embryo despite an initial compensatory effect (Constancia et al., 2002; Sibley et al., 2004). Constancia and colleagues (2005) have conducted physiological assays looking at nutrient transfer in the Igf2 P0+/- mutants. They injected the mother with radioisotopelabelled glucose and MeAIB, an amino acid analogue of the System A transporters. These injections allowed them to deduce the amount and direction of nutrient transfer into the fetus. By comparing the nutrient transfer levels between the mother and Igf2 P0+/- embryos, they have concluded that there is a transient increase in nutrient exchange around E16, which is partially caused by an increase in expression of glucose transporter GLUT2 genes (Slc38a4 and Slc3a2).  1.7 Imprinted genes and fetal growth Mice with uniparental disomy (UPD) or knockouts of imprinted genes tend to exhibit overgrowth, undergrowth, and/or behavioural abnormalities. The null mice of paternallyexpressed Mest and Peg3 exhibit embryonic growth restriction and stunting of the placenta 10  (Lefebvre et al., 1998; Li et al., 1999). Interestingly, null females display abnormal maternal behaviour (Lefebvre et al., 1998; Li et al., 1999). One of the most important growth factors for embryonic development is located directly downstream to H19. Insulin-like growth factor 2 (Igf2) is amongst one of the first genes identified to be imprinted (DeChiara et al., 1991). This gene is expressed from the paternal allele and deletion of Igf2 leads to fetal growth restriction as well as stunting of the placenta. Conversely, the over-expression of Igf2 results in fetal overgrowth and placentomegaly (DeChiara et al., 1991; Ferguson-Smith et al., 1991). The maternally expressed imprinted gene Igf2r codes for a receptor that sequesters IGF2. Mouse knockout of Igf2r shows overgrowth (Lau et al., 1994). Grb10 codes for an adaptor protein that binds to insulin receptor, whilst Dlk1 codes for a ligand that functions in the notch signalling pathway (Baladron et al., 2005; Giovannone et al., 2003; Moore and Haig, 1991). Grb10 and Dlk1 have reciprocal expression, with the former predominantly expressed from the maternal allele and the latter from the paternal allele. These two genes also have intriguing reciprocal phenotypes when knocked out in mice. Newborn Grb10-null mice are 30% heavier than their wild-type littermates but the weight differences between null and wild type diminish postnatally (Smith et al., 2007). On the other hand, mice inheriting the null paternal allele of Dlk1 are dwarf at birth and exhibit postnatal catch-up growth (Moon et al., 2002). The embryonic and placental phenotypes of these mouse knockout studies suggest that genomic imprinting plays a key role in the proper development of the mouse embryo. A few human disorders have also been associated to abnormalities of specific imprinting regions. For example, patients with Prader-Willi (PWS) and Angelman Syndrome (AS) have maternal or paternal microdeletions of 15q11-q13, respectively (OMIM 17620, OMIM 105830). Though the patients of these two syndromes are missing the same chromosomal region, they exhibit different abnormalities due to the effect of parent-of-origin specific imprinting. Patients 11  of these two syndromes suffer from mental retardation as well as other congenital physical and behavioural abnormalities. Moreover, changes in both DNA methylation and histone modifications have been observed in UPD(14) syndromes, Beckwith-Wiedermann syndrome (BWS), PWS and AS. Hypomethylation of the paternal imprints at the DLK-GTL2 DMR leads to the development of UPD(14)mat-like phenotypes since this region is usually methylated on the paternal chromosome. Loss of methylation at this DMR leads to an epigenetic switch and this region of paternal chromosome 14 now behaves more like the maternal chromosome. Conversely, hypermethylation of maternal chromosomes will result in UPD(14)pat-like syndrome. The most relevant syndromes linking genomic imprinting to growth are BWS and Silver Russell Syndrome (SRS). Patients with BWS exhibit pre- and postnatal overgrowth whereas SRS patients exhibit IUGR and postnatal growth restriction (Abu-Amero et al., 2008; Temple, 2007). BWS is caused by genetic and epigenetic aberrations at H19 DMR or KvDMR1, the two separate imprinting control regions on human chromosome 11, and maternal loss of CDKN1C (Lim and Ferguson-Smith, 2010). Some BWS patients exhibit loss of imprinting at the IGF2 locus due to improper methylation at the H19 DMR (Bliek et al., 2001). Half of BWS patients exhibit aberrant methylation at KvDMR1, the promoter of the non-coding RNA KCNQ1OT1 (Bliek et al., 2001). Imprinting of CDKN1C is maintained through the KvDMR1, and hypomethylation at this site causes biallelic KCNQ1OT1 expression and CDKN1C silencing. Maternally inherited mutations in the coding sequence of CDKN1C can also contribute to the development of BWS phenotypes in a subset of patients (Hatada et al., 1996; Romanelli et al., 2009). Additional methylation differences have been discovered in BWS patients at imprinting control regions of PLAGL1, IGF2R, MEST, and GNAS (Bliek et al., 2009). SRS is also associated with mutations as well as hypomethylation at H19 (loss of IGF2) and 12  hypomethylation at KvDMR1 (biallelic CDKN1C) on chromosome 11 (Bliek et al., 2006; Eggermann et al., 2006; Gicquel et al., 2005; Guo et al., 2008; Penaherrera et al., 2010). These human disorders combined with mouse knockout experiments affirm the involvement of genomic imprinting in embryonic growth and development. It has also been demonstrated in mice and humans that some imprinted genes are only imprinted in the placenta, and many of these genes are highly expressed in the placenta during embryogenesis (Coan et al., 2005). Therefore, imprinted genes found in the placenta have been designated as the top candidates in many IUGR studies (Bourque et al., 2010; Diplas et al., 2009; Guo et al., 2008; Jager et al., 2009; McMinn et al., 2006).  1.8 Imprinted genes and IUGR Though imprinted genes have seemed the most likely candidates for the regulation of fetal growth, the associations between them and IUGR are still under debate. One study has specifically looked at the differential expression of imprinted genes between normal and IUGR placentae in humans by microarray (McMinn et al., 2006). The authors discovered that 7% of all differentially expressed genes were imprinted genes. PHLDA2 and CDKN1C were found to be upregulated, whilst MEST, MEG3, GATM, GNAS, PLAGL1, and IGF2 were found to be downregulated. PHLDA2 and PLAGL1, were also found to be affected in a different IUGR study using quantitative Real-Time PCR (qRT-PCR) to look at imprinted genes' expression (Diplas et al., 2009). In contrast, IGF2 was found to be downregulated in studies looking at methylation and expression differences in small-for-gestational placentae (SGA) (Guo et al., 2008; Lo et al., 2002). Moreover, Bourque et al. (2010) assessed 44 imprinted genes' expression using the expression microarray and has found IGF2 to be differentially expressed in IUGR placental 13  villous samples. In their study, they mentioned that since they obtained their samples several hours after birth, expression analyses of placental tissues may be affected by rapid RNA degradation. In addition to tissue handling differences, varying methods of assaying gene expression, tissue sampling, as well as determination of statistical significance may be the cause of these conflicting results. In addition to looking at expression differences, some groups have also looked at DNA methylation differences between normal and growth restricted placental samples (Bourque et al., 2010; Guo et al., 2008). Guo et al. (2008) only found one out of 24 SGA placentae to have lossof-imprinting at the imprinting control region of H19/IGF2 (ICR1). In the study conducted by Bourque and colleagues (2010) using genome-wide methylation array technology, 7 of the 13 IUGR placentae have significant hypomethylation at ICR1 when compared to normal controls. These results indicate that though imprinted genes appear to be excellent candidates for IUGR, only the well-characterized IGF2 is associated with the complication. Nevertheless, several imprinted genes have been reported to regulate IGF2 expression (Abu-Amero et al., 1998; Cattanach et al., 2004; Gabory et al., 2009; Varrault et al., 2006). Slight alteration in their expression may have an impact on IGF2, leading to impaired fetal growth. It is even suggested that a group of imprinted genes may function in a network to facilitate embryonic growth (Arima et al., 2005; Varrault et al., 2006).  1.9 Imprinted gene network This idea of an imprinted gene network (IGN) was explored by Varrault and colleagues (2006) by looking for coexpressed genes with Plagl1/Zac1 in multiple mouse array datasets. They have observed a significant over-representation of imprinted genes coexpressed with 14  Plagl1, which lead them to further promote the theory of an IGN that is first suggested in Arima et al. (2005). Varrault et al. (2006) have confirmed a change in expression levels of a few of the imprinted genes (Igf2, H19, Cdkn1c, Dlk1) from the IGN in Plagl1-transfected cell line as well as the livers of the Plagl1-null mice. Furthermore, they demonstrate that Plagl1 is a regulator of the H19/Igf2 locus due to its ability to bind to a shared enhancer. This observation is not the first demonstration of imprinted genes working together, nor is it likely to be the last. The imprinted gene Igf2r codes for a receptor that binds and sequester insulin growth factor II (Igf2), which is produced from the imprinted Igf2 (Czech et al., 1989; Filson et al., 1993). The decline of gene expression of the genes in IGN corresponds to the growth deceleration of multiple organs in postnatal somatic tissues (Lui et al., 2008). Gabory et al. (2009) have shown that the H19 RNA levels affects the expression level of five of the genes in the IGN. Our group has uncovered a curious relationship between Phlda2 and Ascl2 expression in the developing mouse placenta (R.O.-M., A. B. B. and L.L, submitted). Tunster et al. (2010) have also observed this interplay between these two genes.  1.10 Rationale and experimental approach The research thus far on the effect of imprinted genes perturbation suggests that if there is an IGN, it most likely affects embryonic growth. In conjunction with the effects of some genes in the placenta, we propose the idea that genes within this IGN are affected in terms of their expression level in mouse models of IUGR. Our research is part of a larger collaboration that is looking at perturbation of epigenetic modifications in complicated pregnancies in humans. Recent studies looking at imprinted gene perturbations have indicated that in humans, IGF2 is the main gene that shows a difference in expression in SRS as well as idiopathic IUGR (Bourque 15  et al., 2010; Guo et al., 2008). There is more uncertainty as to whether or not epigenetic modifications are affected at that locus. Using mouse models of IUGR in my research, I hope to provide a more stringent look at the effects of IUGR on imprinted gene expression in the placenta. My IUGR models will be in a more controlled environment than studies done with the human population as well as address the contribution of genetic variations to the effects of IUGR. 1.10.1 Research hypothesis Idiopathic IUGR is mainly caused by disruption of proper placental function and the main group of genes that are affected by this complication are the imprinted genes. Expression levels of imprinted genes are different in IUGR when compared to normal mouse placental samples. 1.10.2 Research objectives My research is aimed at assessing whether or not imprinted genes are affected by IUGR and identifying those that are implicated. The goal was to explore multiple models of IUGR to find the common genes that are affected. Three potential mouse models of IUGR were studied: mouse KO of Mmp2 (non-imprinted gene) and Mest (imprinted gene), and surgically-induced IUGR mouse model. The main method of analysis of gene expression is microarray as well as qRT-PCR. We also wish to characterize placental phenotypes that are previously uncharacterized to determine the potential cause of IUGR in some of these mouse models. Through my research I hope to identify candidate genes implicated in the etiology of IUGR, which will provide some insight into human studies in IUGR. Some of these candidates may become important diagnoses of IUGR in complicated pregnancies. I also wish to assess the 16  role of imprinted genes in IUGR and whether or not they are the most affected group of genes in the disease.  17  Chapter 2: Materials and methods 2.1 Animal work The Mmp2 knockout (Mmp2tm1Itoh) was created by Itoh et al. (1997) (MGI:2386252). Two Mmp2+/- males and two Mmp2+/- females were transferred from the colony room of Dr. Chris Overall and the line was expanded in our mouse facility. The Mest knockout line (Mesttm1Lef) was already available in-house (Lefebvre et al., 1998) (MGI:2181803). These two lines of mice were housed in a windowless room in one of the animal facility (D. H. Copp) at University of British Columbia. Approximately four mice occupied each polycarbonate cage (floor surface area 500cm2) with stainless steel cage tops. The C57BL/6J mice used for the surgical model were obtained from The Jackson Laboratory and maintained at the Centre for Disease Modelling at UBC. The mice were housed in ventilated cages (floor surface area 610cm2). The on-site animal technicians provided food and water as needed. The cages were changed once a week. Mice were randomly mated with discovery of vaginal plug noted as E0.5. Mmp2deficient conceptuses were obtained by heterozygous crosses, whereas loss of Mest was obtained in the progeny of wild-type CD-1 outbred females (UBC Animal Care Centre) crossed with Mest KO heterozygous males. Embryonic dissections were made on E14.5 for Mmp2 and Mest lines. The pregnant female was removed to a separate cage that was placed in a polycarbonate chamber that could attach to the CO2 tank and sacrificed as per UBC Animal care SOP 009E4. The animal remained in the chamber for 15 minutes and then removed. The two uterine horns were dissected from the female and conceptuses were removed. Embryos and placentae were weighed using an analytical balance. Newborn pups were weighed using a bench-top balance at postnatal day 1,  18  around 20 days after discovery of vaginal plug. Individual embryo and placenta were preserved differently depending on the uses. 2.2 Genotyping Yolk sac was used for genotyping the mice. A small piece of yolk sac was placed in a 1.5mL microtube.1.25µL of 20mg/mL Proteinase K (Roche) and 50µL ProK solution (50mM KCl, 10mM Tris-HCl pH 8.3, 2mM MgCl 2, 0.1mg/mL gelatin, and 0.45% Tween). 1.5µL of yolk-sac lysate in 15µL of dH2O was incubated at 95°C for 5min, then held at 85°C for loading of master mix. Cycling was programmed for 35 cycles of 95°C 30s, 59°C (Tm) 30s, 72°C 30s. Each reaction contained 2.5µL of 10X PCR buffer, 2.5µL of MgSO4, 2.0µL of 2.5mM dNTPs, 0.1µL of Tsg DNA Polymerase, and 2µL of 10mM genotyping primers in a volume of 25µL. The PCR buffer, magnesium, and Tsg DNA Polymerase were from BioBasics. Primers used for genotyping or sexing the Mmp2, Mest and C57Bl/6J embryos are listed in Table 2.1.  19  Table 2.1 Primers for genotyping, qRT-PCR, and Mmp2 ISH probe construction. All primers are listed in the 5' to 3' orientation and are obtained from Integrated DNA Technologies. Genotyping Mest+/+ Mest-/Mmp2+/+ Mmp2 +/Male Mmp2 ISH probe qRT-PCR Cdkn1c H19 Igf2 Igf2r  0702a: GAA ACC GAG AAA CAG ATT GGA 0702c: TCC CAG TGG ATC ACC TGA GC 0702a and IRES1: AGA CCG CGA AGA GTT TGT CCT C Mmp2 F2: CTG GCG CTT AGG AAA CAC TC Mmp2 R2: AGC TAG GAG TTC CGG CTT CT F2 and PGK4: CC AAA GAA CGG AGC CGG TTG Zfy1a: GAC TAG ACA T GTC TTA ACA TCT GTCC Zfy1b: CCT ATT GCA TGG ACA GCA GCT TAT G Mmp2 F3: ATG GCC CCG ATC TAC ACC TA Mmp2 R3: TTC CAA ACT TCA CGC TCT TGA Cdkn1c F1: CGA ACA GGC AGG CAA GCT Cdkn1c R1: GCT GTT CTG CTG GCT GAT TG H19 F1: CGT ATG AAT GTA TAC AGC GAG TGT G H19 R1: ACA CGG CCA CAC CCA GTT Igf2 F: GCT TGT TGA CAC GCT TCA GTT TG Igf2 R2: CCG GAA GTA CGG CCT GAG AG Igf2r F: GCA CAG AAT CCA GAC TAG CAT TAC A Igf2r R: CCT CCT TAT CAG CCT TAA ATA TGT CTT TCT T  Peg3  Peg3 F: GCC GAG TCA TAC CAG AAT GTT Peg3 R: ACC TCG ATG AGT GGC CTT G  Phlda2  Phlda2 F2: TCA GCG CTC TGA GTC TGA AA Phlda2 R2: CAG CAA GCA CGG GAA TAT CT  Slc38a4  Slc38a4 F1: TTT GAT ACG GCT CTT CTC ATG GT Slc38a4 R1: CAG CAG TGT GAT CAC CGA AGT AC  Gapdh  G3pdh F: ACC ACA GTC GCC ATC AC G3pdh R: TCC ACC ACC CTG TTG CTG  Ppia*  Ppia F1: CGC GTC TCC TTC GAG CTG TTT G Ppia R1: TGT AAA GTC ACC ACC CTG GCA CAT  Gm5155  Gm5155 3' end F2: CCT TTT GTG CGA GTG ACT GAC A Gm5155 3' end R2: TGA AGA GCC AAC GGA TGG A  D93  D93 F1: GAG CAA GTT TCA GGA CTC AAG GA D93 R1: GAG GAC CCA AAA GCC TGT CA  Loc802  LOC802 F1: GTG ACA AAT GGC ACC AAT GC LOC802 R1: GTG CTC AGA AGG CGC AAT T  633Rik  633RIK F1: TAC GAG GGC CTG TTC GAT AAG 633RIK R1: TCA AAC GCC GAC CAG ATT TCC  Vstm2l  Vstm2l F1: TGG GAC AAC CAC GTC TCC G Vstm2l R1: CTG GTT GGA GGC CCA CGT  Prl2c1  Prl2c1 F1: AGA CAA AAG CCC CAC GAG AT Prl2c1 R1: TCC TGA TTT CAG AAG AGC TTC ATA G  * Primer sequences obtained from Mamo and colleagues (Mamo et al., 2007). 20  2.3 RNA extraction For RNA collection, embryos and placentae were preserved on dry ice then transferred to -70°C. Invitrogen’s TRIzol Reagent was used to extract RNA from whole placenta or embryo according to the manufacturer’s protocol. The tissues were homogenized using motorized pestles in 1mL of TRIzol reagent. 500µL of chloroform was added to separate the organic layer from the aqueous layer. The aqueous layer contained the RNA and was removed into a new tube. 100% isopropanol was used to precipitate the RNA, which was subsequently washed with 75% ethanol (EtOH). Pellets were allowed to air-dry and re-suspended in 100µL of DEPC-treated distilled water (DEPC-dH2O). Total RNA samples were stored in -70°C freezer until further uses.  2.4 Hematoxylin and Eosin staining Placentae were fixed in 4% paraformaldehyde (PFA) in scintillation vials and left at 4°C overnight. The tissues were rinsed in 0.85% NaCl solution for 50min, then in 100% Nacl/100% EtOH for 35min. Two 35min of 70% EtOH washes followed. All of the steps were done on ice. The vials were stored in fresh 70% EtOH overnight at 4°C. The tissues were subjected to five 30min EtOH washes (85%, 90%, 95%, 100%, 100%) at room temperature. Then two 30min xylene washes were used to harden the tissues for sectioning. The vials were filled with paraffin using Shandon Histocentre 3 (Thermo Electron Corporation) and were incubated at 65°C for three hours. Fresh paraffin was changed every hour. The tissues were embedded in silicone molds (VWR, 1560-215) according desired orientation and sectioned after the molds had hardened. Paraffin blocks were stored at room temperature until sectioning. Paraffin sections were made using Leica’s RM2255 microtome. The microtome was set for automatic sectioning until the blade was close to the centre of the placenta, then 10µm sections were made by hand.  21  The paraffin sections were placed on Fisherbrand Superfrost slides and dried overnight in the 37°C room. The sections were re-fixed in xylene for 2 x 10min, then re-hydrated going from 100% EtOH (100%, 95%, 90%, 80%, 70%, 50%, 30%) to water at 2min intervals. The slides were stained in hematoxylin (Sigma-Aldrich) for 15min, then they were placed under running water for 15min. The slides were dehydrated at 2min intervals using the same increase in EtOH concentrations, going from 30% to 100%. The slides were stained in eosin (Sigma-Aldrich) for 30sec and then washed twice with 100% EtOH. Lastly, the slides were placed in xylene for 2 x 10min and then coverslips were mounted onto the slide with Entellan (Harleco). Subsequent analyses of the tissues were made using a light microscope (Leica MS5). Pictures were taken at 0.65X magnification using the QImaging colour camera (Micropublisher) and the QCapture software.  2.5 In situ hybridization All solutions used are listed at the end of this section. 2.5.1 Probe preparation Mmp2 probes were RT-PCR amplified from E14.5 placental total RNA using Mmp2 F3 and R3 primers then cloned into the pGEM-T vector (Promega) (Suppl Fig. 3). The method in which cDNA was generated is listed in Section 2.6. The orientation of the insert was determined by digesting with SacI, then the correctly oriented plasmids were sent for sequencing to the Sanger DNA sequencing services (McGill University and Genome Quebec). One of the plasmid preparations had no mismatches and it was linearized with NotI (for sense probe) and NcoI (for antisense probe). 22μg of plasmid were linearized with 50 units of enzymes overnight. 100μL of phenol chloroform (Invitrogen) was added and centrifuged. 9μL of 3M sodium acetate (NaAc) 22  and 270μLof EtOH were added to the top 90μL of digestion reaction, washed with 70% EtOH, and resuspended in 20μL DEPC water, making the concentration ~1μg/μL. 1μL of the linearized plasmid and 40 units of RNA Polymerase (Roche) were added to 17μL of the labelling master mix and incubated at 37°C for two hours. T7 and SP6 RNA polymerases were used for creation of sense and anti-sense probes, respectively. A 15 minute incubation at 37°C followed the addition of 1μL of RNase-free DNase (Promega). 100μL of 4M LiCl mix and 40μL of DEPC water was added to the labelling reaction and incubated in -20°C overnight. The RNA was pelleted, washed, and resuspended in 100μL of DEPC water. 1μL was run on a RNase-free agarose gel and the plasmid concentration was determined to be approximately 1μg/μL. The probe was stored at -80°C until use. 2.5.2 Tissue preparation and cryosectioning The placentae were fixed with 4% PFA overnight then transferred to a 30% sucrose solution for another overnight incubation at 4°C. The placentae were embedded in silicone molds using O.C.T. (Tissue-Tek) and frozen on dry ice. The molds were wrapped in Saran Wrap and aluminium foil and stored in -80°C until sectioning. 10μm sections were made using the Leica CM 3050S cryostat and adhered onto Fisherbrand Superfrost slides. Sections were placed at the bottom of the slides. The cryosections were immediately used in the following fixation step. 20% PFA was thawed a day prior to the first day of ISH. Cryosections were thawed in a Wheaton dish at 50°C for 20 minutes. The sections then went through the following sequence of PFA fixes and 1X PBS washes: 4% PFA Fix (20 min), PBS (2 x 5 min), TE/ProK (1 min) to increase porosity of the sections, PBS (5min), 4% PFA Fix (5 min), DEPC water (1 min), TEA/AA (10 min) adds acetyl groups to the functional groups on the tissue and slides to reduce background, and three 5-minute washes in 1X PBS.  23  2.5.3 Prehybridization and hybridization Sections for sense and antisense probes were placed in different slide mailers (Fisher Scientific). 