PRENATAL ALCOHOL EXPOSURE PROGRAMS STEADY-STATE GENE EXPRESSION AND THE GENE EXPRESSION RESPONSE TO INFLAMMATION IN THE ADULT RAT BRAIN

by Katarzyna Anna Stepien

B.Sc., The University of Guelph, 2009

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Medical Genetics)

THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)

July 2013

© Katarzyna Anna Stepien, 2013 Abstract

Prenatal alcohol exposure results in alterations in numerous physiological systems, including neuroendocrine and neuroimmune systems. The purpose of this study was to determine whether prenatal ethanol exposure results in long-term alteration of neural gene expression, particularly in genes related to neuroendocrine and neuroimmune function. Utilizing a well-established animal model of prenatal ethanol exposure, ethanol was administered to pregnant Sprague-Dawley dams throughout gestation in a liquid diet fed ad libitum (36% calories derived from ethanol). Maltose-dextrin was isocalorically substituted for ethanol in a liquid control diet for a pair-fed group, and a control group received a pelleted control diet ad libitum. In young adulthood, an adjuvant-induced arthritis paradigm was utilized, where female offspring were injected with either saline or complete Freund’s adjuvant, to induce an inflammatory response and elucidate dysregulated neuroimmune pathways. Gene expression was analyzed in the prefrontal cortex and hippocampus at both the peak and resolution of arthritis using whole genome gene expression microarrays. Within saline-injected animals, prenatal alcohol exposure alone resulted in significant changes in gene expression in both the prefrontal cortex and hippocampus. Included were multiple genes related to, cell death, transcriptional regulation, neuronal signaling and neurodevelopment. Among the genes involved in neurodevelopment, Acs13 has also been shown to be variably methylated in humans according to in utero exposure to environmental factors. Prenatal alcohol exposure also altered the gene expression response to adjuvant-induced arthritis. Many genes showed a significantly different pattern of expression in ethanol-exposed animals compared to both pair-fed and control, in both prefrontal cortex and hippocampus. These genes were either differentially up- or downregulated in ethanol-exposed compared to control animals or failed to show the adjuvant-induced change in regulation shown by controls. As well, several of these genes were mediators of the response to immune or stress challenge, such as Lcn2 and Bhlhe40. Genes found to be differentially expressed in this study are potential mediators contributing to the long-term alterations in neuroendocrine and neuroimmune function observed in prenatal alcohol exposure.

ii Preface

The animal work described in this thesis (breeding, handling, and termination) was conducted primarily by Xinqi Zhang, with assistance from other members of the Weinberg lab. The tissues used for gene expression analysis were dissected by Tamara Bodnar, Linda Ellis, and Kasia Stepien. Subsequent lab work, analyses, and writing were conducted primarily by Kasia Stepien, with some help from Sarah Neumann in optimization of the RNA extraction and amplification protocols.

All animal work approved by the UBC Animal Care Committee and conducted under the Animal Care Certificate number A05-1187.

iii Table of Contents

Abstract ...... ii Preface ...... iii Table of Contents ...... iv List of Tables ...... vi List of Figures ...... vii List of Abbreviations ...... viii Acknowledgements ...... ix Dedication ...... x Chapter 1: Introduction ...... 1 1.1 Prenatal alcohol exposure and Fetal Alcohol Spectrum Disorders ...... 1 1.1.1 Clinical significance of prenatal alcohol exposure ...... 1 1.1.2 Animal models of prenatal alcohol exposure ...... 2 1.2 Fetal programming by prenatal alcohol exposure ...... 3 1.2.1 Fetal programming: the developmental origins of health and disease ...... 3 1.2.2 Long term effects of PAE on the stress response ...... 4 1.2.3 Long term effects of PAE on the immune response ...... 6 1.3 Alterations in neural gene expression: a potential mechanism of fetal programming by prenatal alcohol exposure ...... 7 1.4 Rationale and thesis objectives ...... 9 Chapter 2: Materials and Methods ...... 10 2.1 Breeding and prenatal alcohol exposure ...... 10 2.2 Induction of adjuvant-induced arthritis ...... 10 2.3 Termination of animals ...... 10 2.4 Tissue dissection and RNA extraction ...... 11 2.5 Microarray assaying of whole genome gene expression ...... 11 2.6 Data pre-processing, quality control, and exploratory data analysis ...... 11 2.7 Differential expression analysis ...... 12 2.8 Gene ontology and pathway analysis ...... 12 2.9 Validation of microarray results ...... 13 Chapter 3: Results ...... 15 3.1 Overview of analyses and main findings ...... 15 3.2 Exploratory data analysis...... 16 3.3 Prenatal alcohol exposure alters steady-state gene expression in PFC and HPC ...... 26 3.3.1 Genes altered by PAE at a steady-state level of gene expression ...... 26 iv 3.3.2 GO categories altered by PAE at a steady-state level of gene expression ...... 27 3.3.3 Validation of gene expression changes by RT-qPCR ...... 27 3.3.4 Genes showing common, graded, or differential effects of PAE and pair-feeding ...... 28 3.4 Prenatal alcohol exposure alters the gene expression response to an inflammatory challenge in PFC and HPC ...... 44 3.4.1 Incidence and severity of adjuvant-induced arthritis ...... 44 3.4.2 Genes differentially altered in PAE compared to control animals in response to adjuvant exposure ...... 44 3.4.3 GO categories differentially altered in PAE compared to control animals in response to adjuvant exposure ...... 45 Chapter 4: Discussion and Conclusion ...... 50 4.1 Effects of prenatal alcohol exposure on steady-state gene expression ...... 50 4.2 Effects of prenatal alcohol exposure on the neural response to adjuvant-induced arthritis ...... 52 4.3 Overlapping effects of prenatal alcohol exposure and pair-feeding on steady-state gene expression...... 54 4.4 Unique effects of pair-feeding on steady-state gene expression...... 55 4.5 Limitations and future directions ...... 55 References ...... 57 Appendix A Supplementary Tables...... 65

v List of Tables

Table 2.1. Final number of animals in each treatment condition...... 14 Table 3.1. Correlation of overall expression profies among all arrays and among replicates arrays ...... 20 Table 3.2. Number of surrogate variables generated by SVA for each analysis group ...... 25 Table 3.3. Genes differentially expressed in prefrontal cortex of E vs both PF and C animals...... 33 Table 3.4. Genes differentially expressed in hippocampus of E vs both PF and C animals...... 34 Table 3.5. Sequences of primers used for RT-qPCR ...... 34 Table 3.6. Microarray expression results for common reference genes in PFC of Day 16 Saline animals...... 35 Table 3.7. Microarray expression results for common reference genes in HPC of Day 16 Saline animals...... 35 Table 3.8. Genes showing common change in expression in prefrontal cortex of E and PF animals compared to C animals...... 37 Table 3.9. Genes showing common change in expression in hippocampus of E and PF animals compared to C animals...... 38 Table 3.10. Genes differentially expressed in prefrontal cortex among all 3 contrasts...... 39 Table 3.11. Genes differentially expressed in hippocampus among all 3 contrasts...... 40 Table 3.12. Genes differentially expressed in PFC of PF animals vs both E and C animals...... 42 Table 3.13. Genes differentially expressed in HPC of PF animals vs both E and C animals...... 43 Table 3.14. Genes altered in PFC of E animals in response to adjuvant at peak of inflammation...... 48 Table 3.15. Genes altered in HPC of E animals in response to adjuvant at peak of inflammation...... 48 Table A.1. Candidate genes involved in the etiology of FASD, catalogued in Neurocarta ...... 65 Table A.2. Accessions and probe sequences for differentially expressed genes ...... 67

vi List of Figures

Figure 2.1. Experimental model ...... 14 Figure 3.1. Overview of analyses and main findings for gene expression analysis ...... 15 Figure 3.2. Correlation heatmap of expression profiles among all array in the PFC dataset...... 18 Figure 3.3. Correlation heatmap of expression profiles among all array in the HPC dataset...... 19 Figure 3.4. Spatial artefacts on the PFC outlier arrays...... 20 Figure 3.5. Correlation among hybridization replicate arrays in PFC...... 21 Figure 3.6. Evidence for batch effects in Principal Components Analysis...... 22 Figure 3.7. RNA integrity varied between dissection batches...... 23 Figure 3.8. Proportion of variance of Principal Components...... 23 Figure 3.9. Gene expression correlation before and after quantile normalization...... 24 Figure 3.10. Density plots of p-value distributions...... 30 Figure 3.11. Genes showing a significant effect of prenatal diet at Day 16 post-saline injection...... 31 Figure 3.12. Venn diagrams of the number of probes significantly altered in each prenatal treatment contrast...... 32 Figure 3.13. RT-qPCR expression levels for genes altered by prenatal alcohol exposure...... 36 Figure 3.14. Biological Processes altered by prenatal treatment...... 41 Figure 3.15. P-value distributions for response to adjuvant exposure...... 46 Figure 3.16. Effects of Adjuvant exposure on gene expression at the peak of inflammation...... 47 Figure 3.17. Biological Processes altered in the response to adjuvant exposure...... 49

vii List of Abbreviations

ACTH – adrenocorticotropic hormone AA – adjuvant-induced arthritis ARBD – alcohol-related birth defects ARND – alcohol-related neurodevelopmental disorder BAL – blood alcohol level BLAST – Basic Local Alignment Search Tool C – control (group) cDNA – complementary DNA CNS – central nervous system E – ethanol-exposed (group) FAS – Fetal Alcohol Syndrome FASD – Fetal Alcohol Spectrum Disorder FDR – false discovery rate GO – Gene Ontology GR – glucocorticoid receptor HPA – hypothalamic pituitary adrenal (axis) HPC – hippocampus KEGG – Kyoto Encyclopedia of Genes and Genomes mRNA – messenger RNA NCBI – National Center for Biotechnology Information PAE – prenatal alcohol exposure PF – pair-fed (group) PFC – prefrontal cortex SVA – surrogate variable analysis

viii Acknowledgements

Thank you to my advisors, Dr. Joanne Weinberg and Dr. Michael S. Kobor, for their scientific guidance, mentorship, and persistent motivation, without which this research would not have been possible. To Joanne, I owe additional thanks for her moral support and guidance.

Thank you to Dr. Paul Pavlidis, for the collaborative support and wealth of experience that helped guide me through the microarray analysis.

Thank you to my fellow students and colleagues in the Weinberg and Kobor labs. The friendships developed have been just as valuable as the professional support and assistance.

Finally, thank you to Gregory Baute, for continually providing encouragement and perspective.

ix Dedication

To all the courageous individuals living and working with FASD, and to the hope that one day this will be a condition known only in the past

x Chapter 1: Introduction

1.1 Prenatal alcohol exposure and Fetal Alcohol Spectrum Disorders

1.1.1 Clinical significance of prenatal alcohol exposure

Prenatal alcohol exposure (PAE) is a leading cause of neurodevelopmental disorder in the western world. PAE has well-documented lasting detrimental effects on growth, physiology, and multiple neurological domains, including cognitive function, self-regulation, and adaptive functioning. Maternal alcohol consumption during pregnancy can result in a spectrum of effects, which are collectively known as fetal alcohol spectrum disorders (FASD). The adverse effects of PAE on human development were first described by the French pediatrician Paul Lemoine in 1968 (Lemoine et al. 1968) and shortly thereafter in North America by Jones and colleagues (D. W. Smith et al. 1973; Kenneth L Jones & D. W. Smith 1973), who coined the term Fetal Alcohol Syndrome (FAS). FAS lies at the more severe end of the FASD spectrum, and is characterized by a distinct set of facial abnormalities, growth deficiencies, and central nervous system abnormalities (Chudley et al. 2005). Also on the spectrum are alcohol related effects, namely alcohol-related birth defects (ARBD) and alcohol-related neurodevelopmental disorders (ARND), which each share some commonality with FAS (Stratton et al. 1996). ARBD is characterized by physical congenital abnormalities resulting from prenatal alcohol exposure, whereas ARND consists of neurodevelopmental, behavioural, and/or cognitive abnormalities (Chudley et al. 2005; Stratton et al. 1996).

In spite of the fact that FASD is a preventable condition, the prevalence is quite high. In the USA, the prevalence of FASD has been estimated to be at least 9.1 per 1000 live births, or approximately 1% (Sampson et al. 1997). More recent studies based on in-school assessments of children have suggested that the prevalence in the typical western population is even as high as 2-5%, or even more given that ARND is likely to be under-diagnosed (May et al. 2009; May et al. 2011). Given that 5.4% of women in Ontario and 7.2% in British Columbia reported drinking during a recent pregnancy in the 2007/2008 Canadian Community Health Survey (Thanh & Jonsson 2010), this prevalence rate is not implausible in Canada. The estimated annual cost to Canada of FASD, based on a prevalence of 1% of the population, is $5.3 billion (Stade et al. 2009). FASD is therefore a pressing public health concern, and an important area of research for the identification of biomarkers and elucidation of the etiology that might help diagnose and provide appropriate care to affected individuals.

1 1.1.2 Animal models of prenatal alcohol exposure

Animal models of prenatal alcohol exposure have been crucial to understanding its effects on the developing fetus. Historically, animal models were necessary to demonstrate that ethanol is in fact a potent teratogen, as cases of FAS are frequently wrought with confounds such as malnutrition, exposure to other illicit substances used by the mother, and low socioeconomic status. Some of the first models of PAE following identification of FAS were developed with rats and mice (Bond & Di Giusto 1976; Chernoff 1977; Randall et al. 1977; Abel & Dintcheff 1978; Brown et al. 1979), as well as sheep (Mann et al. 1975; Kirkpatrick et al. 1976). These studies were among the first to demonstrate that ethanol exposure itself is teratogenic to the fetus, resulting in increased mortality, growth retardation, malformations, and behavioural changes in offspring. Correlations between maternal blood alcohol level and severity of outcomes were established (Chernoff 1977; Randall & Taylor 1979), and seminal work by Sulik and colleagues demonstrated that animal models of PAE could replicate specific components of FAS, in this case the unique facial dysmorphology (Sulik et al. 1981). As the field grew, models expanded to include many other organisms in addition to rodents and sheep, and to attempt to identify the effects of timing, dose, and pattern of exposure on the development of FASD. As our understanding of alcohol’s diverse teratogenic effects has grown over the past decades, the focus of FASD research has extended to examining the molecular mechanisms of teratogenicity, identifying biomarkers of exposure and FASD, and understanding how genes and the environment may alter vulnerability to alcohol’s teratogenic effects.

Many different models of prenatal alcohol exposure have been developed for the rodent, to target different patterns, doses, and times of exposure. In the present study, a well-established model for chronic prenatal alcohol exposure was used (Weinberg et al. 2008). This model uses a liquid diet containing 36% calories derived from ethanol, fed ad libitum to pregnant Sprague-Dawley dams for the entire duration of gestation. This model has many benefits, as it does not require stressful handling of the dam for administration of ethanol, it targets the full term of prenatal development in the rat (equivalent to the first two trimesters of human gestation), and it produces moderate blood alcohol levels (BALs) in the dam that are realistic exposures relative to human exposure. One complication of this model, however, is that ethanol-consuming dams reduce their food intake relative to what they would eat in an equivalent non-alcoholic diet. To control for this, two important factors are controlled. First, the diet has been developed to be nutritionally complete after taking into account an anticipated reduced food intake (Weinberg 1985). Second, a pair-fed group is added in addition to an ad libitum fed control group (Weinberg 1984). This pair-fed group receives a liquid control diet that is isocaloric to the ethanol diet (with maltose-dextrin substituted for ethanol), fed in an amount equivalent to that consumed by an ethanol-fed partner (g/kg body weight/day of gestation). This provides control for the effects of reduced food intake in the ethanol-fed dams. However, pair-feeding also has unique effects in itself. Notably, because pair-fed animals receive a reduced ration of food, they are

2 typically hungry, and tend to eat their daily ration of food quickly. As a result, pair-feeding may in itself be a mild stressor (Vieau et al. 2007), not experienced by the ad libitum fed groups. In addition, time of feeding becomes a stronger factor than the light/dark cycle for entraining the circadian rhythm of the HPA axis (Gallo & Weinberg 1981), and feeding therefore needs to occur close to the onset of the dark-cycle to maintain a normal circadian rhythm of corticosterone in pair-fed animals. Relatively rapid consumption of daily rations, followed by long periods without food, may also impose unique metabolic effects. Therefore having both a pair-fed group and an ad libitum fed control group is important, and effects in offspring of ethanol-fed dams should be compared to both and interpreted with care.

1.2 Fetal programming by prenatal alcohol exposure

In addition to the well-established effects of prenatal alcohol exposure on cognitive function and morphological development, it has lasting impacts on many physiological systems. Two systems susceptible to long-term alteration by alcohol exposure include the immune system and the hypothalamic-pituitary- adrenal (HPA) axis (Bodnar & Weinberg 2013). While prenatal alcohol exposure may directly alter the developmental trajectories of the brain, through its toxic impact on structural and developmental trajectories during brain development, how it induces lasting changes in such physiological systems is less clear. These systems are intimately interconnected, and work together with the central nervous system to regulate the body’s response to challenge (Glaser & Kiecolt-Glaser 2005). A recent concept that has entered into FASD research is that of fetal programming as a mechanism to induce lasting change, and it is possible that programming of the central nervous system by PAE may impact both HPA and immune function

1.2.1 Fetal programming: the developmental origins of health and disease

Fetal programming refers to the ability of environmental factors to transmit signals to the developing fetus, promoting changes in development and developmental trajectories, resulting in lasting alterations in structure or function of physiological and behavioral systems (Gluckman et al. 2008). Many environmental factors, such as stress, undernutrition, endocrine disruptors, or infection, have the ability to alter risk for disease in adulthood when an individual is exposed as a fetus. The concept of fetal programming arose principally from studies of human epidemiology, in which Barker and colleagues noted that low birthweight was associated with increased risk of hypertension, type II diabetes, and cardiovascular disease in adulthood (Barker & Osmond 1986; Barker et al. 1989; Hales et al. 1991). It is theorized that the flexibility of the fetal developmental program may provide advantages, allowing the developing organism to adapt to the environment into which it will be born. For example, many species, both vertebrate and invertebrate, exhibit predator-induced polyphenism, which is the ability to develop in to an alternate adult form in the presence of predators, that provides protection and increases chances of survival (Gilbert 2012). The water flea Daphnia

3 pulex develops defensive head morphology when predatory larvae share their environment, decreasing their chances of being consumed (Krueger & Dodson 1981). In humans, undernutrition during fetal development is hypothesized to result in a “thrifty phenotype” that is better able to cope with a nutritionally limited postnatal environment (Hales & Barker 1992). However, such adaptations may be detrimental in the case where the prenatal and postnatal environments do not match well, and the phenotype may then be maladaptive. Similarly, the programmability of physiological systems may be hijacked by teratogens, such as alcohol, to the detriment of the developing individual. These are only a few examples of the many emerging concepts in fetal programming and its effects on development.

1.2.2 Long term effects of PAE on the stress response

The HPA axis, or stress response system, is one physiological system that has been shown to be vulnerable to alteration by the prenatal and early postnatal environment, including prenatal alcohol exposure. The HPA axis is one of the main mediators of the body’s response to stress. In response to a variety of stressful stimuli, the HPA axis produces a hormonal cascade that ultimately results in the release of glucocorticoids into circulation, which promote appropriate responses to stimulate survival (reviewed in Myers et al. 2012). In more detail, when the HPA axis is activated by a stressor, corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP) are released from the paraventricular nucleus (PVN) of the hypothalamus. CRH and AVP both act on the anterior pituitary gland to stimulate the release of adrenocorticotropic hormone (ACTH) into the systemic circulation. ACTH then acts on the adrenal glands to promote the release of glucocorticoids from the adrenal cortex. Glucocorticoids have rapid effects that promote survival, such as increasing gluconeogenesis, as well as suppressing energetically expensive processes not immediately needed, such as reproduction and immune function (Myers et al. 2012). Importantly, glucocorticoids also provide negative feedback to multiple levels of the HPA axis, promoting timely and efficient resolution of the stress response.

