Early-Life Viral Infection Impacts Brain Development in the Piglet

EARLY-LIFE VIRAL INFECTION IMPACTS BRAIN DEVELOPMENT IN THE PIGLET

MATTHEW SCHREIBER CONRAD

DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Neuroscience in the Graduate College of the University of Illinois at Urbana-Champaign, 2014

Urbana, Illinois

Doctoral Committee:

Professor Rodney W. Johnson, Chair, Director of Research Professor Janice M. Juraska Associate Professor Justin S. Rhodes Associate Professor Bradley P. Sutton

ABSTRACT

During infancy and childhood, almost every person will impacted by infection. Neonates, due to their immature immune system, are especially vulnerable to respiratory infection and are more prone to hospitalization. Peripheral immune activation in adulthood has been shown to have neurobehavioral effects, but little is known about how it can impact brain development.

Gliogenesis, synaptogenesis, and myelination peak during the neonatal period and are all prone to environmental insults, including infection. Therefore, the broad aim of this research was to characterize the impact of early-life respiratory viral infection on brain growth and development.

Previous work in our lab has shown that porcine reproductive and respiratory syndrome virus (PRRSV) causes microglia activation, neuroinflammation, and cognitive deficits in piglets.

This work further expands on these findings to look at changes to brain development from a macro and microscopic perspective. In order to non-invasively track macro changes to brain development, magnetic resonance imaging (MRI) procedures were developed to measure brain region volumes in the piglet. Using these methods, the normal brain development of the domestic pig was characterized in order to compare brain growth trajectories to what is known for humans. Additionally, advanced analysis methods for determining gray and white matter changes using voxel based morphometry is now possible due to creation of an averaged piglet brain and MRI-based atlas. This method, along with diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MR-Spec), allow for in vivo quantification of brain growth and development.

Using these imaging methods, the impact of neonatal PRRSV infection was characterized. Reductions in gray and white matter in the primary visual cortex were found in the PRRSV piglets. Additionally, there was a trend for a decrease in fractional anisotropy in the corpus callosum suggesting either delayed or disrupted myelination in the PRRSV piglets. MR-

Spec analysis revealed a significant change in three metabolites in the hippocampus. N-

ii acetylaspartate, creatine, and myo-inositol were all decreased in PRRSV piglets showing that infection caused an energy imbalance and may have impacted neuron and astrocyte health. Collectively, these data show that early-life infection can affect grey and white matter development and cause metabolite disturbances.

In addition to the regional changes to brain development above, changes at the cellular level were explored. Since PRRSV causes deficits in hippocampal-dependent learning and memory, changes to neurogenesis and neuron morphology were assessed as they have been shown to contribute to hippocampal learning. A sexual dimorphism was found in the number of surviving newly divided cells (BrdU+) where males had increased levels compared to females. PRRSV infection caused a significant reduction in males only. Cell fate analysis showed that 80% of the newly divided cells form neurons in controls and this is reduced to 57% in PRRSV piglets. Fifteen percent of the newly formed cells become microglia and this is stable across sex and PRRSV infection. A sexual dimorphism was also found in dentate granule neuron morphology where males had more complex dendritic trees than females in the outer cell layer.

PRRSV infection caused a change in the dendritic tree shape where the initial branch point was further away from the cell soma in the inner granule cell layer. No differences were found in dendritic spine density. Taken together, PRRSV causes changes to neurogenesis and neuron morphology, which may contribute to the deficits in learning and memory.

In order to mitigate the changes to brain development caused by PRRSV, minocycline was used to prevent microglia activation. Minocycline is a second generation tetracycline antibiotic which has been shown to reduce microglia activation

iii and provide protection in rodent models of neuroinflammation. Three weeks of high dose minocycline administration failed to provide neuroprotection in the PRRSV model.

Minocycline treatment caused increased microglia activation and pro-inflammatory cytokine production. The high dose of minocycline may have led to increased intracranial pressure or bilirubin-induced neuroinflammation causing the neuroinflammation. These findings suggest that high dose chronic minocycline treatment is not appropriate attenuating microglia activation in neonates.