4mL of hybridization buffer covered the placental sections at the bottom of the slides in a full slide mailer. The mailers were incubated overnight at 55°C. Separate slides used for different probes into different mailers and 15μL of probes were added to 5mL of hybridization buffer, heated at 80°C for 5 min, and iced for 5 min.. Additional hybridization buffer was added to cover the slides and left in 55°C for hybridization reaction overnight. The slides were then subjected to a series of SSC and 1X RNA washes: 5X SSC (15 min), 0.2X SSC (60 min), 1X RNA at 37°C (10 min), 400μL of boiled 10mg/mL RNaseA in 1X RNA (30 min), 1X RNA (5 min), 2X SSC (10 min), and 0.2X SSC (10 min). The RNase A was used to digest away the endogenous single-stranded RNA to increase the signal strength. The RNA probes that were bound to Mmp2 sequences would not be affected since the nucleic complexes are double stranded. RNase A was boiled to inactivate contaminating DNases. 2.5.4 Antibody reaction and mounting The antibody reaction occurred in the following steps in a slide mailer: NT at 20°C (5 min), 1% blocking in NT at 37°C (60 min), addition of 4μL of 1:2000 dilution of anti DIG-AP antibody to blocking (60 min), NT washes at 20°C (3 x 20 min), NTMT wash (10 min), NTMTL wash (5 min), and addition of 37.75μL of 100mg/mL NBT (Roche) and 35μL of 50mg/mL BCIP (Roche) for the colour reaction. The mailers were placed in a box overnight and stored in a cabinet to prevent light from entering the slide mailers. The slides were washed 3X 15 min with 1X PBS at 20°C and fixed in a Formaldehyde Mounting fix for 2 hours at room temperature. Slides were stained for a few seconds using Nuclear Fast Red (Sigma-Aldrich), then dehydrated with 70%, 90%, and 3X 100% EtOH for a minute each. The slides were transferred to xylene, then mounted with Entellan (Harleco). 24  Table 2.2 ISH Solutions DEPC water 10X PBS 20% PFA  0.5mL of diethylpyrocarbonate in 500mL ddH2O 2g KCl, 2g KH2PO4, 80g NaCl, 1.45g Na2HPO4 Up to 1L dH2O Dissolve 100g paraformaldehyde in 1.9mL 10N NaOH by stirring the solution on a hot plate in the fumehood and make up to 500mL DEPC water  10N NaOH 12g in 30mL DEPC water 10mg/mL Yeast 100mg of yeast tRNA (Roche) in 10mL of DEPC water tRNA 0.5M EDTA, pH 186.1g EDTA in 1L of dH2O 8.0 1mL DEPC can be added before autoclaving 1M TRIS 121.1g in 1L of dH2O pH 7.0, 8.0, and 9.5 4M LiCl 3.39g in 20mL DEPC dH2O 3M NaAc pH 5.2 24.61g of sodium acetate in 100mL dH 2O 100μL DEPC before autoclaving Denhardt's 2g Ficoll 400, 2g PVP, 2g BSA Up to 200mL DEPC dH2O 20X SSC 175.3g NaCl, 88.2g sodium citrate pH 7.0 Up to 1L Can treat with DEPC 30% sucrose 15g in 50mL 1X PBS and syringe filtered with 45mm Stericups Labelling master 2μL of 10X transcription buffer (Roche) mix 1μL of 40 units/μL of Ribonuclease inhibitor (Promega) 2μL of 10X DIG-RNA labelling mix (Roche) 12μL of DEPC dH2O 17μL per reaction 4% PFA Fix 20mL of 10X PBS, 40mL 20% PFA, 140mL dH 2O TE/ProK 10mL of 1M TRIS pH 8.0, 2mL of 0.5M EDTA pH 8.0, 800μL of 10mg/mL Proteinase K, 188mL dH2O TEA/AA 3.72g Triethanolamine, 448μL 10N NaOH, 625 μL to 200mL with dH 2O Hybridization 50mL of 100% Formamide buffer 25mL of 20X SSC 10mL of Denhardt's 2.5mL of Yeast tRNA 5mL of 10mg/mL single stranded fish sperm DNA 7.5mL DEPC dH2O 10X RNA wash 118.4g NaCl, 50mL of 1M Tris pH 7.5, 50mL 0.5M EDTA in 500mL NT 100mL of 1M Tris pH 7.5 and 30mL of 5M NaCl in 870mL of dH 2O 1% blocking 2g of blocking (Roche) in 200mL of NT NTMT 5mL of 5M NaCl 25mL of 1M TRIS pH 9.5 12.5mL of 1M MgCl 250μL of Tween 20 NTMTL 0.03g in 200mL of NTMT FA Mounting fix 20mL of 37% formaldehyde and 20mL of 10X MEM buffer in 160mL dH 2O 10X MEM buffer 20.9g MOPS, 0.76g of EGTA, 1mL of 1M MgSO 4, 2 pellets of NaOH Up to 100mL dH2O then filter sterilized 1M MgSO4 2.64g MgSO4 in 10mL of ddH2O 25  2.6 Quantitative real-time polymerase chain reaction 10µL of total RNA was used in a 20µL DNase treatment solution based on Promega’s protocol. The reaction contained 2.5µL 10X RQI Buffer, 1µL of DNaseI, and 1µL of RNasin (Promega). The reaction was placed in a 37°C thermomixer for 1hr and the DNase was inactivated at 65°C for 30min. DNase-treated RNA (D+ RNA) was checked against crude RNA on RNase-free agarose gel to ensure that the treatment was successful. The D+ RNA was used in first-strand complementary DNA synthesis. The first solution, containing 2µL of N15 (10ng/µL), 2µL of 10mM dNTPs, 8µL of D+RNA, and 7µL of DEPCdH2O, was incubated at 65°C for 5min. Then the second solution, which contained 6.5µL of DEPC-dH2O, 8µL 5X FirstStrand Buffer (Invitrogen), 4µL of 0.1mM DTT, and 0.5µL RNasein (Promega), was added to the first solution. The solutions were mixed by pipetting and 19µL was removed to another PCR tube as a negative control for reverse transcriptase (RT-). The final cDNA mix (RT+) was incubated at 42°C for 2min before the addition of 1µL of SuperScript II (Invitrogen) reverse transcriptase to RT+ tubes. RT+ and RT- tubes were incubated at 42°C for another 50min. Both the RT+ and RT- reactions were checked by qRT-PCR and melt-curve analysis using endogenous control gene prior to running differential expression assays. The standard 25µL PCR reaction volume was used for the qRT-PCR. The reaction contained 2.5µL of 10X PCR buffer, 2.5µL of MgSO4, 2.0µL of 2.5mM dNTPs, 1.0µL of 10X SyberGreen, 0.2µL of Tsg DNA Polymerase, 1µL of 10mM gene-specific primers, and 1µL of cDNA. The cycling program used was set at 95°C for 5min and 95°C 30s, 55-60°C (Tm) 30s, 72°C 30s, 82-86°C (reading temperature) 1s for 35 cycles using Bio-Rad’s Opticon II. The sequences for gene-specific primers are listed in Table 2.2. The amplification data was exported as an Microsoft Excel file for Ct and amplicon’s efficiency analysis using the LinRegPCR ver11.3 software (Ruijter et al., 2009). The Cts from 26  the technical triplicates were averaged. The fold change was determined using the following formulas: Corrected Ct (cCt) = Ct x log(efficiency of amplicon) Fold change = 2(amplicon cCt – endogenous cCt) The fold changes of the samples from each cohort were averaged and the standard deviation was plotted as the error bar.  2.7 Unilateral ovariectomy C57BL/6J females (from The Jackson Laboratory) were given subcutaneous injection of 5mg/kg ketoprofen and anesthetized with isoflurane after an hour. After the females reached a surgical plane of anesthesia, the area around the incision site was shaved and cleaned with 70% EtOH. A small longitudinal incision (<1cm) was made in the skin at the dorsal midline of the last rib with fine dissection scissors. The incision was positioned over the ovary and another small incision in the body wall was made with forceps. A loop of absorbable suture was tied between the oviduct and the ovary and the ovary was excised. The incision in the body wall was closed with a single stitch of absorbable suture and the incision in the skin was closed with nonabsorbable suture. The females were monitored for a week for signs of stress before mating.  2.8 Illumina microarray Approximately 40µg of crude RNA was cleaned using the Qiagen RNeasy Mini Kit. The cleaned RNA was re-suspended in 50µL of RNase-free water. Concentration of the RNA was estimated using the Nanodrop Spectophotometer ND-1000. The final concentration that was sent for microarray analysis was ~200ng/µL. The samples were sent to the Functional Genomics Platform at the Innovation Centre at McGill University 27  (http://www.gqinnovationcenter.com/service). The quality of the RNA was assessed using the BioAnalyzer, then cDNA was synthesized using the TotalPrep RNA amplification kit. The platform used for expression profiling was the Illumina MouseRef8.0 that could evaluate 8 samples on a single BeadChip. A total of 14 samples were profiled on two BeadChips. The raw signal, bead detection p-value and bead standard deviation results were collected into a text file by McGill and made available for download on Nanuq. All of the statistical analyses on the Illumina data were completed using the microarray analysis software FlexArray, developed by Michal Blazejczyk from McGill. Flexarray utilized the lumi package, which transforms the raw data using variance-stabilizing transform (VST) into expression values similar to but superior than a log2 transformation (Du et al., 2008; Lin et al., 2008). Significance of differential expression between WT versus IUGR samples was determined using the Empirical Baysian methods: Wright & Simon and Cyber-T (Murie et al., 2009). The p-values from these two methods were corrected for false discovery rate (Benjamini and Hochberg, 1995).  2.9 Statistical analysis The Student's t test function in Microsoft Excel 2007 was used to find significance of differential expression by qRT-PCR and to find significance of weight differences between WT and IUGR cohorts. Hierarchal clustering of gene expression data from Illumina MouseRef8.0 BeadChip was completed using the MultiExperiment Viewer program (MeV) (Chu et al., 2008). The principal component analysis was a built-in application of FlexArray that generates a PCA plot for viewing the clustering pattern of data generated from the different microarrays (in our case different biological replicates) on the same BeadChip.  28  Chapter 3: Exploration and characterization of IUGR mouse models 3.1 Introduction There are three mouse models of IUGR in which I was interested in looking for gene expression differences in the placenta: Mmp2 knock out (KO), Mest KO, and surgically-induced IUGR. Mmp2 is a member of a very large family of matrix metalloproteinases. It is mainly expressed from implantation in the mouse to E16.5. The site of Mmp2 mRNA is in decidual cells on the maternal side. Its expression pattern is reciprocal to a close relative Mmp9, which has been identified as being expressed early in trophoblast giant cells (Alexander et al., 1996). Mice null for Mmp2 exhibit a significant postnatal growth restriction starting from day 10, as well as small size at birth (Itoh et al., 1997). This is an interesting candidate for IUGR since these proteinases are thought to digest matrix of maternal uterine cells, which eventually results in successful implantation of the conceptuses into the uterine wall (Bischof and Campana, 2000). In the case of improper implantation, it is possible that maternal blood supply may be affected, contributing to growth restriction. In addition, Mmp2 has also been implicated in skeletal development as well as neovasculogenesis (Itoh et al., 1998; Mosig et al., 2007). Mmp2 expression has often been used as an assay to indicate angiogenesis in cancer metastasis studies (Foda and Zucker, 2001). It is also a downstream effector of VEGF, which relates the gene to a hypoxia response (Garzetti et al., 1999; Sounni et al., 2002). All of these findings suggest that Mmp2 may impact placental development. The Mmp2 KO mice were kindly provided to us from a researcher at the Life Sciences Institute, Dr. Chris Overall. Mest/Peg1 is a paternally-expressed gene with a differentially-methylated region in its promoter (Lefebvre et al., 1997). Expression of Mest/Peg1 highlights the fetal capillaries in the 29  labyrinthine (Mayer et al., 2000; Oh-McGinnis et al., 2010). The PEG1 protein is a hydrolase of unknown function. Our group has made the mouse KO of Mest/Peg1 and documented IUGR in the mutant conceptuses (Lefebvre et al., 1998). The KO embryos (Mest+/-) exhibit the growth phenotype from E15.5 to birth with corresponding small placentae. Other than this growth restriction, no other obvious organ or placental defects are observed in the KO embryos. This model is of particular interest to us since Mest mutants and the Plagl1 KO display similar embryonic IUGR, and the authors have discovered changes in imprinted gene expression in in the null mutants (Varrault et al., 2006). The surgical model is a hemiovariectomy model of IUGR. This model is based on the observation that if blood supply is restricted to the conceptus, then the embryos will develop an IUGR phenotype. Uterine ligation animal models, which is a method to restrict blood supply to the placenta by ligating the uterine vessel, have been demonstrated to exhibit IUGR in mouse, rat, and sheep (Andersen et al., 1988; Coe et al., 2008; Newnham et al., 1986; Vileisis et al., 1982; Vuguin, 2007). In our case, we will not ligate the artery that directly feeds the placenta to induce IUGR, but create a crowded situation where the conceptuses in the middle of the horn will receive less blood supply than those on the outside edges of the horn (Fig. 3.1) (Coe et al., 2008; Vom Saal and Dhar, 1992). The reason behind the reduced blood supply is twofold. First, the mouse uterus is divided into two horns, with a membrane separating the horns; thus, removal of one ovary means that the remaining ovary will ovulate every cycle, resulting in all the conceptuses to be implanted in one horn. Second, the uterine blood supply is bidirectional with the main artery splitting into two large branches supplying each end of the horn. These vessels join in the middle of the horn forming a loop, with smaller branches coming off the loop to provide blood to each conceptus. In a crowded situation, the conceptuses implanted in the middle  30  of the horn will be subjected to lower blood pressure and receive less blood (Coe et al., 2008; Vom Saal and Dhar, 1992).  Fig. 3.1 Basis of the "Crowded Uterine Horn" mouse model of IUGR. Taken from Coe et al., 2008 with permission from John Wiley & Sons.  3.2 Results 3.2.1 Mmp2 expression in the developing placenta There have been several studies documenting the function of MMP2 in the adult as well as during early embryogenesis (prior to E8.5). In short, Mmp2 expression in the periimplantation uterus is restricted to the site of the developing maternal decidua (Alexander et al., 1996; Das et al., 1997). Only one group has looked at Mmp2 expression in the placenta at later stages. Teesalu et al. (1999) validated the presence of Mmp2 in the maternal decidua and further demonstrated the emergence of its expression in the labyrinthine at E16.5. Interestingly, their Northern blot data shows that Mmp2 expression actually begins to decline from E13.5 and is nonexistent by E18.5 (Teesalu et al., 1999).  31  Here, I wished to further elucidate the role of Mmp2 in the late stage placenta by examining its expression pattern in more detail. For this, a colorimetric in situ hybridization (ISH) was performed on E14.5 wild-type placental sections using an anti-sense Mmp2 riboprobe (Fig. 3.2). I was able to detect sparse staining in a few fetal vessels in the labyrinthine of the placenta (Fig. 3.2B) and distinct staining in the yolk sac remnants of the sections (Fig. 3.2C,D). However, there was no staining in the maternal decidua near the top of the sections (Fig.3.2A).  Fig. 3.2 Expression of Mmp2 mRNA in E14.5 Mmp2+/+ placentae by ISH. (A) Upper half of the placenta. (B) Lower half of the placenta. Scale bar: 0.25mm for (A) and (B). (C) Yolk sac. Scale bar: 0.13mm (B). (D) Fetal endothelial cell. Scale bar: 0.062mm. Sense probe not shown. dc: decidua, GCs: giant cells; sp: spongiotrophoblast layer; lab: labyrinthine layer.  32  3.2.2 Mmp2-deficient placentae exhibit normal morphology Since Mmp2 has been postulated to be involved in early implantation, specifically in the digestion of the matrix that surrounds uterine cells, it would be expected to observe an abnormal placental morphology in Mmp2-deficient conceptuses (Alexander et al., 1996; Bischof and Campana, 2000; Teesalu et al., 1999). To assess the morphology of the placenta, H&E staining was performed on paraffin sections of E14.5 placentae. There was also no observable difference in the size of each placental layer nor gross disorganization of blood vessels in the labyrinthine between the Mmp2+/+, Mmp2+/-, and Mmp2-/- (Fig. 3.3).  Fig.3.3 Placental morphology of E14.5 Mmp2 placentae by H&E. (A) Mmp2+/+ (B) Mmp2+/-. (C) Mmp2-/-. n=2 for each genotype. All scale bars = 1mm.  3.2.3 Mmp2-/- conceptuses do not exhibit IUGR Itoh et al. (1997) have demonstrated that the Mmp2-/- mice exhibit a clear postnatal growth restriction. They also observed that the Mmp2-null mice have low birth weights. However, they did not present any data showing this growth difference between the null mice and their WT littermates at birth. I needed to confirm whether the Mmp2-/- mice are in fact growth restricted during development in order to validate the Mmp2 KO mice as a model for 33  IUGR to test for differential gene expression. Therefore, I set up intercrosses between Mmp2 heterozygotes and measured the weights of Mmp2+/+, Mmp2+/-, and Mmp2-/- mice at E14.5 and postnatal day 1 (P1). I found that Mmp2-/- conceptuses exhibited no significant weight difference when compared to WT at either stage (Fig.3.4A and Fig. 3.4B). There was also no difference between WT and null placental weights at E14.5 (Fig.3.4C).  34  Fig. 3.4 Weight comparisons between Mmp2+/+, Mmp2+/-, and Mmp2-/-. (A) Embryonic weights of E14.5 conceptuses (Mmp2+/+: n = 25; Mmp2+/-: n = 42; Mmp2-/-: n = 27) from 11 litters . P = 0.85. (B) Placental weights of E14.5 conceptuses (Mmp2+/+: n = 7; Mmp2+/-: n = 17; Mmp2-/-: n = 10) from 4 litters. P = 0.84. (C) Newborn pup weights at postnatal day 1 (Mmp2+/+: n = 8; Mmp2+/-: n = 24; Mmp2-/-: n = 10) from 5 litters . P = 0.89. All weight comparisons were subjected to Student's t test. 35  3.2.4 Imprinted gene expression in Mest mutants The IUGR phenotype of the Mest-deficient mice has been previously described (Lefebvre et al., 1998). Thus I analyzed the gene expression of imprinted genes by quantitative real-time PCR (qRT-PCR) of E14.5 placental samples to see if there is differential expression between WT and Mest+/- littermates. The imprinted genes that were first tested for differential expression between Mest+/+ and Mest+/- placentae are genes that have been implicated in human studies of IUGR or regulators of those genes. H19, CDKN1C, and IGF2 have been directly involved in syndromes such as Beckwith-Wiedemann and Silver-Russell Syndrome (Bliek et al., 2006; Gicquel et al., 2005; Lam et al., 1999; Romanelli et al., 2009; Zhang et al., 1997). Igf2r is a direct regulator of Igf2 (Czech et al., 1989; Filson et al., 1993). Slc38a4 is one of the genes that code for a system A transporter, which is crucial for fetal-maternal nutrient exchange (Mackenzie and Erickson, 2004). In a placental-knockout of Igf2, researchers only identified increased expression of Slc38a4 out of the three System A transporters (Constancia et al., 2005). PEG10 is located in the region on human chromosome 7 and UPD of chromosome 7 is implicated in 10% of SRS patients (Kozak et al., 1997; Penaherrera et al., 2010). The mouse knockout of Peg3 exhibit similar IUGR phenotypes as Mest+/-, though PEG3 specifically has not been implicated in human growth-related syndromes (Li et al., 1999). I assessed the imprinted gene expression of four maternally-expressed genes (H19, Cdkn1c, Phlda2, Igf2r) and four paternally-expressed genes (Peg10, Igf2, Peg3, Slc38a4). None of the genes showed significant difference in expression between the WT cohort (n=3) and the Mest+/- cohort (n=3) but H19, Cdkn1c, Igf2r, and Slc38a4 appear to be decreased, whereas Phlda2 appears to be increased in the Mest+/- placentae (Fig.3.5).  36  A  B  Fig. 3.5 Comparison of candidate imprinted gene expression between E14.5 WT (Mest+/+) and IUGR (Mest+/-) placentae by qRT-PCR. (A) Gene expression of maternally-expressed genes. (B) Gene expression of paternally-expressed genes. All qRT-PCR was performed on three different placentae for each genotype with three technical replicates per placenta. Expression is represented as a ratio of imprinted gene Ct over Gapdh Ct. The value of each cohort (Mest+/+ and Mest+/-) is the average of the three biological replicates. Error bars represent the standard deviation between the biological replicates in each cohort (P > 0.5 for all genes). 37  3.2.5 Variability in gene expression by qRT-PCR Despite a trend for difference between the WT and Mest-/- cohorts, these differences were not significant due to the variability in gene expression between biological replicates within the same cohort. Most notably, Phlda2 and Igf2r exhibited the greatest variability of expression within a particular cohort (Fig.3.5A). Looking more specifically at the expression level of individual replicate, it is apparent that Igf2r expression of one of the WT replicates is very different from the rest and Phlda2 expression is variable in the Mest+/- cohort (Fig. 3.6A). The small sample size (n = 3) makes it difficult to determine what is the actual level of gene expression in Phlda2 since each replicate has an expression level very different from one another. Expression differences between biological replicates is common, but we have particularly noticed it in the placenta (Fig.3.5B). Pidoux et al. (2004) have reached the same conclusion in human placental samples. It is possible that differences in gene expression in IUGR may be masked by this variability in gene expression. Therefore, we decided to eliminate some of the causes of variability and re-assess gene expression. We were also interested in seeing if this variability we observed in the placenta is manifested in the embryo. I analyzed WT C57BL/6J samples and compared the variability of expression of two control genes (Gapdh and Ppia) and imprinted IUGR candidates (H19, Cdkn1c, Phlda2, Igf2r, Igf2, and Slc38a4) between embryo and placenta and found that, with the exception of Igf2 and Cdkn1c, the rest of the imprinted genes and Gapdh have more variable expression in the placenta (Fig.3.5C).  38  A  B  C  Fig. 3.6 Variability in gene expression. (A) Igf2r and Phlda2 expression in placental replicates (n = 3 for each genotype). Error bar represents the standard deviation of three technical triplicates of each biological replicate. (B) Variation of imprinted gene expression in Mest+/+ (blue) and Mest+/- (red) placentae. The Mest IUGR mouse model is on the outbred CD-1 background. (C) Gene expression variation of control genes and imprinted genes in the embryo (blue) and the placenta (purple) of WT C57BL/6J conceptuses. Variation is presented using weighted standard deviation of gene expression (SD/expression). All of the genes' expression level is relative to a new qRT-PCR control, Ppia, which was determined to be a better qRT-PCR control than Gapdh (Mamo et al., 2007). Ppia's expression level is relative to Gapdh.  