The HPA axis, however, does not operate in isolation. Its activity is regulated in part by higher brain regions, such as the prefrontal cortex (PFC), the hippocampus (HPC), and the amygdala (reviewed in Jankord & Herman 2008). The limbic system is particularly important in mediating the response to anticipatory stressors, such as stressors requiring evaluation of complex environmental information and comparison with memories or instincts (versus stressors that present a direct homeostatic threat). Both the HPC and prelimbic PFC provide inhibitory input to the HPA axis, helping to control the extent of the stress response and promote resolution following a stressor. The hippocampus has relatively high levels of glucocorticoid receptors in the rodent brain (Reul & De Kloet 1985), which are the main receptor involved in negative feedback of the stress response mediated by glucocorticoids. Lesion of the hippocampus has been shown to elevate glucocorticoid levels, disrupt diurnal rhythms of glucocorticoid secretion, and prolong the glucocorticoid response to

4 stressors (Jankord & Herman 2008). Other brain regions, such as the amygdala and infralimbic PFC, provide stimulatory input to the HPA axis, promoting onset or extending the stress response (Jankord & Herman 2008). The regulation imparted by multiple limbic regions, and the variability in their roles and the stimuli to which they respond, provides a flexibility and adaptability to the stress response that is integral to responding appropriately to specific types of stressors. Importantly, many of these brain regions, such as the hippocampus and PFC, are highly susceptible to the teratogenic effects of alcohol exposure, providing a target for alcohol to affect the development of the stress response. Of interest, both the PFC and HPC are susceptible to the teratogenic effects of alcohol, and have been shown to be altered in individuals with FASD (Norman et al. 2009) and in animal models of PAE (Gil-Mohapel et al. 2010). In addition to the roles of these two brain regions in behaviour, learning, and memory, changes induced by PAE in these regions may contribute to dysregulation of the HPA axis.

Exposure to many different factors has been shown to alter development and set-point of the stress response. Prenatal exposures to maternal stress, excess glucocorticoids, undernutrition, and alcohol, have all been shown to increase the responsiveness of the stress response in later life. PAE has been shown to have long- term effects on the stress response in both humans and animal models. Infants exposed prenatally to alcohol have been found to have elevated basal and post-stress levels of cortisol (Ramsay et al. 1996; Jacobson et al. 1999; Haley et al. 2006), though normal levels of plasma cortisol have also been documented (Root et al 1975). Animal models have corroborated the more common findings of elevated HPA activity and elucidated many of the levels at which the HPA axis is dysregulated. While basal levels of stress hormones such as glucocorticoids are often not altered in PAE animals, they are typically elevated compared to control animals following exposure to a stressor. This holds true for a variety of stressors and animal models, and for both male and female offspring. Increased levels of glucocorticoids, ACTH, or β-endorphin have been demonstrated in PAE animals in response to stressors such as foot shock, cold stress, restraint stress, drug exposures, and immune challenges (reviewed in Hellemans et al. 2010). Interestingly, responses can be sexually dimorphic depending on the endpoint measured and the stressor. For example, hyper-responsiveness to prolonged or intense restraint and cold stress occurs primarily in male rats (Weinberg 1992; Giberson et al. 1997; Kim et al. 1999), whereas hyper-responsiveness to acute restraint stress, and acute ethanol or morphine exposure occurs primarily in females (Taylor et al. 1982; Taylor et al. 1983; Taylor et al. 1988; Weinberg 1985; Gallo & Weinberg 1986).

Though changes in basal levels of these hormones are not typically evident, central regulation of the stress response has been shown to be altered in PAE animals. Studies have particularly shown that PAE animals have increased central drive to the HPA axis. Increased levels of CRH have been found in the hypothalamus of rats at both weaning and in adulthood (Redei et al. 1989; Lee et al. 1990; Lee & Rivier 1996; Gabriel et al. 2005). Studies using adrenalectomy to remove the corticosterone feedback signal have uncovered further

5 evidence of central dysregulation of basal HPA axis activity. Following adrenalectomy, PAE animals demonstrate increased basal ACTH (Glavas et al. 2001), and increased CRH mRNA (Glavas et al. 2007). In addition to increased drive to the HPA axis, PAE animals have been shown to exhibit deficits in negative feedback to the HPA axis (Hellemans et al. 2010).

1.2.3 Long term effects of PAE on the immune response

The immune system is another physiological system that is susceptible to alteration by PAE. Children and infants exposed prenatally to alcohol have an increased susceptibility to both minor and major infections, including upper respiratory tract infections, pneumonia, recurrent otitis media, and sepsis (Johnson et al. 1981; Streissguth et al. 1985; Gauthier et al. 2004; Gauthier et al. 2005; Church & Gerkin 1988). Gauthier and colleagues have studied infection risk in newborns, and have found that among very low birth weight premature infants, exposure to alcohol resulted in a 15-fold increase in incidence of sepsis shortly after birth (Gauthier et al. 2004). In full-term newborns, alcohol exposure at a level of 7 or more drinks per week in the second and third trimesters increased risk of neonatal infection 3- to 4-fold (Gauthier et al. 2005). Increased susceptibility to immune challenges is reflected at a cellular level in children with FASD, with lower counts of lymphocytes and reduced mitogen-stimulated proliferative responses compared to unexposed children (Johnson et al. 1981).

Animal models have substantiated these observations. Many studies have found increased susceptibility to immune challenges and alterations in adaptive immunity (Bodnar & Weinberg 2013). Changes in rodents, for example, include delayed thymic ontogeny (Ewald & Walden 1988), decrease thymus size, weight, and counts of T cells in the thymus (Ewald & Frost 1987; Ewald & Walden 1988; Weinberg & Jerrells 1991), as well as delays in B cell development (Moscatello et al. 1999). Many studies have also shown that alcohol exposed offspring show decreased proliferative responses of T cells in response to mitogens such as concanavalin A and interleukin-2 (IL-2) in the near-term fetus (Ewald & Frost 1987) as well as into adolescence and adulthood (Norman et al. 1989; Redei et al. 1989; Weinberg & Jerrells 1991; Norman et al. 1991; Jerrells & Weinberg 1998). In line with the observations of immunosuppressive effects of PAE, increased susceptibility to bacterial and viral infections have been observed in non-human primates (Grossmann et al. 1993) and in rodents (Seelig et al. 1996; Gauthier et al. 2009; McGill et al. 2009).

While a large body of research has demonstrated that PAE can have immunosuppressive effects, it has also been shown to increase the inflammatory response. Recent work from our lab has demonstrated that PAE increased the incidence, severity, and course of inflammation in a rat model of adjuvant-induced arthritis (Zhang et al. 2012). Increased levels of proinflammatory cytokines have been observed in PAE offspring in response to LPS (Zhang et al. 2005), and embryos exposed to alcohol in vitro show increased levels of the

6 proinflammatory cytokines TNFα, and IL-6 (Vink et al. 2005). This increase in proinflammatory cytokine profiles (Zhang et al. 2005; Vink et al. 2005) may contribute to the increased risk for developing conditions such as arthritis (Zhang et al. 2012). The impact of alcohol exposure in utero on the immune response is therefore not simply a case of immunosuppression, and multiple levels of immune regulation are likely affected.

Stress, both physical and psychological, is another factor well known to increase susceptibility to infection and disease (reviewed in Glaser & Kiecolt-Glaser 2005), and it is possible that the increased HPA axis activity observed in PAE animals contributes to their increased susceptibility to immune dysfunction. The immune system, the HPA axis, and the central nervous system are intimately connected. Bidirectional communication exists among all three systems, and they share many ligands and receptors (Glaser & Kiecolt- Glaser 2005). For example, glucocorticoids produced by the HPA axis can inhibit immune functions. Similarly, cytokines produced by immune cells can in turn stimulate HPA activity. It therefore is possible that alterations in both immune and HPA responsiveness in alcohol-exposed offspring may have a common neuroendocrine origin, and changes in one system may feed back on the others. For example, interactive effects of stress and prenatal alcohol exposure have been documented on immune function, where postnatal exposure to stress has been shown to exacerbate the effects of PAE on immunity. In male PAE rats, chronic intermittent stress exposure in adulthood reduced specific T cell populations relative to controls (Giberson and Weinberg 1995), and in PAE females, one day of cold stress was found to increase mitogen-induced lymphocyte proliferation compared to stressed controls (Giberson et al. 1997). Consistent with the adverse effects of prenatal alcohol exposure on the risk for arthritis in female rats (Zhang et al. 2012), chronic stress has been shown to increase the risk for adjuvant-induced arthritis in male rats (Seres et al. 2002), both of which may be due in part to increased HPA activity (as is often seen with prenatal alcohol exposure).

1.3 Alterations in neural gene expression: a potential mechanism of fetal programming by prenatal alcohol exposure

The development and continuing function of the brain depends very much on the coordinated and appropriate expression of its transcriptome (the entire set of RNA molecules transcribed in the brain). During embryonic development, expression of signaling molecules called morphogens generates inductive signaling pathways that guide development of the organism (Nahmad & Lander 2011). The specific pattern and timing of their expression generates concentration gradients that induce different cellular responses, typically changes in gene expression, depending on the morphogen concentration, length of exposure, and transduction of the signal (Nahmad & Lander 2011). The sonic hedgehog (Shh) gene, for example, encodes a peptide hormone that induces different patterns of gene expression and thus different patterns of cellular differentiation in the developing vertebrate nervous system and limbs in a concentration and time-dependent manner (Nahmad &

7 Lander 2011). In neural progenitor cells, the concentration of Shh and the length of exposure results in changes in the cellular transcriptome, such as induction of transcription factors like Pax-6 (Ericson et al. 1997; Dessaud et al. 2007), which contributes to the establishment of spatially and functionally distinct neural progenitor domains. Distinct transcriptional profiles are exhibited across brain regions and cell types, during fetal development and in adulthood (Nelson et al. 2006; Oldham et al. 2008; Johnson et al. 2009), and changes in gene expression in the brain can result in cascades of cellular, physiological, and behavioural changes.

The developing brain is the organ most susceptible to the teratogenic effects of ethanol. The effects of PAE on functions related to cognition and behaviour are profound, but the brain also plays an integral role in many physiological functions. As described above, the central nervous system is intimately connected to the immune system and HPA axis. Ethanol has been consistently shown to alter the neural transcriptome during development, and changes in gene expression in the brain may therefore cascade into changes in the stress response and immune response. Only a small number of studies of genome-wide gene expression in the brain have been conducted on animals exposed prenatally to alcohol. These have largely examined the immediate/acute effects of alcohol on the developing embryo or fetus (Da Lee et al. 2004; Hard et al. 2005; Green et al. 2007; Zhou, Chen, et al. 2011). Studies examining changes in the whole embryo or the whole fetus have found changes in genes involved in neural specification, development, apoptosis, and growth factor expression (Da Lee et al. 2004; Zhou, Chen, et al. 2011). Studies specifically looking at expression in the developing fetal brain have found similar changes, in genes and pathways related to energy metabolism, cell- cell adhesion, cytoskeletal remodeling, cell proliferation, differentiation, and apoptosis, and neuronal growth and survival (Green et al. 2007; Hard et al. 2005). Global gene expression has also been analyzed in vitro in neural cell culture, identifying alterations in cell cycle signaling and cellular adhesion (Hicks et al. 2010). The pathways that most consistently appear to be altered in genome wide surveys of PAE effects are related to cellular adhesion, cell survival, and growth and development. However, no specific genes appear to be consistently altered across genome-wide studies of PAE, which is likely a testament to the diverse teratogenic effects of alcohol.

One study has been conducted examining the effects of PAE genome-wide on the neural transcriptome in adulthood. Kleiber and colleagues (Kleiber et al. 2012) used microarrays to look at genome-wide gene expression in the whole brains of adult male mice that had been exposed prenatally to alcohol. Interestingly, the effects of PAE were found to be subtle, with very few genes (less than 10) showing fold-changes in expression greater than 1.3-fold. Kleiber and colleagues therefore reduced the stringency of their analyses to examine genes with even smaller fold-changes, and found that overall, altered genes were enriched for developmental processes. It appears therefore that prenatal ethanol exposure has subtle but lasting effects on the brain transcriptome overall. It may, however, have larger effects that can only be observed at a subregion-

8 specific level, that are washed out by a global approach. Additionally, while basal levels of gene expression may show little change, challenges such as stress or immune insult may play on underlying differences in PAE animals and exacerbate/uncover changes in the transcriptome.

1.4 Rationale and thesis objectives

The present study aimed to identify the long-term effects of prenatal alcohol exposure on gene expression in the rat brain, at both a basal level and in response to an inflammatory challenge. The PFC and HPC were investigated in this study, due to their involvement in the regulation of the HPA axis, and to the known interaction between HPA and immune function. These two brain regions are also susceptible to the teratogenic effects of alcohol, and have been shown to be altered in individuals with FASD (reviewed in Norman et al. 2009) and animal models of prenatal alcohol exposure. In animal models, global analysis of gene expression has largely been limited to the acute effects of alcohol exposure during development (Da Lee et al. 2004; Hard et al. 2005; Green et al. 2007; Zhou, Zhao, et al. 2011). One recent study examined global changes in gene expression in the brains of adult male mice, and found subtle, but significant, long-term effects of prenatal alcohol exposure, mostly in genes related to neurodevelopment (Kleiber et al. 2012). Because gene expression changes in the adult PAE brain have not been widely studied, let alone in response to an inflammatory challenge, a global approach to analyzing gene expression was taken to identify long-term effects of PAE. This approach will provide an unbiased assessment of the persistent effects of alcohol on the neural transcriptome, and provide insight into how the neuroendocrine and neuroimmune systems are impacted by this early life insult.

9 Chapter 2: Materials and Methods

2.1 Breeding and prenatal alcohol exposure

Rats were obtained from the Animal Care Center at the University of British Columbia, and were group- housed 1-2 weeks before breeding, with ad libitum access to standard lab chow (Jamieson’s Pet Food Distributors, Ltd., Delta, BC, Canada). Details of the procedures for breeding and handling have been published previously (Glavas et al. 2007). Briefly, females and males were co-housed in stainless steel cages with mesh front and floors, and wax paper under the cages was checked daily for the presence of vaginal plugs, which indicated day 1 of gestation. Thereafter pregnant dams were singly housed, and were assigned to one of three groups – a control group (C; fed laboratory chow ad libitum), pair-fed group (PF; liquid-control diet, with maltose-dextrin isocalorically substituted for ethanol, in the amount consumed by an ethanol- consuming partner, matched for g/kg body weight/day of gestation), or an ethanol-fed group (E; ad libitum access to liquid ethanol diet, with 36% calories derived from ethanol). All animals had ad libitum access to water, and diets were fed from gestation days 1-21 (Weinberg/Kiever Ethanol Diet #710324, Weinberg/Kiever Control Diet #710109, Dyets Inc., Bethlehem, PA). After gestational day 21, all animals were given laboratory chow and water ad libitum. Litters were weighed and culled at birth to 5 males and 5 females, when possible. Following weaning (on postnatal day 22) female offspring were group-housed by litter (2-3 rats per cage) until the start of testing.

2.2 Induction of adjuvant-induced arthritis

Details of the postnatal induction of adjuvant-induced arthritis in these animals have been previously published (Zhang et al. 2012). Briefly, female offspring (50-65 days of age) from the C, PF, and E groups were divided into two postnatal treatment groups, in which animals received an intradermal injection at the base of the tail of either 0.1 ml of 12 mg/ml suspension of complete Freund’s adjuvant (CFA) (Adjuvant group), or 0.1ml saline (Saline group). All animals were single-housed after injection, and monitored for clinical signs of arthritis, from the onset of AA through to resolution. To evaluate clinical signs of arthritis, animals were lightly anesthetized with isofluorane, and paws were scored for severity of redness and swelling, on days 7, 10, and every other day afterwards, following injection (results published previously in Zhang et al., 2012).

2.3 Termination of animals

Animals were terminated in two cohorts for analysis of gene expression. One cohort was terminated on day 16 post-injection (at the peak of adjuvant-induced arthritis), and another on day 39 post-injection (during 10 resolution phase of arthritis). Each cohort contained 27 adjuvant-injected animals (9 each of C, PF, and E) and 15-18 saline-injected animals (5-6 each of C, PF, and E). For termination, animals were singly removed from their colony room, exposed to CO2 for 30 seconds, and then quickly decapitated. Brains were rapidly removed, immediately frozen on dry ice, wrapped in foil, and stored at -70 °C.

2.4 Tissue dissection and RNA extraction

Brains were gradually thawed to 4 °C, and the prefrontal cortex (PFC) and hippocampus (HPC) were dissected using RNase-free technique. Dissected tissues were placed in RNAlater and stored at -20 °C. Total RNA and DNA were simultaneously extracted from the dissected tissues using the Qiagen AllPrep DNA/RNA Mini kit, and a DNase digestion step was included in the RNA extraction process. RNA integrity was determined using the Agilent BioAnalyzer mRNA Nano assay, and no samples were excluded due to low RNA integrity. Only one sample was excluded at this stage (PF hippocampus from the Day 39 Adjuvant group), due to suspected contamination.

2.5 Microarray assaying of whole genome gene expression

To generate cRNA for microarray analysis, 250 ng of total RNA from each sample was amplified using the Ambion Illumina TotalPrep RNA Amplification kit, in batches of ~24 samples at a time. Samples were distributed across amplification batches such that batch was not confounded with experimental treatment group. Gene expression was analyzed for both the PFC and HPC using the Illumina RatRef-12 Expression BeadChip microarray, which has 12 arrays per chip. PFC and HPC samples were run separately on different dates, due a limit of processing 8 chips (96 arrays) per batch. 750 ng of cRNA was applied to each array, with one sample per array. Experimental groups were counter-balanced across the arrays, such that chip batch was not confounded with any experimental treatment group, amplification batch, or dissection batch. Additional arrays included an amplification replicate and several hybridization replicates, yielding a total of 96 arrays run per tissue. Microarrays were scanned on the Illumina iScan, and bead-level expression data was collected.

2.6 Data pre-processing, quality control, and exploratory data analysis

The bioconductor package beadarray (Dunning et al. 2007) was used to collapse bead-level data into probe- level data, and log2-transform the resultant expression data. Spatial artifacts were identified using the BASH algorithm (Cairns et al. 2008), and were masked prior to calculating the summarized expression values for each probe. Pairwise Pearson correlations were calculated to compare correlation of quantile-normalized expression profiles; 2 PFC arrays were identified as outliers (one replicate array and one PF Day 39 sample). The outlier samples and all replicate samples were removed, and the final dataset for each tissue consisted of

11 86 unique arrays. The original expression data (before quantile normalization) was then filtered to remove control probes and any probes without evidence for expression in at least one sample (i.e. with no detection p- value <0.05 in comparison to negative control probes). After filtering, 20215 probes remained in the PFC dataset, and 20069 probes remained in the HPC dataset (out of a total 23350 probes). The filtered, log2- transformed gene expression profiles were then quantile-normalized across arrays within each tissue. Principal components analysis (PCA) was used to look for expression heterogeneity in the microarray data that was attributable to batch effects, namely differences among tissue dissection batches, RNA extraction batches, RNA amplification batches, and among beadchips. Principal components were also compared to prenatal treatment group and adjuvant treatment group to discern whether there were strong signals in gene expression due to experimental treatments. Batch effects were identified in the expression data using PCA, therefore surrogate variable analysis (sva) (Leek & Storey 2007) was used to generate surrogate variables representative of expression heterogeneity from sources other than the experimental treatments (such as batch effects).

2.7 Differential expression analysis

Gene expression analysis was conducted using the package limma (Smyth 2005) in the statistical program R. The surrogate variables generated with sva were included in linear modeling of gene expression, conducted with limma, which uses a moderated F-statistic and moderated t-statistic to denote significant expression changes. Limma was used to model gene expression changes in two ways: 1) effects of prenatal treatment alone (among the Saline-treated animals), and 2) interaction of prenatal treatment with the response to an inflammatory adjuvant (among both Saline- and Adjuvant-treated animals). In each model, a moderated F- statistic was generated for each probe. F-statistic p-values were corrected for multiple testing using Benjamini-Hochberg correction, and the false-discovery rate (FDR) was controlled at <25% (q-value <0.25). Within the probes with FDR <25%, significant contrasts of interest (e.g. significant effect of ethanol exposure compared to controls) were denoted as having a moderated t-statistic p-value <0.05. As the majority of probes on the RatRef-12 beadchip were designed based on transcripts in RefSeq with only provisional annotation, the sequences for significant probes were queried against the current RefSeq database for Rattus norvegicus to establish the most current identity of the target transcripts.