Together, these novel data show that early-life respiratory viral infection can impact brain development. Inflammation can disrupt sensitive developmental processes, some in a sexually dimorphic manner. Long-term studies are needed to see if these changes are permanent or if the brain is able to recover.

ACKNOWLEDGEMENTS

I would like to give many thanks to my doctoral advisor, Dr. Rodney Johnson. He has given me a tremendous amount of support during my graduate school career. The willingness to let follow my passions with developing new MRI techniques has allowed me to step out of my comfort zone and gain a small amount of expertise in a field that always fascinated me. I appreciate all of the mentoring and the support to travel across the world to present work from our lab. I would also like to thank Dr. Bradley Sutton for all of his assistance with the MRI work.

I have gained so much knowledge and it will be extremely helpful for my clinical training. I also owe a debt of gratitude to Dr. Janice Juraska and Dr. Justin Rhodes for their mentoring.

Through discussions of data, study design, and techniques, they have contributed greatly to my training. I am also very grateful for my undergraduate advisor, Dr. Andrea Neely, who taught me the basics of scientific research and inspired me to get my MD/PhD.

I would like to extend a special thanks to Dr. Ryan Dilger who took me in under his wing when I first started in the lab. You have taught me so many things over the years including how to handle piglets and sows. You were always available to bounce ideas off of and to help with a problem. Your friendship, both in and out of the workplace, has been incredible.

Also, thank you to all of the other graduate students in the lab. I have learned so much from you throughout the years and look forward to becoming collaborators in the future. Help from the undergraduate students was invaluable as it took a small army to conduct these studies. Also, thank you to my friends in the neuroscience and medical scholars program. I have met so many great people and look forward to continuing the friendships in the future.

Finally, I have to give a huge thank you to my wife, Marti, and my family. I still think that it is fate that brought me to Illinois and that I met you my first day being here. Your unwavering

v love and support has been my inspiration and motivation. Thank you to my parents, Lauree and

Dan, and brother, Greg, for always supporting me. You still think that going to school for this long of a time is crazy (and probably right), but you have always supported me.

TABLE OF CONTENTS

CHAPTER 1 GENERAL INTRODUCTION AND JUSTIFICATION ...... 1 1.1 References ...... 3

CHAPTER 2 LITERATURE REVIEW ...... 5 2.1 Neurodevelopment of the Domestic Pig ...... 5 2.2 Neuroimaging in the Piglet ...... 7 2.3 The Piglet as a Model for Human Neurodevelopment ...... 10 2.4 Piglet Models in Neonatal Research ...... 11 2.5 Microglia and the Developing Brain ...... 14 2.6 Immune to Brain Communication ...... 17 2.7 Pathogenesis of Porcine Reproductive and Respiratory Syndrome Virus ...... 19 2.8 PRRSV and Neuroinflammation ...... 20 2.9 Inflammation and Neurogenesis ...... 22 2.10 Inflammation, Neuron Morphology, and Dendritic Spines ...... 24 2.11 Minocycline and Neuroinflammation ...... 26 2.12 Summary and Objectives ...... 27 2.13 References ...... 30

CHAPTER 3 EARLY POSTNATAL RESPIRATORY VIRAL INFECTION INDUCES STRUCTURAL AND NEUROCHEMICAL CHANGES IN THE NEONATAL PIGLET BRAIN ... 42 3.1 Abstract ...... 42 3.2 Introduction ...... 43 3.3 Materials and Methods ...... 44 3.4 Results ...... 50 3.5 Discussion ...... 53 3.6 Figures and Tables ...... 58 3.7 References ...... 66

CHAPTER 4 EARLY POSTNATAL RESPIRATORY VIRAL INFECTION ALTERS HIPPOCAMPAL NEUROGENESIS, CELL FATE, AND NEURON MORPHOLOGY IN THE NEONATAL PIGLET ...... 70 4.1 Abstract ...... 70 4.2 Introduction ...... 71 4.3 Materials and Methods ...... 72 4.4 Results ...... 80 4.5 Discussion ...... 83 4.6 Figures ...... 88 4.7 References ...... 96