39  3.2.6 Crowded uterine horn Two C57BL/6J females had one ovary removed in order to induce IUGR in the embryos via the crowded uterine horn method (Coe et al., 2008). Female 1 was plugged by a C57BL/6J male twice. She did not appear pregnant on the day of the first planned dissection (E15.5) and was allowed to go to term. She delivered one abnormal pup on E23.5 with the aid of on-site veterinarian. The pup was large and had an elongated neck. It had no discernible facial features but limbs and digits were present. Female 1 was dissected 15.5 days after the second plug and had a litter of ten embryos, all located in one uterine horn (Fig. 3.7A). Weight data indicated that all except one of the embryos and placentae were smaller than the previous normal C57BL/6J litter (Fig. 3.7B). Average weight of the crowded embryos was 0.397 ± 0.025g, which was significantly different (p < 0.001) from the average weight of WT embryos at E15.5 (0.452 ± 0.026g). The bottom 5th percentile of normal Bl6 was <0.409g (n = 15), therefore five embryos were deemed as IUGR. These IUGR embryos did not exhibit gross morphological difference when compared to WT embryos (data not shown). Female 2 was plugged by a C57BL/6J male twice. She appeared pregnant on the day of her first planned dissection (E15.5) but was not dissected because she had pregnancy bulges on both sides, which indicated that she was carrying pups in both horns. She delivered a litter of 7 pups on E20 or postnatal day 1 (P1). Pups had an average weight of 1.26 ± 0.07g and none had observable difference in size (Suppl. Fig. 1B). IUGR cut-off was pre-determined to be <1.1g, which represented the bottom 10th percentile of P1 pups (n=41) and none of the pups in this litter was found to be under 1.1g. Female 2 carried a litter of 10 conceptuses for her second pregnancy and all of the conceptuses appeared to be implanted in one horn. The average embryonic weight was 0.235 ± 0.015g. This weight average was closer to E14.5 embryonic  40  weight (Mmp2+/+), which was 0.254 ± 0.030g. Due to the uncertainly of embryonic stage, the second litter from Female 2 was not used.  41  Fig. 3.7 Growth phenotype of surgical model. (A) All implantations were located in the uterine horn that is still attached to the ovary. Uterine horn was dissected from Female 1 at E15.5 (n=10). (B) Comparison of the embryonic and placental weights from crowded uterine horn versus normal E15.5 conceptuses (n = 11). (C) Embryonic and placental weight of E14.5 litter from hemiovariectomized Female 2. Implantations were also found only in one horn. 42  3.3 Discussion 3.3.1 First IUGR model: Mmp2 -/- conceptuses do not exhibit IUGR The selection of this model was based on an extensive search throughout the Mouse Genome Informatics for mouse models that exhibit prenatal growth defect. Mmp2 was amongst one of the five mouse knockouts selected that would have live mice with an IUGR phenotype and that were purchasable from repositories. It became the first candidate when we realized we could easily obtain live Mmp2 knockout mice for breeding from a fellow researcher at UBC. Mmp2’s role in implantation and the growth phenotype observed by Itoh et al. (1997) further suggested that this model may be a suitable model of IUGR due to placental dysfunction. However, Itoh et al. (1997) did not conduct any prenatal analysis on the Mmp2-/- conceptuses. This omission lead me to confirm the status of IUGR and placental dysfunction in the KO mice. At the end of ten litters, there was no difference in embryonic weights between the three genotypes (Fig. 3.4). At the same time as I was collecting embryos for weight measurements, I also conducted morphometric analyses by H&E on Mmp2 placentae to assess if placental dysfunction was the cause of the supposed IUGR, as noted by Itoh et al. (1997). I did not observe any gross morphological differences between Mmp-/- and its littermates (Fig. 3.3). There is the possibility that differences may exist if I assess size and amount of branching of fetal vessels in the labyrinthine by doing placental casts, but in conjunction with the lack of IUGR, any findings would not be worthwhile for our purpose. It is possible that since there are sixteen other matrix metalloproteinases in the mouse, there may be a compensatory effect from these other proteases, or their inhibitors. For example, the target proteases of Timp3, which includes multiple MMPs, show no change in enzyme activity in Timp3-/- tissue-derived culture (Fogarasi et al., 2008). 43  The observation of Mmp2 being expressed in the maternal decidua during early embryogenesis suggests that instead of looking for IUGR in Mmp2-/- embryos, it is more likely that breeding Mmp2-/- females may lead to IUGR. I had collected one litter from Mmp2-/females and did not observe any difference in weight measurements when compared to litters born from Mmp2+/- females. Nevertheless, there is still a possibility of IUGR as one litter is statistically insignificant. An issue with studying IUGR effects in Mmp2-/- female pregnancies is that both of the Mmp2+/- and Mmp2-/- conceptuses may be affected if the null mutant mother does not the produce the metalloproteinase important for implantation, then there will be no phenotypically normal embryos within the same litter to act as WT control. For gene expression in the mouse system, the best method for comparison is to compare IUGR placenta with phenotypically normal littermates since there may be a greater amount of variation in gene expression between litters (Pidoux et al., 2004). This expression variation may mask slight differences in gene expression between IUGR and WT placentae that may be important in IUGR. It is possible to still use this strain if Mmp2-/- conceptuses from Mmp2-/-female mother are more affected than their Mmp2+/- littermates. However, it will be more cost effective to obtain a strain of mice that has well-documented IUGR such as Akt1tm1Mbb, which is suggested to be caused by a lack of glycogen cells important for continual invasion of the labyrinthine leading to placental insufficiency. (Cho et al., 2001; Yang et al., 2003). In this case IUGR would not have to be confirmed before doing a differential expression assessment.  3.3.2. Second IUGR model: Mest/Peg1 IUGR model has high variation in gene expression Several mouse KO of imprinted genes exhibit IUGR. One of these is the Mest/Peg1 mouse knockout. There is a 10% reduction in the Mest+/- embryo by E18.5 and it is one of the genes in the imprinted gene network proposed by Varrault et al. (2006), therefore making this a 44  suitable IUGR model to study. All the three layers of the mouse placenta appear to retain proper thickness in the Mest+/- placenta, though these placentae are smaller in comparison to their WT littermates (Lefebvre et al., 1998). The 10% reduction in placental size reflected the 10% growth restriction in the Mest+/- embryos. The cause of the IUGR observed in the Mest+/- embryos is unclear. The gene codes for a hydrolase, but it has an unknown function. The gene is highly expressed in all mesoderm derivatives (Kaneko-Ishino et al., 1995). Placental expression of Mest at E14.5 is restricted to the fetal blood vessels. This may indicate a function for Mest in the development of fetal vasculature though no obvious disturbance in vasculature is observed in the Mest+/- placenta with in situ hybridization (personal communication). It may be necessary to construct Mest+/- placental casts in order to see if there are differences in branching of fetal blood vessels in the labyrinthine. If there is reduction in branching in these placentae, it may explain the IUGR observed in the embryo as one of the main cause of IUGR in human. Alternatively, the placental vasculature might not show any morphological abnormalities in the Mest mutants but rather could exhibit a functional defect in transport properties. Out of the eight imprinted genes assessed, none of them had significant expression difference between the Mest+/+ and Mest+/- placentae as determined by Student's t-test. This may be explained by the small sample size used as well as the accompanied variation in expression between biological replicates. There was only a sample size of three for each cohort (Mest+/+ and Mest+/-). Therefore, if there is even one outlier in the cohort, the overall expression would be affected. This was the case for the Igf2r assay that showed WT sample #2 was an outlier (Fig.3.6A). Placental sample 2 was a clear outlier in all three cases. If it was removed then Igf2r expression between WT and IUGR cohort would be comparable. An increase in sample size would be necessary for accurate assessment of differential expression and subsequent determination of role of some imprinted genes in IUGR. In the Mest+/- IUGR 45  cohort, Phlda2 exhibits an increase in expression that is similarly observed by three other studies (Apostolidou et al., 2007; Diplas et al., 2009; McMinn et al., 2006). But the variation of gene expression in Phlda2 between biological replicates was substantial in my study, especially in the IUGR placental samples (Fig.3.6A). Nonetheless, my microarray study has also identified Phlda2 to be over-expressed in a separate IUGR model, indicating its involvement as a general regulator of growth (Apostolidou et al., 2007).  3.3.3 Placenta has greater variation in gene expression than the embryo With the previous discovery that gene expression between biological replicates can vary significantly, we became interested in the issue of expression variation. It turned out that our concurrent study of imprinting expression in human IUGR was also plagued by this situation. Bourque et al. (2010) documented that their expression study done on placental tissues had great amount of variation in expression. They were concerned with the difference in the amount of RNA degradation for each placenta than the amount of time for labour in each individual (Avila et al., 2010). Another group had also documented that different sampling sites within the same human placenta was a contributor to variability as well. They also noted that variability between placentae from different pregnancies was greater than the variability from sampling the same placenta (Pidoux et al., 2004). The great inter-individual variability in human placental expression may be attributed to genetic heterogeneity in the human population. In our case, the Mest KO is on an outbred background (CD1). This means that embryos of the same litter will also be genetically heterogeneous. This may have explained partly for the variability I observed in my imprinted gene analyses between Mest+/+ and Mest+/- cohort though I observed similar variability in expression in WT C57BL/6J samples (Fig.3.). Another cause of variation can be due to pipetting 46  error during qRT-PCR, but the technical triplicates generally have standard deviation of less than 0.5% of the expression Ct (data not shown). A more interesting idea is that expression variability may be an intrinsic characteristic of the placenta. Development in the embryo is specific at each stage since organs need to properly form in order for it to be able to survive throughout development and adulthood; this may require a tighter control in gene expression. Comparatively, the placenta is a tissue that is discarded after birth and regulation do not need to be as stringent. This can be seen in the observation of normal pregnancy outcome with polyploid placenta whereas polypoloidy has never been observed in fetus that survives (Kalousek, 1994). Considering the function of the placenta is to provide nutrients for the embryo, each embryo may have different needs due to the amount of blood supply it obtain or other factors. Thus programming in the placenta may need to be more relaxed to reflect this difference in nutrient demand. This can result in a difference in gene expression between biological replicates.  3.3.4 Third IUGR model: crowded uterine horn can lead to development of IUGR The average litter size at E14.5 or E15.5 of C57BL/6J females was eight embryos in my hands. Removal of one ovary in a female would result in the conceptuses to be all implanted in one uterine horn (Coe et al., 2008). The average embryonic weight of crowded litters at E14.5 and E15.5 were reduced by 8% and 12.5% respectively when compared to WT litters. This suggested that crowding does affect embryonic weights, most likely affecting the entire litter. For the litter collected at E15.5, some of the embryos did exhibit IUGR. The other litter that was collected at E14.5 instead E15.5 and only had one IUGR embryo, even though it carried the same number of conceptuses. This suggests the development of IUGR in the surgical model may be just beginning at E14.5, which is a similar time for the development of IUGR for other mouse mutants (Constancia et al., 2005; Constancia et al., 2002; Lefebvre et al., 1998). The growth 47  phase in mouse development begins at E14.5, as signalled by significant invasion of the labyrinthine vasculature into the maternal decidua. This invasion will result in an increase in surface area for nutrient exchange in the placenta. The embryos increase in size significantly between this point until birth. It is possible that the embryos collected at E14.5 may become more growth restricted by E15.5, as the fetal-maternal blood flow becomes increasingly important. A key difference between our surgical-induction of IUGR is that the embryos on the side of the uterine horn are smallest, not the ones in the middle as had been suggested by other studies (Coe et al., 2008; Vom Saal and Dhar, 1992). In fact, this is true even for normal litters (Fig. 3.7). Coe et al. (2008) actually have depicted this in their figure of the crowded model (Fig. 3.1). I hypothesize that these embryos are smaller because they are the ones that get pushed further up in the body cavity during mouse gestation. This may physically prevent those embryos from growing normally. This is supported by the consistent observation of the smaller embryos occupying the ends of the horn in normal C57BL/6J litters. The crowded horn phenomenon would be affected by the number of conceptuses in a single litter. The second litter from Female 2 showed that the conceptuses were implanted in one horn, indicating that the surgery was successful. Nevertheless, the first litter from Female 2 did not contain any IUGR pups. This inconsistency could be explained by the litter size between litter one and litter two. The average litter size for C57BL/6J is 6.2±0.2 (Nagasawa et al., 1973). Litter one had a litter size of seven, whilst litter two had a litter size of ten. Consequently, I considered the second litter from Female 2 to be crowded since there were ten conceptuses all in one horn and this was reflected in their decreased weights (Fig. 3.7). It may be that blood pressure difference within the less-crowded horn (litter 1 from Female 2) was not enough to  48  result in weight differences. More litters would need to be collected and weighed in order to observe how litter size may contribute to the degree of IUGR in this surgical model. The next chapter presents the results from whole-genome expression studies done to compare the expression levels between normal C57BL/6J litter and the litter from female 2 that exhibited IUGR (Fig. 3.7A,B).  49  Chapter 4: Effects of surgically-induced IUGR on whole-genome expression 4.1 Introduction We constructed a mouse model of IUGR by performing hemiovariectomy in C57BL/6J females. All the conceptuses were implanted in one horn and the IUGR phenotype results due to a reduction of blood supply to the crowded embryos (Coe et al., 2008; Vom Saal and Dhar, 1992). We wished to observe the effects of crowding on gene expression in the embryo and the placenta. We utilized the Illumina BeadChip Array technology to obtain a profile of gene expression at the desired embryonic stage. Normally IUGR in the mouse is observed from E15.5 onwards, as examplified in the Mest KO (Lefebvre et al., 1998). This is the stage we chose to obtain RNA from the embryo and the placenta to conduct microarray gene expression analysis. Illumina uses a probe-cDNA hybridization technique often use in spotted microarrays, except that probes are attached to microbeads. The probes are 50 bases and are linked to an "address" sequence for identification (Kuhn et al., 2004). The probes hybridize to biotin-labelled cDNA derived from total RNA from samples of interest. Each transcript is represented by an average of 30 beads with attached probes placed randomly on the array, and each gene is usually represented by an average of 2-3 transcripts. The background controls are probes of random sequences that are not complementary to any location in the mouse genome. The raw signals are then normalized to the background signal (Kuhn et al., 2004). The Illumina BeadChip technology allows for the simultaneous assessment of wholegenome expression for up to eight biological replicates per beadchip. Since we have observed expression variation between biological replicates by qRT-PCR, increasing the number of biological replicates is desirable (www.illumina.com). By utilizing the Illumina platform, we can assess differential expression between WT and IUGR samples using multiple biological 50  replicates at a lower cost as compared to other microarrays that only allow one sample to be assessed per slide. We have completed two MouseRef8 Beadchips for this project. The first Beadchip slide contained normal C57BL/6J E15.5 embryo and placental samples, with four biological replicates per tissue. The second Beadchip slide was used to assess expression of three biological replicates of surgically-induced IUGR samples. Results from the Illumina array are interpreted using the FlexArray software, developed by Michael Blazejczyk at McGill Innovation Centre (http://genomequebec.mcgill.ca/FlexArray). The FlexArray software employs the lumi package that normalizes the raw signal obtained from the Illumina Beadchips. The lumi package uses a method called variance stabilization transformation (VST), which is an enhanced version of a Log2 transformation of the raw signal (Lin et al., 2008). FlexArray has built-in statistical analyses to assess the significance of differential expression between WT and IUGR samples; it also generates different plots to visualize the data.  4.2 Results 4.2.1 Gene expression in wild type C57BL/6J embryos and placentae As expected, gene expression patterns between E15.5 embryo and the placenta are really different (Fig. 4.1). Out of the 18,138 genes assessed, there are 9031 genes that exhibit significant differential expression (p<0.05) between the placenta and the embryo at E15.5. Fold change is expressed as the ratio of a gene's expression in the placenta to its expression in the embryo. Since microarray technology is comparative, even genes that are not expressed (not detected by PCR) will have a raw signal. Therefore, traditional methods have used an arbitrary 2- to 10-fold to distinguish gene expression differences. The genes with fold change differences of over 10 are defined as placenta-specific genes, and under 0.1 to be embryo-specific genes 51  (Tanaka et al., 2000). According to these criteria, our analysis uncovered 60 placental-specific genes and 39 embryo-specific genes (Suppl. Table 1).  Fig. 4.1 Gene expression in E15.5 WT C57BL/6J embryo and placenta by Illumina expression profiling system. Scatter plot of mean expression generated by Flexarray software. Illumina dot plot of gene expression of a subset of genes. The y- and x-axis shows the VSTtransformed value of gene expression signal from the placenta and the embryo, respectively. Samples size = 4 for each tissue. The genes that exhibit significant difference in expression between the embryo and the placenta are represented by the black dots, whilst the red dots represent the genes that exhibit similar expression in the embryo as well as the placenta.  52  4.2.2 Variability in gene expression in the placenta vs in the embryo The intriguing observation of higher variations of gene expression in the placenta introduced in 3.2.5 prompted us to pursue this question of whether or not gene expression is truly more variable in the placenta than in the embryo on a genome-wide scale. We used the Illumina platform to compare genome-wide expression between the placenta and the embryo. After normalization against background signal control and data transformation, most genes (over 10,000) have expression values between 7.8 and 8 in embryo and/or placenta (Fig.4.2A), and they have relatively low expression at this developmental stage. The genes were binned into categories based on their expression level. The Illumina data delineates a non-linear positive correlation between expression and variance (Fig.4.2B). It is observed that the variance in gene expression is higher in the placenta than in the embryo in almost all categories. Large difference in the variance between the two tissues are only present in the expressions bins 10-11, 11-12, 1213, and >13 (Fig.4.2B). The placenta exhibits a minimum of 1.4-fold higher variance in these categories. We weighted the variance against expression level to prevent a mathematical confound: since one particular gene has higher expression, leaving more room for variance. I then filtered for genes that have expression level > 10 and almost no difference in expression between the embryo and the placenta (fold change ~ 1) (Fig.4.2C). The number of genes per expression bin was less than 40 genes; thus, only three expression bins were created with a minimum of 15 genes per bin (Fig.4.2C). Since there was almost no difference in fold change between the embryo and the placenta, it was not necessary to divide the variance by the expression. The placenta exhibits a minimum 1.25-fold increase in variance using this filtering method.  53  Fig. 4.2 Comparison of gene expression variation between E15.5 WT embryos and placentae. (A) The number of genes in each expression level. Expression level ranges from 7.8 to 15.5. (B) Variances in gene expression increase with an increase in expression level. The variance in the embryo plateaus when expression level reaches 10 but the placenta continues to exhibit increased variance at higher expression levels. Expression variance is weighted by dividing the expression variance of a gene with its expression level. (C) Variance is assessed for genes that have similar expression (> 10) in the embryo and the placenta. Both (B) and (C) show a higher variance in the placenta than in the embryo. 