2.8 Gene Ontology and pathway analysis

Gene Ontology (GO) analysis was conducted to identify any “Biological Processes” annotated in the rat that were enriched for the effects of prenatal diet and postnatal adjuvant exposure. Over-representation analysis was conducted using the ontology and pathway analyzer RatMine in the Rat Genome Database (Rat Genome Database n.d.; Dwinell et al. 2009), to look for enrichment of GO categories, KEGG pathways, and disease

12 phenotypes within lists of top differentially expressed genes. Additional GO analysis was conducted using the gene-score resampling (GSR) method in the program ermineJ (Lee et al. 2005), which allows the entire list of analyzed genes to be evaluated for the effects of a treatment. T-test p-values were used to order the gene list from most to least significant genes for each comparison of interest (e.g. expression change in PFC in E vs. C animals), and GSR was used to identify gene sets enriched towards the significant end of the list. Gene sets were limited to Biological Processes GO categories that were annotated with 5-200 genes only, and had representative genes in the filtered expression datasets. A custom gene set of candidate FASD genes was also included (Table S1), curated from the online database Neurocarta (Portales-Casamar et al. 2013). This limited the number of GO categories analyzed to 4072 in PFC, and 4037 in HPC. Correction for multiple testing was performed using the Benjamini-Hochberg method, and the false-discovery rate was controlled at 1%. Where large numbers of GO categories were found to be significant, GO categories were mapped to their parent GO Slim terms using the program CateGOrizer (Zhi-Liang et al. 2008) to determine the most common types of altered functions.

2.9 Validation of microarray results

A selection of differentially expressed genes was validated using reverse-transcription quantitative real time PCR (RT-qPCR). RT-qPCR was conducted for both PFC and HPC, with the same RNA used for microarray analysis. Samples were selected from the dissection batch with higher RNA integrity, giving n=3 for C and E in each tissue. PF samples were not included due to RNA quality and sample replicate limitations. Primers were designed using NCBI Primer-BLAST, using well-established guidelines for RT-qPCR primer design (Nolan et al. 2006). Where possible, primers were designed close to the probe site, and towards the 3' end of the RNA transcript. Multiple reference genes were used to normalize RT-qPCR expression data, as per best practice (Nolan et al. 2006). Reference genes were chosen based on stability of expression across treatment groups demonstrated in the microarray data. Three reference genes with high expression levels and no evidence for expression differences across group (F-statistic p-value >0.05) were selected for each tissue. The geometric mean of the cycle threshold (Ct) values for the three reference genes was calculated to generate a normalization factor (reference gene index) for each sample (Vandesompele et al. 2002). Expression levels relative to the reference gene index were averaged for each treatment group, and a two-tailed Student’s t-test was conducted to test for differences between groups (Schmittgen & Livak 2008). To compare overall similarity between expression results with the two methods, fold-changes were calculated for RT-qPCR results and correlated to fold-changes from the microarray data.

13

Figure 2.1 Experimental model

Table 2.1. Final number of animals in each treatment condition. Prenatal Adult Treatment Euthanization Timepoint n Treatment (~PND 60) (days post-injection) 16 5 Saline 39 6 Ethanol 16 9 Adjuvant 39 9 16 5 Saline 39 6 Pair-fed 16 9 Adjuvant 39 8 16 5 Saline 39 6 Control 16 9 Adjuvant 39 9 Total 86

14 Chapter 3: Results

3.1 Overview of analyses and main findings

An overview of the analyses performed in this study and the main findings as related to changes in gene expression in the PFC and HPC of animals prenatally exposed to ethanol is outlined in Figure 3.1. These results are elaborated upon in subsequent sections.

Figure 3.1. Overview of analyses and main findings for gene expression analysis

15

3.2 Exploratory data analysis

Exploratory data analysis was performed to evaluate the variability of the microarray data, and to identify any batch effects that might confound differential expression analysis. As described in the methods, each brain region was analyzed separately. Within brain regions, each sample was run on its own microarray, and microarrays exist in groups of 12 on a single beadchip. Pearson correlations were calculated among samples (ie. microarrays) within each tissue, and were plotted as heatmaps to visualize variability and outliers (Figure 3.2 and Figure 3.3). Overall samples were highly correlated, as the mean of all sample correlations was 0.97 for both PFC and HPC (Table 3.1). The degree of correlation of microarrays within beadchips appears to vary, where arrays on some beadchips showed higher than average correlations to each other (eg. chip 5398636033, excepting the arrays K and L), and others were more variable (eg. chip 5398636011). Two microarrays in the prefrontal cortex data set had a majority of correlations less than 0.94 with other arrays, and appeared to be outliers. These two arrays appeared to have large spatial artefacts, as apparent from large swathes of outliers across one end of each array (Figure 3.4). Despite the fact that outlier beads are masked during preprocessing and not included when calculating expression levels, these extensive artefacts likely contribute to the decreased correlation of the arrays, and these arrays were removed from further analyses. Groups of replicate arrays had consistently higher correlations than the group of overall samples (Table 3.1 and Figure 3.5), which suggests that the microarray assays were carried out with reproducibility. However, given the high level of correlation among all samples, expression differences among treatment groups are likely to be subtle. This is supported by the fact that clustering the correlation heatmaps by treatment groups did not result in patterns that showed like samples were more highly correlated than others (data not shown).

Principal components analysis (PCA) was used to identify which experimental variables broadly contributed most to variation in gene expression. Within the first five principal components, the factors that appear to contribute most to expression variability were processing batches, particularly dissection batch and RNA amplification batch. Prenatal treatment group and adjuvant treatment, on the other hand, did not appear to correlate with any of the first five principal components (data not shown). One factor that may have contributed to the variability between dissection batches is a significant difference between dissection groups in RNA integrity. RNA integrity was significantly lower (p <0.001) in the first dissection batch than in the third (one sample was dissected at an intermediate time point) (Figure 3.7). This may be a result of different individuals conducting the dissection between the two batches, or longer storage of the first batch as dissected tissue prior to RNA extraction. However, given that RNA integrity was consistently high, the contribution of RNA integrity differences is likely small. Along that line, the magnitude of variance contributed by each principal component to the expression data in each tissue appears to be relatively small, no larger than 6% in the HPC and 3% in PFC (Figure 3.8). Global expression did not appear to vary between treatment groups, and

16 the expression data thus was quantile-normalized, reducing large systemic differences between arrays (Fig 3.8) and preparing the data for analysis of gene expression.

For all susbsequent analyses, gene expression data was divided to analyze the effects of prenatal treatment alone on steady-state gene expression among the Saline-treated animals only, and the interaction of prenatal treatment with the response to an inflammatory adjuvant among both Saline- and Adjuvant-treated animals. Each tissue (PFC and HPC) and euthanization time point (Day 16 and Day 39 post-injection) were analyzed separately. Surrogate variable analysis was conducted to adjust for batch effects in each of these groups of expression data, and the number of surrogate variables generated for each group of data is summarized in Table 3.2.

17

Figure 3.2. Correlation heatmap of expression profiles among all arrays in the PFC dataset.

18

Figure 3.3. Correlation heatmap of expression profiles among all arrays in the HPC dataset.

19

Figure 3.4. Spatial artefacts on the PFC outlier arrays. Spatial artefacts were identified on arrays 5398636033_K and 5398636033_L (two arrays on the same chip), by visualizing the distribution of outlier probes on each array using the R package beadarray.

Table 3.1. Correlation of overall expression profiles among all arrays and among replicates arrays a) Correlations in the PFC microarray dataset Quantiles for Pearson correlations among samples 0% 25% 50% 75% 100% n Overall 0.90 0.97 0.97 0.98 1.00 96 Hybridization replicate group 0.96 0.97 0.98 0.98 1.00 9 Amplification replicate 0.98 0.98 0.99 1.00 1.00 2 b) Correlations in the HPC microarray dataset Quantiles for Pearson correlations among samples 0% 25% 50% 75% 100% n Overall 0.93 0.96 0.97 0.97 1.00 96 Hybridization replicate group 1 0.96 0.97 0.97 0.98 1.00 4 Hybridization replicate group 2 0.97 0.97 0.99 0.99 1.00 4 Hybridization replicate group 3 0.96 0.96 0.98 0.99 1.00 4 Hybridization replicate group 4 0.96 0.97 0.97 0.98 1.00 4 Replicate average 0.96 0.97 0.98 0.99 1.00

20

Figure 3.5. Correlation among hybridization replicate arrays in PFC. Heatmap of Pearson correlations among replicates of a single PFC sample hybridized to multiple microarrays.

21

Figure 3.6. Evidence for batch effects in Principal Components Analysis. PCA showed that the first few principal components (PCs) of expression data correlated with processing batches in both the PFC (a) and HPC (b). In PFC, PC2 varied with dissection and extraction batch, and PC3 varied with amplification batch. In HPC, PC1 varied with dissection batch, and PC2 varied with amplification batch. None of the first five principal components appeared to vary with prenatal or adjuvant treatment.

22

Figure 3.7. RNA integrity varied between dissection batches. RNA integrity varied significantly between tissue dissection batches, with significantly lower RNA integrity numbers (RIN) in the first (older) dissection batch compared to the third (p<0.0001 for both PFC and HPC). Only one sample was dissected in batch 2.

Figure 3.8. Proportion of variance of Principal Components. Plot of the relative magnitude of each principal component in the expression data for PFC and HPC. Each principal component appeared to have a small contribution to variability in the expression data.

23

Figure 3.9. Gene expression correlation before and after quantile normalization. MA plots showing the distribution of log-correlation of gene expression between five arrays, where the y-axis graphs the intensity ratio, M (ie. log of correlation), and x-axis graphs the average intensity. Given the assumption that most genes will not be differentially expressed, the majority of points should fall along 0 (ie. the log of a correlation equal to 1). Array 5 demonstrates systematic differences from the other 4 arrays, given a distribution of expression values around 1 rather than 0. This systematic difference is largely alleviated by quantile normalization.

24

Table 3.2. Number of surrogate variables generated by SVA for each analysis group Surrogate variables generated for each dataset to be analyzed in limma, representing expression heterogeneity not attributable to experimental treatments.

Time point Number of surrogate variables Analysis group (days post-injection) PFC HPC Day 16 6 2 Saline only Day 39 4 6 Day 16 15 8 Saline and Adjuvant Day 39 11 12

25 3.3 Prenatal alcohol exposure alters steady-state gene expression in PFC and HPC

The effects of PAE on unchallenged levels of gene expression were examined in animals in the saline-injected condition, at both termination time points (Day 16 post-injection, at ~75 days of age, and Day 39 post- injection, at ~95 days of age). To identify changes specific to ethanol exposure, we looked for genes with significantly different levels of mRNA in E compared to both PF and C animals, and preferably where PF and C did not differ from each other. According to p-value distributions, gene expression was significantly different in E compared to both C and PF females at Day 16 post-injection, in both PFC and HPC, as distributions were enriched towards zero for the contrasts of E vs C, and E vs PF (Figure S1). However, significant effects of prenatal ethanol exposure versus both C and PF were not apparent at Day 39 post- injection in either brain region. The effects of prenatal treatment on gene expression were subtle, as correlation between samples was very high (Table 3.1). A false-discovery rate of 25% was used in order to capture a moderate number of expression differences among groups (moderated F-statistic q-value < 0.25, Benjamini-Hochberg FDR). Only 2 probes in the Day 39 post-saline injection tissues met a 25% FDR, and these were not unique effects of alcohol exposure (data not shown), so subsequent analyses focused on effects in Day 16 tissues.

3.3.1 Genes altered by PAE at a steady-state level of gene expression

At a 25% FDR, significant effects of prenatal treatment were found for 80 probes in PFC and 30 probes in HPC at Day 16 post-saline injection (Figure 3.11). The number of genes altered in each of the three comparisons (E vs C, E vs PF, and PF vs C) is shown in Figure 3.12. In both PFC and HPC, more than a third of the probes (43% in PFC, 37% in HPC) showed significant effects of prenatal alcohol exposure against both control groups, with mRNA levels significantly different in E compared to both PF and C animals (moderated t-test p < 0.05). In a subset of these probes (15 in PFC, 4 in HPC) (Figure 2), mRNA levels for PF and C animals also did not differ from each other (p >0.05), and thus represent changes unique to the effects of ethanol (Table 3.3 and Table 3.4). These uniquely altered genes had a number of annotated functions in common (though no significant enrichment of GO categories was found within the list of genes). Several genes are involved in neurodevelopment (Tcf4, Ap1s2, Acsl3, and Cnih2), some of which have been implicated in human neurodevelopmental disorders (Tcf4 with Pitt-Hopkins syndrome, and Ap1s2 with X- linked mental retardation, respectively). Other common functions include regulation of cell death (Dusp6 and Atp6ap1) and cell differentiation (Dusp6, Med28, Ndfip1, Tcf4), regulation of transcription (Med28, H2afv, Tcf4, Rnasek), and roles in neuronal signaling, particularly with regards to AMPA receptor activity (Ppp1r14a and Cnih2).

26 3.3.2 GO categories altered by PAE at a steady-state level of gene expression

Enrichment analysis was conducted in RatMine to find functions or pathways enriched within the differentially expressed genes in the Day 16 post-saline injection group. No enrichment of GO categories, KEGG pathways, or disease phenotypes was found within the genes significantly altered by prenatal diet overall (genes in Figure 3.11), or by prenatal alcohol exposure (genes in Table 3.3 and Table 3.4). Using ermineJ, a gene-score resampling method was used to analyze the effects of a treatment on the entire set of >20,000 genes analyzed, rather than just the top genes. 7.4% (in PFC) to 14.6% (in HPC) of Biological Processes analyzed were found to be altered in at least one treatment comparison at an FDR of 1% (Figure 3.14a). Six processes were altered in PFC of E vs both PF and C animals: “positive regulation of cell projection organization”, “cellular metal ion homeostasis”, “divalent inorganic cation homeostasis”, “cellular divalent inorganic cation homeostasis”, “response to virus”, and “regulation of intracellular transport”. Many more processes (79) were altered in HPC in E vs PF and C, none of which overlapped with the 6 altered in PFC. These processes were most commonly involved in metabolism (24%), cell communication (18%), development (18%), transport (15%), and signal transduction (10%) (as determined by mapping the 79 Biological Processes to their parent GOslim terms using CateGOrizer). At an FDR of 10%, five processes overlapped between PFC and HPC in response to ethanol exposure: “positive regulation of neuron differentiation”, “dorsal/ventral pattern formation”, “circadian rhythm”, “regulation of lymphocyte differentiation”, and “regulation of lipase activity” (Figure 3.14a).

3.3.3 Validation of gene expression changes by RT-qPCR

RT-qPCR was used to validate the microarray expression results. Of the 19 probes showing unique differential expression with prenatal alcohol exposure (Table 3.3 and Table 3.4), 17 aligned to an existing RNA sequence in the Rattus norvegicus Refseq RNA database and were specific to their intended targets (the exceptions being ILMN_1372701 and ILMN_1374168, which align only to obsolete Rattus norvegicus nucleotide sequences). Specific RT-qPCR primers were successfully designed for 15 of the 17 genes (Table 3.5; the exceptions being Rps8 and Rpl7, which as a result were not analyzed), and RT-qPCR analysis was conducted using the same samples that were used for microarray analysis. Three stable, highly expressed genes were used as reference genes to normalize candidate mRNA levels in each tissue , and these were Pgk1, Sdha, and Hprt1 in PFC (Table 3.6), and Pgk1, Sdha, and Actb in HPC (Table 3.7).

Three genes showed significant increased gene expression (p <0.05) (Figure 3.13a), and two genes showed a trend for decreased expression in E animals (p <0.10) (Figure 3.13b) with RT-qPCR. Importantly, the direction of change was consistent with the microarray results, and fold-changes in expression (E samples/C samples) between RT-qPCR results and microarray results for these genes were strongly correlated

27 (r=0.92(3), p<0.03) (Figure 3.13c). The remaining genes showed no significant change in gene expression, though a significant positive correlation existed between RT-qPCR fold-changes and microarray fold-changes for all 15 tested genes (r=0.68(13), p<0.02).

3.3.4 Genes showing common, graded, or differential effects of PAE and pair-feeding

Of the remaining genes that were found to be altered by prenatal treatment at Day 16 post-saline injection, a variety of prenatal group effects were observed. These genes either had common or graded effects of ethanol exposure and pair-feeding (compared to control animals), opposite effects of ethanol exposure and pair- feeding, or unique effects of pair-feeding alone (Figure 3.11).

Many genes showed similar changes in both E and PF compared to C animals. Given that both E and PF treatments result in decreased caloric consumption relative to controls, and both are known to have effects on the HPA axis, this is not entirely unexpected. Nearly half of significant probes in the PFC had the same level of expression in PF and E, but not C, animals and several others showed the same pattern in HPC (Table 3.8 and Table 3.9; 39/80 probes in PFC and 4/30 probes in HPC). No GO categories were enriched within this set of genes, but many are annotated to be involved in anatomical structure development (Chn1, Igfbp7, Mapkapk2, Ndrg2, Nme2, Nrxn3, Satb1, and Sep15). Similarly, many genes exhibited graded effects of prenatal treatment, where the effects of ethanol exposure were greater than the effects of pair-feeding (i.e., E>PF>C), or the effects of pair-feeding exposure were greater than the effects of ethanol exposure, (i. e. PF>E>C) (Table 3.10 and Table 3.11). Conversely, a handful of genes were altered in opposite directions by ethanol exposure and pair-feeding, relative to control animals, mostly in the HPC (Table 3.10 and Table 3.11; 2/80 probes in PFC, 6/30 probes in HPC).

Pair-feeding also had unique effects on gene expression, where mRNA levels differed in PF compared to E and C animals, particularly in the HPC (Table 3.12 and Table 3.13). Many of these genes are involved in small molecule metabolism (Ak2, Aqp1, Enpp2, Lrp1, Retsat, Tlr), transport (Agp1, Lrp1, Park7, Slco1a5, Ttr), signal transduction (Igfbp2, Lrp1, Park7, Ppp1r1b, Sostdc1), and the response to stress (Aqp1, F5, Igfbp2, Park7), among other biological processes. GO analysis was conducted for the effects of pair-feeding as described above for the effects of prenatal alcohol exposure. Within the list of these genes, no enrichment of GO categories, KEGG pathways, or disease phenotypes was found. Using gene-score resampling, however, many processes were found to be uniquely altered in PF animals at an FDR of 1%. In the PFC, 18 processes were altered at an FDR of 1%, and were most commonly involved in cell organization and biogenesis (28%), metabolism (22%), and transport, similar to ethanol exposure. In the HPC, 49 processes were altered, and similar to the unique effects of ethanol exposure, these were most commonly involved in metabolism (25%), and development (13%). Interestingly, the processes altered in PF animals in the PFC included the curated list

28 of candidate FASD genes from Neurocarta, suggesting these genes are not necessarily specific to prenatal ethanol exposure. This group of candidate genes was in fact enriched in E animals as well at a higher (10%) FDR, and may therefore be related to other common underlying mechanisms such as stress. At an FDR of 10%, two other processes altered by pair-feeding overlapped between brain regions: “negative regulation of neuron projection development” and “positive regulation of epithelial cell migration” (Figure 3.14c).

29

Figure 3.10. Density plots of p-value distributions. Density plots of p-value distributions for gene expression differences among prenatal treatment groups, within the saline-treated conditions, in a) PFC at Day 16 post-saline injection, b) PFC at Day 39 post-saline injection, c) HPC at Day 16 post-saline injection, and d) HPC at Day 39 post-saline injection. The greatest effects of prenatal ethanol exposure on gene expression p-values were exhibited at D16 in PFC, followed by D16 HPC, as exhibited by enrichment of p-values towards zero for the ethanol contrasts (E-C, E-PF). No change in p-values was apparent in Day 39 PFC, and only a pair-fed effect was apparent in Day 39 HPC.

30

Figure 3.11 Genes showing a significant effect of prenatal diet at Day 16 post-saline injection. 31 (Previous page) In the prefrontal cortex (a), 84 genes were differentially expressed in response to at least one prenatal treatment. In the hippocampus (b), 30 genes were differentially expressed in response to at least one prenatal treatment. F-statistic q-value <0.25.

Figure 3.12. Venn diagrams of the number of probes significantly altered in each prenatal treatment contrast. 80 genes were altered in PFC and 30 in HPA at Day 16 post-saline injection. The number of probes with unique effects in ethanol-exposed versus both PF and C animals are highlighted in grey, and listed in Table 3.3 and 3.4. Probes with altered by both ethanol exposure and pair-feeding (intersection on the left of each diagram) are listed in Table 3.8 and 3.9. Probes differentially expressed among all three prenatal treatment groups (center of Venn diagrams) are listed in Table 3.10 and 3.11. Probes with a unique effect of pair- feeding (intersection on the left) are listed in Table 3.12 and 3.13. Moderated F-statistic q-value <0.25, moderated t-statistic p-value <0.05.

32 Table 3.3. Genes differentially expressed in prefrontal cortex of E vs both PF and C animals. Bold = p <0.05. na = probe had no specific alignment to current RefSeq RNA database.