CHAPTER 5 MINOCYCLINE ADMINISTRATION DOES NOT ATTENUATE MICROGLIA ACTIVATION AND NEUROINFLAMMATION INDUCED BY PORCINE REPRODUCTIVE AND RESPIRATORY SYNDROME VIRUS ...... 100 5.1 Abstract ...... 100 5.2 Introduction ...... 101 5.3 Materials and Methods ...... 103 5.4 Results ...... 107 5.5 Discussion ...... 109 5.6 Figures ...... 113 5.7 References ...... 118

vii

CHAPTER 6 SUMMARY AND SIGNIFICANCE...... 121

APPENDIX A MAGNETIC RESONANCE IMAGING OF THE NEONATAL PIGLET BRAIN . 124 A.1 Abstract ...... 124 A.2 Introduction ...... 125 A.3 Materials and Methods ...... 127 A.4 Results ...... 131 A.5 Discussion ...... 132 A.6 Figures and Tables ...... 136 A.7 References ...... 141

APPENDIX B BRAIN GROWTH OF THE DOMESTIC PIG (SUS SCROFA) FROM 2 TO 24 WEEKS OF AGE: A LONGITUDINAL MRI STUDY ...... 144 B.1 Abstract ...... 144 B.2 Introduction ...... 145 B.3 Materials and Methods ...... 147 B.4 Results ...... 149 B.5 Discussion ...... 151 B.6 Figures and Tables ...... 156 B.7 References ...... 161

APPENDIX C AN IN VIVO THREE-DIMENSIONAL MAGNETIC RESONANCE IMAGING- BASED AVERAGED BRAIN AND ATLAS OF THE NEONATAL PIGLET (SUS SCROFA) .. 164 C.1 Abstract ...... 164 C.2 Introduction ...... 165 C.3 Materials and Methods ...... 167 C.4 Results and Discussion ...... 171 C.5 Figures and Tables ...... 174 C.6 References ...... 179

viii

CHAPTER 1

GENERAL INTRODUCTION AND JUSTIFICATION

Environmental insults such as malnutrition, traumatic stress, and infection during sensitive developmental periods can affect neural cells and circuits, slow cognitive development, and increase the risk for behavioral disorders (Shah, Doyle et al. 2008).

However, despite the high susceptibility of the human neonate to infection and a recent prospective study suggesting a correlation between neonatal infection and adverse neurodevelopmental outcomes, the ramifications of postnatal infection on brain and cognitive development are poorly understood (Stoll, Hansen et al. 2004). The potential effects of postnatal infection on brain growth and cognitive development are of particular interest because the human brain undergoes extraordinary growth during the first year of life and disruption of developmental processes may have long-lasting or permanent effects on brain structure and function.

Supporting this notion, inflammation has been shown to inhibit neurogenesis (Ekdahl,

Claasen et al. 2003). Furthermore, epidemiological studies have linked childhood infection with increased risk of schizophrenia, autism, and other affective disorders (Bode, Ferszt et al. 1993;

Rantakallio, Jones et al. 1997; Atladottir, Thorsen et al. 2010). Likewise, maternal infection during pregnancy as well as infection in preterm infants is a risk factor for neurodevelopmental deficits and mental illness (Stoll, Hansen et al. 2004; Boksa 2010). Studies in rodents have shown that adult offspring from infected dams have behavioral abnormalities reminiscent of those seen in humans with psychiatric illnesses (Fatemi, Reutiman et al. 2008). The effects of infection on behavioral development are not caused by the pathogen per se, but rather the host’s inflammatory response against it (Bilbo and Schwarz 2009). Peripheral infection is well known to activate microglia, which we suspect creates an inflammatory environment that affects neurodevelopmental processes including neurogenesis. Abnormal immune activation is seen in

1 psychiatric illnesses, including increased circulating proinflammatory cytokines and activated microglia (van Berckel, Bossong et al. 2008; Ashwood, Krakowiak et al. 2011). In addition to abnormal cytokine profiles, deficits in learning and memory as well as abnormal hippocampal morphology have been reported in association with pre- and postnatal infection (Fatemi, Earle et al. 2002; Bilbo, Rudy et al. 2006). Similar structural changes and abnormal hippocampal volumes have been reported in schizophrenia and autism disorders (Nelson, Saykin et al. 1998;

Rosoklija, Toomayan et al. 2000).