54  4.2.3 Most variable genes in the placenta and embryo Even though the placenta appears to have more variability in gene expression, both tissues have highly variable genes (Table 4.1). The arbitrary cut-off for high variance was having expression variance > 0.2 (Fig. 4.3). There are 29 genes that exhibit the highest variance in the placenta, which represent 0.1% of the total number of genes included on the Illumina array. In the embryo, the most variable genes represent 0.066% (12 genes) of the total number of genes. The variances of these genes were tested for equality of variance (F Test) to determine if the variance of a gene in the placenta is significantly different from the variance of the same gene in the embryo. Specifically genes that have variance ratio of placenta variance/embryo variance >1 as placenta-variable genes and <1 as embryo-variable genes. Out of the 29 placenta-variable genes, only 17 have significant difference in variance between the placenta and the embryo: Gzmg, Cxcl1, Prl8a6, Aqp1, Rnu6, Prlpn/Prl7b1, Apom, Apoa4, Gzmd, Eraf/Ahsp, Slc4a1, Spp2, Spink3, Apoa1, Afp, Apoa2, and Ttr. Most of these genes code for proteins that are located in the extracellular compartment. They function in transport of substances, regulation of protein activity, and regulation of blood vessel size. Out of the 12 embryo-variable genes, 9 genes have significant difference in variance: Myl7, Asprv1, Nppa, Lor, Krt10, Myl2, Car3, Akr1b7, and Pnliprp1. These embryo-variable genes play a role in metabolism or code for proteins that make up the cytoskeleton.  55  Table 4.1 Genes with the most variability in expression in E15.5 C57BL/6J embryos and placentae. Variable genes characterized by their biological functions and components, and their chromosomal locations. Exprs. = VST-transformed expression level. Var. = Variance of expression. U = chromosome location unknown. Tissue  Gene  Exprs.  Var.  mChr  Embryo  Myl2 Asprv1  11.74 9.529  0.277 0.211  5 6  Myl7  8.991  0.204  11  11.58 11.85 8.942 10.95 8.966 10.41 9.513 12.91 10.35 9.807 12.63 10.83 9.364 10.97 8.583 8.478 8.919 10.13 12.99 8.940 8.650 8.491 9.263  0.273 0.228 0.224 0.485 0.405 0.212 0.394 0.283 0.388 0.214 0.284 1.53 0.800 0.389 0.405 0.211 0.202 0.333 0.251 0.330 0.201 0.212 0.218  11 3 17 19 6 4 3 7 11 13 13 9 1 U 3 3 14 14 13 9 12 5 16  GO Component Cytoskeleton Myosin Membrane Cytoskeleton Myosin Cytoskeleton intermediate filaments Membrane Extracellular space Extracellular space Cytoplasmic Cytoplasmic Cytoplasmic Hemoglobin Hemoglobin Extracellular Space Extracellular Space Extracellular Space Extracellular Space Cytoplasmic Unknown Extracellular Space Unknown Unknown Extracellular Space Extracellular Space Extracellular Space Extracellular Space Extracellular Space  Alas2 Ttr Slc4a1 Tfrc Afp Apom  11.67 11.64 10.66 12.19 11.08 8.916  0.397 1.92 0.418 0.560 1.56 0.303  X 18 11 16 5 17  Mitochondria Extracellular Space Membrane Membrane Extracellular Space Extracellular Space  Apoa2 Hba-X Hbb-Y Aqp1 Rnu6 Loc674706 Wfdc2 Stfa1 Spink3  10.84 9.666 12.09 9.759 10.86 10.39 11.67 8.321 10.58  1.84 0.439 0.683 0.246 0.249 0.257 0.258 0.270 1.306  1 11 7 6 U U 2 16 18  Extracellular Space Hemoglobin Hemoglobin Membrane Unknown Unknown Extracellular Space Cytoplasmic Extracellular Space  Krt10 Lor Clp2 Pnliprp1 Akr1b7 Nppa Car3 Hbb-Y Hba-X Placenta Prl8a6 Prlpn Apoa1 Spp2 Eraf/Ahsp S100a9 Fgg Gzmg Gzmd Prl8a2 Apoa4 Serpina1b Cxcl1 Kng1  GO Process Cell differentiation - muscle Cell differentiation - skin Cytoskeletal function  Formation of keratin Keratinization of the skin Lipid catabolism Lipid catabolism Lipid metabolism Modulate size of blood vessel Single-carbon metabolism Transport of oxygen Transport of oxygen (placental hormone) (placental hormone) Angiogenesis Bone metabolism Cell differentiation -blood cells Cell movement - leukocyte Clot formation Cytolysis Cytolysis Hypoxic response Immune response - innate Immune response - non-humoral Inflammation Modulate size of blood vessel Regulate formation of hemoglobin Thyroid hormone synthesis Transport - anions Transport - vesicle mediated Transport of copper Transport of lipids Transport of lipids and beta fatty acids Transport of oxygen Transport of oxygen Transport of water Unknown Unknown Unknown Unknown Unknown  56  Fig.4.3 Whole-genome comparison of variance in gene expression between E15.5 C57BL/6J embryo and placenta. Scatter plot of expression variance illustrates that the maximal variance is under 0.5 in the embryo and under 2 in the placenta.  57  4.2.4 Imprinted gene expression and variability in C57BL/6J embryos and placentae The Illumina platform assessed the expression of 18,138 genes, including 71 known imprinted genes. Many imprinted genes are important regulators of embryonic development, as suggested by their high expression level during development. Blcap, Pon2, Snurf, H19, Igf2, Cd81, Cdkn1c, Nap1l4, Dcn, Grb10, Dlk1, and Igf2r have relatively high expression (expression >10) in both the embryo and the placenta. Rhox5 and Slc38a4 are only highly expressed in the placenta whereas Mest, Ndn,and Commd1 are predominantly highly expressed in the embryo (Fig. 4.4A). The variability in imprinted gene expression is also higher in the placenta than in the embryo, though none of the imprinted genes have variance > 0.08 (Fig. 4.4B).  58  Fig. 4.4 Imprinted genes' expression and variability in E15.5 WT embryo and placenta. (A) Expression profile of 71 imprinted genes, organized by chromosomes from left to right. Expression level is the VST-transformed raw signal. (B) Variance of imprinted genes was weighted against the expression. The table represents the number of genes in each expression bin on the x axis.  59  4.2.5 Differential expression between wild-type and IUGR Though the placenta seemed to exhibit more variance in gene expression between biological replicates than in the embryo, overall the variance is relatively small (10 -2) using the Illumina platform. Thus we profiled the gene expression of the embryonic and placental RNA samples derived from E15.5 conceptuses from the surgical IUGR model. There were five embryos that exhibited IUGR but only two of them were males (Fig. 3.7A,B). There was one other male that did not meet the IUGR cut-off but we believed that the crowding effect would affect all of the conceptuses, which prompted us to include that male sample as well. The samples were sent to McGill for expression profiling using the Illumina MouseRef8.0 BeadChip. Clustering analyses of microarray data group together samples (biological replicates) with similar properties. Four clusters of the samples are clearly displayed using a hierarchal clustering method to group the different RNA samples (Fig. 4.5). As expected, the embryonic samples cluster independently from the placental samples. Furthermore the IUGR samples cluster separately from the WT samples. The IUGR embryo cluster exhibits larger within-group variation than the other clusters. One of the IUGR placental samples (LLE052) actually clustered with the IUGR embryonic samples (Suppl. Fig. 2). The tissue type of LLE052 may have been labelled incorrectly when I was preparing the RNA samples to send to McGill for expression profiling. LLE049 is the corresponding embryo sample to LLE052. LLE049 and LLE052 are not used when looking for differential expression between WT and IUGR cohorts. Two Empirical Bayesian methods (Wright & Simon and cyber-T) and BenjaminiHochberg false discovery rate were used to find genes that have significant differential expression (p<0.05) between E15.5 wild-type (n=4) and IUGR samples (n=2). 1770 genes were found to have differential expression in the IUGR placenta cohort. In the embryo, 2039 genes have differential expression in the IUGR embryo cohort. There are 799 genes that are 60  differentially expressed in both tissues, with 84 genes exhibiting > 2-fold difference between WT and IUGR samples. In the placenta, only 16 genes have a > 2-fold difference (Fig. 4.6C, Suppl. Table 2). 42 genes have > 2-fold difference only in the embryo (Fig. 4.6B, Suppl. Table 2), while 26 genes have > 2-fold difference in both the embryo and the placenta (Fig. 4.6A,D, Suppl. Table 2). These genes are involved in transport of proteins or ions, intracellular signalling, cellular processes (differentiation, proliferation, death, and metabolism), anatomical structural development, nucleotide-associated activities, immune or stress response, and cell adhesion (Fig.4.6B-D).  Fig. 4.5 Clustering of WT and IUGR samples. Four independent clusters are illustrated using the hierarchal clustering function on MeV. The embryo samples group significantly from the placental samples and accounts for the majority of the differences between the four groups.  61  Fig. 4.6 Differentially expressed genes with > 2-fold different in IUGR. (A) Venn diagram illustrating the common genes that are differentially expressed in both IUGR embryo and placentae. (B) The functional categories of the 68 genes that are differentially expressed only in the embryo. The categories are based on gene ontology of the biological processes of the genes involved. The unknown category include genes with known molecular function but unknown biological process. (C) The functional categories of the 42 genes differentially expressed only in the placenta. (D) The functional categories of the 26 genes that are differentially expressed in both tissue. The functional categories of genes differentially expressed only in the placenta. IUGR samples size = 2 for each tissue. St. Dev. = Structural development.  62  4.2.6 Imprinted gene expression in IUGR samples The majority of imprinted genes analyzed do not exhibit significant differential expression between WT and IUGR samples. The genes that are differentially expressed in the placenta are Sfmbt2, Phlda2, Cdkn1c, Cobl, Zrsr1, Dlk1, Slc38a4, Slc22a3, and Xlr4c (Table 4.2). In the embryo, H19, Igf2, Dlk1, Pde4d, and Slc38a4 are differentially expressed between WT and IUGR samples. The fold change ratio of IUGR to WT and gene ontology of these imprinted genes are illustrated in Table 4.2.6. Dlk1 exhibits the greatest increase in expression in both IUGR embryos and placentae for imprinted genes. It is also the imprinted gene that has the highest increase in expression in IUGR placentae (Suppl. Table 2). 71 genes previously demonstrated to be imprinted in mice are assayed on Illumina MouseRef8.0. We have hypothesized that imprinted genes as a whole will be more affected in IUGR because many of them have been shown to be essential to embryonic growth and development. Therefore I have conducted a chi-square test to see if imprinted genes are overrepresented in the group of genes that are differentially expressed in IUGR (Table 4.3). Moreover, I have also done the chi-square test on potential candidates that have recently been suggested to be imprinted in the mouse (Gregg et al., 2010). In both categories, imprinted genes are not over-represented out of the differentially expressed genes in IUGR embryo cohort, but the number of imprinted genes that are actually differentially expressed are slightly more than expected in the IUGR placenta cohort. However, the chi-squared test shows that this slight difference in the IUGR placenta cohort is not statistically significant.  63  Table 4.2 Differentially expressed imprinted genes in IUGR samples. The imprinted genes are specified by gene ontologies of biological and molecular function. Tissue Embryo  Placenta  WT Gene mChr Exprs.  IUGR Fold Exprs. change  GO Biological Process Regulate expression of Igf2 Promote growth Transport of amino acid Generate force in smooth muscle Embryonic skeletal development Transport of amino acid Regulation of transcription  GO Molecular Function ncRNA that interacts with DNA Growth factor activity  H19  7  14.99  14.52  0.7214  Igf2  7  15.68  15.28  0.7575  Slc38a4  15  8.340  8.059  0.8662  Pde4d  13  7.982  8.133  1.055  Dlk1  12  13.45  14.07  1.534  Slc38a4  15  12.96  11.34  0.3253  Sfmbt2  2  8.811  8.235  0.6709  Slc22a3  17  8.617  8.126  0.7116  Transport  Transfer of quaternary ammonium group  Xlr4c  X  8.273  8.034  0.8472  Unknown  Unknown  Cobl Zrsr1 Phlda2  11 11 7  7.964 7.942 8.797  7.996 8.002 9.300  1.023 1.043 1.418  Neural tube closure Unknown Glycogen storage  Cdkn1c  7  14.77  15.36  1.509  Cell cycle arrest  Protein interaction Interact with zinc Unknown Cyclin-dependent cell cycle inhibitor  Dlk1  12  10.75  12.25  2.841  Embryonic skeletal development  Uptake amino acid Hydrolase activity Binding to calcium Uptake amino acid Unknown  Binding to calcium  Table 4.3 Chi-square test of known imprinted genes and imprinting candidates. The statistical test compares the number of imprinted genes or candidates that are actually differentially expressed in IUGR samples versus the expected number of imprinted genes or candidates. Confirmed or candidate imprinted genes  Embryo (actual)  Imprinted Genes  5  7.98  9  6.93  Gregg et al. (2010) Candidates  27  28.9  33  25.1  Chi square (p-value)  Embryo (expected) Placenta (actual) Placenta (expected)  0.533  0.208  64  4.3 Discussion 4.3.1 Genome-wide expression analysis using Illumina MouseRef8.0 Beadchip My goals to conduct genome-wide analysis were two-fold: 1) Provide wild-type (WT) control for comparison with IUGR samples and 2) address the issue of variability of gene expression in the embryo and the placenta. Another advantage was that we could use the expression data to uncover unknown genes that may play a role in either the embryo or the placenta at E15.5. We employed an arbitrary cut-off value for a gene as expressed at 9 after variance-stabilized transformation of raw intensity level since the majority of the genes with expression < 9 did not exhibit any ISH staining in E14.5 embryo sections submitted to GenePaint (www.genepaint.org). As expected, the gene expression pattern in the placenta is very different from that of the embryo (Fig.4.1). Several members of the prolactin and cathepsin families are amongst the highest-expressed genes in the placenta. Prolactins are hormones that bind to maternal targets, which result in alteration of maternal physiology (Lin et al., 2000). The cathepsin family mostly functions in the hydrolysis of peptide bonds. These proteases are thought to likely function in placental vasculature remodelling (Simmons et al., 2007; Varanou et al., 2006). These genes are only highly expressed in the placenta and not in the embryo (Suppl. Table 1). Ribosomal genes are the highest-expressed genes in the embryo. The high levels of ribosomal expression suggests that protein anabolism occurs extensively in the embryo, probably due to the metabolic requirements of rapidly dividing cells during the growth phase. These genes also exhibit high expression in the developing placenta. The embryo-specific genes (Fold change > 10) code for actin, myosin, and troponin proteins that compose the cytoskeleton. Some of these molecules like ACTA1 and MYLPF, are important for skeletal development, which only occurs in the embryo (Garner et al., 1989; Wang et al., 2007). Expression pattern of imprinted genes in our 65  array data correlate well with previously known expression levels (Fig. 4.4A) (Schulz et al., 2008).  4.3.2 Variability of gene expression is more significant in the placenta than in the embryo Many studies have indicated that gene expression is not consistent from one biological replicate to another. Pidoux and colleagues (2003) demonstrated that in the placenta there was greater variation of expression between placentae from different individuals than within the same placenta. In our study, the variance observed in both the embryo and the placenta was on the order of 10-2. From the principal component plot, the largest difference (97%) of the WT samples was between the embryo cohorts versus the placental cohort. It was advantageous to see that variation between biological replicates was not as large as qRT-PCR of the Mest samples, since it could mask the difference in expression between normal and IUGR samples. Fig.4.2B shows a positive correlation between expression and variance. This is expected since genes at lower expression level (Expression < 8) were likely not expressed or had a very low level of expression. If the gene was not expressed, then its expression from one biological replicate to another would not alter. Interestingly, the placenta cohort displayed a higher variance even for these genes, indicating that gene expression in the placenta is more variable. Moreover, all of the most variable genes have expression level < 12, indicating that the highest expressed genes do not necessarily have the highest variance (Table 4.1). Though the difference in variances between the embryo and placenta cohort were small, I consistently found the placenta to have expression just slightly more variable than that of the embryo, especially at higher levels of expression (Fig.4.2E-F). This result is in agreement with qRT-PCR data (Fig.3.6B) The function of the placenta can explain the observation of greater variability in gene expression. We have used littermates as biological replicates but even if the littermates are 66  within the same environment, there may be differences in the amount of nutrient exchange between the littermates. Several studies have indicated that blood flow to each conceptus may be different depending on their location in the uterine horn (Coe et al., 2008; Vom Saal and Dhar, 1992). The variation in blood flow may be due to the bi-furicated nature of the blood flow in the uterine artery, which branches out to supply each placenta in the murine uterine horns. Vom Saal and Dhar (1992) have found that the blood pressure supplying the conceptuses at the ends of each uterine horn is higher than the blood pressure supplying those implanted in the middle of the horn. Accordingly, the number of conceptuses and their position in the horn can affect the level of blood each placenta receives causing it to be different. So the placenta of the middle conceptuses may show altered gene expression level as a result of compensation for the lower level of blood supply (Vom Saal and Dhar, 1992). It has been demonstrated that the placenta can respond to the altered intrauterine environment by changing gene expression. For example, gene expression in the placenta is altered in response to hypoxia (Genbacev et al., 1996). Moreover, alteration of solute carriers' gene expression to meet embryonic nutritional demand has been demonstrated in the placentalspecific mouse knockout of Igf2, which codes for an essential embryonic growth factor (Constancia et al., 2005). Therefore, gene expression in the placenta may be altered to meet the needs of the embryo even in regular pregnancy, which can lead to variability in gene expression between littermates. This is supported by our finding that 7/29 of the most variable genes in the placenta have transport function when only 2 genes would be expected given the number of transport genes represented on the array (Table 4.1). The next question to ask is why the embryo does not display the same variability. The major developmental events in the embryo require very specific levels of signalling between different cell types. For example, neural tube closure is governed by specific signalling between 67  neural crest cells and neighbouring cell types. Perturbation of signalling at this stage in development can result in openings along the dorsal midline, which impacts the survival and welfare of the embryo (Copp et al., 2003; Detrait et al., 2005). Differences in gene expression in an embryo can lead to malformation or even death. Due to this sensitivity embryonic gene expression may be more specific than that of the placenta, as the placenta may possibly tolerate volatile gene expression since it will be disposed after birth. It is also possible that the placenta evolved to have more adaptive gene regulatory mechanism to specifically address nutrient demands (Coan et al., 2008; Coan et al., 2010). Genomic imprinting is a mechanism that is believed to have evolved to regulate nutritional demands of the embryo (Wolf and Hager, 2006). In the placenta, there are more genes that have been identified with imprinted expression than in the embryo. Moreover, imprinted gene regulation by histones, which is more volatile than gene regulation by DNA methylation has been suggested to play a more significant role in the placenta than in the embryo (Lewis et al., 2004; Umlauf et al., 2004; Wagschal et al., 2008). We found that 23/71 (33%) imprinted genes have significant difference in variance between placenta and embryo, 15 of which exhibit more variance in the placenta. This signifies that the overall trend of higher placental variance is also seen for imprinted genes (Fig. 4.4). It will be interesting to see if the genes that have high variability will be confirmed with expression profiling of additional WT litters.  4.3.3 Comparison of whole-genome expression between normal C57BL/6J and IUGR samples Performing hemiovariectomy on C57BL/6J females resulted in all the conceptuses being implanted in one horn (Fig. 3.7A). The decreased size of placenta created a crowding effect 68  where some of the embryos had lower embryonic weights than embryos in a normal pregnancy. There were only two E15.5 male embryos from the crowded horn that had weights in the bottom 5th percentile. I believed that all the embryos were subjected to the crowding effect, thus we included the last E15.5 male in our second microarray. Plate-to-plate variation in older microarray studies created problems for researchers since it was difficult to assess if the differences observed between normal and diseased samples were real or due to technical artifact. It was suggested that all microarray studies use three to five biological replicates for each cohort (http://discover.nci.nih.gov/microarrayAnalysis). Modern array equipment the plate-to-plate variation has been reduced. In our study, we conducted two Illumina Beadchips, each contained eight samples. All the WT samples were on one Beadchip whilst all the crowded samples were on the second Beadchip. However, though the normalization procedures used by the Flexarray software enables me to directly compare the expression levels between the WT and crowded samples, there may still be chip-to-chip variation that can influence the current findings presented in this thesis. Principal component analysis was used to quickly visualize clustering of the samples (Suppl. Fig. 2). Four clusters were observed: WT embryo, WT placenta, crowded embryo, and crowded placenta. One difficulty encountered was that one of the crowded placental sample (LLE052) actually clustered with the crowded-embryo samples. This was perhaps due misplacement or labeling error. Fortunately LLE052 was the male placentae whose embryonic weight was not in the bottom 5th percentile. Interestingly, the corresponding male embryo (LLE049) also clustered slightly further from the two IUGR embryos. Since the purpose of this study was to look for genes differentially expressed in IUGR, both samples LLE049 and LLE052 were not used in subsequent statistical analyses for assessing significance of differential expression between WT and IUGR samples. 