Fold change Average Gene Symbol Gene Name F p-value q-value Ethanol- Ethanol- Pairfed- Expression Control Pairfed Control H2afv Rattus norvegicus similar to H2A 10.6 18.7 4.8E-05 0.11 0.65 0.76 0.86 histone family, member V isoform 1 (LOC685909) Tcf4 transcription factor 4 11.2 11.4 7.0E-04 0.23 0.67 0.66 1.01 Rnasek ribonuclease, RNase K 13.2 11.1 8.0E-04 0.23 0.68 0.57 1.19 Ppp1r14a protein phosphatase 1, regulatory 10.0 12.6 4.1E-04 0.23 0.68 0.64 1.05 (inhibitor) subunit 14A Rps8 ribosomal protein S8 13.0 11.1 7.9E-04 0.23 0.69 0.74 0.93 ILMN_1372701 na 9.4 11.3 7.3E-04 0.23 0.71 0.79 0.90 ILMN_1374168 na 9.1 10.7 9.4E-04 0.25 0.77 0.73 1.05 Pex11g peroxisomal biogenesis factor 11 7.0 11.5 6.7E-04 0.23 0.82 0.71 1.16 gamma Ndfip1 Nedd4 family interacting protein 1 11.4 12.1 5.1E-04 0.23 1.32 1.37 0.97 Acsl3 acyl-CoA synthetase long-chain 10.2 12.2 4.9E-04 0.23 1.36 1.36 1.00 family member 3 Dusp6 dual specificity phosphatase 6 9.9 12.5 4.4E-04 0.23 1.41 1.21 1.17 Rpl7 ribosomal protein L7 11.6 13.7 2.7E-04 0.22 1.44 1.36 1.05 Med28 mediator complex subunit 28 9.2 11.1 7.9E-04 0.23 1.48 1.29 1.15 Atp6ap1 ATPase, H+ transporting, lysosomal 11.0 10.6 9.8E-04 0.25 1.50 1.35 1.11 accessory protein 1 Ap1s2 adaptor-related protein complex 1, 9.7 12.4 4.6E-04 0.23 1.60 1.35 1.19 sigma 2 subunit

33 Table 3.4. Genes differentially expressed in hippocampus of E vs both PF and C animals. Bold = p <0.05.

Gene Symbol Average F p-value q-value Fold change Expression Ethanol- Ethanol- Pairfed- Control Pairfed Control Cnih2 cornichon homolog 2 (Drosophila) 11.1 16.0 8.1E-05 0.14 0.61 0.60 1.01 Caap1 caspase activity and apoptosis 9.2 15.2 1.1E-04 0.14 0.68 0.71 0.95 inhibitor 1 LOC688637 similar to WD repeat domain 36 8.8 15.4 1.0E-04 0.14 1.46 1.36 1.08 Rgs3 regulator of G-protein signaling 3 9.1 14.6 1.4E-04 0.15 1.71 1.83 0.93

Table 3.5. Sequences of primers used for RT-qPCR Gene Type Accession Forward primer (5'-3') Reverse primer (5'-3') Actb Reference NM_031144.2 CTGCCCTGGCTCCTAGCACCAT CTCAGTAACAGTCCGCCTAGAAGCA Hprt1 Reference NM_012583.2 TGTGGCCAGTAAAGAACTAGCAGACGTT GTGCAAATCAAAAGGGACGCAGCAACA Pgk1 Reference NM_053291.3 AGTCCTTCCTGGGGTGGATGCTCT AGGGTTCCTGGTGCTGCGTCTT Sdha Reference NM_130428.1 TGCCAGGGAAGATTACAAGGTGCGG AGAGGGTGTGCTTCCTCCAGTGTTC Acsl3 Target NM_057107.1 ACTCCCGAAACTGGTCTGGTGACTGATG ATCCGCTCAATGTCTGCCTGGTAGTGT Ap1s2 Target NM_001127531.2 TGTCACTGCCTAGTCGTCGGA GCCAACCAATGCCACTTTGCTTCAG Atp6ap1 Target NM_031785.1 GGGTTAAGAATGAGCGGTACACTGGGG ACTTCTGGCTTCTTGACAGGCAATCCTT Dusp6 Target NM_053883.2 GTGGGATGCGACAGGTTGTGAGGA ACACCACGAACATCATGGAGCAAGTGAA H2afv Target NM_001106019.1 CTGATCGGAAAGAAGGGGCAGCAGA CACACACAGTGAGGACAGCAGGTCA Med28 Target NM_001107217.1 TGCAGCACAAGAAGCCAGCCGA GGTCTGCTTCAGAGGTGCAGGTATGTT Ndfip1 Target NM_001013059.1 ACTGGCTCTGGTGGGTGTTCTTGGT AGAACTCTGGTCCTGGGGAGATTTGAGA Pex11g Target NM_001105902.1 AACGAGACTCAGATTCCCAGAGCGG ATTTGAGCCCCTTTCCCACCCCA Ppp1r14a Target NM_130403.1 GACGAGCTGCTGGAATTGGACAGTGA GGACGAAGTCCTCTGTGGGATTCAGG Rnasek Target NM_001137561.2 TTGGGACTGTTACCCTGGCGAGAC TCCAGGGGTTGGGCAGCAGTTT Tcf4 Target NM_053369.1 AGAGAAGGTGTCCTCAGAGCCTCCC GGTGGCAACTTGGACCCTTTCACATC Cnih2 Target NM_001025132.1 GGGCCAGGCAAAGCTCTAAACAGGG GGCCCAAATTCCCCTGAAACGGACA Loc688637 Target XM_001067706.2 AGAGGCCATGCGGAGCTTTTTGAGT AAATCACGCTTTCTGTCCAGCATCACCC MCG125002 Target NM_001034154.1 TCTAGCCCAAAGGAACCCAAAGCGG GGCTGAACGTCTTCTGGTGGAGGA Rgs3 Target NM_019340.1 TGGCACATGAACGGTAATAGGAGAGCC TGGGACCAGCAAATGCCCTGAAACT

34 Table 3.6. Microarray expression results for common reference genes in PFC of Day 16 Saline animals. Genes in bold were used as reference genes for RT-qPCR.

Fold change Average Adjusted Symbol Probe_ID F P-value E:C PF:C E:PF Expression p-value Polr2a_mapped ILMN_1372495 1.00 0.95 1.05 7.71 0.24 0.79 0.94 Tbp ILMN_1349379 1.04 0.99 1.05 8.04 0.29 0.75 0.93 Ubc ILMN_1350494 1.06 0.98 1.08 13.83 0.33 0.72 0.93 Pgk1 ILMN_1369074 1.04 1.11 0.94 12.3 0.52 0.6 0.89 Sdha ILMN_1357678 1.12 0.99 1.13 10.53 0.83 0.45 0.84 Hmbs ILMN_1353365 1.05 0.94 1.11 8.13 0.96 0.4 0.82 Actb ILMN_1355039 0.89 0.88 1.02 12.13 1.12 0.35 0.79 Hprt1 ILMN_1367708 1.14 1.09 1.05 11.73 1.17 0.33 0.79 Gusb ILMN_1350544 1.06 1.14 0.93 7.78 1.24 0.31 0.78 H2A.1 ILMN_1372198 0.90 0.98 0.92 7.41 1.33 0.29 0.77 Gapdh ILMN_1649859 1.19 0.96 1.23 13.33 1.61 0.23 0.73 Tfrc ILMN_1360908 1.16 1.13 1.03 7.13 2.02 0.16 0.7 Actb ILMN_2038799 0.84 0.93 0.91 13.74 2.12 0.15 0.68 Actb ILMN_2038798 0.80 0.76 1.06 11.87 2.89 0.083 0.63 B2m ILMN_1368656 1.15 0.87 1.33 12.99 3.62 0.049 0.57 Ywhaz ILMN_1373913 0.76 1.12 0.68 13.53 4.93 0.02 0.49

Table 3.7. Microarray expression results for common reference genes in HPC of Day 16 Saline animals. Genes in bold were used as reference genes for RT-qPCR.

Fold change Average Adjusted Symbol Probe ID F P-value E:C E:PF PF:C Expression p-value Sdha ILMN_1357678 1.00 1.01 0.99 10.24 0.01 0.99 1 Gusb ILMN_1350544 0.96 0.96 1.00 7.83 0.18 0.84 0.98 Pgk1 ILMN_1369074 0.91 0.99 0.92 12.36 0.46 0.64 0.96 Gapdh ILMN_1649859 1.11 0.93 1.19 12.95 0.64 0.54 0.95 Tfrc ILMN_1360908 1.00 0.92 1.09 7.01 0.65 0.54 0.95 Actb ILMN_1355039 1.17 0.93 1.27 12.27 0.9 0.42 0.93 Actb ILMN_2038798 1.18 0.87 1.36 11.87 1.19 0.33 0.91 Ubc ILMN_1350494 1.26 1.16 1.09 13.49 1.29 0.3 0.91 Polr2a_mapped ILMN_1372495 1.02 0.90 1.14 7.83 1.57 0.23 0.89 Hmbs ILMN_1353365 0.88 1.06 0.83 8.22 1.68 0.21 0.88 Actb ILMN_2038799 1.21 0.90 1.34 13.81 2.12 0.15 0.85 Tbp ILMN_1349379 0.88 1.09 0.81 8.32 2.49 0.11 0.83 H2A.1 ILMN_1372198 1.01 1.19 0.85 7.83 3 0.07 0.78 Ywhaz ILMN_1373913 0.77 0.82 0.94 13.58 3.23 0.06 0.76 Hprt1 ILMN_1367708 0.78 1.06 0.74 11.66 3.49 0.05 0.75 B2m ILMN_1368656 0.98 1.23 0.80 12.96 4.31 0.03 0.72

35

Figure 3.13. RT-qPCR expression levels for genes altered by prenatal alcohol exposure. Three genes were significantly upregulated in E animals (Med28 and Acsl3 in PFC; LOC688637 in HPC) (a), and two showed down regulation that approached significance in PFC (b). Fold-changes in expression were positively correlated between microarray and RT-qPCR results (c). ** = p<0.01, * = p<0.05, # = p<0.1.

36 Table 3.8. Genes showing common change in expression in prefrontal cortex of E and PF animals compared to C animals. Bold = p<0.05. na = probe had no specific alignment to current RefSeq RNA database.

Fold change Average q- Gene Symbol Gene Name F p-value Ethanol/ Ethanol/ Pairfed/ Expression value Control Pairfed Control Rpusd1 RNA pseudouridylate synthase domain containing 1 9.4 11.9 5.7E-04 0.23 0.68 0.93 0.73 Nme2 NME/NM23 nucleoside diphosphate kinase 2 11.5 10.7 9.7E-04 0.25 0.70 0.89 0.79 Klhl24 kelch-like 24 (Drosophila) 9.6 10.7 9.4E-04 0.25 0.69 0.91 0.75 Ndrg2 N-myc downstream regulated gene 2 13.5 11.2 7.6E-04 0.23 0.71 0.95 0.75 ILMN_1370609 na 7.1 12.3 4.8E-04 0.23 0.60 1.10 0.55 Rasl10a RAS-like, family 10, member A 8.5 13.3 3.2E-04 0.23 0.74 1.14 0.64 ILMN_1359879 na 9.2 10.9 8.6E-04 0.24 0.76 1.17 0.65 Grik5 glutamate receptor, ionotropic, kainate 5 9.4 12.6 4.3E-04 0.23 1.33 0.84 1.58 RGD1309651 similar to 1190005I06Rik protein 7.7 14.0 2.4E-04 0.22 1.31 0.88 1.49 ILMN_1368369 na 7.6 11.1 7.8E-04 0.23 1.38 0.92 1.50 Satb1 SATB homeobox 1 10.2 12.2 5.0E-04 0.23 1.42 0.91 1.55 Tmem178b transmembrane protein 178B 9.5 15.5 1.4E-04 0.16 1.37 0.87 1.56 ILMN_1356747 na 14.1 10.7 9.7E-04 0.25 1.41 0.97 1.45 ILMN_1351805 na 12.1 12.0 5.4E-04 0.23 1.45 0.95 1.53 Mapkapk2 mitogen-activated protein kinase-activated protein kinase 2 8.4 11.9 5.7E-04 0.23 1.33 0.97 1.38 Igfbp7 insulin-like growth factor binding protein 7 11.9 11.2 7.6E-04 0.23 1.43 0.99 1.45 Nrxn3 neurexin 3 10.7 16.5 9.8E-05 0.13 1.46 0.86 1.69 Ywhaq tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein, theta polypeptide 11.8 11.4 7.0E-04 0.23 1.38 1.01 1.37 ILMN_1352779 na 7.1 11.7 6.0E-04 0.23 1.35 1.00 1.36 Gpkow G patch domain and KOW motifs 7.1 10.9 8.8E-04 0.24 1.37 1.04 1.32 LOC685828 hypothetical protein LOC685828 7.6 12.6 4.3E-04 0.23 1.47 0.99 1.49 Gabrr2 gamma-aminobutyric acid (GABA) A receptor, rho 2 7.1 12.5 4.3E-04 0.23 1.37 1.01 1.36 Chn1 chimerin (chimaerin) 1 13.3 11.6 6.5E-04 0.23 1.51 1.08 1.40 RGD1565784 RGD1565784 9.4 11.8 5.9E-04 0.23 1.37 1.05 1.30 ILMN_1366825 na 9.9 13.1 3.4E-04 0.23 1.42 1.01 1.41 Rpl27-l1 ribosomal protein L27-like 1 9.9 11.5 6.8E-04 0.23 1.40 1.16 1.21 ILMN_1366004 na 8.2 11.8 5.8E-04 0.23 1.42 1.11 1.28 ILMN_1359502 na 8.5 13.0 3.7E-04 0.23 1.53 1.06 1.44 ILMN_1366169 na 9.3 16.6 9.3E-05 0.13 1.55 0.98 1.58 ILMN_1367588 na 9.4 13.2 3.3E-04 0.23 1.61 1.18 1.37 ILMN_1359650 na 8.2 14.1 2.3E-04 0.22 1.48 1.10 1.34

37 Fold change Average q- Gene Symbol Gene Name F p-value Ethanol/ Ethanol/ Pairfed/ Expression value Control Pairfed Control RGD1309730 similar to RIKEN cDNA B230118H07 9.2 16.7 9.2E-05 0.13 1.57 1.02 1.53 Sep15 selenoprotein 15 11.3 18.0 5.9E-05 0.12 1.55 1.00 1.55 Hint3 histidine triad nucleotide binding protein 3 9.0 16.2 1.1E-04 0.14 1.52 1.07 1.43 ILMN_1352441 na 13.1 15.0 1.7E-04 0.18 1.74 1.19 1.46 ILMN_1366381 na 10.5 16.5 9.9E-05 0.13 1.55 1.07 1.45 Psma7 proteasome (prosome, macropain) subunit, alpha type 7 11.8 25.5 6.9E-06 0.06 1.68 0.86 1.97 ILMN_1368258 na 10.5 19.8 3.4E-05 0.10 1.76 1.05 1.68 LOC301193 similar to Discs large homolog 5 (Placenta and prostate DLG) (Discs large protein P-dlg) 10.5 21.4 2.1E-05 0.07 1.74 1.04 1.67

Table 3.9. Genes showing common change in expression in hippocampus of E and PF animals compared to C animals. Bold = p<0.05.

Fold change Average q- Gene Symbol Gene Name F p-value Ethanol/ Ethanol/ Pairfed/ Expression value Control Pairfed Control LOC100360417 RUN and SH3 domain containing 1-like 10.4 12.8 2.9E-04 0.21 1.65 1.10 1.50 ATP synthase, H+ transporting, mitochondrial F1 Atp5a1 complex, alpha subunit 1, cardiac muscle 13.1 13.5 2.2E-04 0.20 1.35 0.91 1.48 Acsl1 acyl-CoA synthetase long-chain family member 1 9.5 13.0 2.7E-04 0.21 1.34 0.91 1.47 Sqle squalene epoxidase 10.0 14.1 1.7E-04 0.17 1.44 1.01 1.42

38

Table 3.10. Genes differentially expressed in prefrontal cortex among all 3 contrasts. (p <0.05 for each contrast). Fold change Average q- Gene Symbol Gene Name F p-value Expression value Ethanol/ Ethanol/ Pairfed/ Control Pairfed Control Baiap2 BAI1-associated protein 2 10.8 10.6 9.8E-04 0.25 0.69 0.83 0.83 Lxn latexin 8.3 15.8 1.2E-04 0.15 0.77 1.42 0.54 Tuba1a tubulin, alpha 1A 14.1 25.2 7.5E-06 0.06 0.77 1.38 0.56 Tom1 target of myb1 homolog (chicken) 8.2 10.7 9.3E-04 0.25 0.83 1.17 0.71 Sumf1 sulfatase modifying factor 1 8.3 12.8 3.8E-04 0.23 1.17 1.45 0.81 Acat1 acetyl-CoA acetyltransferase 1 8.3 12.5 4.3E-04 0.23 1.20 0.80 1.49 Dynlrb1 dynein light chain roadblock-type 1 12.2 11.4 6.8E-04 0.23 1.21 0.79 1.54 Rnd2 Rho family GTPase 2 11.0 11.0 8.1E-04 0.23 1.22 0.85 1.43 Epn1 Epsin 1 8.7 22.7 1.4E-05 0.06 1.24 0.75 1.66 LOC100361558 histone H3.3B-like 11.4 11.8 5.9E-04 0.23 1.24 0.79 1.56 Acly ATP citrate lyase 10.0 11.4 7.1E-04 0.23 1.25 0.78 1.60 Peo1 progressive external ophthalmoplegia 1 8.7 13.9 2.6E-04 0.22 1.26 0.80 1.58 Hbb-b1 hemoglobin, beta adult major chain 10.8 24.7 8.6E-06 0.06 1.28 1.73 0.74 Anxa4 annexin A4 9.8 13.7 2.7E-04 0.22 1.29 0.84 1.54 Ckb creatine kinase, brain 12.8 16.5 9.8E-05 0.13 1.34 0.72 1.86 Scd stearoyl-Coenzyme A desaturase 1 13.2 12.6 4.1E-04 0.23 1.36 0.78 1.75 LOC501223 similar to Discs large homolog 5 (Placenta and prostate DLG) (Discs 11.2 13.9 2.5E-04 0.22 1.55 1.21 1.27 large protein P-dlg) Rps27l3 ribosomal protein S27-like 3 7.4 18.5 5.0E-05 0.11 1.65 1.21 1.37 LOC363320 similar to Discs large homolog 5 (Placenta and prostate DLG) (Discs 9.4 22.6 1.5E-05 0.06 1.90 1.26 1.50 large protein P-dlg) E

39

Table 3.11. Genes differentially expressed in hippocampus among all 3 contrasts. (p <0.05 for each contrast). Fold change Average Gene Symbol Gene Name F p-value q-value Ethanol/ Ethanol/ Pairfed/ Expression Control Pairfed Control Phlpp1 PH domain and leucine rich repeat protein phosphatase 1 10.2 12.7 3.1E-04 0.21 0.78 0.63 1.24 RGD1565117 similar to 40S ribosomal protein S26 9.6 22.4 9.5E-06 0.04 1.24 1.61 0.77 Trpv4 transient receptor potential cation channel, subfamily V, 7.4 27.0 2.6E-06 0.02 1.26 1.84 0.69 member 4 Agap1 ArfGAP with GTPase domain, ankyrin repeat and PH domain 1 10.1 12.7 3.0E-04 0.21 1.30 0.81 1.59 Mgp matrix Gla protein 9.0 12.9 2.8E-04 0.21 1.42 1.97 0.72 Col8a1 collagen, type VIII, alpha 1 7.9 17.2 5.2E-05 0.12 1.46 2.47 0.59 Igf2 insulin-like growth factor 2 11.8 13.4 2.3E-04 0.20 1.63 2.80 0.58 E

40

Figure 3.14. Biological Processes altered by prenatal treatment. Venn diagrams demonstrating the number of Biological Processes significant for each contrast in Day 16 Saline animals, and overlap of processes between different contrasts for PFC and HPC at FDR <1% (a). FDR was increased to 10% to identify Biological Processes that showed overlapping changes in both tissues, specific to prenatal alcohol exposure (b) and pair-feeding (c). FDR <10%.

41

Table 3.12. Genes differentially expressed in PFC of PF animals vs both E and C animals. Bold = p <0.05. na = probe had no specific alignment to current RefSeq RNA database.