Thus, understanding how infection affects brain and cognitive development may be key to preventing and treating mental illness, which affects more than 25 million Americans each year (Kessler, Chiu et al. 2005). The broad working hypothesis is that neonatal infection affects brain growth and development, leaving behind a “finger print” that leads to behavioral disorders later in life. The goal of this research was to determine if neonatal infection results in changes to multiple levels of brain structure. First, MRI was used to determine regional changes to grey and white matter development including neurochemical concentration analysis of the hippocampus. Secondly, focus was placed cellular changes in the hippocampus including neurogenesis, cell fate, and neuron morphology. Lastly, studies were conducted to test the potential therapeutic value of minocycline for protection against changes induced by infection.

1.1 References Ashwood, P., P. Krakowiak, et al. (2011). "Elevated plasma cytokines in autism spectrum disorders provide evidence of immune dysfunction and are associated with impaired behavioral outcome." Brain Behav Immun 25(1): 40-45.

Atladottir, H. O., P. Thorsen, et al. (2010). "Association of Hospitalization for Infection in Childhood With Diagnosis of Autism Spectrum Disorders: A Danish Cohort Study." Arch Pediatr Adolesc Med 164(5): 470-477.

Bilbo, S. D., J. W. Rudy, et al. (2006). "A behavioural characterization of neonatal infection- facilitated memory impairment in adult rats." Behav Brain Res 169(1): 39-47.

Bilbo, S. D. and J. M. Schwarz (2009). "Early-life programming of later-life brain and behavior: a critical role for the immune system." Front Behav Neurosci 3: 14.

Bode, L., R. Ferszt, et al. (1993). "Borna disease virus infection and affective disorders in man." Arch Virol Suppl 7: 159-167.

Boksa, P. (2010). "Effects of prenatal infection on brain development and behavior: A review of findings from animal models." Brain, Behavior, and Immunity 24(6): 881-897.

Ekdahl, C. T., J.-H. Claasen, et al. (2003). "Inflammation is detrimental for neurogenesis in adult brain." Proceedings of the National Academy of Sciences 100(23): 13632-13637.

Fatemi, S. H., J. Earle, et al. (2002). "Prenatal viral infection leads to pyramidal cell atrophy and macrocephaly in adulthood: implications for genesis of autism and schizophrenia." Cell Mol Neurobiol 22(1): 25-33.

Fatemi, S. H., T. J. Reutiman, et al. (2008). "Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: implications for genesis of neurodevelopmental disorders." Schizophr Res 99(1-3): 56-70.

Kessler, R. C., W. T. Chiu, et al. (2005). "Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication." Arch Gen Psychiatry 62(6): 617-627.

Nelson, M. D., A. J. Saykin, et al. (1998). "Hippocampal Volume Reduction in Schizophrenia as Assessed by Magnetic Resonance Imaging: A Meta-analytic Study." Arch Gen Psychiatry 55(5): 433-440.

Rantakallio, P., P. Jones, et al. (1997). "Association between central nervous system infections during childhood and adult onset schizophrenia and other psychoses: a 28-year follow- up." Int J Epidemiol 26(4): 837-843.

Rosoklija, G., G. Toomayan, et al. (2000). "Structural Abnormalities of Subicular Dendrites in Subjects With Schizophrenia and Mood Disorders: Preliminary Findings." Arch Gen Psychiatry 57(4): 349-356.