69  One statistical method used to determine the significance of differential expression between WT and IUGR samples may return a list of genes that differs from the list of differentially-expressed genes generated when another statistical method was used. The best approach to find genes that would be more representative of real differential expression was to use more than one statistical method. Thus the data was subjected to the Empirical Baysian tests EB (Wright & Simon) and cyber-T tests, two of the most common tests used on microarray data (Murie et al., 2009). The Significance Analysis of Microarray (SAM) was also conducted but this test did not return any genes that showed differential expression. This was most likely due to the small sample size of the IUGR cohort (n=2) which does not meet the stringency of SAM. Nevertheless, it was still worthwhile to identify candidate genes so when more samples are used, I could compare those findings that would be analyzed using SAM with the ones I have identified in the current study. Overlapping the results of the two tests showed 1770 genes to be differentially expressed in the IUGR placentae, while 2039 genes were differentially expressed in IUGR embryos. The maximum fold difference was slightly less than 6 in the embryo, and less than 3.5 in the placenta. The majority of the genes exhibited fold difference of less than 1.5. Considering more genes are expressed (Expression > 9) in the placenta than in the embryo (3595 versus 3569), it was surprising to find more genes to be differentially expressed in the embryo. Looking at the clustering of the IUGR placental cohort versus the IUGR embryo cohort provided some insight. The two IUGR placental samples cluster closer than the two IUGR embryos, which indicated that the overall variance in the IUGR placental cohort was less. With a sample size of two, it is impossible to determine if one of the IUGR embryo sample used was an outlier. If an outlier was present, then the accuracy of the differential expression data would decrease. Indeed the limited samples size was a major issue in this study. Nonetheless, the goal of the microarray study was 70  to find candidates that may play a role in IUGR. All of the interesting candidates would need to be further re-evaluated by qRT-PCR, or even quantitative sequencing. The sample size would need to be increased during these validations. To narrow down the candidate genes that are affected in IUGR, an arbitrary 2-fold cutoff was employed. This reduced the number of genes altered in IUGR to less than 100. The major categories that are implicated were transport of substances, cellular metabolism, signalling, and nucleotide metabolism (Fig. 4.6). It was found that 12% of genes (5/42) exhibit differential expression in the IUGR placenta cohort are involved in transport, whereas 11% of differentially expressed genes in the IUGR embryo cohort are involved in transport. This was not surprising since transport molecules are the most affected because at this stage of development, transport genes encompass the largest class of genes that are expressed in our WT microarray data (13% in embryo and 15% in the placenta). Though transport genes were not particularly enriched, there was still a candidate that was of particularly interest: Npc2. Npc2 is highly expressed in both the embryo and the placenta and exhibits a 2-fold under-expression in IUGR embryos, and a 2.7-fold under-expression in IUGR placentae. The NPC2 mutation in humans represents the cause of 5% of the lipid storage disease known as Niemann-Pick disease (Millat et al., 2001). NPC2 functions together with NPC1 to facilitate intracellular transport of lipids in lysosomes (Sleat et al., 2004). It has also been documented that IUGR is a phenotype in a subset of individuals who experience fetal onset of the disease (Spiegel et al., 2009). A couple of interesting candidates in IUGR embryos function in cellular differentiation: Gap43 and Stmn1. Gap43 exhibits a 2.3-fold down regulation in IUGR embryos. It codes for a growth-associated protein whose main function is the development of the optic nerve (Strittmatter et al., 1995). Stmn1 exhibits a 6-fold under-expression in IUGR embryo cohort. It is found that down-regulation of Stmn1 affects cell motility (Jin et al., 2004; Ozon et al., 2002). It 71  has been suggested that the expression of Stmn1 is associated with early migration of trophoblast cells and differentiation into syncytiotrophoblasts (Yoshie et al., 2008). Two collagen-encoding genes, Col1a1 and Col3a1, are over- and under-expressed in IUGR embryos, respectively. These collagen proteins contribute to structural integrity of the extracellular matrix and both contribute to blood vessel development (Liu et al., 1997; Rahkonen et al., 2004). A mutation in human COL3A1 has been linked to aortic rupture. Homozygous mouse mutants that survive have reduced body size and most die within six months due to rupturing of blood vessels (Liu et al., 1997). Col1a1 also functions in bone development and mutations in COL1A1 are thought to be the main cause of Osteogenesis Imperfecta in humans (OMIM 166200, OMIM 166210). None of the non-imprinted genes found to be differentially expressed in human studies comparing normal versus IUGR term placentae were significantly differentially expressed in this study. A few genes (IGFBP1, IGF-I, CRH, and LEP) were consistently shown to have different expression level in human IUGR placentae (Economides et al., 1989; Lee et al., 2010; McMinn et al., 2006; Struwe et al., 2010). Other Igfbp genes (Igfbp3 and Igfbp4) do exhibit differential expression in IUGR embryos but do not meet the 2-fold cut-off. In the IUGR placental cohort, Prl8a9 shows a 3.3-fold down-regulation when compared to WT cohort. Expression of Prl8a9 is the highest in the spongiotrophoblast from E12.5 until birth (Simmons et al., 2008). Not much is known about the specific function of Prl8a9, but it is likely to aid in the invasion of maternal decidua since it is highly expressed in the spongiotrophoblast, the placental layer suggested to house the precursors of the invasive glycogen cells in the murine placenta (Georgiades et al., 2002; Rossant and Cross, 2001). On a separate note, this gene is murine-specific so it may not play a very significant role in human IUGR. 72  Three other genes that are of interest are Vhl (FC = 0.3), Hspa5 (FC = 0.4), and Gpx5 (FC = 2.2). Gene ontologies indicate that they are involved in stress response. Vhl has been found to code for a protein product that regulates oxygen-dependent degradation of hypoxia-induced factor 1 (HIF-1) (Maxwell et al., 1999). Absence of the VHL protein results in defective vasculature in the labyrinthine of the murine placenta (Gnarra et al., 1997; Tang et al., 2006). The connection of IUGR to regulators of placental vasculature indicates that impaired vasculogenesis may be involved in IUGR. HSPA5 is an endoplasmic reticulum (ER) chaperone protein that is involved in signalling pathways activated upon ER stress (Kaufman, 1999). Hspa5 exhibits an increase in expression upon ER stress and has been suggested to attenuate apoptosis (Oyadomari et al., 2002). GPX5 is a gluthathione peroxidase that is implicated in oxidative stress response (MGI Gene Ontology J:72247). Researchers have found that disrupting Gpx5 in the mouse results in abnormal sperm production. The abnormal sperm exhibit an increase in DNA fragmentation upon treatment with hydrogen peroxide, which indicates that the null mutants have problems when responding to oxidative stress (Chabory et al., 2009). Though some of the candidates discussed here have functions that may contribute to IUGR, a very important issue must be addressed before we can conclude if these genes are truly candidates for IUGR. Pidoux et al. (2003) has found that there is more variation in expression between different human placenta than within the same placenta. In the mouse system littermates have been shown to have relatively low variance in expression (Fig. 4.2). The best approach to solve this problem is to conduct expression profiling of additional WT samples from multiple different litters than the ones used in our current study. Genes that are found to be differentially expressed between WT litters can then be cross-referenced to the list of IUGR candidates and be excluded from the candidate list.  73  In addition, more IUGR samples are also needed to provide a samples size that is much greater than two as in the current study to perform the Significance Analysis of Microarray (SAM) to find candidate IUGR genes. I propose to start with 8 biological replicates for each cohort (32 samples in four categories: WT embryo and placenta, IUGR embryo and placenta) and analyze those first to see if there are any commonalities between those findings and the results presented in this thesis. Once those candidates have been identified then I can pursue the interesting candidates using qRT-PCR, which enables me to analyze more samples due to the lower cost. Only after all these steps are taken can we present the final list of genes differentially expressed in IUGR. If structural developmental genes and stress response genes still play a role in IUGR embryo and placenta, then it suggests that a different mechanism may be contributing to IUGR in the two tissues. It is not clear whether or not the changes in expression level of these genes are causing the IUGR or an adaptive response to the stress. It may be a combination of both since our study only takes a "snapshot" of gene expression at a specific stage in development.  4.3.4 Imprinted candidates of IUGR Recently the topic of epigenetics in IUGR has been explored in several studies (Apostolidou et al., 2007; Bourque et al., 2010; Diplas et al., 2009; Guo et al., 2008; McMinn et al., 2006; Penaherrera et al., 2010). Given the involvement of genomic imprinting in embryonic growth and development, imprinted genes are thought to be the most likely candidates that are affected in IUGR. Indeed several studies have looked at differential expression of imprinted genes in human IUGR placentae (Apostolidou et al., 2007; Bourque et al., 2010; Diplas et al., 2009; Guo et al., 2008; McMinn et al., 2006). We have found four imprinted genes to exhibit significant fold change in the embryo of the crowded IUGR model. H19 (FC = 0.72), Igf2 (FC = 74  0.76), and Slc38a4 (FC = 0.87) are under-expressed in IUGR embryos and Dlk1 (FC = 1.5) is over-expressed. IGF2 has been repeatedly reported to be under-expressed and hypomethylated in human IUGR (Abu-Amero et al., 1998; Antonazzo et al., 2008; Bourque et al., 2010). Igf2 codes for an insulin-like growth factor that is important in regulating embryonic growth. Mouse KO of Igf2 (Igf2+/-) has been found to exhibit a 40% reduction in birth weight (DeChiara et al., 1990). The function of Igf2 has largely been explored in the placenta. The placental-specific knockout of Igf2 (Igf2 P0+/-) display a milder phenotype than Igf2+/-, the null mutant of Igf2 since the gene is paternally expressed. Comparison of these two mouse mutants has increased the understanding of the function of the fetal Igf2 and placental Igf2. The fetal Igf2 functions in labyrinthine cells from both the trophoblast and epiblast lineage (fetal endothelial cells), whilst the placental Igf2 only functions in the labyrinthine cells from the trophoblast lineage (Constancia et al., 2000; Redline et al., 1993). Igf2 P0+/- is constructed through deleting the placental-specific promoter of Igf2. Constancia and colleagues (2005) has also demonstrated that in Igf2 P0+/-, Slc38a4 expression is up-regulated in the placenta to increase amino acid transport which they suggest is an adaptive response to the need to maintain nutrient transport to the embryos.  Accordingly,  since Igf2 is under-expressed in IUGR, then Slc38a4 should increase. This differs from the decreased expression observed for both IUGR embryo and placenta in this study. Though Slc38a4 shows differential expression, there is a 10% reduction in IUGR embryos. This result needs to be verified with increased sample size. It is also possible that Igf2 may not affect Slc38a4 in the embryo, or affect Slc38a4 expression in a different manner. Though H19 has not shown differential expression in IUGR in previous studies, it is found to be biallelically expressed indicating a loss of imprinting (LOI) in IUGR placentae (Diplas et al., 2009; Guo et al., 2008). H19 codes for a non-coding RNA that directly regulates 75  the expression of Igf2 through epigenetic mechanisms (Drewell et al., 2002; Leighton et al., 1995). Maternally-inherited double KO of H19 and Igf2 has shown a compounded effect that leads to embryonic lethality (Eggenschwiler et al., 1997). Furthermore, targeted disruption to H19 only affects embryonic and placental growth. Depending on whether the mutant is inherited maternally or paternally, overgrowth or IUGR will occur (Drewell et al., 2000; Thorvaldsen et al., 1998). The inter-relatedness of H19, Igf2, and Slc38a4 supports the idea of an imprinted gene network (IGN) (Arima et al., 2005; Varrault et al., 2006). Though the fold change difference of these genes in IUGR is not greater than 2-fold, the compound effect of irregular gene expression can lead to the development of IUGR. Human studies cannot look at differential gene expression in IUGR fetus except using cord blood, so our study is the first study to look at the genes that may be altered in expression using whole IUGR embryos. As mentioned, the genes that contribute to IUGR will not necessarily coincide with those that are differentially expressed in the placenta. For imprinted genes, only Slc38a4 and Dlk1 exhibit differential expression in both the embryo and the placenta. Igf2 in our IUGR placental cohort is over-expressed by 1.2-fold, though the difference was not significant (p < 0.5). Dlk1 exhibits increased expression in both IUGR embryo (FC = 1.5) and placenta (FC > 2.8). Dlk1 is normally highly expressed in both the embryo and the placenta (Schmidt et al., 2000; Yevtodiyenko and Schmidt, 2006). Its expression in the embryo is high at E12.5 in most mesodermally-derived tissues, as well as in the pituitary, adrenal gland, and pancreas. This trend changes to high expression only in the pituitary, adrenal gland, and skeletal muscle by E16.5. In the placenta, Dlk1 is expressed in endothelial linings of labyrinthine vessels (Yevtodiyenko and Schmidt, 2006). The expression pattern signifies its role in embryonic development, as well as its role in nutrient exchange in the placenta. Dlk1-/- mouse mutants exhibit IUGR, perinatal 76  mortality, and skeletal defects involving the rib. Dlk1 is expressed from the paternal allele, yet paternal inheritance of the null mutation only results in postnatal growth restriction (Moon et al., 2002). This suggests that both the parental alleles of Dlk1 are essential for normal development in the mouse. DLK1 is one of the genes from the imprinting region (IG-DMR) on human chromosome 14. Uniparental disomy of chromosome 14 (UPD14), which exhibit various congenital phenotypes including growth restriction, is attributed to aberration of epigenetic features of IG-DMR (Kagami et al., 2008; Temple et al., 2007). Recently DLK1 has been discovered to exhibit loss of imprinting (LOI) in Silver-Russell Syndrome or IUGR (Azzi et al., 2009; Guo et al., 2008). The over-expression of Dlk1 in our IUGR samples may occur through LOI. This could be evaluated in the mouse by doing bisulphite sequencing on F1 offspring of C57BL6/J and Mus musculus castaneus mice for instance. Two other genes were found to be under-expressed by greater than 1.4-fold in the IUGR placental cohort. The gene Sfmbt2 (FC = 0.67) is a newer addition to the imprinting family in the mouse. It is the first gene discovered in a known imprinted region on mouse chromosome 2 (mChr2); the region is considered imprinted since maternal duplication of the proximal region result in early embryonic lethality (Cattanach et al., 2004; Kuzmin et al., 2008). Not much is known about the function of this gene except it is part of the polycomb group and is hypothesized to be important in maintenance of trophoblast stem cells during early embryonic development (Kuzmin et al., 2008). Slc22a3 (FC = 0.71) belongs to the organic cation transporter family. Along with Igf2r and Slc22a2, Slc22a3 is regulated in cis by the Air noncoding RNA in the murine placenta (Nagano et al., 2008). The organic cationic family is important in controlling the amount of neurotransmitters like norepinephrine or epinephrine in the extracellular matrix. No obvious defects are observed in Slc22a3-null mice except for reduced activity of this uptake system in the adult heart as well as in the embryos (Zwart et al., 77  2001). The uptake system also exhibits high activity in the placenta, and Slc22a3 shows high expression in the labyrinthine (Verhaagh et al., 1999; Zwart et al., 2001). It is suggested that redundancy of other transporters in the family (Slc22a1 and Slc22a2) is the reason why reduced uptake activity is not observed in other organs of Slc22a3-null mice (Verhaagh et al., 1999; Zwart et al., 2001). SLC22A3 expression is also observed in first-trimester and term human placenta (Verhaagh et al., 1999). No association between embryonic growth has been demonstrated with Slc22a3, but considering its high expression in the labyrinthine, it is likely to function in nutrient exchange. Slc38a4 is under-expressed in IUGR placentae by 3-fold when compared to WT. As previously mentioned this is different from my expectation since in the IUGR Igf P0+/-, Slc38a4 was observed to increase at E16 (Constancia et al., 2005). Another study has also found an inverse relationship between birth weight and this System A amino acid transporter activity in human (Godfrey et al., 1998). SLC38A4 is an essential transporter responsible for amino acid transport that is required since amino acids account for 50% of all nitrogen and carbon required for fetal growth (Fowden and Forhead, 2004). Considering the function of this protein, it is possible to interpret that the decreased expression is contributing to growth restriction. Interestingly, Slc38a4 expression in the Mest KO model, though statistically insignificant, do exhibit a decreased expression in the Mest+/- cohort (Fig. 3.7). In contrast to the Igf2 P0+/- mice that do not exhibit IUGR until E16, our crowded embryos have already begun to be growth restricted by E14.5 (Suppl. Fig. 1A). It is possible that earlier on there may have been an adaptive increase in expression of Slc38a4 to compensate for the reduced blood supply of the crowded IUGR placenta. Constancia et al. (2005) has documented that by E19 of Igf2 P0+é-, Slc38a4 expression is not different. This does not completely explain the decrease in Slc38a4  78  expression but the discovery of differential expression of this gene does signify that it may be an important regulator of growth. Cdkn1c (FC = 1.5) and Phlda2 (FC = 1.4) display over-expression in the IUGR placentae. CDKN1C is involved in the overgrowth syndrome known as Beckwith-Wiedemann Syndrome (BWS) (OMIM 130650). The function of CDKN1C is to inhibit the cell cycle so that when it is down regulated abnormal overgrowth of different tissues will result (Chellappan et al., 1998; Zhang et al., 1997). Mouse mutants homozygous for null mutation of the gene exhibit growth restriction as well as some phenotypes characteristic of BWS (Zhang et al., 1997). However, other groups have not observed the same BWS-related phenotypes (Takahashi and Nakayama, 2000). Only a subset of BWS patients have maternally inherited mutations in CDKN1C, indicating the cause of BWS is multifactorial (Lam et al., 1999). Additionally several groups of researchers have found an association between CDKN1C and preeclampsia (Enquobahrie et al., 2008; Kanayama et al., 2002; Romanelli et al., 2009). Preeclampsia is a maternal hypertensive disorder that if present in the mother, the fetus has a higher probability of having IUGR. PHLDA2 is the only imprinted gene that has consistently been found to be differentially expressed in human IUGR placentae (Apostolidou et al., 2007; Diplas et al., 2009; McMinn et al., 2006). The function of PHLDA2 has been recently suggested to regulate glycogen storage in the mouse placenta (Tunster et al., 2010). Glycogen-containing cells are important in late mouse gestation for the expansion of the labyrinthine into maternal decidua, an important event for maximization of nutrient transfer (Georgiades et al., 2002). Mouse KO of Phlda2 have placentomegaly as well as a 13% decrease in fetal size (Frank et al., 2002). The over-expression observed in our IUGR placental samples also agreed with the pattern of these studies (Fig. 3.7 and Table 4.2.5). McMinn and colleagues (2006) have suggested under-perfusion in the placenta 79  leading to IUGR may induce activation of genes to restrict placental growth to compensate for the lack of blood. Even though all of the data we have obtained regarding differential expression needs to be verified, our differentially expressed imprinted genes correlate well with their roles in IUGR. The candidate that is thought to play the largest role in IUGR, Igf2, did not appear altered in our data. The one imprinted gene that warrants further functional studies to identify its role in the placenta is Sfmbt2 since its function is unknown. Also further analysis of Phlda2 in human IUGR samples may prove to be fruitful as a good candidate for screening for IUGR since it has repeatedly been reported to be differentially expressed. Though several of these imprinted genes are interesting candidates for IUGR, our main purpose was to evaluate the role of imprinted genes as a whole in IUGR. We hypothesized that imprinted genes would be over-represented in those genes that are differentially expressed in IUGR. We approached this question by conducting a chi-square test which compared the number of imprinted genes that were found to be differentially expressed versus the expected number of imprinted genes that should be differentially expressed based on the percentage of imprinted genes assayed on the Illumina array. Aside from the genes that were known to be imprinted, we included the imprinting candidates that came out of the recent research from Gregg et al. (2010). Based on the chi-square test we could not conclude that imprinted genes are over-represented (Table 4.3). However, there was a trend that suggested that imprinted genes may be more important in the placenta as the actual number of differentially expressed imprinted genes were more than expected. This agreed with the function of the placenta as a tissue that mediates embryonic growth, as well as the observation that many imprinted genes only demonstrate parent-of-origin expression in the placenta (Lewis et al., 2006; Mizuno et al., 2002).  