Fold change Average Gene Symbol Gene name F p-value q-value Expression Pairfed/ Ethanol/ Ethanol/ Control Pairfed Control ILMN_1358743 na 8.0 11.5 6.8E-04 0.23 0.81 1.44 1.17 Lrp1 low density lipoprotein receptor-related 8.8 12.4 4.6E-04 0.23 1.38 0.69 0.95 protein 1 ILMN_1361625 na 9.8 11.4 6.9E-04 0.23 1.58 0.72 1.13 Ak2 adenylate kinase 2 9.9 13.4 3.1E-04 0.23 1.69 0.67 1.13 ILMN_1359487 na 10.6 11.0 8.3E-04 0.24 1.53 0.76 1.16 Ppp1r1b protein phosphatase 1, regulatory 10.8 11.7 6.2E-04 0.23 1.47 0.72 1.06 (inhibitor) subunit 1B Park7 parkinson protein 7 12.2 12.5 4.4E-04 0.23 0.70 1.38 0.97

42 Table 3.13. Genes differentially expressed in HPC of PF animals vs both E and C animals. Bold = p <0.05. na = probe had no specific alignment to current RefSeq RNA database.

Fold change Average Gene Symbol Gene Name F p-value q-value Pairfed/ Ethanol/ Ethanol/ Expression Control Pairfed Control ILMN_1351851 na 8.9 12.4 3.5E-04 0.24 0.77 1.49 1.15 Sostdc1 sclerostin domain containing 1 8.7 26.4 3.1E-06 0.02 0.28 4.40 1.23 Nt5dc2 5'-nucleotidase domain 7.8 14.6 1.4E-04 0.15 0.68 1.70 1.16 containing 2 Retsat retinol saturase (all trans 8.4 15.8 8.7E-05 0.14 1.68 0.49 0.82 retinol 13,14 reductase) ILMN_1356875 na 10.0 22.2 1.0E-05 0.04 0.46 2.68 1.22 Aqp1 aquaporin 1 7.6 15.9 8.6E-05 0.14 0.63 1.75 1.10 Igfbp2 insulin-like growth factor 10.0 14.7 1.4E-04 0.15 0.47 3.45 1.62 binding protein 2 Lxn latexin 8.3 15.4 1.0E-04 0.14 0.57 1.69 0.96 Ttr transthyretin 11.7 19.3 2.5E-05 0.08 0.20 9.87 1.96 Slco1a5 solute carrier organic anion 8.2 15.4 1.0E-04 0.14 0.43 2.57 1.11 transporter family, member 1a5 Glb1l galactosidase, beta 1-like 7.4 18.6 3.2E-05 0.09 0.68 1.47 1.00 Epn3 epsin 3 7.5 12.8 2.9E-04 0.21 0.76 1.46 1.11 F5 coagulation factor V 8.6 17.8 4.3E-05 0.11 0.33 4.27 1.39 (proaccelerin, labile factor) Cox8b cytochrome c oxidase, subunit 7.4 13.5 2.1E-04 0.20 0.71 1.48 1.06 VIIIb Enpp2 ectonucleotide 12.7 27.2 2.5E-06 0.02 0.47 2.45 1.16 pyrophosphatase/phosphodies terase 2

43 3.4 Prenatal alcohol exposure alters the gene expression response to an inflammatory challenge in PFC and HPC

To examine the effects of prenatal alcohol exposure on the response to an inflammatory challenge, gene expression profiles were compared between saline-exposed and adjuvant-exposed animals. The greatest effect of adjuvant exposure on gene expression was seen at Day 16 post-adjuvant injection (Figure 3.15), at the peak of inflammation (Zhang et al. 2012), with 59 genes altered in PFC and 13 altered in HPC in response to adjuvant in at least one prenatal group (FDR <25%) (Figure 3.16). According to p-value distributions, there was also some evidence for differential expression in PF and E animals during the resolution phase of arthritis in response to adjuvant (Figure 3.15). However, only 2 genes were differentially expression at this time point at an FDR <25%, so analyses focused on expression changes at the peak of inflammation.

3.4.1 Incidence and severity of adjuvant-induced arthritis

Data on incidence and severity of adjuvant-induced arthritis in C, PF and E offspring have been reported previously (Zhang et al. 2012). Briefly, E animals had an increase in the incidence, course, and severity of adjuvant-induced arthritis, and demonstrated a blunted lymphocyte proliferative response to the mitogen concanavalin A during the induction phase of adjuvant-induced arthritis. Additionally, E animals had higher basal ACTH levels during the induction phase compared to PF and C animals.

3.4.2 Genes differentially altered in PAE compared to control animals in response to adjuvant exposure

The dominant response to adjuvant in the brain appeared to be an up-regulation of mRNA levels at the peak of inflammation, in many cases across all three prenatal groups (Figure 3.16). Notably, this was observed in both PFC and HPC, and 8 genes were altered in common in both tissues (again, mostly up-regulated in the adjuvant condition); these were Lcn2, Vwf, Hba-a2, Csda, Asah3l, S100a8, Slc38a5, and MGC72973. However, a subset of these differentially expressed genes (8 in PFC, and 4 in HPC) demonstrated a significantly different response to adjuvant in E animals compared to both C and PF (i.e. a significant effect of the interaction between prenatal alcohol exposure and adjuvant exposure in adulthood) (Table 3.14 and Table 3.15). For the majority of these genes, C and PF animals showed a significant up-regulation of expression in response to adjuvant, but E animals showed no change in expression levels between the saline and adjuvant conditions. These genes were largely multifunctional, with functions in cell growth and proliferation (Ghrhr, Ctgf, Sgk1, Vwf), cell adhesion and structural organization (Ctgf, Flna, Vwf, Sgk1), cell death (Ctgf, Lcn2, Sgk1), response to stress (Ctgf, Lcn2, Sgk1, Vwf), and the immune response (Lcn2,

44 Bhlhe40). In general, these functions are involved in the cellular response to immunological or stressful stimuli.

3.4.3 GO categories differentially altered in PAE compared to control animals in response to adjuvant exposure

As with genes altered in the unchallenged, saline-injected animals, none of the genes altered in response to adjuvant were significantly enriched for GO categories, KEGG pathways, or disease phenotypes. With gene- score resampling, however, many Biological Process categories were found to be altered in response to adjuvant exposure within each prenatal treatment group, at an FDR of 1%. In both PFC and HPC, E animals had the fewest uniquely altered categories (8% in PFC, and 11% in HPC), C animals had the most uniquely altered categories (25% in PFC and 30% in HPC), and PF animals were intermediate (24% in PFC and 21% in HPC) (Figure 3.17a). Four specific processes overlapped between PFC and HPC for unique ethanol effects (Figure 3.17b): “regulation of epithelial cell proliferation”, “positive regulation of epithelial cell proliferation”, “cellular protein complex assembly”, and “regulation of hormone level”. In categories that were common between PF and C but not E animals (in what might be considered the normal response to adjuvant exposure), 6 categories overlapped between PFC and HPC (Figure 3.17c): “response to organic nitrogen”, “actin filament-based process”, “actin cytoskeleton organization”, “regulation of cell morphogenesis”, “developmental growth”, and “mRNA metabolic process”.

45

Figure 3.15. P-value distributions for response to adjuvant exposure. Plot of p-value distributions for gene expression differences within prenatal treatment groups in response to adjuvant exposure at the peak of adjuvant-induced arthritis (16 days post-adjuvant injection) and during the resolution phase of adjuvant-induced arthritis (39 days post-injection). PFC = prefrontal cortex, HPC = hippocampus.

46

Figure 3.16. Effects of Adjuvant exposure on gene expression at the peak of inflammation.

47

(Previous page) Effects of Adjuvant exposure on gene expression at the peak of inflammation (Day 16 post-injection). 59 probes demonstrated significant changes in expression among treatment groups in prefrontal cortex (a). 13 genes demonstrated significant changes among treatment groups in the hippocampus (b). A subset of genes in either tissue demonstrated unique alterations in response to prenatal alcohol exposure.

Table 3.14. Genes altered in PFC of E animals in response to adjuvant at peak of inflammation. Genes with a significantly different response to Adjuvant in Ethanol-exposed compared to both Control and Pair-fed animals (p <0.05) in prefrontal cortex at peak of inflammation. Bold = p <0.05. na = probe had no specific alignment to current RefSeq RNA database.

Average p- q- Fold change (Adjuvant/Saline) Gene Symbol Gene Name F Expression value value Control Pairfed Ethanol ILMN_1351665 na 7.0 7.9 3.4E-04 0.17 0.80 0.80 1.12 Ghrhr growth hormone releasing hormone receptor 7.0 8.2 2.5E-04 0.14 0.87 0.78 1.23 ILMN_1354124 na 6.9 7.1 7.0E-04 0.24 0.94 0.99 1.34 ILMN_1364624 na 8.4 7.2 6.3E-04 0.24 1.22 1.06 0.51 ILMN_1372588 na 11.1 8.7 1.7E-04 0.13 1.38 1.10 0.67 ILMN_1351971 na 11.9 9.8 6.5E-05 0.08 1.40 1.23 0.71 Flna filamin A, alpha 8.6 7.1 7.1E-04 0.24 1.33 1.27 0.99 Bhlhe40 basic helix-loop-helix family, member e40 9.5 8.1 2.8E-04 0.15 1.42 1.45 1.02

Table 3.15. Genes altered in HPC of E animals in response to adjuvant at peak of inflammation. Genes with a significantly different response to Adjuvant in Ethanol-exposed compared to both Control and Pair-fed animals (p <0.05) in hippocampus at peak of inflammation. Bold = p <0.05.

Average q- Fold change (Adjuvant/Saline) Gene Symbol Gene Name F p-value Expression value Control Pairfed Ethanol Sgk1 serum/glucocorticoid regulated kinase 1 11.4 9.1 1.1E-04 0.18 1.63 1.67 1.01 Vwf von Willebrand factor 8.9 15.6 7.3E-07 0.00 1.76 1.70 1.06 Lcn2 lipocalin 2 7.4 18.6 1.1E-07 0.00 1.55 1.92 1.03 Ctgf connective tissue growth factor 10.4 11.4 1.6E-05 0.05 1.77 2.14 0.85

48

Figure 3.17. Biological Processes altered in the response to adjuvant exposure. Venn diagrams demonstrating the number of Biological Processes significantly altered in the response to Adjuvant within each prenatal treatment group, and the overlap of processes enriched between groups (a). Many Biological Processes showed changes specific to prenatal alcohol exposure, and several overlapped between tissues (b). Other processes were common to the PF and C response to adjuvant, and several overlapped between tissues (c). FDR <1%.

49 Chapter 4: Discussion and Conclusion

In this study, prenatal ethanol exposure was shown to cause lasting changes in the expression of many genes in both the PFC and HPC of adult female rats. These genes had many shared functions, including roles in neurodevelopment, cell signaling, cell death, and transcriptional regulation. Additionally, prenatal alcohol exposure altered the response to an inflammatory challenge for many other genes in these brain regions, particularly at the peak of inflammation. The majority of these genes play a role in immune function and cellular responses to stressful stimuli. Notably, while the inflammatory challenge elicited overlapping changes in gene expression between both brain regions, the effects of alcohol differed between brain regions. The results of this study support the premise that prenatal alcohol exposure has lasting effects on neurological and neuroimmune function at the level of gene expression, and provides several new candidate genes that may contribute to the etiology FASD.

4.1 Effects of prenatal alcohol exposure on steady-state gene expression

The effects of prenatal ethanol exposure in itself were analyzed within the animals that were unchallenged in adulthood (ie. the saline-injected animals). In both PFC and HPC, a number of genes – 15 in PFC and 4 in HPC – showed unique effects of ethanol, where E animals differed from control animals. These genes had many overlapping functions, including roles in neurodevelopment, regulation of cell death and cell differentiation, regulation of transcription, and roles in neuronal signaling, particularly with regards to AMPA receptor activity. Overall, however, the effects of prenatal alcohol exposure on steady-state gene expression in the PFC and HPC were subtle. This finding is consistent with the one previous study examining global gene expression in the adult PAE brain (Kleiber et al. 2012),which found very low fold-changes in gene expression in brains of adult male mice exposed prenatally to alcohol. In fact, only 8 genes in the study by Kleiber et al. showed consistent fold-changes higher than 1.3-fold, and this appeared to be in the absence of multiple- testing correction (therefore the false discovery rate is unknown). In the present study, 19 genes showed unique effects of prenatal alcohol exposure (15 in PFC and 4 in HPC) at a 25% FDR versus both PF and C animals (where PF and C did not differ from each other), with fold-changes in expression ranging from approximately 1.3-fold to 1.8-fold. Many of these genes are involved in neurodevelopment (Tcf4, Ap1s2, Cnih2, and Acsl3), as expected, given ethanol’s known neuroteratogenic effects. While long-term effects of prenatal alcohol exposure on neurodevelopmental genes were also identified in the genome-wide expression analysis by Kleiber et al (Kleiber et al. 2012), none of the genes altered by PAE overlapped between these two studies. This could be attributed to differences between the brain regions analyzed, species differences, and sex of the animals. Additionally, these genes have not been identified in any genome-wide expression studies that looked at immediate rather than long-term effects of alcohol exposure during early development

50 (Da Lee et al. 2004; Hard et al. 2005; Green et al. 2007; Zhou, Zhao, et al. 2011, Downing et al. 2012), nor in any targeted studies of the impact of prenatal alcohol exposure on the transcriptome. This is thus a novel discovery of the impacts of prenatal alcohol exposure on the brain.

Notably, significant changes in basal gene expression in response to prenatal ethanol exposure were only evident at 16 days but not 39 days post-saline injection. It is possible that differences between the two time points may have arisen due to handling associated with being the control group for the adjuvant-injected animals. These animals were injected with saline around 60 days of age, and were anesthetized with isofluorane every few days for measurement of paw volumes to compare with adjuvant-injected animals (Zhang et al. 2012). Saline injection and the stress of associated handling may have exacerbated underlying differences among prenatal groups in the earlier time point (16 days post-injection). Conversely, additional exposures to isofluorane may have normalized or masked subtle differences among groups by the later time point (39 days post-injection), as isofluorane exposure has been shown to alter the neural transcriptome in the rat (Ponomarev et al. 2010). Either of these factors may contribute to the fact that differences in basal gene expression were seen in the first time point, but not the second.

RT-qPCR was used to validate the expression results for 15 of the 19 PAE-altered genes, and the fold- changes between E and C samples had a significant positive correlation between the RT-qPCR and microarray results. Expression changes were significant for 3 of these genes (Acsl3, LOC688637, and Med28), and approached significance for another two (Ppp1r14a and Pex11g). Failure to validate all genes at a significant level could be due to the small n tested (limited to 3 each of E and C animals, due to batch issues), combined with trying to validate relatively small fold-changes in expression (less than two-fold). Replication of these findings in larger, independent studies would properly validate their susceptibility to programming by prenatal alcohol exposure.

One of the validated neurodevelopmental genes, Acsl3, is an acyl-coA synthetase that is highly expressed in the brain, and is important for lipid metabolism. Acsl3 is normally highly expressed in the rat brain during early postnatal development, with expression decreasing into adulthood (Fujino et al. 1996). It was found to be upregulated in PFC of adult ethanol-exposed animals in our study, and therefore could possibly be escaping the typical age-dependent downregulation observed (Fujino et al. 1996). DNA methylation and mRNA levels of Acsl3 have been shown to be altered by prenatal exposure to air pollutants in human fetal umbilical cord blood (Perera et al. 2009), therefore it is plausible that other noxious prenatal exposures, such as alcohol exposure, could also reprogram its expression in other tissues, such as the brain. Another gene, Ppp1r14a, was validated to be down-regulated in the PFC of ethanol-exposed animals. Ppp1r14a is a substrate for Cdk5, a kinase with function important in neuronal signaling and neuronal development (Schnack et al. 2008), and endogenous Ppp1r14a has been shown to be required for long-term depression

51 (LTD) in cerebellar Purkinje cells (Eto et al. 2002). Prenatal alcohol exposure has been shown to alter LTD in the cerebellum and hippocampus (Izumi et al. 2005; Servais et al. 2007), and downregulation of Ppp1r14a could contribute to this impairment.

LOC688637 was the only gene validated in the HPC, and was upregulated in ethanol-exposed animals in this study. LOC688637 is homologous to the human and mouse Wdr36 gene. Wdr36 has been previously shown to be expressed in both human and mouse brain tissue (Monemi et al. 2005), and it appears to be important to apoptosis, synaptic function, and immune function. Wdr36 knockout is embryonic lethal in the mouse (Gallenberger et al. 2011), and its deletion, depletion, and mutation have all been shown to be proapoptotic in the mouse embryo, human cell lines, and multiple types of retinal neuronal cells (Gallenberger et al. 2011; Chi et al. 2010). Chi et al also demonstrated that mutation of Wdr36 results in synapse disruption between retinal neurons (Chi et al. 2010). In regards to immune function, expression of Wdr36 correlates highly with expression of IL2, appears to be involved in T-cell activation (Mao et al. 2004), and the Wdr36 gene region has been repeatedly linked to asthma in genome-wide association studies (Gudbjartsson et al. 2009; Moffatt et al. 2010; Hirota et al. 2011). While LOC688637 was not found to be significantly altered in response to adjuvant in this study, it may still play a role in PAE-related immune dysfunction.

4.2 Effects of prenatal alcohol exposure on the neural response to adjuvant-induced arthritis

Additional differences in gene expression emerged between the ethanol-exposed and control groups when the animals were challenged with an inflammatory adjuvant. Overall, we found that ethanol-exposed animals showed less evidence for changes in gene expression in response to an immune challenge than their control counterparts. According to p-value distributions, C animals showed changes at the peak of AA in the PFC, PF animals showed changes during resolution of AA in the HPC, and E animals showed little change at either time point in either tissue. At a 25% FDR, the main response to adjuvant was for increased expression of genes at the peak of arthritis, many of which are involved in the immune response or response to cellular stress. While this response was particularly strong in control animals, in a number of cases this increase in expression was seen across all three diet groups. However, ethanol-exposed animals differed from controls in a number of genes, where they failed to mount a typical expression response to the inflammatory stimulus. In most of these cases, control animals showed an increase in mRNA levels in response to adjuvant, whereas ethanol-exposed animals showed no change between the saline and adjuvant conditions. This corresponds well to the null p-value distributions observed in ethanol-exposed animals.

The majority of the genes that ethanol-exposed animals failed to regulate in response to adjuvant exposure (Ghrhr, Ctgf, Sgk1, Vwf, Flna, Lcn2, and Bhlhe40) were primarily involved in functions related to the immune response and cellular responses to stressful stimuli. The CNS and HPA axis play an important role in

52 responding to immune challenges, as all three systems share bidirectional communication, with shared ligands and receptors (Bodnar & Weinberg 2013). In this case, the CNS may be upregulating immune-related genes in response to sensing a peripheral inflammatory stimulus (reviewed in Ousman & Kubes 2012), or in response to local neuroinflammation, which can occur in adjuvant-induced arthritis (Liu et al. 2012). Animals prenatally exposed to alcohol may be failing to sense these immune changes or to launch the appropriate recovery program, and this may contribute to the prolonged inflammatory response seen in ethanol-exposed animals (Zhang et al. 2012). The gene Lcn2, for example, was upregulated in response to adjuvant in the HPC of control, but not ethanol-exposed, animals. Lcn2 has been shown to be upregulated in response to neuroinflammation or injury, and is thought to be a mediator of reactive astrocytosis and glial cell death (Lee et al. 2009). Of relevance, adjuvant-induced arthritis has been shown to alter glial cell morphology in the CA1 region of the hippocampus, with astrocytes developing larger bodies and thicker processes (Liu et al. 2012), characteristics of reactive astrocytes that may be mediating repair and recovery from neuroinflammation (reviewed in Sofroniew 2005). Lack of Lcn2 upregulation in the HPC of ethanol-exposed animals may therefore be reflective of improper cellular activation in response to neuroinflammation, though it is not known whether this expression change is occurring globally in the hippocampus or within specific cell types, such as glia. Similarly, these animals failed to upregulate Bhlhe40 in the PFC at the peak of AA. Bhlhe40- deficient mice have been shown to have impairments in T-cell proliferation and in elimination of activated T and B cells (Sun et al. 2001). A lack of upregulation in ethanol-exposed animals may therefore be impairing their ability to clear activated, proinflammatory lymphocytes. Additionally, Bhlhe40 -/- mice develop characteristics of autoimmune disease as they age (Sun et al. 2001), which may be a product of their impaired ability to clear activated lymphocytes, and which might parallel the sustained inflammatory response seen in PAE rats (Zhang et al. 2012).