Shah, D. K., L. W. Doyle, et al. (2008). "Adverse neurodevelopment in preterm infants with postnatal sepsis or necrotizing enterocolitis is mediated by white matter abnormalities on magnetic resonance imaging at term." J Pediatr 153(2): 170-175, 175 e171.

Stoll, B. J., N. I. Hansen, et al. (2004). "Neurodevelopmental and growth impairment among extremely low-birth-weight infants with neonatal infection." JAMA 292(19): 2357-2365. van Berckel, B. N., M. G. Bossong, et al. (2008). "Microglia Activation in Recent-Onset Schizophrenia: A Quantitative (R)-[11C]PK11195 Positron Emission Tomography Study." Biological Psychiatry 64(9): 820-822.

CHAPTER 2

LITERATURE REVIEW

2.1 Neurodevelopment of the Domestic Pig

The process of neurodevelopment begins very early in embryogenesis and is relatively conserved across mammalian species. Much is known about pig fetal development as it is a standard for mammalian embryology (Patten 1931). After fertilization, embryo implantation occurs in the uterine wall of the sow by the 14th day of pregnancy (Book and Bustad 1974). The pig has a flat embryonic disk and the neural plate is formed on the dorsal surface of the embryo from ectodermal tissue at 14 days. Soon after formation, the neural plate invaginates and becomes folded forming the neural tube. Neural tube closure is similar to humans as it is initiated at multiple points along the anteroposterior axis (van Straaten, Peeters et al. 2000).

After neural tube closure, roughly 17 days post-conception, the anterior portion dilates and enlarges to form the prosencephalon, mesencephalon, and rhombencephalon. By 20-22 days, the brain has divided into five regions including the telencephalon, diencephalon, mesencephalon, metencephalon, and myelencephalon. At this time, cortical development begins with the formation of the preplate and subsequent increases in neurogenesis and neuronal migration. One critical protein for cortical formation and organization is Reelin. Reelin expression is highest in the Cajal-Retzius cells within the marginal zone. The highest levels occur in the pig at E60 and correspond to peak neurogenesis (Nielsen, Sondergaard et al.

2010). Expression patterns of Reelin in the pig are very similar to humans.

Much of the brain development characterization in late gestation in the pig has been conducted in larger regional areas with a few studies on cellular development. Early studies by

Dickerson and Dobbing, and confirmed by Pond et al., characterized brain weight and cholesterol content from the 52nd day of gestation through 3 years (Dickerson and Dobbing

1967; Pond, Boleman et al. 2000). Both humans and pigs have a perinatal growth spurt with the brain being 27% and 25% of adult weight at birth respectively (Dobbing and Sands 1979).

This is dissimilar to other species such as the rhesus monkey and rodents which have prenatal and postnatal growth spurts respectively. Passingham compared regional brain development between many mammals and found that ungulates, including the pig, and primates had very similar brain growth during gestation, but ungulates build their body at a higher growth rate

(Passingham 1985). Comparison of subareas of the brain between the pig and rhesus monkey revealed similar proportions in all but two areas. The pig has a larger paleocortex, including amygdala, and the rhesus monkey has a proportionally higher amount of neocortex. Magnetic resonance imaging studies have characterized the normal postnatal brain development of the pig from 2- to 24-weeks of age (Conrad, Dilger et al. 2012). During this time, the whole brain volume increases 120-130% with a sexual dimorphism in regional growth similar to humans.

Stereological studies conducted by Jelsing et al. found significant postnatal gliogenesis in both domestic and Göttingen pigs, but the Göttingen minipig also had significant neurogenesis

(Jelsing, Nielsen et al. 2006). The authors concluded that the domestic pig may be a better model for developmental hypotheses than the Göttingen minipig. An extensive list of neuroanatomical studies of the porcine brain, including functional cortical mapping, has been previously published (Lind, Moustgaard et al. 2007).