80  Chapter 5: General discussion The topic of intrauterine growth restriction has been explored for decades due to its relationship with many perinatal diseases and mortality. The recent interest in exploring how epigenetics contributes to pregnancy complications has prompted many researchers to assess the extent of epigenetic aberration, which can lead to changes in gene expression in cases of IUGR (Bourque et al., 2010; Enquobahrie et al., 2008; Guo et al., 2008). Our work has looked at genome-wide expression, with particular emphasis on imprinted genes, and how it is affected in IUGR using the mouse as a model system. Besides studying differential expression in IUGR placenta, our work is the first to explore differential expression in the embryo during IUGR. There are many advantages to studying gene expression in the murine system. The major cause of IUGR has been attributed to placental dysfunction. In particular, vascular problems in the placenta can lead to disruption in blood flow, subsequently affecting fetal-maternal nutrient and gas exchange (Cetin and Alvino, 2009; Cox and Marton, 2009; Jansson et al., 1993; Jansson et al., 2002; Roos et al., 2007). Compared to human IUGR studies where the samples have a variety of causes resulting in IUGR, we can know the cause of IUGR in the mouse. This thesis has outlined a surgical procedure where blood flow to some conceptuses in the uterus is restricted, resulting in IUGR. Another benefit with using the mouse system is that variation due to heterogeneity in the population can be avoided. Many lab strains of mice are inbred, which means their genetic composition will be almost identical, therefore minimizing variation due to differences in their genetic code. This is particularly useful in gene expression studies since smaller sample size can have the same power as human studies that require a large sample size just to be representative of the genetically heterogenous population. At least one study has indicated large variation in gene expression between placental samples from different individuals 81  (Pidoux et al., 2004). Human placental RNA samples are also subject to rapid RNA degradation due to indeterminate amount of time before RNA preservation procedures can be done postpartum (Avila et al., 2010). This can create false positives for differential expression in human IUGR studies. Studies involving human placentae also have sampling issues where they can only take a "core" of the placenta for RNA extraction, which may not be representative of the condition of the entire human placenta. In our study we extract RNA from the entire mouse placenta without any sampling bias. Another advantage of studying IUGR in the mouse is that we can explore how IUGR impacts embryonic gene expression, whereas human studies cannot due to ethical concerns. Diagnoses of pregnancy complications have often divided cases into early-onset or lateonset. In general, the earlier the onset , the more severe the maternal and/or fetal phenotypes will be. Early obstruction in placental function often results in more severe fetal phenotype (Cox and Marton, 2009). The majority of these cases will also result in the appearance of additional phenotypes other than IUGR (Cox and Marton, 2009; Genbacev et al., 1996). Since the focus of our study is on IUGR specifically and not in conjunction with other malformations, we have focused on late-onset IUGR. The final growth phase of human fetal development occurs in the last trimester (>27 weeks), but late-onset IUGR is defined as after 32 weeks gestation. Mouse gestation differs from human gestation in that mouse gestation is divided into two terms, whilst in humans there are three terms. Even so the mouse system still has a comparable growth phase that begins around E14.5. Similarly mouse mutants that exhibit only IUGR, the phenotype is often not observed until E15.5 in mouse gestation (Constancia et al., 2002; Lefebvre et al., 1998). Gene expression profiles in IUGR studies may provide insight to the factors involved in the regulation of placental function and embryonic growth. In particular, imprinted genes are 82  shown to be crucial to proper embryonic growth and development. Growth-related disorders such as Silver-Russell Syndrome (SRS) and Beckwith-Wiedemann Syndrome are linked to mutations or epigenetic aberration of several imprinted genes (Frost and Moore, 2010; Lim and Ferguson-Smith, 2010). Some of these same imprinted genes have also been identified to be differentially expressed in human IUGR studies (Abu-Amero et al., 1998; Antonazzo et al., 2008; Bourque et al., 2010; Guo et al., 2008). These observations suggest that some of these imprinted genes may work in concert to regulate development (Arima et al., 2005; Gabory et al., 2009; Varrault et al., 2006). Moreover, differentially-expressed genes implicated in IUGR can potentially be used in the diagnosis of IUGR, possibly through procedures similar to the Triple Screen Test (http://www.sogc.org/health/pregnancy-prenatal_e.asp#triple). Earlier monitoring and intervention in complicated pregnancies have been successful at alleviating disease symptoms and preventing mortality, which in turn reduce costs to our healthcare system in the long run. The next section summarizes the findings of our study of IUGR in the mouse. Three proposed models of IUGR were explored in the hope of identifying genes that contribute to the etiology of IUGR, but only samples from the surgical model are used in the microarray screen to identify candidates that contribute to IUGR in the murine embryo and placenta. This study has identified non-imprinted IUGR candidate genes and discussed the relevance of their function to embryonic growth and development. Finally, we explored in depth the involvement of genomic imprinting to IUGR.  5.1 Summary of results The grant this project was funded under hypothesizes that epigenetic abnormalities in the placenta contribute to pregnancy complications (preeclampsia and IUGR). The original 83  involvement of our lab in the grant was to study the function in vivo of the gene candidates pulled out from array-based screens. However, we decided to go straight ahead to study the relationship of IUGR and imprinting in the mouse model. The basis for my project stemmed from the idea of the imprinted gene network (IGN) first proposed in Arima et al. (2005). An actual network was drawn out through meta-analysis of co-expression of mouse genes using publicly available microarray data (Varrault et al., 2006). This proposed imprinted gene network included 15 imprinted genes: Gnas, Dcn, H19, Igf2, Igf2r, Plagl1, Sgce, Cdkn1c, Mest, Ndn, Peg3, Gatm, Grb10, Meg3, and Dlk1. Varrault et al. (2006) suggested these genes may function in a network to regulate embryonic growth and development. Subsequently we decided to explore imprinted gene expression in three mouse models of IUGR: one non-imprinted mouse knockout Mmp2, Mest mouse KO, and surgically-induced model. Mmp2-/- was suggested by the authors to exhibit IUGR, but no birthweight data was shown (Itoh et al., 1997). Therefore, I decided to characterize placental phenotype and weight differences in Mmp2-/- conceptuses. There was no significant difference in weight between Mmp2+/+ and Mmp2-/- embryos and placenta, nor was there any obvious morphological difference in the null placenta. This suggested that we could not use Mmp2 as a model for IUGR. Since the Mest mouse knock-out was previously demonstrated to exhibit IUGR, we started looking at differential expression of some imprinting candidates in Mest+/- placental samples. However, we did not find any significant differences between WT and IUGR placental samples. Moreover we observed significant variation in gene expression between littermates. Variation in gene expression was expected, but not to the extent that we had observed. In accordance to theories suggesting the adaptive trait of the placenta, we proposed to that there could be more variation in gene expression in the placenta than in the embryo (Coan et al., 2008; Coan et al., 2010; Constancia et al., 2005). Illumina expression profiling (MouseRef8.0) enabled 84  us to assess expression variation in four biological replicates of WT inbred C57BL/6J embryos and placentae. Through various techniques that binned genes by expression level, we were able to demonstrate that gene expression variation was slightly higher in the placenta than in the embryo. Even though the variance in expression was higher in the placenta, the embryonic and placental samples still formed independent clusters. The variance observed was also on the order of 10-2 by Illumina, which indicated that we could still look for candidates of IUGR by microarray. Therefore, we decided to induce IUGR in C57BL/6J females via hemiovariectomy. Weight measurements of this surgical model indicated that some of the embryos did exhibit lower weight than embryos from WT females, indicating the presence of IUGR. Subsequently, we conducted analysis of differential gene expression by Illumina using IUGR samples from the surgical model. Comparison between WT and IUGR cohort indicated that transport genes were the most affected in both the embryo and the placenta. This was expected because at this developmental stage, nutrient transport activity is essential for growth. There were also differences in the gene functions between the genes found to be differentially-expressed in IUGR embryo versus those found differentially-expressed in IUGR placenta. In the embryo, developmental and cellular differentiation genes were found to be differentially expressed, while more stress response genes were differentially expressed in the placenta. The involvement of these genes correlated well with the function of the embryo and the placenta. Development of organs through cellular differentiation occurred mainly in the embryo. Several studies have demonstrated that stress response contributes to pre-eclampsia and IUGR (Burton et al., 2009). Four and six imprinted candidates of IUGR were identified in the embryo and the placenta, respectively. We could not compare the results from IUGR embryos with other human IUGR studies since there was no gene expression analyses on human fetus affected only by 85  IUGR. Nevertheless the genes that were found to be differentially expressed in the embryo (H19, Igf2, Slc38a4, and Dlk1) have all been demonstrated to be involved in embryonic growth (Antonazzo et al., 2008; Bliek et al., 2006; Bourque et al., 2010; Constancia et al., 2005; Moon et al., 2002) In the placenta, Phlda2 emerged as a promising candidate that may be used as a diagnostic tool in the future as it was repeatedly reported to be over-expressed in IUGR placentae in human studies (Apostolidou et al., 2007; Diplas et al., 2009; McMinn et al., 2006). The polycomb group gene Sfmbt2 was found to be under-expressed by 1.5-fold in IUGR placenta and had not been previously associated with IUGR. This gene may be an interesting candidate to study for growth-related phenotypes since its function has not been characterized. Verifications of the IUGR candidates still need to be completed via qRT-PCR. The IUGR sample size can be increased with three additional female IUGR samples. Hopefully with increased sample size, the natural variation between biological replicates will be smaller, which may allow us to detect even small changes between WT and IUGR samples. In contrast to our hypothesis that genomic imprinting is the most important group of genes involved in IUGR, we found that imprinted genes were not over-represented amongst those genes found to be differentially expressed in IUGR for both the embryo and the placenta.  5.2 Future directions Immediate experiments can be done to evaluate the onset of IUGR and the extent of growth restriction in the surgical model used here. This may prove to be useful in separating cause from consequence. Assuming that we can pinpoint the onset of IUGR, we can conduct another microarray analysis to see which genes are differentially expressed between WT and IUGR samples. Then we can cross-reference to our list of candidate genes to see if there is any overlap. Genes that are found on both lists may likely be the cause of the IUGR phenotype in the 86  surgical model. Morphometric and blood flow assessment of the IUGR placentae in the surgical models can aid in understanding the physiological cause of IUGR. Moreover, assessing the genes we have already found to be differentially expressed at E15.5 just prior to birth (E18.5) may further identify those genes that are responding to IUGR. Constancia and colleagues (2005) have compared the amount of nutrient transfer between two embryonic stages (E15 and E19) in the placental KO of Igf2. They wanted to know what is contributing to the observed increase in placental efficiency, which they suggest is an adaptive response to meet the nutrient demands of the embryo in Igf2 P0+é- (Constancia et al., 2005). The genes that we may identify to be responding to E18.5 may also be involved in adapting to adverse developmental conditions, though we will need to know the gene functions relatively well to make that supposition. We can also design experiments to assess the gene function of some of the candidates we have found. For example, Sfmbt2 currently has not been mutated in mouse, thus we could knockout the gene and see if there are any placental and/or embryonic abnormalities. We can also determine some of the general regulators of IUGR by assessing expression difference of our IUGR genes in other mouse models of IUGR. We can begin with looking at differential expression in the Mest knockout. To address the issue of expression variability due to heterogeneity of the littermates, we can increase the sample size or breed the KO mice back onto the inbred C57BL/6J background. Lastly, we can explore the molecular mechanism that causes differential expression of imprinted genes in IUGR. As an example, we can determine if the observed increase in Dlk1 expression is due to it being biallelically expressed or due to an increase in monoallelic expression. This can be examined by inducing IUGR using C57BL/6J x Mus musculus castaneus females and look to see if both alleles are present in polymorphisms within the gene.  87  References Abu-Amero, S., Monk, D., Frost, J., Preece, M., Stanier, P., and Moore, G.E. (2008). The genetic aetiology of Silver-Russell syndrome. J Med Genet 45, 193-199. Abu-Amero, S.N., Ali, Z., Bennett, P., Vaughan, J.I., and Moore, G.E. (1998). Expression of the insulin-like growth factors and their receptors in term placentas: a comparison between normal and IUGR births. Mol Reprod Dev 49, 229-235. Alexander, C.M., Hansell, E.J., Behrendtsen, O., Flannery, M.L., Kishnani, N.S., Hawkes, S.P., and Werb, Z. (1996). Expression and function of matrix metalloproteinases and their inhibitors at the maternal-embryonic boundary during mouse embryo implantation. Development 122, 17231736. Andersen, L.L., Poulsen, O.M., and Hau, J. (1988). Monitoring of murine feto-placental growth by pregnancy-specific serum proteins in an experimentally induced fetal growth retardation model. In Vivo 2, 367-370. Antonazzo, P., Alvino, G., Cozzi, V., Grati, F.R., Tabano, S., Sirchia, S., Miozzo, M., and Cetin, I. (2008). Placental IGF2 expression in normal and intrauterine growth restricted (IUGR) pregnancies. Placenta 29, 99-101. Apostolidou, S., Abu-Amero, S., O'Donoghue, K., Frost, J., Olafsdottir, O., Chavele, K.M., Whittaker, J.C., Loughna, P., Stanier, P., and Moore, G.E. (2007). Elevated placental expression of the imprinted PHLDA2 gene is associated with low birth weight. J Mol Med 85, 379-387. Arima, T., Kamikihara, T., Hayashida, T., Kato, K., Inoue, T., Shirayoshi, Y., Oshimura, M., Soejima, H., Mukai, T., and Wake, N. (2005). ZAC, LIT1 (KCNQ1OT1) and p57KIP2 (CDKN1C) are in an imprinted gene network that may play a role in Beckwith-Wiedemann syndrome. Nucleic Acids Res 33, 2650-2660. Avila, L., Yuen, R.K., Diego-Alvarez, D., Penaherrera, M.S., Jiang, R., and Robinson, W.P. (2010). Evaluating DNA methylation and gene expression variability in the human term placenta. Placenta 31, 1070-1077. Azzi, S., Rossignol, S., Steunou, V., Sas, T., Thibaud, N., Danton, F., Le Jule, M., Heinrichs, C., Cabrol, S., Gicquel, C., et al. (2009). Multilocus methylation analysis in a large cohort of 11p15related foetal growth disorders (Russell Silver and Beckwith Wiedemann syndromes) reveals simultaneous loss of methylation at paternal and maternal imprinted loci. Hum Mol Genet 18, 4724-4733. Baladron, V., Ruiz-Hidalgo, M.J., Nueda, M.L., Diaz-Guerra, M.J., Garcia-Ramirez, J.J., Bonvini, E., Gubina, E., and Laborda, J. (2005). dlk acts as a negative regulator of Notch1 activation through interactions with specific EGF-like repeats. Exp Cell Res 303, 343-359. BC Vital Statistics Agency. Selected Vital Statistics and Health Status Indicators, 133rd Annual Report. 2004. Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Stat Soc B 57, 289-300. 88  Benyo, D.F., Miles, T.M., and Conrad, K.P. (1997). Hypoxia stimulates cytokine production by villous explants from the human placenta. J Clin Endocrinol Metab 82, 1582-1588. Bischof, P., and Campana, A. (2000). Molecular mediators of implantation. Baillieres Best Pract Res Clin Obstet Gynaecol 14, 801-814. Bliek, J., Maas, S.M., Ruijter, J.M., Hennekam, R.C., Alders, M., Westerveld, A., and Mannens, M.M. (2001). Increased tumour risk for BWS patients correlates with aberrant H19 and not KCNQ1OT1 methylation: occurrence of KCNQ1OT1 hypomethylation in familial cases of BWS. Hum Mol Genet 10, 467-476. Bliek, J., Terhal, P., van den Bogaard, M.J., Maas, S., Hamel, B., Salieb-Beugelaar, G., Simon, M., Letteboer, T., van der Smagt, J., Kroes, H., et al. (2006). Hypomethylation of the H19 gene causes not only Silver-Russell syndrome (SRS) but also isolated asymmetry or an SRS-like phenotype. Am J Hum Genet 78, 604-614. Bliek, J., Verde, G., Callaway, J., Maas, S.M., De Crescenzo, A., Sparago, A., Cerrato, F., Russo, S., Ferraiuolo, S., Rinaldi, M.M., et al. (2009). Hypomethylation at multiple maternally methylated imprinted regions including PLAGL1 and GNAS loci in Beckwith-Wiedemann syndrome. Eur J Hum Genet 17, 611-619. Bourque, D.K., Avila, L., Penaherrera, M., von Dadelszen, P., and Robinson, W.P. (2010). Decreased placental methylation at the H19/IGF2 imprinting control region is associated with normotensive intrauterine growth restriction but not preeclampsia. Placenta 31, 197-202. Burton, G.J., Yung, H.W., Cindrova-Davies, T., and Charnock-Jones, D.S. (2009). Placental endoplasmic reticulum stress and oxidative stress in the pathophysiology of unexplained intrauterine growth restriction and early onset preeclampsia. Placenta 30 Suppl A, S43-48. Cattanach, B.M., Beechey, C.V., and Peters, J. (2004). Interactions between imprinting effects in the mouse. Genetics 168, 397-413. Cetin, I., and Alvino, G. (2009). Intrauterine growth restriction: implications for placental metabolism and transport. A review. Placenta 30 Suppl A, S77-82. Cetin, I., Ronzoni, S., Marconi, A.M., Perugino, G., Corbetta, C., Battaglia, F.C., and Pardi, G. (1996). Maternal concentrations and fetal-maternal concentration differences of plasma amino acids in normal and intrauterine growth-restricted pregnancies. Am J Obstet Gynecol 174, 15751583. Chabory, E., Damon, C., Lenoir, A., Kauselmann, G., Kern, H., Zevnik, B., Garrel, C., Saez, F., Cadet, R., Henry-Berger, J., et al. (2009). Epididymis seleno-independent glutathione peroxidase 5 maintains sperm DNA integrity in mice. J Clin Invest 119, 2074-2085. Chellakooty, M., Vangsgaard, K., Larsen, T., Scheike, T., Falck-Larsen, J., Legarth, J., Andersson, A.M., Main, K.M., Skakkebaek, N.E., and Juul, A. (2004). A longitudinal study of intrauterine growth and the placental growth hormone (GH)-insulin-like growth factor I axis in maternal circulation: association between placental GH and fetal growth. J Clin Endocrinol Metab 89, 384-391.  