Animals exposed prenatally to alcohol may be failing to launch an appropriate neuroimmune response to an inflammatory insult, which may be resulting in the increased susceptibility and impaired recovery from adjuvant exposure seen in these animals (Zhang et al. 2012). Liu and colleagues have recently shown that rats of a similar age to the animals in this study show an anti-inflammatory microglial cell morphology in response to adjuvant-induced arthritis, whereas older rats showed a pro-inflammatory state in response to adjuvant-induced arthritis (Liu et al. 2012). Given that E animals failed to upregulate a number of potentially anti-inflammatory genes in response to adjuvant (eg. Lcn2, Bhlhe40), and that these animals had a prolonged course and increased severity of adjuvant-induced arthritis (Zhang et al. 2012), PAE may be inducing a lasting pro-inflammatory state in the brain. Similar to prenatal exposure to infection, prenatal alcohol exposure has the potential to expose the fetus to elevated levels of cytokines, as chronic alcohol consumption has been shown to increase levels of proinflammatory cytokines (Crews et al. 2006). Interestingly, in the unchallenged state, few genes involved in the inflammatory response or immune response appear to be dysregulated (the exception being LOC688637). It was only when presented with an inflammatory challenge

53 that evidence emerges for a dysregulated neuroimmune system. This is a common theme in stress-related FASD research, where certain deficits in the HPA axis are only revealed in the presence of repeated or chronic stress.

4.3 Overlapping effects of prenatal alcohol exposure and pair-feeding on steady-state gene expression

Many genes were similarly altered in ethanol-exposed and pair-fed animals compared to ad libitum-fed controls. For a few probes in the PFC, there was a gradient effect of prenatal treatment on gene expression, where ethanol exposure had the greatest effect on expression, but pair-feeding similarly increased or decreased expression relative to controls (in the genes Baiap2, LOC501223, LOC363320, and RGD1564290). This is not surprising, as there are common effects of prenatal ethanol exposure and pair-feeding on dams compared to control dams (such as decreased caloric intake, and activation of the stress response), yet there ethanol exposure still provides unique effects. Baiap2, the best annotated of the four genes, was downregulated in ethanol-exposed and pair-fed animals, compared to controls. Baiap2 is important to formation of dendritic spines (Choi et al. 2005), and knockout mice demonstrate impairments in learning and memory (Kim et al. 2009). While neither LOC501223 nor LOC363320 are on the rat genome reference assembly, both are annotated as similar to Discs large homolog 5 (DLG5), which has functions in maintenance of cell structure (Nakamura et al. 1998) and cell-cell contact (Wakabayashi et al. 2003).

Many additional genes were expressed at the same level in ethanol-exposed and pair-fed animals, with both being different from ad libitum-fed animals, particularly in PFC (39/80 probes in PFC and 4/30 probes in HPC). These genes could therefore be sensitive to the common components of ethanol exposure and pair- feeding in the developing fetus, but insensitive to the other effects of ethanol. These common effects include the reduced caloric availability and the potentially stressful effects of the feeding paradigms on the pregnant dam. Similarly, some genes demonstrated a graded effect where both groups were similarly altered relative to controls, but the effect in response to pair-feeding was larger than the effect in response to ethanol exposure. In such cases, certain common effects of pair-feeding and ethanol exposure may be exacerbated in pair-fed animals, with a resultant larger lasting change in expression. Interestingly, for some genes, ethanol exposure and pair-feeding altered gene expression in opposite directions relative to controls. It is possible that these genes are prone to fetal programming by a variety of environmental factors, and are programmed differentially depending on the type of exposure.

54 4.4 Unique effects of pair-feeding on steady-state gene expression

Pair-feeding also had unique effects on adult basal gene expression, where 22 genes were different only in pair-fed animals compared to both E and C groups. This was especially the case for the hippocampus, where 50% of the genes with a significant effect of prenatal treatment were altered only in pair-fed animals. The majority of these genes were involved in processes such as small molecule metabolism, signal transduction, and response to stress. This is not surprising, for while both ethanol-fed and pair-fed dams receive the same amount of calories, the pair-fed dams are generally hungry. Not only is this decreased food intake likely stressful for the dams (Harris & Seckl 2011), but they tend to eat their daily ration quickly, which may have unique metabolic effects associated with disordered eating. The fact that the effects of pair-feeding were the dominant effect seen in the hippocampus suggests that this brain region may be particularly susceptible to fetal programming in response to metabolic and stress-related environmental factors. While pair-feeding is an important control to account for the effects of reduced food intake in this paradigm of prenatal alcohol exposure, it is clearly also a treatment in itself that can result in unique changes in gene expression, and therefore discretion must be exercised when interpreting results in this experimental paradigm.

4.5 Limitations and future directions

In this study, we used a careful experimental design, rigorous statistical methods, and a well-established platform for analyzing gene expression. Regardless, there are several limitations of this study. One limitation is that relatively large brain regions (whole PFC and whole HPC) with heterogeneous cell populations were assayed. This may have contributed to the subtlety of the effects of alcohol exposure observed. It is possible that larger expression changes may exist in smaller sub-regions, or specific cell types, but these changes may have been masked or washed out by the signal from the rest of the tissue. Future studies should examine smaller, more specific regions of the brain, and would benefit even further if divided into specific cell types, such as glia and neurons. Large sample sizes will also increase the power to detect small effects. Another limitation is that the RT-qPCR replication of gene expression was conducted using the same tissue samples, and not in an independent population. Future studies in an independent cohort of animals should be conducted to replicate these findings in a targeted fashion. A final issue is that the animals used to examine steady-state levels of gene expression were not entirely unchallenged, but had been injected with saline and exposed to isofluorane to serve as controls for adjuvant-injected animals. It is possible that additional changes in gene expression, or entirely different ones, might emerge in a population of PAE animals that has not been manipulated in adulthood.

While this study has given us an insight into some of the gene expression changes induced by prenatal alcohol exposure on a genome-wide level in the brain, it is only the tip of the iceberg. We are just beginning to

55 understand the molecular basis of the long-term neuroendocrine and neuroimmune effects of PAE. Future studies should focus first on replicating these findings, and then expand upon them to examine changes in additional brain regions, and under other challenged conditions. Another issue to explore is whether epigenetic changes underlie these differences in gene expression, and at what point they arise. It is possible that persistent changes in DNA methylation at these genes could serve as a biomarker for PAE and aid in diagnosing FASD. Ultimately, it is hoped that this study and others like it will contribute not only to understanding the effects of alcohol exposure on the developing brain, but will one day help alleviate the challenges faced by individuals with FASD.

56 References

Abel, E.L. & Dintcheff, B.A., 1978. Effects of prenatal alcohol exposure on growth and development in rats. The Journal of Pharmacology and Experimental Therapeutics, 207(3), pp.916–921.

Barker, D.J.P. & Osmond, C., 1986. Infant mortality, childhood, nutrition, and ischaemic heart disease in England and Wales. The Lancet, 327(8489), pp.1077–1081.

Barker, DJP et al., 1989. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease. BMJ, 298(6673), pp.564–567.

Bodnar, T. & Weinberg, J., 2013. Prenatal Alcohol Exposure : Impact on Neuroendocrine – Neuroimmune Networks,

Bond, N.W. & Di Giusto, E.L., 1976. Effects of prenatal alcohol consumption on open-field behaviour and alcohol preference in rats. Psychopharmacologia, 46(2), pp.163–5.

Brown, N.A., Goulding, E.H. & Fabro, S., 1979. Ethanol embryotoxicity: direct effects on mammalian embryos in vitro. Science (New York, N.Y.), 206(4418), pp.573–5.

Cairns, J.M. et al., 2008. BASH: a tool for managing BeadArray spatial artefacts. Bioinformatics (Oxford, England), 24(24), pp.2921–2.

Chernoff, G.F., 1977. The fetal alcohol syndrome in mice: an animal model. Teratology, 15(3), pp.223–9.

Chi, Z.-L. et al., 2010. Mutant WDR36 directly affects axon growth of retinal ganglion cells leading to progressive retinal degeneration in mice. Human molecular genetics, 19(19), pp.3806–15.

Choi, J. et al., 2005. Regulation of dendritic spine morphogenesis by insulin receptor substrate 53, a downstream effector of Rac1 and Cdc42 small GTPases. The Journal of neuroscience : the official journal of the Society for Neuroscience, 25(4), pp.869–79.

Chudley, A.E. et al., 2005. Fetal alcohol spectrum disorder: Canadian guidelines for diagnosis. Canadian Medical Association Journal, 172(5), pp.S1–S21.

Church, M. & Gerkin, K., 1988. Hearing disorders in children with fetal alcohol syndrome: findings from case reports. Pediatrics, 82(2), pp.147–154.

Crews, F.T. et al., 2006. Cytokines and alcohol. Alcoholism, clinical and experimental research, 30(4), pp.720–30.

Dessaud, E. et al., 2007. Interpretation of the sonic hedgehog morphogen gradient by a temporal adaptation mechanism. Nature, 450(7170), pp.717–20.

Downing, C. et al., 2012. Gene expression changes in C57BL/6J and DBA/2J mice following prenatal alcohol exposure. Alcoholism, clinical and experimental research, 36(9), pp.1519–29.

Dunning, Mark J et al., 2007. beadarray: R classes and methods for Illumina bead-based data. Bioinformatics (Oxford, England), 23(16), pp.2183–4.

57 Dwinell, M.R. et al., 2009. The Rat Genome Database 2009: variation, ontologies and pathways. Nucleic acids research, 37(Database issue), pp.D744–9.

Ericson, J. et al., 1997. Pax6 controls progenitor cell identity and neuronal fate in response to graded Shh signaling. Cell, 90(1), pp.169–80.

Eto, M. et al., 2002. Cerebellar long-term synaptic depression requires PKC-mediated activation of CPI-17, a myosin/moesin phosphatase inhibitor. Neuron, 36(6), pp.1145–58.

Ewald, S.J. & Frost, W.W., 1987. Effect of prenatal exposure to ethanol on development of the thymus. Thymus, 9(4), pp.211–5.

Ewald, S.J. & Walden, S.M., 1988. Flow cytometric and histological analysis of mouse thymus in fetal alcohol syndrome. Journal of Leukocyte Biology, 44, pp.434–440.

Fujino, T. et al., 1996. Molecular characterization and expression of rat acyl-CoA synthetase 3. The Journal of biological chemistry, 271(28), pp.16748–52.

Gabriel, K.I. et al., 2005. Postnatal handling does not normalize hypothalamic corticotropin-releasing factor mRNA levels in animals prenatally exposed to ethanol. Brain research. Developmental brain research, 157(1), pp.74–82.

Gallenberger, M. et al., 2011. Lack of WDR36 leads to preimplantation embryonic lethality in mice and delays the formation of small subunit ribosomal RNA in human cells in vitro. Human molecular genetics, 20(3), pp.422–35.

Gallo, P. V & Weinberg, J., 1981. Corticosterone rhythmicity in the rat: interactive effects of dietary restriction and schedule of feeding. The Journal of nutrition, 111(2), pp.208–18.

Gallo, P. V & Weinberg, J., 1986. Organ growth and cellular development in ethanol-exposed rats. Alcohol (Fayetteville, N.Y.), 3(4), pp.261–7.

Gauthier, T.W. et al., 2009. In utero ethanol exposure impairs defenses against experimental group B streptococcus in the term Guinea pig lung. Alcoholism, clinical and experimental research, 33(2), pp.300–306.

Gauthier, T.W. et al., 2005. Maternal Alcohol Abuse and Neonatal Infection. Alcoholism: Clinical & Experimental Research, 29(6), pp.1035–1043.

Gauthier, T.W., Manar, M.H. & Brown, L.A.S., 2004. Is maternal alcohol use a risk factor for early-onset sepsis in premature newborns? Alcohol (Fayetteville, N.Y.), 33(2), pp.139–45.

Giberson, P.K. et al., 1997. The effect of cold stress on lymphocyte proliferation in fetal ethanol-exposed rats. Alcoholism, clinical and experimental research, 21(8), pp.1440–7.

Gilbert, S.F., 2012. Ecological developmental biology: environmental signals for normal animal development. Evolution & development, 14(1), pp.20–28.

Gil-Mohapel, J. et al., 2010. Hippocampal cell loss and neurogenesis after fetal alcohol exposure: insights from different rodent models. Brain research reviews, 64(2), pp.283–303.

58 Glaser, R. & Kiecolt-Glaser, J.K., 2005. Stress-induced immune dysfunction: implications for health. Nature Reviews Immunology, 5(3), pp.243–251.

Glavas, M.M. et al., 2007. Effects of prenatal ethanol exposure on basal limbic-hypothalamic-pituitary- adrenal regulation: role of corticosterone. Alcoholism, clinical and experimental research, 31(9), pp.1598–610.

Glavas, M.M. et al., 2001. Effects of prenatal ethanol exposure on hypothalamic-pituitary-adrenal regulation after adrenalectomy and corticosterone replacement. Alcoholism, clinical and experimental research, 25(6), pp.890–7.

Gluckman, P.D. et al., 2008. Effect of in utero and early-life conditions on adult health and disease. The New England journal of medicine, 359(1), pp.61–73.

Green, M.L. et al., 2007. Reprogramming of genetic networks during initiation of the Fetal Alcohol Syndrome. Developmental Dynamics, 236(2), pp.613–631.

Grossmann, A. et al., 1993. Immune function in offspring of nonhuman primates (Macaca nemestrina) exposed weekly to 1.8 g/kg ethanol during pregnancy: preliminary observations. Alcoholism: Clinical …, 17(4), pp.822–827.

Gudbjartsson, D.F. et al., 2009. Sequence variants affecting eosinophil numbers associate with asthma and myocardial infarction. Nature genetics, 41(3), pp.342–7.

Hales, C. et al., 1991. Fetal and infant growth and impaired glucose tolerance at age 64. British Medical …, 303, pp.1019–1022.

Hales, C.N. & Barker, D.J.P., 1992. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia, 35(7), pp.595–601.

Haley, D.W., Handmaker, N.S. & Lowe, J., 2006. Infant stress reactivity and prenatal alcohol exposure. Alcoholism, clinical and experimental research, 30(12), pp.2055–64.

Hard, M.L. et al., 2005. Gene-expression analysis after alcohol exposure in the developing mouse. Journal of Laboratory and Clinical Medicine, 145(1), pp.47–54.

Harris, A. & Seckl, J., 2011. Glucocorticoids, prenatal stress and the programming of disease. Hormones and behavior, 59(3), pp.279–89.

Hellemans, K.G.C. et al., 2010. Prenatal alcohol exposure: fetal programming and later life vulnerability to stress, depression and anxiety disorders. Neuroscience and biobehavioral reviews, 34(6), pp.791–807.

Hicks, S.D., Middleton, F. a & Miller, M.W., 2010. Ethanol-induced methylation of cell cycle genes in neural stem cells. Journal of neurochemistry, 114(6), pp.1767–80.

Hirota, T. et al., 2011. Genome-wide association study identifies three new susceptibility loci for adult asthma in the Japanese population. Nature genetics, 43(9), pp.893–6.

Izumi, Y. et al., 2005. A single day of ethanol exposure during development has persistent effects on bi- directional plasticity, N-methyl-D-aspartate receptor function and ethanol sensitivity. Neuroscience, 136(1), pp.269–79.

59 Jacobson, S.W., Bihun, J.T. & Chiodo, L.M., 1999. Effects of prenatal alcohol and cocaine exposure on infant cortisol levels. Development and psychopathology, 11(2), pp.195–208.

Jankord, R. & Herman, J.P., 2008. Limbic regulation of hypothalamo-pituitary-adrenocortical function during acute and chronic stress. Annals of the New York Academy of Sciences, 1148, pp.64–73.

Jerrells, T.R. & Weinberg, J., 1998. Influence of ethanol consumption on immune competence of adult animals exposed to ethanol in utero. Alcoholism, clinical and experimental research, 22(2), pp.391–400.

Johnson, M.B. et al., 2009. Functional and evolutionary insights into human brain development through global transcriptome analysis. Neuron, 62(4), pp.494–509.

Johnson, S. et al., 1981. Immune deficiency in fetal alcohol syndrome. Pediatric research, 15(6), pp.908–11.

Jones, Kenneth L & Smith, D.W., 1973. Recognition of the fetal alcohol syndrome in early infancy. The Lancet, 302(7836), pp.999–1001.

Kim, C.K. et al., 1999. Effects of Prenatal Ethanol Exposure on Hypothalamic-Pituitary-Adrenal Responses to Chronic Cold Stress in Rats. Alcoholism: Clinical and Experimental Research, 23(2), pp.301–310.

Kim, M.-H. et al., 2009. Enhanced NMDA receptor-mediated synaptic transmission, enhanced long-term potentiation, and impaired learning and memory in mice lacking IRSp53. The Journal of neuroscience : the official journal of the Society for Neuroscience, 29(5), pp.1586–95.

Kirkpatrick, S.E. et al., 1976. Acute effects of maternal ethanol infusion on fetal cardiac performance. American journal of obstetrics and gynecology, 126(8), pp.1034–7.

Kleiber, M.L. et al., 2012. Long-term alterations to the brain transcriptome in a maternal voluntary consumption model of fetal alcohol spectrum disorders. Brain research, 1458, pp.18–33.

Krueger, D.A. & Dodson, S.I., 1981. Embryological induction and predation ecology in Daphnia pulex. Limnology and Oceanography, 26(2), pp.219–223.

Lee, H.K. et al., 2005. ErmineJ: tool for functional analysis of gene expression data sets. BMC bioinformatics, 6(1), p.269.

Da Lee, R. et al., 2004. Differential gene profiles in developing embryo and fetus after in utero exposure to ethanol. Journal of toxicology and environmental health. Part A, 67(23-24), pp.2073–84.

Lee, Shinrye et al., 2009. Lipocalin-2 is an autocrine mediator of reactive astrocytosis. The Journal of neuroscience : the official journal of the Society for Neuroscience, 29(1), pp.234–49.

Lee, Soon et al., 1990. Effect of ethanol on the activity of the hypothalamic-pituitary-adrenal axis of the offspring: importance of the time of exposure to ethanol and possible modulating mechanisms. Molecular and Cellular Neuroscience, 1(2), pp.168–177.

Lee, Soon & Rivier, C., 1996. Gender differences in the effect of prenatal alcohol exposure on the hypothalamic-pituitary-adrenal axis response to immune signals. Psychoneuroendocrinology, 21(2), pp.145–155.

Leek, J.T. & Storey, J.D., 2007. Capturing heterogeneity in gene expression studies by surrogate variable analysis. PLoS genetics, 3(9), pp.1724–35. 60 Lemoine, P. et al., 1968. Les enfants des parents alcooliques: anomalies observées apropos de 127 cas. Ouest Medical, 21, pp.476–482.

Liu, X. et al., 2012. Age-dependent neuroinflammatory responses and deficits in long-term potentiation in the hippocampus during systemic inflammation. Neuroscience, 216, pp.133–42.

Mann, L.I. et al., 1975. Placental transport of alcohol and its effect on maternal and fetal acid-base balance. American journal of obstetrics and gynecology, 122(7), pp.847–44.

Mao, M. et al., 2004. T lymphocyte activation gene identification by coregulated expression on DNA microarrays. Genomics, 83(6), pp.989–99.

May, P.A. et al., 2009. Prevalence and epidemiologic characteristics of FASD from various research methods with an emphasis on recent in-school studies. Developmental disabilities research reviews, 15(3), pp.176–92.

May, P.A. et al., 2011. Prevalence of children with severe fetal alcohol spectrum disorders in communities near Rome, Italy: new estimated rates are higher than previous estimates. International journal of environmental research and public health, 8(6), pp.2331–51.

McGill, J. et al., 2009. Fetal exposure to ethanol has long-term effects on the severity of influenza virus infections. The Journal of Immunology, 182(12), pp.7803–7808.

Moffatt, M.F. et al., 2010. A large-scale, consortium-based genomewide association study of asthma. The New England journal of medicine, 363(13), pp.1211–21.

Monemi, S. et al., 2005. Identification of a novel adult-onset primary open-angle glaucoma (POAG) gene on 5q22.1. Human molecular genetics, 14(6), pp.725–33.

Moscatello, K.M. et al., 1999. Effects of in utero alcohol exposure on B cell development in neonatal spleen and bone marrow. Cellular immunology, 191(2), pp.124–30.

Myers, B., McKlveen, J.M. & Herman, J.P., 2012. Neural Regulation of the Stress Response: The Many Faces of Feedback. Cellular and molecular neurobiology, pp.683–694.

Nahmad, M. & Lander, A.D., 2011. Spatiotemporal mechanisms of morphogen gradient interpretation. Current opinion in genetics & development, 21(6), pp.726–31.

Nakamura, H. et al., 1998. Identification of a novel human homolog of the Drosophila dlg, P-dlg, specifically expressed in the gland tissues and interacting with p55. FEBS letters, 433, pp.63–67.

Nelson, S.B., Hempel, C. & Sugino, K., 2006. Probing the transcriptome of neuronal cell types. Current opinion in neurobiology, 16(5), pp.571–6.

Nolan, T., Hands, R.E. & Bustin, S. a, 2006. Quantification of mRNA using real-time RT-PCR. Nature protocols, 1(3), pp.1559–82.