In addition to neuroanatomical characterization of brain development in the pig, immunohistochemistry has been used to describe neurotransmitter system architecture and function. The two most well characterized systems in the pig are the serotonergic and dopaminergic systems. Tryptophan hydroxylase serotonergic (5-HT) neuron distribution in the medulla is very similar to the human infant, with a few differences such as additional clusters on the ventral surface in the arcuate nucleus (Niblock, Luce et al. 2005). Age-related differences are also seen in both the pig and human, where in humans the distribution of 5-HT cells change while receptor binding changes in the pig. Distributions of dopaminergic cells in the pig have

6 been found to be very similar between human and non-human primates (Lind, Moustgaard et al.

2007). Like humans, the highest amount of dopamine was found in the substantia nigra and ventral tegmental area with a gradient distribution of dopamine receptor concentration (Larsen,

Bjarkam et al. 2004; Rosa-Neto, Doudet et al. 2004). Less is known about adrenergic and noradrenergic systems, but it has been shown that highest levels occur in the diencephalon and brainstem, similar to other mammals (Agarwal, Chandna et al. 1993). Interestingly, the pig has a 10-fold higher expression of monoamine oxidase B (MAO-B) than MAO-A, enzymes that catabolize catecholamines (Anderson, Lupo et al. 2005). This distribution of MAO isoforms is seen in humans, but not rodents.

There are several differences in pregnancy and embryogenesis between humans and pig with the majority attributed to different timescales of development. Embryonic implantation occurs earlier in humans than pigs, by the 6th day after ovulation (Book and Bustad 1974). The placental layers differ where humans have a discoid hemochorial placenta allowing direct contact between the fetal chorion and maternal blood (Book and Bustad 1974). Pigs have a diffuse epitheliochorial placenta. In addition, humans usually have one child while pigs are multiparous. The gestational length for humans is longer, 40-weeks compared to 114 days. The time course for intrauterine development is slightly different in humans and pigs, with the human having a more advanced development when corrected for percent of gestation (Book and

Bustad 1974). But within the developmental stages, the fetus is very similar. This difference allows the human fetus to have more time in the third fetal stage.

2.2 Neuroimaging in the Piglet

Magnetic resonance imaging (MRI) is a safe, non-ionizing radiation modality that is ideal for in vivo neuroimaging both clinically and for research. Different MRI sequences allow imaging and quantification of brain region anatomy, function, and neurochemical content and are very useful

7 for serial studies including tracking neurodevelopment. The use of MRI in the pig has lagged behind research using humans, non-human primates, and rodents, but these techniques are easily transferable to a pig model. Due to the size of the piglet, clinical strength magnets with head coils designed for humans are able to be used with piglets without modification. Rodents, however, require specially designed magnets at higher field strength due to the size of the animal. Thus, the piglet represents an ideal animal model for using clinical MRI scanners.

One of the most common uses for MRI is to image the structure of the brain. A combination of both T1- and T2-weighted imaging sequences are used clinically to examine brain structure and to identify abnormalities or pathologies. These sequences are also very useful for research. One of the first studies using MRI in pigs was conducted in 1993 where the authors acquired structural images from pigs placed in stereotaxic conditions and then compared the images to postmortem histological sections (Marcilloux, Felix et al. 1993). This work was further expanded upon and refined in 2000 in order to correlate histological data onto the MR images (Sørensen, Bjarkam et al. 2000). This was an important step as prior to 2000, only a histologically based atlas of the domestic pig was available, and this paved the way for future work in creating digital MR-based structural atlases of the brain (Felix, Leger et al. 1999).

One of the first MR-based atlases was constructed in the Göttingen minipig using 22 subjects and labeling 34 regions of interest (Watanabe, Andersen et al. 2001). This study opened the door to cross-modality analysis such as regional quantification of radiotracer uptake in positron emission topography (PET) studies (Rosa-Neto, Gjedde et al. 2004). Unfortunately, this atlas is not appropriate for analysis of the domestic pig brain due to the size and morphology differences between the Göttingen minipig brain and the domestic pig. A high resolution MR- based atlas for the adult domestic pig was constructed in 2009 by Saikali et al. and is a very good resource for studies using adult pigs (Saikali, Meurice et al. 2010). This atlas is not appropriate for studies in piglets though due once again to size differences between the neonatal and adult brain. Previous studies using MRI in piglets used manual segmentation