89  Chellappan, S.P., Giordano, A., and Fisher, P.B. (1998). Role of cyclin-dependent kinases and their inhibitors in cellular differentiation and development. Curr Top Microbiol Immunol 227, 57-103. Cho, H., Thorvaldsen, J.L., Chu, Q., Feng, F., and Birnbaum, M.J. (2001). Akt1/PKBalpha is required for normal growth but dispensable for maintenance of glucose homeostasis in mice. J Biol Chem 276, 38349-38352. Choong, C.S., Patole, S., and Whitehall, J. (2000). Congenital varicella syndrome in the absence of cutaneous lesions. J Paediatr Child Health 36, 184-185. Chu, V.T., Gottardo, R., Raftery, A.E., Bumgarner, R.E., and Yeung, K.Y. (2008). MeV+R: using MeV as a graphical user interface for Bioconductor applications in microarray analysis. Genome Biol 9, R118. Coan, P.M., Angiolini, E., Sandovici, I., Burton, G.J., Constancia, M., and Fowden, A.L. (2008). Adaptations in placental nutrient transfer capacity to meet fetal growth demands depend on placental size in mice. J Physiol 586, 4567-4576. Coan, P.M., Burton, G.J., and Ferguson-Smith, A.C. (2005). Imprinted genes in the placenta--a review. Placenta 26 Suppl A, S10-20. Coan, P.M., Vaughan, O.R., Sekita, Y., Finn, S.L., Burton, G.J., Constancia, M., and Fowden, A.L. (2010). Adaptations in placental phenotype support fetal growth during undernutrition of pregnant mice. J Physiol 588, 527-538. Coe, B.L., Kirkpatrick, J.R., Taylor, J.A., and vom Saal, F.S. (2008). A new 'crowded uterine horn' mouse model for examining the relationship between foetal growth and adult obesity. Basic Clin Pharmacol Toxicol 102, 162-167. Constancia, M., Angiolini, E., Sandovici, I., Smith, P., Smith, R., Kelsey, G., Dean, W., Ferguson-Smith, A., Sibley, C.P., Reik, W., et al. (2005). Adaptation of nutrient supply to fetal demand in the mouse involves interaction between the Igf2 gene and placental transporter systems. Proc Natl Acad Sci U S A 102, 19219-19224. Constancia, M., Dean, W., Lopes, S., Moore, T., Kelsey, G., and Reik, W. (2000). Deletion of a silencer element in Igf2 results in loss of imprinting independent of H19. Nat Genet 26, 203-206. Constancia, M., Hemberger, M., Hughes, J., Dean, W., Ferguson-Smith, A., Fundele, R., Stewart, F., Kelsey, G., Fowden, A., Sibley, C., et al. (2002). Placental-specific IGF-II is a major modulator of placental and fetal growth. Nature 417, 945-948. Copp, A.J., Greene, N.D., and Murdoch, J.N. (2003). The genetic basis of mammalian neurulation. Nat Rev Genet 4, 784-793. Cox, P., and Marton, T. (2009). Pathological assessment of intrauterine growth restriction. Best Pract Res Clin Obstet Gynaecol 23, 751-764. Cross, J.C., Werb, Z., and Fisher, S.J. (1994). Implantation and the placenta: key pieces of the development puzzle. Science 266, 1508-1518. 90  Czech, M.P., Lewis, R.E., and Corvera, S. (1989). Multifunctional glycoprotein receptors for insulin and the insulin-like growth factors. Ciba Found Symp 145, 27-41; discussion 42-24. Das, S.K., Yano, S., Wang, J., Edwards, D.R., Nagase, H., and Dey, S.K. (1997). Expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in the mouse uterus during the peri-implantation period. Dev Genet 21, 44-54. DeChiara, T.M., Efstratiadis, A., and Robertson, E.J. (1990). A growth-deficiency phenotype in heterozygous mice carrying an insulin-like growth factor II gene disrupted by targeting. Nature 345, 78-80. DeChiara, T.M., Robertson, E.J., and Efstratiadis, A. (1991). Parental imprinting of the mouse insulin-like growth factor II gene. Cell 64, 849-859. Detrait, E.R., George, T.M., Etchevers, H.C., Gilbert, J.R., Vekemans, M., and Speer, M.C. (2005). Human neural tube defects: developmental biology, epidemiology, and genetics. Neurotoxicol Teratol 27, 515-524. Diplas, A.I., Lambertini, L., Lee, M.J., Sperling, R., Lee, Y.L., Wetmur, J., and Chen, J. (2009). Differential expression of imprinted genes in normal and IUGR human placentas. Epigenetics 4, 235-240. Drewell, R.A., Arney, K.L., Arima, T., Barton, S.C., Brenton, J.D., and Surani, M.A. (2002). Novel conserved elements upstream of the H19 gene are transcribed and act as mesodermal enhancers. Development 129, 1205-1213. Drewell, R.A., Brenton, J.D., Ainscough, J.F., Barton, S.C., Hilton, K.J., Arney, K.L., Dandolo, L., and Surani, M.A. (2000). Deletion of a silencer element disrupts H19 imprinting independently of a DNA methylation epigenetic switch. Development 127, 3419-3428. Du, P., Kibbe, W.A., and Lin, S.M. (2008). lumi: a pipeline for processing Illumina microarray. Bioinformatics 24, 1547-1548. Economides, D.L., Proudler, A., and Nicolaides, K.H. (1989). Plasma insulin in appropriate- and small-for-gestational-age fetuses. Am J Obstet Gynecol 160, 1091-1094. Eggenschwiler, J., Ludwig, T., Fisher, P., Leighton, P.A., Tilghman, S.M., and Efstratiadis, A. (1997). Mouse mutant embryos overexpressing IGF-II exhibit phenotypic features of the Beckwith-Wiedemann and Simpson-Golabi-Behmel syndromes. Genes Dev 11, 3128-3142. Eggerding, F.A., Schonberg, S.A., Chehab, F.F., Norton, M.E., Cox, V.A., and Epstein, C.J. (1994). Uniparental isodisomy for paternal 7p and maternal 7q in a child with growth retardation. Am J Hum Genet 55, 253-265. Eggermann, T., Schonherr, N., Meyer, E., Obermann, C., Mavany, M., Eggermann, K., Ranke, M.B., and Wollmann, H.A. (2006). Epigenetic mutations in 11p15 in Silver-Russell syndrome are restricted to the telomeric imprinting domain. J Med Genet 43, 615-616. Enders, A.C. (1965). A Comparative Study of the Fine Structure of the Trophoblast in Several Hemochorial Placentas. Am J Anat 116, 29-67. 91  Enquobahrie, D.A., Meller, M., Rice, K., Psaty, B.M., Siscovick, D.S., and Williams, M.A. (2008). Differential placental gene expression in preeclampsia. Am J Obstet Gynecol 199, 566 e561-511. Fang, J., Mao, D., Smith, C.H., and Fant, M.E. (2006). IGF regulation of neutral amino acid transport in the BeWo choriocarcinoma cell line (b30 clone): evidence for MAP kinasedependent and MAP kinase-independent mechanisms. Growth Horm IGF Res 16, 318-325. Ferguson-Smith, A.C., Cattanach, B.M., Barton, S.C., Beechey, C.V., and Surani, M.A. (1991). Embryological and molecular investigations of parental imprinting on mouse chromosome 7. Nature 351, 667-670. Filson, A.J., Louvi, A., Efstratiadis, A., and Robertson, E.J. (1993). Rescue of the T-associated maternal effect in mice carrying null mutations in Igf-2 and Igf2r, two reciprocally imprinted genes. Development 118, 731-736. Foda, H.D., and Zucker, S. (2001). Matrix metalloproteinases in cancer invasion, metastasis and angiogenesis. Drug Discov Today 6, 478-482. Fogarasi, M., Janssen, A., Weber, B.H., and Stohr, H. (2008). Molecular dissection of TIMP3 mutation S156C associated with Sorsby fundus dystrophy. Matrix Biol 27, 381-392. Fowden, A.L., and Forhead, A.J. (2004). Endocrine mechanisms of intrauterine programming. Reproduction 127, 515-526. Frank, D., Fortino, W., Clark, L., Musalo, R., Wang, W., Saxena, A., Li, C.M., Reik, W., Ludwig, T., and Tycko, B. (2002). Placental overgrowth in mice lacking the imprinted gene Ipl. Proc Natl Acad Sci U S A 99, 7490-7495. Frost, J.M., and Moore, G.E. (2010). The importance of imprinting in the human placenta. PLoS Genet 6, e1001015. Gabory, A., Ripoche, M.A., Le Digarcher, A., Watrin, F., Ziyyat, A., Forne, T., Jammes, H., Ainscough, J.F., Surani, M.A., Journot, L., et al. (2009). H19 acts as a trans regulator of the imprinted gene network controlling growth in mice. Development 136, 3413-3421. Garner, I., Sassoon, D., Vandekerckhove, J., Alonso, S., and Buckingham, M.E. (1989). A developmental study of the abnormal expression of alpha-cardiac and alpha-skeletal actins in the striated muscle of a mutant mouse. Dev Biol 134, 236-245. Garzetti, G.G., Ciavattini, A., Lucarini, G., Pugnaloni, A., De Nictolis, M., Amati, S., Romanini, C., and Biagini, G. (1999). Expression of vascular endothelial growth factor related to 72kilodalton metalloproteinase immunostaining in patients with serous ovarian tumors. Cancer 85, 2219-2225. Gaunt, M., and Ockleford, C.D. (1986). Microinjection of human placenta. II: Biological application. Placenta 7, 325-331. Genbacev, O., Joslin, R., Damsky, C.H., Polliotti, B.M., and Fisher, S.J. (1996). Hypoxia alters early gestation human cytotrophoblast differentiation/invasion in vitro and models the placental defects that occur in preeclampsia. J Clin Invest 97, 540-550. 92  Genbacev, O., Zhou, Y., Ludlow, J.W., and Fisher, S.J. (1997). Regulation of human placental development by oxygen tension. Science 277, 1669-1672. Georgiades, P., Ferguson-Smith, A.C., and Burton, G.J. (2002). Comparative developmental anatomy of the murine and human definitive placentae. Placenta 23, 3-19. Gerretsen, G., Huisjes, H.J., and Elema, J.D. (1981). Morphological changes of the spiral arteries in the placental bed in relation to pre-eclampsia and fetal growth retardation. Br J Obstet Gynaecol 88, 876-881. Gicquel, C., Rossignol, S., Cabrol, S., Houang, M., Steunou, V., Barbu, V., Danton, F., Thibaud, N., Le Merrer, M., Burglen, L., et al. (2005). Epimutation of the telomeric imprinting center region on chromosome 11p15 in Silver-Russell syndrome. Nat Genet 37, 1003-1007. Giovannone, B., Lee, E., Laviola, L., Giorgino, F., Cleveland, K.A., and Smith, R.J. (2003). Two novel proteins that are linked to insulin-like growth factor (IGF-I) receptors by the Grb10 adapter and modulate IGF-I signaling. J Biol Chem 278, 31564-31573. Godfrey, K.M., Matthews, N., Glazier, J., Jackson, A., Wilman, C., and Sibley, C.P. (1998). Neutral amino acid uptake by the microvillous plasma membrane of the human placenta is inversely related to fetal size at birth in normal pregnancy. J Clin Endocrinol Metab 83, 33203326. 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. Guillemot, F., Caspary, T., Tilghman, S.M., Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Anderson, D.J., Joyner, A.L., Rossant, J., and Nagy, A. (1995). Genomic imprinting of Mash2, a mouse gene required for trophoblast development. Nat Genet 9, 235-242. Guillemot, F., Nagy, A., Auerbach, A., Rossant, J., and Joyner, A.L. (1994). Essential role of Mash-2 in extraembryonic development. Nature 371, 333-336. Guo, L., Choufani, S., Ferreira, J., Smith, A., Chitayat, D., Shuman, C., Uxa, R., Keating, S., Kingdom, J., and Weksberg, R. (2008). Altered gene expression and methylation of the human chromosome 11 imprinted region in small for gestational age (SGA) placentae. Dev Biol 320, 79-91. Hadlock, F.P., Harrist, R.B., Sharman, R.S., Deter, R.L., and Park, S.K. (1985). Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. Am J Obstet Gynecol 151, 333-337. Hatada, I., Ohashi, H., Fukushima, Y., Kaneko, Y., Inoue, M., Komoto, Y., Okada, A., Ohishi, S., Nabetani, A., Morisaki, H., et al. (1996). An imprinted gene p57KIP2 is mutated in Beckwith-Wiedemann syndrome. Nat Genet 14, 171-173. Hoeltzli, S.D., and Smith, C.H. (1989). Alanine transport systems in isolated basal plasma membrane of human placenta. Am J Physiol 256, C630-637.  93  Holcberg, G., Huleihel, M., Sapir, O., Katz, M., Tsadkin, M., Furman, B., Mazor, M., and Myatt, L. (2001). Increased production of tumor necrosis factor-alpha TNF-alpha by IUGR human placentae. Eur J Obstet Gynecol Reprod Biol 94, 69-72. Holmes, R., Montemagno, R., Jones, J., Preece, M., Rodeck, C., and Soothill, P. (1997). Fetal and maternal plasma insulin-like growth factors and binding proteins in pregnancies with appropriate or retarded fetal growth. Early Hum Dev 49, 7-17. Huppertz, B., Frank, H.G., Kingdom, J.C., Reister, F., and Kaufmann, P. (1998). Villous cytotrophoblast regulation of the syncytial apoptotic cascade in the human placenta. Histochem Cell Biol 110, 495-508. Itoh, T., Ikeda, T., Gomi, H., Nakao, S., Suzuki, T., and Itohara, S. (1997). Unaltered secretion of beta-amyloid precursor protein in gelatinase A (matrix metalloproteinase 2)-deficient mice. J Biol Chem 272, 22389-22392. Itoh, T., Tanioka, M., Yoshida, H., Yoshioka, T., Nishimoto, H., and Itohara, S. (1998). Reduced angiogenesis and tumor progression in gelatinase A-deficient mice. Cancer Res 58, 1048-1051. Jager, S., Schonherr, N., Spengler, S., Ranke, M.B., Wollmann, H.A., Binder, G., and Eggermann, T. (2009). LOT1 (ZAC1/PLAGL1) as member of an imprinted gene network does not harbor Silver-Russell specific variants. J Pediatr Endocrinol Metab 22, 555-559. Jansson, T., Wennergren, M., and Illsley, N.P. (1993). Glucose transporter protein expression in human placenta throughout gestation and in intrauterine growth retardation. J Clin Endocrinol Metab 77, 1554-1562. Jansson, T., Ylven, K., Wennergren, M., and Powell, T.L. (2002). Glucose transport and system A activity in syncytiotrophoblast microvillous and basal plasma membranes in intrauterine growth restriction. Placenta 23, 392-399. Jauniaux, E., Gulbis, B., and Burton, G.J. (2003). The human first trimester gestational sac limits rather than facilitates oxygen transfer to the foetus--a review. Placenta 24 Suppl A, S86-93. Jauniaux, E., Jurkovic, D., and Campbell, S. (1995). Current topic: in vivo investigation of the placental circulations by Doppler echography. Placenta 16, 323-331. Jauniaux, E., Watson, A.L., Hempstock, J., Bao, Y.P., Skepper, J.N., and Burton, G.J. (2000). Onset of maternal arterial blood flow and placental oxidative stress. A possible factor in human early pregnancy failure. Am J Pathol 157, 2111-2122. Jin, K., Mao, X.O., Cottrell, B., Schilling, B., Xie, L., Row, R.H., Sun, Y., Peel, A., Childs, J., Gendeh, G., et al. (2004). Proteomic and immunochemical characterization of a role for stathmin in adult neurogenesis. FASEB J 18, 287-299. Johnson, L.W., and Smith, C.H. (1988). Neutral amino acid transport systems of microvillous membrane of human placenta. Am J Physiol 254, C773-780. Jones, H.N., Powell, T.L., and Jansson, T. (2007). Regulation of placental nutrient transport--a review. Placenta 28, 763-774. 94  Kagami, M., Sekita, Y., Nishimura, G., Irie, M., Kato, F., Okada, M., Yamamori, S., Kishimoto, H., Nakayama, M., Tanaka, Y., et al. (2008). Deletions and epimutations affecting the human 14q32.2 imprinted region in individuals with paternal and maternal upd(14)-like phenotypes. Nat Genet 40, 237-242. Kalousek, D.K. (1994). Current topic: confined placental mosaicism and intrauterine fetal development. Placenta 15, 219-230. Kanayama, N., Takahashi, K., Matsuura, T., Sugimura, M., Kobayashi, T., Moniwa, N., Tomita, M., and Nakayama, K. (2002). Deficiency in p57Kip2 expression induces preeclampsia-like symptoms in mice. Mol Hum Reprod 8, 1129-1135. 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. Nat Genet 11, 52-59. Karl, P.I. (1995). Insulin-like growth factor-1 stimulates amino acid uptake by the cultured human placental trophoblast. J Cell Physiol 165, 83-88. Karl, P.I., Alpy, K.L., and Fisher, S.E. (1992). Amino acid transport by the cultured human placental trophoblast: effect of insulin on AIB transport. Am J Physiol 262, C834-839. Koerner, M.V., Pauler, F.M., Huang, R., and Barlow, D.P. (2009). The function of non-coding RNAs in genomic imprinting. Development 136, 1771-1783. Kozak, K.R., Abbott, B., and Hankinson, O. (1997). ARNT-deficient mice and placental differentiation. Dev Biol 191, 297-305. Kuhn, K., Baker, S.C., Chudin, E., Lieu, M.H., Oeser, S., Bennett, H., Rigault, P., Barker, D., McDaniel, T.K., and Chee, M.S. (2004). A novel, high-performance random array platform for quantitative gene expression profiling. Genome Res 14, 2347-2356. Kuzmin, A., Han, Z., Golding, M.C., Mann, M.R., Latham, K.E., and Varmuza, S. (2008). The PcG gene Sfmbt2 is paternally expressed in extraembryonic tissues. Gene Expr Patterns 8, 107116. Lam, W.W., Hatada, I., Ohishi, S., Mukai, T., Joyce, J.A., Cole, T.R., Donnai, D., Reik, W., Schofield, P.N., and Maher, E.R. (1999). Analysis of germline CDKN1C (p57KIP2) mutations in familial and sporadic Beckwith-Wiedemann syndrome (BWS) provides a novel genotypephenotype correlation. J Med Genet 36, 518-523. Lau, M.M., Stewart, C.E., Liu, Z., Bhatt, H., Rotwein, P., and Stewart, C.L. (1994). Loss of the imprinted IGF2/cation-independent mannose 6-phosphate receptor results in fetal overgrowth and perinatal lethality. Genes Dev 8, 2953-2963. Ledbetter, D.H., and Engel, E. (1995). Uniparental disomy in humans: development of an imprinting map and its implications for prenatal diagnosis. Hum Mol Genet 4 Spec No, 17571764.  95  Lee, M.H., Jeon, Y.J., Lee, S.M., Park, M.H., Jung, S.C., and Kim, Y.J. (2010). Placental gene expression is related to glucose metabolism and fetal cord blood levels of insulin and insulin-like growth factors in intrauterine growth restriction. Early Hum Dev 86, 45-50. 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. Nat Genet 20, 163-169. Leighton, P.A., Ingram, R.S., Eggenschwiler, J., Efstratiadis, A., and Tilghman, S.M. (1995). Disruption of imprinting caused by deletion of the H19 gene region in mice. Nature 375, 34-39. Lewis, A., Green, K., Dawson, C., Redrup, L., Huynh, K.D., Lee, J.T., Hemberger, M., and Reik, W. (2006). Epigenetic dynamics of the Kcnq1 imprinted domain in the early embryo. Development 133, 4203-4210. Lewis, A., Mitsuya, K., Umlauf, D., Smith, P., Dean, W., Walter, J., Higgins, M., Feil, R., and Reik, W. (2004). Imprinting on distal chromosome 7 in the placenta involves repressive histone methylation independent of DNA methylation. Nat Genet 36, 1291-1295. Li, L., Keverne, E.B., Aparicio, S.A., Ishino, F., Barton, S.C., and Surani, M.A. (1999). Regulation of maternal behavior and offspring growth by paternally expressed Peg3. Science 284, 330-333. Lim, A.L., and Ferguson-Smith, A.C. (2010). Genomic imprinting effects in a compromised in utero environment: implications for a healthy pregnancy. Semin Cell Dev Biol 21, 201-208. Lin, J., Toft, D.J., Bengtson, N.W., and Linzer, D.I. (2000). Placental prolactins and the physiology of pregnancy. Recent Prog Horm Res 55, 37-51; discussion 52. Lin, S.M., Du, P., Huber, W., and Kibbe, W.A. (2008). Model-based variance-stabilizing transformation for Illumina microarray data. Nucleic Acids Res 36, e11. Liu, X., Wu, H., Byrne, M., Krane, S., and Jaenisch, R. (1997). Type III collagen is crucial for collagen I fibrillogenesis and for normal cardiovascular development. Proc Natl Acad Sci U S A 94, 1852-1856. Lo, H.C., Tsao, L.Y., Hsu, W.Y., Chen, H.N., Yu, W.K., and Chi, C.Y. (2002). Relation of cord serum levels of growth hormone, insulin-like growth factors, insulin-like growth factor binding proteins, leptin, and interleukin-6 with birth weight, birth length, and head circumference in term and preterm neonates. Nutrition 18, 604-608. Lui, J.C., Finkielstain, G.P., Barnes, K.M., and Baron, J. (2008). An imprinted gene network that controls mammalian somatic growth is down-regulated during postnatal growth deceleration in multiple organs. Am J Physiol Regul Integr Comp Physiol 295, R189-196. Mackenzie, B., and Erickson, J.D. (2004). Sodium-coupled neutral amino acid (System N/A) transporters of the SLC38 gene family. Pflugers Arch 447, 784-795. Mamo, S., Gal, A.B., Bodo, S., and Dinnyes, A. (2007). Quantitative evaluation and selection of reference genes in mouse oocytes and embryos cultured in vivo and in vitro. BMC Dev Biol 7, 14. 96  Masuyama, H., Hiramatsu, Y., and Kudo, T. (1996). Effect of epidermal growth factor on placental amino acid transport and regulation of epidermal growth factor receptor expression of hepatocyte in rat. J Perinat Med 24, 213-220. Maxwell, P.H., Wiesener, M.S., Chang, G.W., Clifford, S.C., Vaux, E.C., Cockman, M.E., Wykoff, C.C., Pugh, C.W., Maher, E.R., and Ratcliffe, P.J. (1999). The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399, 271-275. McGrath, J., and Solter, D. (1984). Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell 37, 179-183. McMinn, J., Wei, M., Schupf, N., Cusmai, J., Johnson, E.B., Smith, A.C., Weksberg, R., Thaker, H.M., and Tycko, B. (2006). Unbalanced placental expression of imprinted genes in human intrauterine growth restriction. Placenta 27, 540-549. Miller, J., Turan, S., and Baschat, A.A. (2008). Fetal growth restriction. Semin Perinatol 32, 274280. Mizuno, Y., Sotomaru, Y., Katsuzawa, Y., Kono, T., Meguro, M., Oshimura, M., Kawai, J., Tomaru, Y., Kiyosawa, H., Nikaido, I., et al. (2002). Asb4, Ata3, and Dcn are novel imprinted genes identified by high-throughput screening using RIKEN cDNA microarray. Biochem Biophys Res Commun 290, 1499-1505. Mochizuki, A., Takeda, Y., and Iwasa, Y. (1996). The evolution of genomic imprinting. Genetics 144, 1283-1295. Moon, Y.S., Smas, C.M., Lee, K., Villena, J.A., Kim, K.H., Yun, E.J., and Sul, H.S. (2002). Mice lacking paternally expressed Pref-1/Dlk1 display growth retardation and accelerated adiposity. Mol Cell Biol 22, 5585-5592. Moore, T., and Haig, D. (1991). Genomic imprinting in mammalian development: a parental tugof-war. Trends Genet 7, 45-49. Mosig, R.A., Dowling, O., DiFeo, A., Ramirez, M.C., Parker, I.C., Abe, E., Diouri, J., Aqeel, A.A., Wylie, J.D., Oblander, S.A., et al. (2007). Loss of MMP-2 disrupts skeletal and craniofacial development and results in decreased bone mineralization, joint erosion and defects in osteoblast and osteoclast growth. Hum Mol Genet 16, 1113-1123. Muntener, M., and Hsu, Y.C. (1977). Development of trophoblast and placenta of the mouse. A reinvestigation with regard to the in vitro culture of mouse trophoblast and placenta. Acta Anat (Basel) 98, 241-252. Murie, C., Woody, O., Lee, A.Y., and Nadon, R. (2009). Comparison of small n statistical tests of differential expression applied to microarrays. BMC Bioinformatics 10, 45. Nagano, T., Mitchell, J.A., Sanz, L.A., Pauler, F.M., Ferguson-Smith, A.C., Feil, R., and Fraser, P. (2008). The Air noncoding RNA epigenetically silences transcription by targeting G9a to chromatin. Science 322, 1717-1720.  97  Nagasawa, H., Miyamoto, M., and Fujimoto, M. (1973). [Reproductivity in inbred strains of mice and project for their efficient production (author's transl)]. Jikken Dobutsu 22, 119-126. Neilson, J.P., Munjanja, S.P., Whitfield, C.R. (1984). Screening for small for dates fetuses: a controlled trial. Br Med J (Clin Res Ed) 289, 1179-1182. Nelson, D.M., Smith, S.D., Furesz, T.C., Sadovsky, Y., Ganapathy, V., Parvin, C.A., and Smith, C.H. (2003). Hypoxia reduces expression and function of system A amino acid transporters in cultured term human trophoblasts. Am J Physiol Cell Physiol 284, C310-315. Newnham, J.P., Lam, R.W., Hobel, C.J., Padbury, J.F., Polk, D.H., and Fisher, D.A. (1986). Differential response of ovine placental lactogen levels in maternal and fetal circulations following single umbilical artery ligation in fetal sheep. Placenta 7, 51-64. Nicholls, R.D., Saitoh, S., and Horsthemke, B. (1998). Imprinting in Prader-Willi and Angelman syndromes. Trends Genet 14, 194-200. Ono, R., Nakamura, K., Inoue, K., Naruse, M., Usami, T., Wakisaka-Saito, N., Hino, T., SuzukiMigishima, R., Ogonuki, N., Miki, H., et al. (2006). Deletion of Peg10, an imprinted gene acquired from a retrotransposon, causes early embryonic lethality. Nat Genet 38, 101-106. Oyadomari, S., Koizumi, A., Takeda, K., Gotoh, T., Akira, S., Araki, E., and Mori, M. (2002). Targeted disruption of the Chop gene delays endoplasmic reticulum stress-mediated diabetes. J Clin Invest 109, 525-532. Ozon, S., Guichet, A., Gavet, O., Roth, S., and Sobel, A. (2002). Drosophila stathmin: a microtubule-destabilizing factor involved in nervous system formation. Mol Biol Cell 13, 698710. Penaherrera, M.S., Weindler, S., Van Allen, M.I., Yong, S.L., Metzger, D.L., McGillivray, B., Boerkoel, C., Langlois, S., and Robinson, W.P. (2010). Methylation profiling in individuals with Russell-Silver syndrome. Am J Med Genet A 152A, 347-355. Phillips, P.G., Birnby, L.M., and Narendran, A. (1995). Hypoxia induces capillary network formation in cultured bovine pulmonary microvessel endothelial cells. Am J Physiol 268, L789800. Pidoux, G., Gerbaud, P., Laurendeau, I., Guibourdenche, J., Bertin, G., Vidaud, M., Evain-Brion, D., and Frendo, J.L. (2004). Large variability of trophoblast gene expression within and between human normal term placentae. Placenta 25, 469-473. Platz, E., and Newman, R. (2008). Diagnosis of IUGR: traditional biometry. Semin Perinatol 32, 140-147. Potau, N., Riudor, E., and Ballabriga, A. (1981). Insulin receptors in human placenta in relation to fetal weight and gestational age. Pediatr Res 15, 798-802. Pringle, K.G., Kind, K.L., Thompson, J.G., and Roberts, C.T. (2007). Complex interactions between hypoxia inducible factors, insulin-like growth factor-II and oxygen in early murine trophoblasts. Placenta 28, 1147-1157. 