Norman, A.L. et al., 2009. Neuroimaging and fetal alcohol spectrum disorders. Developmental disabilities research reviews, 15(3), pp.209–17.

Norman, D.C. et al., 1991. Changes with age in the proliferative response of splenic T cells from rats exposed to ethanol in utero. Alcoholism, clinical and experimental research, 15(3), pp.428–32. 61 Norman, D.C. et al., 1989. Diminished proliferative response of con A-blast cells to interleukin 2 in adult rats exposed to ethanol in utero. Alcoholism, clinical and experimental research, 13(1), pp.69–72.

Oldham, M.C. et al., 2008. Functional organization of the transcriptome in human brain. Nature neuroscience, 11(11), pp.1271–82.

Ousman, S.S. & Kubes, P., 2012. Immune surveillance in the central nervous system. Nature Neuroscience, 15(8), pp.1096–1101.

Perera, F. et al., 2009. Relation of DNA Methylation of 59-CpG Island of ACSL3 to Transplacental Exposure to Airborne Polycyclic Aromatic Hydrocarbons and Childhood Asthma. Childhood A Global Journal Of Child Research, 4(2).

Ponomarev, I. et al., 2010. Amygdala transcriptome and cellular mechanisms underlying stress-enhanced fear learning in a rat model of posttraumatic stress disorder. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 35(6), pp.1402–11.

Portales-Casamar, E. et al., 2013. Neurocarta: aggregating and sharing disease-gene relations for the neurosciences. BMC Genomics. 14(1), pp. 129.

Ramsay, D.S., Bendersky, M.I. & Lewis, M., 1996. Effect of prenatal alcohol and cigarette exposure on two- and six-month-old infants’ adrenocortical reactivity to stress. Journal of pediatric psychology, 21(6), pp.833–40.

Randall, C.L., Taylor, J. & Walker, D.W., 1977. Ethanol-induced malformations in mice. Alcoholism, clinical and experimental research, 1(3), pp.219–24.

Randall, C.L. & Taylor, W.J., 1979. Prenatal ethanol exposure in mice: teratogenic effects. Teratology, 19(3), pp.305–11.

Rat Genome Database, RatMine. http://ratmine.mcw.edu/ratmine/begin.do.

Redei, E., Clark, W.R. & McGivern, R.F., 1989. Alcohol exposure in utero results in diminished T-cell function and alterations in brain corticotropin-releasing factor and ACTH content. Alcoholism, clinical and experimental research, 13(3), pp.439–43.

Reul, J.M. & De Kloet, E.R., 1985. Two receptor systems for corticosterone in rat brain: microdistribution and differential occupation. Endocrinology, 117(6), pp.2505–11.

Sampson, P.D. et al., 1997. Incidence of fetal alcohol syndrome and prevalence of alcohol-related neurodevelopmental disorder. Teratology, 56(5), pp.317–26.

Schmittgen, T.D. & Livak, K.J., 2008. Analyzing real-time PCR data by the comparative CT method. Nature Protocols, 3(6), pp.1101–1108.

Schnack, C., Hengerer, B. & Gillardon, F., 2008. Identification of novel substrates for Cdk5 and new targets for Cdk5 inhibitors using high-density protein microarrays. Proteomics, 8(10), pp.1980–6.

Seelig, L.L., Steven, W.M. & Stewart, G.L., 1996. Effects of maternal ethanol consumption on the subsequent development of immunity to Trichinella spiralis in rat neonates. Alcoholism: Clinical and Experimental Research, 20(3), pp.514–522.

62 Seres, J. et al., 2002. Effects of chronic food restriction stress and chronic psychological stress on the development of adjuvant arthritis in male Long Evans rats. Annals of the New York Academy of Sciences, 966, pp.315–319.

Servais, L. et al., 2007. Purkinje cell dysfunction and alteration of long-term synaptic plasticity in fetal alcohol syndrome. Proceedings of the National Academy of Sciences of the United States of America, 104(23), pp.9858–63.

Smith, D.W., Jones, Kenneth L & Ulleland, C.N., 1973. Patterns of malformation in offspring of chronic alcoholic mothers.

Smyth, G.K., 2005. limma: Linear Models for Microarray Data. In R. Gentleman et al., eds. Bioinformatics and Computational Biology Solutions Using R and Bioconductor. Springer New York, pp. 397–420.

Sofroniew, M. V, 2005. Reactive astrocytes in neural repair and protection. The Neuroscientist : a review journal bringing neurobiology, neurology and psychiatry, 11(5), pp.400–7.

Stade, B. et al., 2009. The burden of prenatal exposure to alcohol: revised measurement of cost. The Canadian journal of clinical pharmacology = Journal canadien de pharmacologie clinique, 16(1), pp.e91–102.

Stratton, K.R., Howe, C.J. & Battaglia, F.C., 1996. Fetal alcohol syndrome: diagnosis, epidemiology, prevention, and treatment, Washington, DC: National Academy Press.

Streissguth, Clarren, S K & Jones, K L, 1985. Natural history of the fetal alcohol syndrome: a 10-year follow- up of eleven patients. Lancet, 2(8446), pp.85–91.

Sulik, K.K., Johnston, M.C. & Webb, M.A., 1981. Fetal alcohol syndrome: embryogenesis in a mouse model. Science, 214(4523), pp.936–938.

Sun, H. et al., 2001. Defective T cell activation and autoimmune disorder in Stra13-deficient mice. Nature immunology, 2(11), pp.1040–7.

Taylor, A.N. et al., 1982. Long-term effects of fetal ethanol exposure on pituitary-adrenal response to stress. Pharmacology, biochemistry, and behavior, 16(4), pp.585–589.

Taylor, A.N. et al., 1988. Maternal alcohol consumption and stress responsiveness in offspring. Advances in experimental medicine and biology, 245, pp.311–7.

Taylor, A.N. et al., 1983. Neonatal and long-term neuroendocrine effects of fetal alcohol exposure. Monographs in neural sciences, 9, pp.140–52.

Thanh, N.X. & Jonsson, E., 2010. Drinking alcohol during pregnancy: evidence from Canadian Community Health Survey 2007/2008. Journal of population therapeutics and clinical pharmacology = Journal de la thérapeutique des populations et de la pharamcologie clinique, 17(2), pp.e302–7.

Vandesompele, J. et al., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome biology, 3(7), p.RESEARCH0034.

Vieau, D. et al., 2007. HPA axis programming by maternal undernutrition in the male rat offspring. Psychoneuroendocrinology, 32 Suppl 1, pp.S16–20.

63 Vink, J. et al., 2005. Novel peptides prevent alcohol-induced spatial learning deficits and proinflammatory cytokine release in a mouse model of fetal alcohol syndrome. American journal of obstetrics and gynecology, 193(3 Pt 1), pp.825–9.

Wakabayashi, M. et al., 2003. Interaction of lp-dlg/KIAA0583, a membrane-associated guanylate kinase family protein, with vinexin and beta-catenin at sites of cell-cell contact. The Journal of biological chemistry, 278(24), pp.21709–14.

Weinberg, J., 1985. Effects of ethanol and maternal nutritional status on fetal development. Alcoholism, clinical and experimental research, 9(1), pp.49–55.

Weinberg, J., 1984. Nutritional issues in perinatal alcohol exposure. Neurobehavioral toxicology and teratology, 6, pp.261–269.

Weinberg, J., 1992. Prenatal ethanol effects: sex differences in offspring stress responsiveness. Alcohol (Fayetteville, N.Y.), 9(3), pp.219–23.

Weinberg, J. et al., 2008. Prenatal alcohol exposure: foetal programming, the hypothalamic-pituitary-adrenal axis and sex differences in outcome. Journal of neuroendocrinology, 20(4), pp.470–88.

Weinberg, J. & Jerrells, T.R., 1991. Suppression of immune responsiveness: sex differences in prenatal ethanol effects. Alcoholism, clinical and experimental research, 15(3), pp.525–31.

Zhang, X. et al., 2012. Prenatal alcohol exposure alters the course and severity of adjuvant-induced arthritis in female rats. Brain, behavior, and immunity, 26(3), pp.439–450.

Zhang, X., Sliwowska, J.H. & Weinberg, J., 2005. Prenatal alcohol exposure and fetal programming: effects on neuroendocrine and immune function. Experimental biology and medicine (Maywood, N.J.), 230(6), pp.376–88.

Zhi-Liang, H., Bao, J. & Reecy, J., 2008. CateGOrizer: a web-based program to batch analyze gene ontology classification categories. Online J Bioinformatics.

Zhou, F.C., Zhao, Q., et al., 2011. Alteration of gene expression by alcohol exposure at early neurulation. BMC genomics, 12(1), p.124.

Zhou, F.C., Chen, Y. & Love, A., 2011. Cellular DNA methylation program during neurulation and its alteration by alcohol exposure. Birth defects research. Part A, Clinical and molecular teratology, 91(8), pp.703–15.

64 Appendix A Supplementary Tables

Table A.1. Candidate genes involved in the etiology of FASD, catalogued in Neurocarta Symbol Gene name Probe Abca1 ATP-binding cassette, subfamily A (ABC1), member 1 ILMN_1650701 Abcg1 ATP-binding cassette, subfamily G (WHITE), member 1 ILMN_1354046 Actb actin, beta ILMN_1355039 Actb actin, beta ILMN_2038798 Actb actin, beta ILMN_2038799 Adcy8 adenylate cyclase 8 (brain) ILMN_1350196 Akt1 v-akt murine thymoma viral oncogene homolog 1 ILMN_1353102 Alpl alkaline phosphatase, liver/bone/kidney ILMN_1372113 Apoe apolipoprotein E ILMN_1367529 Atoh1 atonal homolog 1 (Drosophila) ILMN_1368168 Bad BCL2-associated agonist of cell death ILMN_1369751 Bcl2 B-cell CLL/lymphoma 2 ILMN_1366150 Bcl2l1 Bcl2-like 1 ILMN_1355163 Bcl2l1 Bcl2-like 1 ILMN_1365285 Bdnf brain-derived neurotrophic factor ILMN_1360447 Cacna1c calcium channel, voltage-dependent, L type, alpha 1C subunit ILMN_1370304 Casp3 caspase 3 ILMN_1349218 Cat catalase ILMN_1369530 Ccnd1 cyclin D1 ILMN_1350372 Ccnd2 cyclin D2 ILMN_1362471 Chat choline O-acetyltransferase ILMN_1363883 Creb1 cAMP responsive element binding protein 1 ILMN_1649829 Creb1 cAMP responsive element binding protein 1 ILMN_1376791 Cyba cytochrome b-245, alpha polypeptide ILMN_1366276 Dlg4 discs, large homolog 4 (Drosophila) ILMN_1650748 Duox1 dual oxidase 1 ILMN_1367874 E2f1 E2F transcription factor 1 ILMN_1360877 Egfr epidermal growth factor receptor ILMN_1362571 Erbb2 v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma ILMN_1350020 derived oncogene homolog (avian) Fgfr2 fibroblast growth factor receptor 2 ILMN_1371701 Gad1 glutamate decarboxylase 1 ILMN_1351478 Gfap glial fibrillary acidic protein ILMN_1376423 Gpx1 glutathione peroxidase 1 ILMN_1372510 Gpx3 glutathione peroxidase 3 ILMN_1365802 Gria2 glutamate receptor, ionotropic, AMPA 2 ILMN_1356417 Gria3 glutamate receptor, ionotrophic, AMPA 3 ILMN_1368538 Gria4 glutamate receptor, ionotrophic, AMPA 4 ILMN_1371769 Grin1 glutamate receptor, ionotropic, N-methyl D-aspartate 1 ILMN_1365529 Grin2b glutamate receptor, ionotropic, N-methyl D-aspartate 2B ILMN_1366396 Grm5 glutamate receptor, metabotropic 5 ILMN_1361607 Gsk3b glycogen synthase kinase 3 beta ILMN_1349648 Gsr glutathione reductase ILMN_1352580 Gstm2 glutathione S-transferase mu 2 ILMN_1350896 Gstm3 glutathione S-transferase mu 3 ILMN_1374835 Hoxa1 homeo box A1 ILMN_1353666 Hoxb4 homeo box B4 ILMN_1363620 Hoxd4 homeo box D4 ILMN_1367426 Hoxd4 homeo box D4 ILMN_1353520

65 Symbol Gene name Probe Igf1r insulin-like growth factor 1 receptor ILMN_1374575 Igf2 insulin-like growth factor 2 ILMN_1359301 Igf2r insulin-like growth factor 2 receptor ILMN_1349413 Insr insulin receptor ILMN_1360127 Irs1 insulin receptor substrate 1 ILMN_1360680 L1cam L1 cell adhesion molecule ILMN_1376861 Mapk1 mitogen activated protein kinase 1 ILMN_1349290 Mapt microtubule-associated protein tau ILMN_1354816 Ncf2 neutrophil cytosolic factor 2 ILMN_1365484 Ndufv1 NADH dehydrogenase (ubiquinone) flavoprotein 1 ILMN_1365082 Neurod neurogenic differentiation 1 ILMN_1363838 1 Ngfr nerve growth factor receptor (TNFR superfamily, member 16) ILMN_1365512 Notch1 notch 1 ILMN_1359640 Nox3 NADPH oxidase 3 ILMN_1376975 Noxa1 NADPH oxidase activator 1 ILMN_1365297 Noxo1 NADPH oxidase organizer 1 ILMN_1368197 Ntf3 neurotrophin 3 ILMN_1371735 Ntf4 neurotrophin 4 ILMN_1363013 Ntrk1 neurotrophic tyrosine kinase, receptor, type 1 ILMN_1370831 Ntrk2 neurotrophic tyrosine kinase, receptor, type 2 ILMN_1366426 Ntrk3 neurotrophic tyrosine kinase, receptor, type 3 ILMN_1362434 Plat plasminogen activator, tissue ILMN_1358127 Rac1 ras-related C3 botulinum toxin substrate 1 ILMN_1355225 Rara retinoic acid receptor, alpha ILMN_1368986 Rbp1 retinol binding protein 1, cellular ILMN_1375320 S100b S100 calcium binding protein B ILMN_1373043 Sdha succinate dehydrogenase complex, subunit A, flavoprotein (Fp) ILMN_1357678 Serpine serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1), ILMN_1376417 1 member 1 Serpine serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1), ILMN_2040557 1 member 1 Sod1 superoxide dismutase 1, soluble ILMN_1353544 Sod2 superoxide dismutase 2, mitochondrial ILMN_1367263 Sod3 superoxide dismutase 3, extracellular ILMN_1361581

66 Table A.2. Accessions and probe sequences for differentially expressed genes

Gene Symbol Probe ID Current Accession Probe Sequence H2afv ILMN_1356468 NM_001106019.1 TCCTCACTGTGTGTGACTGGGCAGAGGGTACCAGTCGGTGTGTGGGAAAG Tcf4 ILMN_1369541 NM_053369.1 GGGAGACACAGCGAATCACATGGGTCAGATGTGAAAGGGTCCAAGTTGCC Rnasek ILMN_1373564 NM_001137561.2 AGTCCATCTGTTCCACTCATCTGGTGTCCTTTGGGACTGTTACCCTGGCG Ppp1r14a ILMN_1374409 NM_130403.1 GCAGGGAGGCAGACATGCCAGATGAGGTCAACATCGACGAGCTGCTGGAA Rps8 ILMN_1362384 NM_031706.1 TCTTTCCAGCCAGCGCCGAGCGATGGGCATCTCTCGGGACAACTGGCACA ILMN_1372701 ILMN_1372701 na TCTCTAGGAGCCTTGCCTGTCCAAGTCTATCAGCAGACTGTGTTCCTGTC ILMN_1374168 ILMN_1374168 na CTACTTGGCAGATACAAACTGACCAATGGATGATGTCAGGGAGTCTACAC Pex11g ILMN_1360592 NM_001105902.1 AGACTCAGATTCCCAGAGCGGGAACCACTGGCGGGGGGAGCATCCATAAT Ndfip1 ILMN_1650482 NM_001013059.1 CTCAGCTGCGGGAAGGTATGGGGCCATCTCAGGATTTGGTCTTTCTCTAA Acsl3 ILMN_1368504 NM_057107.1 CCCACTGAAAATTCGTTTGAGCCCTGACCCATGGACTCCCGAAACTGGTC Dusp6 ILMN_1362834 NM_053883.2 ATGCTCGCCCATTCAACGGGTGGGATGCGACAGGTTGTGAGGAAGGGAAA Rpl7 ILMN_1370118 NM_001100534.1 TCCGGCTGGAACCATGGAGGCTGTACCAGAGAAGAAAAAGAAGGTTGCCG Med28 ILMN_1355511 NM_001107217.1 GACATGCCTCAGGGCTCCTTGGCCTACCTTGAGCAGGCATCTGCCAACAT Atp6ap1 ILMN_1359644 NM_031785.1 GGCGGGTGGGGGTTAAGAATGAGCGGTACACTGGGGTTTATTTCTGTGAC Ap1s2 ILMN_1372527 NM_001127531.2 GCAGACTTGACAGCAGCTCCCTATCCTTTCATGTCACTGCCTAGTCGTCG Cnih2 ILMN_1372279 NM_001025132.1 TGGAAGGGGTAGGACTTCCGGTCTTGTCCGTTTCAGGGGAATTTGGGCCC Caap1 ILMN_1349802 NM_001034154.1 GTGCTGATGGAGGACGCCTCTTTGTGGATGTGAGTTTCCTAGTTAAAACT LOC688637 ILMN_1363581 XM_001067706.1 GGACGGCGGGTCAATAGAGGCCATGCGGAGCTTTTTGAGTATGATCGGGG Rgs3 ILMN_1370455 NM_019340.1 GGCAGCTGGGCCTTCTAGACTGACATGACCTTGGAGGGGATGCTGCAGAA ILMN_1358743 ILMN_1358743 na TTTGGATACACCCTGTTTCTCTTCCGGTCCCAGGCACTGCGGGAGCTGCA Lrp1 ILMN_1357522 NM_001130490.1 TTAGTTGAGGGAAGTCACCCCAAGCCCCAGCTCCCACTTTTAGGGGCACG ILMN_1361625 ILMN_1361625 na AAAGCAACCCTAAGAGAACACAAATGCCAGCCCAGGTTACTGTATCCTGC Ak2 ILMN_1359709 NM_001033967.2 GCCTGGCACTGGAAAGCCTTGGGTTCGGTCCTCTACAGTGAAGGGTTAGG ILMN_1359487 ILMN_1359487 na TGGCCTCTACTGCACTCTTCCCACCAGAGAAGCACAGATCCAGGGGCACT Ppp1r1b ILMN_2040370 NM_138521.1 CATTCTGGATGTCGTCCCTTATTGTCCTGTTCCTGCTGGGTGCCTGCAAG Park7 ILMN_1370124 NM_057143.1 GTCCACAGCCCAGTGAACCTCAGGAACTAACGTGTGAAGTAGCCCGCTGC ILMN_1351851 ILMN_1351851 na CATGCTCTGTGTGGGATGGCTCGTGTGCAGCGTAAATCTATCTCGTGTGG Sostdc1 ILMN_1352748 NM_153737.1 ATCCCCCTCGTGTTGACCTCTCTTGGAGTGGAATGCCAGCAATGCAAGGC Nt5dc2 ILMN_1356169 NM_001009271.1 GCCCTGTTCAACGCTCAGTTTGGGAGCATCTTCCGCACCTTCCACAACCC Retsat ILMN_1356474 NM_145084.1 ACTGTTCCCACAGCTGGAAGGCAAGGTGGAGAGTGTGACTGGAGGATCCC ILMN_1356875 ILMN_1356875 na CCTCCTCTCCCACAGGCCCAAGATGTAACCCACCAGTGCCTTTTGTCTTC Aqp1 ILMN_1358325 NM_012778.1 CCCTAGCAGGCACTATACTCACTTCACAGGTCAGGACACTGAGGACCCAT Igfbp2 ILMN_1360048 NM_013122.2 ATAGAGAGGGTGGTGGCACTGGGGATACTGGGTACAGGCTTGGGAATGGG Lxn ILMN_1362583 NM_031655.1 GCACAACAGCCGCCTCCCAAAGGAAGCACCAGCAGAGTAAACAAGACCCC Ttr ILMN_1363307 NM_012681.2 TGTCGTCAGTAACCCCCAGAACTGAGGGACCCAGCCCACGAGGACCAAGA

67 Gene Symbol Probe ID Current Accession Probe Sequence Slco1a5 ILMN_1363789 NM_030838.1 GGAGAGGTGTGCTTTCTACCAAGCCTGACAAGGTGGGTTTGATCTCTGGG Glb1l ILMN_1364521 NM_001127529.2 GGACAATGCGGGGTCCACAACAGACCCTATACGTGCCAAGACCTCTGCTG Epn3 ILMN_1365679 NM_001024791.1 CAAGCTAGGGACTGACTGCATCTTGGGATCGAGGACTACGCCCGCCTAAT F5 ILMN_1371753 XM_222831.4 CAAGAAGGTAACGGCCATCGTAACTCAGGGTTGCAAGTCTCTGTCCTCTG Cox8b ILMN_1374366 NM_012786.1 GGCCAAGGAAAGAGTGCGACCCCGAGAATCATGCCAAGGCTTCCCCCTAT Enpp2 ILMN_1376810 NM_057104.2 AGCGAGATTTAACTTTCTGGGCCTGGGCAGTGTAGTCTTAGCAACTGGTG H2afv ILMN_1356468 NM_001106019.1 TCCTCACTGTGTGTGACTGGGCAGAGGGTACCAGTCGGTGTGTGGGAAAG Tcf4 ILMN_1369541 NM_053369.1 GGGAGACACAGCGAATCACATGGGTCAGATGTGAAAGGGTCCAAGTTGCC Rnasek ILMN_1373564 NM_001137561.2 AGTCCATCTGTTCCACTCATCTGGTGTCCTTTGGGACTGTTACCCTGGCG Ppp1r14a ILMN_1374409 NM_130403.1 GCAGGGAGGCAGACATGCCAGATGAGGTCAACATCGACGAGCTGCTGGAA Rps8 ILMN_1362384 NM_031706.1 TCTTTCCAGCCAGCGCCGAGCGATGGGCATCTCTCGGGACAACTGGCACA ILMN_1372701 ILMN_1372701 na TCTCTAGGAGCCTTGCCTGTCCAAGTCTATCAGCAGACTGTGTTCCTGTC ILMN_1374168 ILMN_1374168 na CTACTTGGCAGATACAAACTGACCAATGGATGATGTCAGGGAGTCTACAC Pex11g ILMN_1360592 NM_001105902.1 AGACTCAGATTCCCAGAGCGGGAACCACTGGCGGGGGGAGCATCCATAAT Ndfip1 ILMN_1650482 NM_001013059.1 CTCAGCTGCGGGAAGGTATGGGGCCATCTCAGGATTTGGTCTTTCTCTAA Acsl3 ILMN_1368504 NM_057107.1 CCCACTGAAAATTCGTTTGAGCCCTGACCCATGGACTCCCGAAACTGGTC Dusp6 ILMN_1362834 NM_053883.2 ATGCTCGCCCATTCAACGGGTGGGATGCGACAGGTTGTGAGGAAGGGAAA Rpl7 ILMN_1370118 NM_001100534.1 TCCGGCTGGAACCATGGAGGCTGTACCAGAGAAGAAAAAGAAGGTTGCCG Med28 ILMN_1355511 NM_001107217.1 GACATGCCTCAGGGCTCCTTGGCCTACCTTGAGCAGGCATCTGCCAACAT Atp6ap1 ILMN_1359644 NM_031785.1 GGCGGGTGGGGGTTAAGAATGAGCGGTACACTGGGGTTTATTTCTGTGAC Ap1s2 ILMN_1372527 NM_001127531.2 GCAGACTTGACAGCAGCTCCCTATCCTTTCATGTCACTGCCTAGTCGTCG Baiap2 ILMN_1365343 NM_057196.1 AGTTCCTGCCTTCTCTCAGGGTCTGGATGACTACGGGGCACGGTCTGTGA Lxn ILMN_1362583 NM_031655.1 GCACAACAGCCGCCTCCCAAAGGAAGCACCAGCAGAGTAAACAAGACCCC Tuba1a ILMN_1354206 NM_022298.1 TAAGTGTGAATGATTTGTCAGAGACCCGAGCCGTCCACTTCACTGATGGG Tom1 ILMN_1351051 NM_001008365.1 GGAGCCTGAAGAGGGCTTTAGTGGCTTATTAGGAAGGGCAATGGTGGCCC Sumf1 ILMN_1360059 NM_001108639.1 GAAGGAAAGCGCTGGAGGAGCTGCCATGAGGGAAATGGACATGTGGCCAG Acat1 ILMN_1373473 NM_017075.1 GGCATGGCTCAGCCGTTAAGAGCACTTGTTGCTACCTGTGTGGTGCATGG Dynlrb1 ILMN_1372238 NM_131910.3 CCACCAAGGAGTGCCTCTGATGATCCGGTCAGTCCCCAGAAGAGCTCAGT Rnd2 ILMN_1356527 NM_001010953.1 GGGTAGGCATCGGAGGCATGAACTTGGATAGGGCAGGTAGGTGTTCGGAA Epn1 ILMN_1351904 NM_057136.1 TACACCGCCAGGAGCCAAGGCTTCCAACCCATTCCTTCCAAGTGGAGCTC LOC100361558 ILMN_1649986 XM_002728043.2 GATTCGCAAGCTCCCCTTTCAGTGTCTGGAGCGAGAAATTGCTCAGGACC Acly ILMN_1366910 NM_016987.2 TGTCAAGGGGAGGAGGGTTGGGGCCATTGTACCCTTAGCCATCGTCACAC Peo1 ILMN_1357847 NM_001107599.1 CAACAAGAGTTCCCTTACCTTCTCCATCCCACCTAAGAGCAAAGCCCGAC Hbb-b1 ILMN_1361935 NM_198776.1 GGGGAAAGGTGAACCCTGTTGAAATTGGCGCTGAGTCCCTTGCCAGTCTG Anxa4 ILMN_1349705 NM_024155.3 GAGTGGACTCGGCCAAGGTTGTCCTGGTAATGAGATGCTCTGGGTGTGGC Ckb ILMN_1370888 NM_012529.2 AAGTGAAGCCGTGGCCCTAGCCACCACCAGGCTGCCGCTTCCTAACTTAT Scd ILMN_1359586 NM_031841.1 CGTGAATCATGGGACAGACTCAAACCGTAGCACTGGGCATGTCGCCTAAC 68 Gene Symbol Probe ID Current Accession Probe Sequence LOC501223 ILMN_1350559 XM_001071000.3 ACCACCAGTAACTGCTTTTCTACTATGTCTCCAAGGCAAGCCACAGGCTA RGD1564290 ILMN_1362582 XM_574121.3 AAGGTTGACATTCGCTGGTAAGCAGCTGGAAGGTGGCCGTACTTTGTCTG LOC363320 ILMN_1363630 XM_001075455.3 GGCCTTACTTTCACCCCAGCTCTCTCCAGCAAGTGTGCACTTTTAGAGGG Cnih2 ILMN_1372279 NM_001025132.1 TGGAAGGGGTAGGACTTCCGGTCTTGTCCGTTTCAGGGGAATTTGGGCCC MGC125002 ILMN_1349802 NM_001034154.1 GTGCTGATGGAGGACGCCTCTTTGTGGATGTGAGTTTCCTAGTTAAAACT LOC688637 ILMN_1363581 XM_001067706.1 GGACGGCGGGTCAATAGAGGCCATGCGGAGCTTTTTGAGTATGATCGGGG Rgs3 ILMN_1370455 NM_019340.1 GGCAGCTGGGCCTTCTAGACTGACATGACCTTGGAGGGGATGCTGCAGAA Phlpp1 ILMN_1362245 NM_021657.1 ACCTCGCCCATGTGCAGTGTGGGCCATTTGCTTAGTGTGCTTCTGTGCAG RGD1565117 ILMN_1360197 XM_235217.1 GTCACCAGGAATCCATCTCGTGAGGACCGAACACCCCCACCACGATTCAG Trpv4 ILMN_1360233 NM_023970.1 CCCGGGCTAGGGTGGGTCTTCTGTACTTTGTAGAGATCGGGGCTGTTGGT Agap1 ILMN_1357547 NM_001108230.1 GGCCTCAGCCACTCCCGATCCACAAAGTCTGAATCACCCAGGTTTCTGTC Mgp ILMN_1351917 NM_012862.1 CTACTTCAGGCAGCGCCGAGGAGCCAAATAAGAGCGCAAGGAAACAGTCG Col8a1 ILMN_1362033 NM_001107100.1 CGGAGACCGGGTGTTCCTCCAAATGCCTTCAGAACAGGCTGCTGGACTCT Igf2 ILMN_1359301 NM_031511.2 CCCATGTCATCCAGCAGTGGCCCCGGGTATTTGCCCCAACTCAGTCCTTT ILMN_1351665 ILMN_1351665 na CTCAGCCCCGTAACTGATGTGGCTTGTGACTGGGTTGTGACTGTGTCGTG Ghrhr ILMN_1368059 NM_012850.1 GCTCTGAAGGGGAGCTCTTGTCAGCAGCCATTATTTGCACTTCCGGTGCA ILMN_1354124 ILMN_1354124 na CTCGGTTTCGGAGAGCCCTGAATTTCTCAACCTTGGTCTCGGTGGGCAGC ILMN_1364624 ILMN_1364624 na TGGCAAGTCCCACACTGTGCCCAAGAAGCTACTGATGTTGGCTGGTATAG ILMN_1372588 ILMN_1372588 na AAAGGCTCTGTGAAGAGGCCGGAATGAACATCTGTGACCCCAGTGCCACG ILMN_1351971 ILMN_1351971 na AATTTCTACTCAGTGTTGGATGGCTTTTTCCTTAATACCCCCACGCCAAC Flna ILMN_1368821 NM_001134599.1 CCTGCGCTGTGTTCACCTGCCTTTGGGCTTTCACTTGGGCAGAGGGAGTT Bhlhe40 ILMN_1374180 NM_053328.1 GCTAAGGTGGTGAGGTAGCCAACACTGGCATGTCTCGGTAGTGGTTTGGG Sgk1 ILMN_1349269 NM_019232.3 GGGTTTTTATGGACCAATGCCCCAGTTGTCAGTCAGAGCCGTTGGTGTTC Vwf ILMN_1352807 XM_001066203.3 TGCAGATGTTCTCCCCGTAATTGTGGCAAGTGAGGCCTGTGCAGCCACGG Lcn2 ILMN_1363606 NM_130741.1 TGAACAGACGGTGAGCGTGGCTGACTGGGATGTGCAGTGGCCTGATGGTT Ctgf ILMN_1364113 NM_022266.2 CCACGAGGAAGTGTTTGCTGCTTCTTTGACTATGACTGGTTTGGGAGGCA LOC363306 ILMN_1352300 XM_343647.2 CACCAATGCCCCCAGGAAAGGCTTTGGTTAAAGAAGGGAGGTACTGAGAT Cib1 ILMN_1370750 NM_031145.1 CTCTCCGAGTTCCAGCACGTCATCTCTCGCTCACCAGACTTTGCCAGCTC LOC498989 ILMN_1363791 XM_574280.1 CCAGGTCTGCTGCCAGGCTCCAAGGGTGGGTCTCTGAGGGGCTAGAAAAT LOC363181 ILMN_1359502 XM_001061883.1 TGGAGGAAGTCAGAGAAGTGTTGATGCACATCAATCAAGAGCTGCTGGTC LOC501089 ILMN_1363630 XM_576504.1 GGCCTTACTTTCACCCCAGCTCTCTCCAGCAAGTGTGCACTTTTAGAGGG Satb1 ILMN_1376648 NM_001012129.1 ACTGCTTGGCGGCCCCAGGTGAAGCGTCAAGGATTGTTGGGTAGAATTTG LOC363433 ILMN_1364278 XM_343755.3 GGATTGAATAGGCTGTACTTTCACCTCAGCTCTCTCCAACAAGTGTGCAC LOC363320 ILMN_1352441 XM_343660.3 GTAACTGCTTGTCTACTATGTCTCCAAGGCAAGCCACAGGCTATAAGCCA Eef1a2l1 ILMN_2039949 NM_033539.1 GAGGCAGACAGTTGCTGTGGGTGTCATCAAAGCCGTGGACAAGAAGGCTG Ndn ILMN_1358752 NM_001008558.1 GGGACTGATGGTCCGTATCGACAAAGAAGGCCCTGGAGAGTTAGCAGGAC LOC360975 ILMN_1354780 XM_573650.1 GGCAGGTCAGAAGCAGATCAATGGATAAGGGCAAGGTGTCCCGAGGAGCC 69 Gene Symbol Probe ID Current Accession Probe Sequence Npc2 ILMN_1352122 NM_173118.1 GGCTGGCCGGGAGTATTACCTCTTCTGTATCTAAGTGCCTCCTGAGTCCC LOC689577 ILMN_1371567 XM_001071243.1 CAGCGTCATCCTCATCCACACCCAGGTGAAACTGTTGGCCTCACCAGCAC Lxn ILMN_1362583 NM_031655.1 GCACAACAGCCGCCTCCCAAAGGAAGCACCAGCAGAGTAAACAAGACCCC LOC501300 ILMN_1352504 XM_576713.1 CTGAGCCCTCTGACAATGACTTACTCTGGGAGAGAAACATCATCCCCTGG RGD1565715_predicted ILMN_1362392 XM_341434.3 GTTGATGATGTCTTGTTGGGCAAGAGGAGAGAAGAGGCTTGAAGACGGGC Per1 ILMN_1353839 XM_340822.2 GACTGTCCGTCTGGTTAAGGCTGCTGACAAGCTGCTGAAGTGGTCTCTCC LOC499079 ILMN_1650602 XM_574363.1 GGCAGGACCCCACGAGCAAACTTGAGCCTTGGAACCACAGAAATAGCAGA Cops4 ILMN_1368937 NM_001004275.1 AAGCAGTGCAGCCTTGAAGCAGTAGCTCCCGTGCCGCCTGGGTCTATGTT Npepl1_predicted ILMN_1373909 XM_001055241.1 GCCAGGGTACGTGTGGTGACTGGCTGTTAGGGACCCATTCTGTGAAGCAG RGD1311122 ILMN_1355336 NM_001037792.1 GTGGCAGTTTATCTGTGGGTGGCAGTTTTCTGTAGTCCTTGACGGTGACG RGD1562162_predicted ILMN_1649986 XR_008163.1 GATTCGCAAGCTCCCCTTTCAGTGTCTGGAGCGAGAAATTGCTCAGGACC Asah3l_predicted ILMN_1364753 XM_001053269.1 GCTTGCCATAGCCCCCACCATTTCCAGGCTCTCTCATTACACAGGAGTCA Arl4a ILMN_1351318 NM_019186.1 TGAAGAGTGTCTACAGCCTGGTTTGCCTGTCTGCCCTCACGGATGCTATT Mrpl37 ILMN_1369643 NM_001004235.1 CTTGCATGGTGCTGTGTGAACCAGGAACCTTCTGGGGCCTGATGCCTCTG Csda ILMN_1355756 XM_001069862.1 ACTAACAACTGCAAAGGGAAGGAGCCCGCACTGTCCATCAAGCTGCGTCC Rasd1 ILMN_1369914 XM_340809.3 TGTGGGGCCAGGACTAACAGGGCATTATCTCGTCTGTGATTGGTGTTGCC Myh11 ILMN_1371040 XM_573030.2 CATTACCCCACCTCTCACCAGGAGTCAACCACAGCCCTGCACAAAGGATG Tmod1 ILMN_1373707 NM_013044.2 CTGCAGGGACAGCCAGCTCCACTCAGCTTCTCCTTGAAACACAACTGCAG Lgi3_predicted ILMN_1367471 XM_224337.4 GGCAAGAATCCTGGGAGAGCCTGTATGGGTGCCAGGAACGTGTTGGTAGC Vwf ILMN_1352807 XM_001066203.1 TGCAGATGTTCTCCCCGTAATTGTGGCAAGTGAGGCCTGTGCAGCCACGG Rap1ga1 ILMN_1372167 XM_233609.4 GCCGTGAGCCAAGTCCTTGTGTGTATCTGTTCACTCTTAGGAGCCACGCC RGD1359349 ILMN_1369723 NM_001007738.1 GGGACCATGTTCAAGAAAACAGCGGCCTGAAGGAAGGCGAAGAACCCTGC Acly ILMN_1366910 NM_016987.1 TGTCAAGGGGAGGAGGGTTGGGGCCATTGTACCCTTAGCCATCGTCACAC S100a8 ILMN_1350690 NM_053822.1 GTTCCTTGCGTTGGTGATAAGGGTGGGCGTGGCAGCTCATAAAGACAGCC RGD1359529 ILMN_1366485 NM_001014193.1 ACTGCACTTTACTGAGGGGTTCGTGTCCAGCATCAGCTCACCTGCCTGAG Lrg1 ILMN_1353943 NM_001009717.1 CAGAGCTGGGGACCTTGTGAGGATGGCAACTGGGGTGCGAGCCAAGGGTA LOC500488 ILMN_1359785 XM_580072.1 ACAGGTGGCATGTACCCTGGCTGAGGTAACATTAGTCATTGCTCTGGGGG LOC287167 ILMN_1376663 NM_001013853.1 GCGCAGAGACCATAGGGAGGTTGTTCATTGTCTTCCCCTCCTCCAAGACC Upp1 ILMN_1370862 NM_001030025.1 ACCATGTGCAGTGCCTGTGGCCTGAAAGCGGCTGTGGTGTGTGTCACTCT Mcfd2 ILMN_1369244 NM_139253.1 GGAAGAGCAGTAGTAGCTGAAAGAGAAACAGCCATAGGTCGTACTTTGCG MGC72973 ILMN_1361935 NM_198776.1 GGGGAAAGGTGAACCCTGTTGAAATTGGCGCTGAGTCCCTTGCCAGTCTG Slc38a5 ILMN_1349808 NM_138854.1 AGTCACTTTCCTGAGTCCCTTCTGCCTGGGACATGGAGGTGGCTGGTCTC Anxa11 ILMN_1376793 NM_001011918.1 GTTTCTGGAGAGAATGGTAGGTGAGCGGGCCACCCGTCTTTGCCTAGGAC Timp3 ILMN_1348821 NM_012886.2 GACCACCTCACACTGTCCCAGCGCAAGGGCCTCAATTACCGCTACCACCT Lcn2 ILMN_1363606 NM_130741.1 TGAACAGACGGTGAGCGTGGCTGACTGGGATGTGCAGTGGCCTGATGGTT Mgp ILMN_1351917 NM_012862.1 CTACTTCAGGCAGCGCCGAGGAGCCAAATAAGAGCGCAAGGAAACAGTCG Hbb ILMN_1353696 NM_033234.1 TGATGATGTTGGTGGCGAGGCCCTGGGCAGGCTGCTGGTTGTCTACCCTT 70 Gene Symbol Probe ID Current Accession Probe Sequence Hba-a2 ILMN_1356639 NM_013096.1 CCCTCCCTTGCACCTATACCTCTTGGTCTTTGAATAAAGCCTGAGTAGGA Nfkbia ILMN_1356628 XM_343065.3 GTTGAACCGCCATAGACTGTAGCTGACCCCAGTGTGCCCTCTCACGTAAG Ddit4 ILMN_1357747 NM_080906.1 GGGGGGATCGGAGCTTCACTACTGACCTGTTCGAGGCAGCTATCTTACAG Slc38a5 ILMN_1349808 NM_138854.1 AGTCACTTTCCTGAGTCCCTTCTGCCTGGGACATGGAGGTGGCTGGTCTC S100a8 ILMN_1350690 NM_053822.1 GTTCCTTGCGTTGGTGATAAGGGTGGGCGTGGCAGCTCATAAAGACAGCC Gpd1 ILMN_1353571 NM_022215.2 ATGAAGGTCAGAGCCATTGGGAAAGGTGAAGTGGGGGAGCCCTGTCATCG Csda ILMN_1355756 XM_001069862.1 ACTAACAACTGCAAAGGGAAGGAGCCCGCACTGTCCATCAAGCTGCGTCC Hba-a2 ILMN_1356639 NM_013096.1 CCCTCCCTTGCACCTATACCTCTTGGTCTTTGAATAAAGCCTGAGTAGGA Snai3_predicted ILMN_1358708 XM_001079335.1 CCTTCTCCCGAATGTCTCTCTTGGTGAGGCACGAGGATGCCGGCTGCTGT MGC72973 ILMN_1361935 NM_198776.1 GGGGAAAGGTGAACCCTGTTGAAATTGGCGCTGAGTCCCTTGCCAGTCTG Asah3l_predicted ILMN_1364753 XM_001053269.1 GCTTGCCATAGCCCCCACCATTTCCAGGCTCTCTCATTACACAGGAGTCA Crtac1 ILMN_2040211 XM_574670.2 ATGGCAAGATGCTGAGCCGAAGTGTGGCCAACAGGGAGATGAACTCGGTG

71