98  Rahkonen, O., Su, M., Hakovirta, H., Koskivirta, I., Hormuzdi, S.G., Vuorio, E., Bornstein, P., and Penttinen, R. (2004). Mice with a deletion in the first intron of the Col1a1 gene develop agedependent aortic dissection and rupture. Circ Res 94, 83-90. Redline, R.W., Chernicky, C.L., Tan, H.Q., and Ilan, J. (1993). Differential expression of insulin-like growth factor-II in specific regions of the late (post day 9.5) murine placenta. Mol Reprod Dev 36, 121-129. Rigano, S., Bozzo, M., Ferrazzi, E., Bellotti, M., Battaglia, F.C., and Galan, H.L. (2001). Early and persistent reduction in umbilical vein blood flow in the growth-restricted fetus: a longitudinal study. Am J Obstet Gynecol 185, 834-838. Rodesch, F., Simon, P., Donner, C., and Jauniaux, E. (1992). Oxygen measurements in endometrial and trophoblastic tissues during early pregnancy. Obstet Gynecol 80, 283-285. Romanelli, V., Belinchon, A., Campos-Barros, A., Heath, K.E., Garcia-Minaur, S., MartinezGlez, V., Palomo, R., Mercado, G., Gracia, R., and Lapunzina, P. (2009). CDKN1C mutations in HELLP/preeclamptic mothers of Beckwith-Wiedemann Syndrome (BWS) patients. Placenta 30, 551-554. Roos, S., Jansson, N., Palmberg, I., Saljo, K., Powell, T.L., and Jansson, T. (2007). Mammalian target of rapamycin in the human placenta regulates leucine transport and is down-regulated in restricted fetal growth. J Physiol 582, 449-459. Rossant, J., and Cross, J.C. (2001). Placental development: lessons from mouse mutants. Nat Rev Genet 2, 538-548. Ruijter, J.M., Ramakers, C., Hoogaars, W.M., Karlen, Y., Bakker, O., van den Hoff, M.J., and Moorman, A.F. (2009). Amplification efficiency: linking baseline and bias in the analysis of quantitative PCR data. Nucleic Acids Res 37, e45. Schmidt, J.V., Matteson, P.G., Jones, B.K., Guan, X.J., and Tilghman, S.M. (2000). The Dlk1 and Gtl2 genes are linked and reciprocally imprinted. Genes Dev 14, 1997-2002. 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. Schuster, V., Janssen, W., Seidenspinner, S., and Kreth, H.W. (1993). [Congenital Epstein-Barr virus infection]. Monatsschr Kinderheilkd 141, 401-404. Sebire, N.J., and Fisher, R.A. (2005). Partly molar pregnancies that are not partial moles: additional possibilities and implications. Pediatr Dev Pathol 8, 732-733. Sekita, Y., Wagatsuma, H., Nakamura, K., Ono, R., Kagami, M., Wakisaka, N., Hino, T., Suzuki-Migishima, R., Kohda, T., Ogura, A., et al. (2008). Role of retrotransposon-derived imprinted gene, Rtl1, in the feto-maternal interface of mouse placenta. Nat Genet 40, 243-248. Sferruzzi-Perri, A.N., Owens, J.A., Pringle, K.G., Robinson, J.S., and Roberts, C.T. (2006). Maternal insulin-like growth factors-I and -II act via different pathways to promote fetal growth. Endocrinology 147, 3344-3355. 99  Sferruzzi-Perri, A.N., Owens, J.A., Standen, P., Taylor, R.L., Heinemann, G.K., Robinson, J.S., and Roberts, C.T. (2007). Early treatment of the pregnant guinea pig with IGFs promotes placental transport and nutrient partitioning near term. Am J Physiol Endocrinol Metab 292, E668-676. Sibley, C.P., Coan, P.M., Ferguson-Smith, A.C., Dean, W., Hughes, J., Smith, P., Reik, W., Burton, G.J., Fowden, A.L., and Constancia, M. (2004). Placental-specific insulin-like growth factor 2 (Igf2) regulates the diffusional exchange characteristics of the mouse placenta. Proc Natl Acad Sci U S A 101, 8204-8208. Simmons, D.G., Fortier, A.L., and Cross, J.C. (2007). Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta. Dev Biol 304, 567-578. Simmons, D.G., Rawn, S., Davies, A., Hughes, M., and Cross, J.C. (2008). Spatial and temporal expression of the 23 murine Prolactin/Placental Lactogen-related genes is not associated with their position in the locus. BMC Genomics 9, 352. Smith, F.M., Holt, L.J., Garfield, A.S., Charalambous, M., Koumanov, F., Perry, M., Bazzani, R., Sheardown, S.A., Hegarty, B.D., Lyons, R.J., et al. (2007). Mice with a disruption of the imprinted Grb10 gene exhibit altered body composition, glucose homeostasis, and insulin signaling during postnatal life. Mol Cell Biol 27, 5871-5886. Sounni, N.E., Devy, L., Hajitou, A., Frankenne, F., Munaut, C., Gilles, C., Deroanne, C., Thompson, E.W., Foidart, J.M., and Noel, A. (2002). MT1-MMP expression promotes tumor growth and angiogenesis through an up-regulation of vascular endothelial growth factor expression. FASEB J 16, 555-564. Spiegel, R., Raas-Rothschild, A., Reish, O., Regev, M., Meiner, V., Bargal, R., Sury, V., Meir, K., Nadjari, M., Hermann, G., et al. (2009). The clinical spectrum of fetal Niemann-Pick type C. Am J Med Genet A 149A, 446-450. Strittmatter, S.M., Fankhauser, C., Huang, P.L., Mashimo, H., and Fishman, M.C. (1995). Neuronal pathfinding is abnormal in mice lacking the neuronal growth cone protein GAP-43. Cell 80, 445-452. Struwe, E., Berzl, G., Schild, R., Blessing, H., Drexel, L., Hauck, B., Tzschoppe, A., Weidinger, M., Sachs, M., Scheler, C., et al. (2010). Microarray analysis of placental tissue in intrauterine growth restriction. Clin Endocrinol (Oxf) 72, 241-247. Surani, M.A., Barton, S.C., and Norris, M.L. (1984). Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature 308, 548-550. Takahashi, K., and Nakayama, K. (2000). Mice lacking a CDK inhibitor, p57Kip2, exhibit skeletal abnormalities and growth retardation. J Biochem 127, 73-83. Tanaka, T.S., Jaradat, S.A., Lim, M.K., Kargul, G.J., Wang, X., Grahovac, M.J., Pantano, S., Sano, Y., Piao, Y., Nagaraja, R., et al. (2000). Genome-wide expression profiling of midgestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc Natl Acad Sci U S A 97, 9127-9132.  100  Teesalu, T., Masson, R., Basset, P., Blasi, F., and Talarico, D. (1999). Expression of matrix metalloproteinases during murine chorioallantoic placenta maturation. Dev Dyn 214, 248-258. Temple, I.K. (2007). Imprinting in human disease with special reference to transient neonatal diabetes and Beckwith-Wiedemann syndrome. Endocr Dev 12, 113-123. Temple, I.K., Shrubb, V., Lever, M., Bullman, H., and Mackay, D.J. (2007). Isolated imprinting mutation of the DLK1/GTL2 locus associated with a clinical presentation of maternal uniparental disomy of chromosome 14. J Med Genet 44, 637-640. 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 Dev 12, 36933702. Tunster, S.J., Tycko, B., and John, R.M. (2010). The imprinted Phlda2 gene regulates extraembryonic energy stores. Mol Cell Biol 30, 295-306. 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. Nat Genet 36, 1296-1300. van der Smagt, J.J., Scheenjes, E., Kremer, J.A., Hennekam, F.A., and Fisher, R.A. (2006). Heterogeneity in the origin of recurrent complete hydatidiform moles: not all women with multiple molar pregnancies have biparental moles. BJOG 113, 725-728. Varanou, A., Withington, S.L., Lakasing, L., Williamson, C., Burton, G.J., and Hemberger, M. (2006). The importance of cysteine cathepsin proteases for placental development. J Mol Med 84, 305-317. Varrault, A., Gueydan, C., Delalbre, A., Bellmann, A., Houssami, S., Aknin, C., Severac, D., Chotard, L., Kahli, M., Le Digarcher, A., et al. (2006). Zac1 regulates an imprinted gene network critically involved in the control of embryonic growth. Dev Cell 11, 711-722. Verhaagh, S., Schweifer, N., Barlow, D.P., and Zwart, R. (1999). Cloning of the mouse and human solute carrier 22a3 (Slc22a3/SLC22A3) identifies a conserved cluster of three organic cation transporters on mouse chromosome 17 and human 6q26-q27. Genomics 55, 209-218. Vileisis, R.A., Fain, J., and Oh, W. (1982). Fatty acid synthesis in rat fetuses with intrauterine growth retardation. Metabolism 31, 217-222. Vom Saal, F.S., and Dhar, M.G. (1992). Blood flow in the uterine loop artery and loop vein is bidirectional in the mouse: implications for transport of steroids between fetuses. Physiol Behav 52, 163-171. Vuguin, P.M. (2007). Animal models for small for gestational age and fetal programming of adult disease. Horm Res 68, 113-123. Wagschal, A., Sutherland, H.G., Woodfine, K., Henckel, A., Chebli, K., Schulz, R., Oakey, R.J., Bickmore, W.A., and Feil, R. (2008). G9a histone methyltransferase contributes to imprinting in the mouse placenta. Mol Cell Biol 28, 1104-1113. 101  Wang, Y., Szczesna-Cordary, D., Craig, R., Diaz-Perez, Z., Guzman, G., Miller, T., and Potter, J.D. (2007). Fast skeletal muscle regulatory light chain is required for fast and slow skeletal muscle development. FASEB J 21, 2205-2214. Wolf, J.B., and Hager, R. (2006). A maternal-offspring coadaptation theory for the evolution of genomic imprinting. PLoS Biol 4, e380. Yang, Z.Z., Tschopp, O., Hemmings-Mieszczak, M., Feng, J., Brodbeck, D., Perentes, E., and Hemmings, B.A. (2003). Protein kinase B alpha/Akt1 regulates placental development and fetal growth. J Biol Chem 278, 32124-32131. Yevtodiyenko, A., and Schmidt, J.V. (2006). Dlk1 expression marks developing endothelium and sites of branching morphogenesis in the mouse embryo and placenta. Dev Dyn 235, 11151123. Yoshie, M., Kashima, H., Bessho, T., Takeichi, M., Isaka, K., and Tamura, K. (2008). Expression of stathmin, a microtubule regulatory protein, is associated with the migration and differentiation of cultured early trophoblasts. Hum Reprod 23, 2766-2774. Zhang, P., Liegeois, N.J., Wong, C., Finegold, M., Hou, H., Thompson, J.C., Silverman, A., Harper, J.W., DePinho, R.A., and Elledge, S.J. (1997). Altered cell differentiation and proliferation in mice lacking p57KIP2 indicates a role in Beckwith-Wiedemann syndrome. Nature 387, 151-158. Zwart, R., Verhaagh, S., Buitelaar, M., Popp-Snijders, C., and Barlow, D.P. (2001). Impaired activity of the extraneuronal monoamine transporter system known as uptake-2 in Orct3/Slc22a3-deficient mice. Mol Cell Biol 21, 4188-4196.  102  Appendix A: Supplementary information A  *  B  Suppl Fig. 1 Weight and size measurements of Surgical IUGR model. (A) Weight comparisons between embryos from E15.5 WT C57BL/6J, hemi-ovariectomized Female 1, and Mmp2+/+. The embryonic weight litter from Female 1 was assessed to be significantly different by t test (p < 0.001) from normal C57BL/6J litter of the same stage. Error bars represent standard deviation of weights of littermates. (B) Comparison of size and weight difference between littermates of P1 litter from hemi-ovariectomized Female 2 and Mmp2+/+.  103  Suppl Fig. 2 Grouping of WT and IUGR samples. Four independent clusters are illustrated using the principal component analysis (PCA) function on FlexArray. The embryo samples group significantly from the placental samples. The only exception is LLE052-1, which is IUGR placental sample that has grouped with the IUGR embryo samples. The IUGR samples also group different from WT samples.  104  Suppl Fig. 3 Plasmid map of pGEM-T vector with Mmp2 ISH probe as insert. The map of the plasmid is generated using A Plasmid Editor software (ApE) made by M. W. Davis (http://biologylabs.utah.edu/jorgensen/wayned/ape/).  105  Suppl Table 1. Genes differentially expressed in embryo versus placenta at E15.5 (Empirical Bayes Wright & Simon adjusted with Bejamini-Hochberg false discovery rate = 0.05) Only genes with 10-fold difference between the embryo and the placenta are listed. Symbol  Fold change Placenta /Embryo  Adjusted p- Definition value  Embryo-specific (FC < 0.1) Acta1 0.01299186 Myl1 0.0188905 MYLPF 0.01997901  4.62E-08 9.03E-08 2.73E-08  TNNC2 ACTC1  0.02548014 0.02595083  1.28E-07 9.29E-08  MYH8  0.02930698  4.56E-07  TNNI2 IGFBP5 AMBP TUBB2B TNNT1  0.03616861 0.04921925 0.05152062 0.05169867 0.05451234  3.02E-07 2.40E-09 1.98E-07 1.47E-08 5.17E-07  SLN ACTN2 LOR MYL2  0.05826783 0.06011932 0.06353214 0.06610431  2.45E-07 5.38E-08 1.04E-05 1.74E-05  PTN GAP43  0.0673354 0.06804696  1.85E-08 3.19E-09  SERPINF1  0.07127633  2.13E-07  D0H4S114 SBK SERPINA1B  0.07186939 0.07198077 0.07279541  1.64E-08 1.77E-08 1.43E-05  CAPN6 LOC100045403  0.07284341 0.07464686  1.42E-08 5.38E-08  6330403K07RIK COL6A1 COL1A1 COL1A2 ATP2A1  0.07644952 0.07763579 0.08120836 0.08505238 0.08513667  1.42E-08 5.66E-08 3.16E-07 1.13E-07 4.41E-07  COX6A2  0.08569458  5.23E-07  AHSG COL5A1  0.08896152 0.08924741  6.79E-06 9.63E-07  actin, alpha 1, skeletal muscle. myosin, light polypeptide 1. myosin light chain, phosphorylatable, fast skeletal muscle. troponin C2, fast. actin, alpha, cardiac muscle 1. myosin, heavy polypeptide 8, skeletal muscle, perinatal. troponin I, skeletal, fast 2. insulin-like growth factor binding protein 5. alpha 1 microglobulin/bikunin. tubulin, beta 2b. troponin T1, skeletal, slow. sarcolipin. actinin alpha 2. loricrin. myosin, light polypeptide 2, regulatory, cardiac, slow. pleiotrophin. growth associated protein 43. serine (or cysteine) peptidase inhibitor, clade F, member 1. DNA segment, human D4S114. SH3-binding kinase 1. serine (or cysteine) preptidase inhibitor, clade A, member 1b. calpain 6. PREDICTED: similar to orthologue of H. sapiens chromosome 21 open reading frame 102 (C20orf102), misc RNA. RIKEN cDNA 6330403K07 gene. procollagen, type VI, alpha 1. collagen, type I, alpha 1 collagen, type I, alpha 2. ATPase, Ca++ transporting, cardiac muscle, fast twitch 1. cytochrome c oxidase, subunit VI a, polypeptide 2, nuclear gene encoding mitochondrial protein. alpha-2-HS-glycoprotein. procollagen, type V, alpha 1.  106  Symbol  Fold change Placenta /Embryo Embryo-specific (FC < 0.1 )continued STFA1 0.09011926 TNNT3 0.09072438 MYL3 0.09089174  Adjusted p- Definition value  5.70E-05 3.55E-07 1.26E-05  stefin A1. troponin T3, skeletal, fast. myosin, light polypeptide 3.  EMID2 STMN2 HIST1H2AH HIST1H2AF SLC4A1  0.09700852 0.09783124 0.09786976 0.09847035 0.09848848  1.47E-08 3.96E-08 4.59E-06 2.65E-06 1.09E-04  EMI domain containing 2. stathmin-like 2. histone cluster 1, H2ah. histone cluster 1, H2af. solute carrier family 4 (anion exchanger), member 1.  GNS LOC100046802 D930020E02RIK SERPINB6B  10.49764 10.56654 10.9562 11.11392  1.80E-07 1.73E-07 4.26E-07 9.28E-07  LOC100041103  11.169  2.73E-07  ADA GPX3 PCGF5 FABP3 GKN2 TCFAP2C KRT8  11.58439 12.13485 12.33669 13.35281 14.43306 14.76052 15.74381  3.10E-07 1.03E-05 7.83E-09 5.33E-08 6.18E-06 7.61E-09 1.03E-07  glucosamine (N-acetyl)-6-sulfatase. PREDICTED: similar to Inhbb protein. RIKEN cDNA D930020E02 gene. serine (or cysteine) peptidase inhibitor, clade B, member 6b. PREDICTED: hypothetical protein LOC100041103. adenosine deaminase. glutathione peroxidase 3, transcript variant 2. polycomb group ring finger 5. fatty acid binding protein 3, muscle and heart. gastrokine 2. transcription factor AP-2, gamma. keratin 8.  SLCO2A1  16.80767  3.45E-08  SERPINB9G  17.28993  1.68E-06  GM2A SLC6A12  17.53276 17.58351  6.70E-08 5.70E-08  GJB2  17.97248  1.27E-07  TNFRSF9  18.27959  8.92E-08  CTSM SLC13A4  18.50106 20.04887  9.18E-09 1.40E-06  SLC38A4 LGALS3 KRT18  20.11131 21.73318 22.2426  5.80E-08 1.92E-07 6.34E-08  tumor necrosis factor receptor superfamily, member 9, transcript variant 1. cathepsin M. solute carrier family 13 (sodium/sulfate symporters), member 4. solute carrier family 38, member 4. lectin, galactose binding, soluble 3. keratin 18.  PRL4A1 PLAC8 CAR4 PSG16  22.9778 23.09103 23.13739 23.19711  2.34E-06 3.78E-08 2.65E-07 1.69E-07  prolactin family 4, subfamily a, member 1. placenta-specific 8. carbonic anhydrase 4. pregnancy specific glycoprotein 16.  Placenta-specific (FC > 10)  solute carrier organic anion transporter family, member 2a1. serine (or cysteine) peptidase inhibitor, clade B, member 9g. GM2 ganglioside activator protein. solute carrier family 6 (neurotransmitter transporter, betaine/GABA), member 12. gap junction protein, beta 2.  107  Symbol  Fold change Placenta /Embryo Placenta-specific (FC > 10) continued RHOX6 24.72437 PRL2C5 25.37251 PRLPN 27.94913  Adjusted p- Definition value  3.55E-07 2.73E-08 7.03E-06  reproductive homeobox 6. prolactin family 2, subfamily c, member 5. prolactin family 7, subfamily b, member 1.  PRL2A1 TAF7L  28.20164 30.31489  3.55E-07 1.21E-11  CEACAM14 TPBPB PRL8A2 CTS3  31.02065 31.63772 32.65128 38.49477  2.76E-08 9.82E-08 5.57E-06 1.88E-07  prolactin family 2, subfamily a, member 1. TAF7-like RNA polymerase II, TATA box binding protein (TBP)-associated factor. CEA-related cell adhesion molecule 14. trophoblast specific protein beta. prolactin family 8, subfamily a, member 2. cathepsin 3.  RARRES2  39.80894  2.93E-07  CTSR RHOX5 PSG19 OTTMUSG00000000651 PSG18  43.80073 44.52213 44.61459 46.145 53.50965  2.71E-08 6.39E-10 1.31E-10 1.88E-07 5.77E-08  retinoic acid receptor responder (tazarotene induced) 2. cathepsin R. reproductive homeobox 5. pregnancy specific glycoprotein 19. predicted gene, OTTMUSG00000000651. pregnancy specific glycoprotein 18.  PSG25 CEACAM12 SCT GHRH CTS6 PSG26 PSG23  71.48119 73.25169 74.85725 94.85656 100.2826 109.1666 111.8458  2.36E-09 7.93E-09 2.87E-08 4.78E-08 1.25E-10 1.22E-08 1.32E-09  pregnancy-specific glycoprotein 25. CEA-related cell adhesion molecule 12. secretin. growth hormone releasing hormone. cathepsin 6. pregnancy-specific glycoprotein 26. pregnancy-specific glycoprotein 23.  RHOX9 PRLPC3 PRL2B1 PSG27 PRL2C4 LOC381852  123.7153 129.8168 149.9389 153.0084 158.1028 164.0082  2.40E-09 2.36E-09 2.56E-09 4.03E-12 1.12E-08 4.03E-12  PRL3B1 PRL2C3 CTSQ CEACAM11  176.3235 187.1391 187.901 189.0166  2.92E-08 3.01E-08 0 2.13E-09  CTSJ  190.1107  3.58E-09  reproductive homeobox 9. prolactin-like protein C 3. prolactin family 2, subfamily b, member 1. pregnancy-specific glycoprotein 27. prolactin family 2, subfamily c, member 4. similar to carcinoembryonic antigen-related cell adhesion molecule 3. prolactin family 3, subfamily b, member 1. prolactin family 2, subfamily c, member 3. cathepsin Q. carcinoembryonic antigen-related cell adhesion molecule 11. cathepsin J.  108  Suppl Table 2. Differentially expressed genes in E15.5 IUGR samples (EB Wright & Simon and cyber-T both adjusted with Benjamini-Hochberg false discovery rate = 0.05). Fold change ratio is presented as raw signal intensity of WT Bl6 samples over the raw signal intensity of IUGR samples. Symbol  mChr  Fold change IUGR vs non-IUGR GO Biological Process  Embryo (> 2-fold difference) BC030476 DEK OTTMUSG00000007855 PTPRE ATP5A1  15 13 4 7 18  0.291509 0.345345 0.346944 0.351035 0.351811  Unknown (Interact with DNA) Unknown Dephosphorylation; signalling Cellular biosynthesis  VPS26B COL1A2 GAP43 ATP5E SETX LOC100044087 2210412D01RIK  9 6 16 2 2 10 7  0.376267 0.376994 0.379969 0.397288 0.405491 0.415231 0.415548  Transport of proteins Signalling Nervous system development ATP synthase; cellular metabolism DNA damage Unknown Unknown  NDUFB9 DUSP7 SDHD CRYGA PRF1 ATP6V1A  15 9 9 1 10 16  0.42548 0.443806 0.44595 0.456604 0.457993 0.460898  COL3A1 GNB2L1 EIF2S3Y NDUFB5 1110002B05RIK AFP AI314180  1 11 Y 3 12 5 4  0.464736 0.469798 0.472128 0.472291 0.472334 0.478028  ETC Dephosphorylation; Signalling Transport of iron Development of the eye Cell death ATP synthase; cellular metabolism Blood vessel formation; Digestive system development Phosphorylation; Signalling Translation ETC Unknown Transport of copper Unknown  THSD4 CCNG2 ZC3H15 TOMM70A ALDH6A1 LOC100048622  9 5 2 Unknown 12 Unknown  0.478174 0.482467 0.483429 0.48619 0.487987 0.490724  Hydrolysis of carbon bonds Cell cycle Signalling Signalling Cellular metabolism Unknown  9130005N14RIK NDUFB2 GLTP  5 6 5  0.491016 0.498272 2.003372 2.018775 2.018933  Unknown ETC Transport of glycolipid  KRT2-1 HIST1H2BF  Unknown 13  0.461502  Unknown DNA packaging  109  Symbol  mChr  Embryo (> 2-fold difference) continued EG433923 5 LOC100046918 Unknown PTMS 6 HIST1H2BH TACSTD2 DMKN Placenta (> 2-fold difference)  13 6 7  Fold change IUGR vs non-IUGR GO Biological Process 2.043626 2.120295 2.161152 2.170868  Unknown Unknown Immune reponse  2.274034 3.163002  DNA packaging Unknown Cell Differentiation  PRL8A9 ALDH1A3 SLC38A4  13 7 15  0.307782 0.313594 0.325339  Signaling (hormone) Anatomical Structure Development Transport of amino acids  1600029D21RIK HSPA5 SERPINB6B CEACAM13 PRL3C1 PTPRA GKN1  9 2 13 7 13 2 6  0.400365 0.404304 0.425737 0.437663 0.449103 0.4509 0.464278  Unknown ER stress response Unknown Unknown Signalling (hormone) Dephosphorylation; signalling Cell proliferation  MGAT4A SRP14 DDX6 NID2 GPX3 DLK1  1 2 9 14 9 12  0.465189 0.48053 0.496593 2.054392 2.190024 2.841104  Carbonhydrate synthesis Repress translation DNA unwinding Cell adhesion Cellular metabolism Embryonic skeletal development  > 2-fold difference in both Target ID VAPA COL1A1  mChr 17 11  Fold change in Fold change in IUGR embryo IUGR placenta GO Biological Process 0.401791 0.418898 (Structural molecule) 2.089283 2.154194 Skeletal development  VHL  6  0.465995  0.289318  PTGES3 ACADSB IDH2  10 7 7  0.315851 0.326914 0.460251  0.38242 0.38917 0.472428  STMN1 HUWE1  4 X  0.169451 0.278906  0.495621 0.400182  H1F0 GKAP1 KLF6  15 13 13  0.301818 0.400632 0.407415  0.463918 0.38971 0.349256  BACH1  16  0.411779  0.347331  Angiogenesis; blood endothelial migration; hypoxic response Cell proliferation; Fatty acid synthesis Fatty acid β-oxidation TCA Cycle Cellular component organization; Cellular metabolic process DNA packaging Signalling Signalling Regulate transcription 110  Symbol  mChr  Fold change IUGR vs non-IUGR GO Biological Process  > 2-fold difference in both continued TRRAP 5 POLR2G 19 PAIP2 18  0.49338 0.402921 0.421414  0.441657 0.422382 0.485043  CALR POM121 KPNA3 KCTD3 NPC2 RP23-297J14.5  8 5 14 1 12 11  2.022648 0.357 0.320142 0.43465 0.430276 0.461336  2.568425 0.47367 0.392815 0.418345 0.374513 0.370887  LOC100041703 LOC668837  2 14  0.367259 0.205998  0.406665 0.336597  ZBED4 DDX3X LOC100044779  15 X Unknown  0.411543 0.299251 0.33  0.431737 0.327402 0.341703  Chromatin modification Transcription Repress translation Actin organization; repress translation; regulate meiosis Transport of protein Transport of protein Transport of potassium ion Transport of cholesterol Unknown Unknown Unknown Unknown Unknown Unknown  111  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items