8 methods for brain region volume estimation (Conrad, Dilger et al. 2012). Using these techniques, a longitudinal study was conducted characterizing the normal brain growth and development of the domestic pig from 2- to 24-weeks of age showing many similarities and a few differences in brain growth compared to humans (Conrad, Dilger et al. 2012). Additionally, brain growth curves for the domestic pig have also been made using a cross-section study design (Winter, Dorner et al. 2011). Recently, an averaged brain and atlas for the domestic pig was constructed that will allow for more sophisticated analysis techniques, such as voxel based morphometry, to be conducted in the piglet (http://pigmri.illinois.edu). The use of these MRI techniques will allow for hypothesis driven research in the pig such as quantification of brain structure changes due to environmental, nutritional, and inflammatory insults.

In addition to T1- and T2-weighted structural imaging, diffusion tensor imaging (DTI) and diffusion-weighted imaging (DWI) provides information about brain microstructure and white matter tract structure. Diffusion tensor imaging uses diffusion properties of water molecules to produce contrast by quantifying the magnitude, degree of anisotropy, and orientation of diffusion anisotropy (Alexander, Lee et al. 2007). Diffusion anisotropy and the principle diffusion directions are also used for white matter tractography (Alexander, Lee et al. 2007). These measures can track microstructure changes the brain and are very useful for developmental and aging studies, as well as clinically looking at brain pathologies. The three parameters that are generated from diffusion tensor imaging are the three diffusion tensor eigenvalues, mean diffusivity (MD), and fractional anisotropy (FA), which collectively describes the direction of diffusion and magnitude of anisotropy. One of the first papers to use DWI in the piglet characterized apparent diffusion coefficients (ADC), now reported as MD, in both grey and white matter after perinatal-asphyxia (Thornton, Ordidge et al. 1997). The changes in MD and FA values during normal development have also been characterized in domestic pigs and found similar trends in changes in FA over time but dissimilar trends in MD compared to humans

(Winter, Dorner et al. 2011). The differences in MD levels may be due to different timing of

9 water content changes in the pig. DTI and DWI provide robust information about brain microstructure and white matter development/integrity and are well suited for use in a pig model.

MRI also allows for in vivo quantification of brain metabolites. 1H magnetic resonance spectroscopy (MRS) is the most commonly used spectroscopy technique and analyzes the signal from hydrogen protons that are bound to different chemicals. Detectible metabolites include N-acetyl-aspartate, creatine, choline, myo-inositol, as well as neurotransmitters such as glutamate and glutamine to name a few. Spatial resolution is reduced in MRS compared to

MRI, thus spectra that are obtained are for larger, more encompassing areas (Blüml 2013).

There are two methods used for MRS, single voxel MRS and chemical shift imaging. Single voxel MRS quantifies metabolites in one region of the brain and is able to quantify more metabolites, including lower concentration metabolites, due to a better signal to noise ratio

(S/N). Chemical shift imaging is able to acquire spectra from multiple voxels, but at the expense of S/N, meaning less detection of low concentration metabolites. MRS has been used in pigs in both brain trauma and hypoxia-ischemia research to detect changes to metabolites in the brain

(Smith, Cecil et al. 1998; Munkeby, De Lange et al. 2008).

A lesser used MRI modality in pigs is functional magnetic resonance imaging (fMRI). fMRI detects changes in brain activity by analyzing differences in the blood-oxygen-level dependent (BOLD) response , which detects increases to blood flow to a region. Studies have been conducted using conscious pigs in order to look at regional activation in response to pain and visual stimulation (Fang, Lorke et al. 2005; Fang, Li et al. 2006). Even though it has had limited use in pigs, fMRI is a valuable tool for translational imaging studies.

2.3 The Piglet as a Model for Human Neurodevelopment

Workman et al. have constructed a very good resource allowing for comparison of 271 neurodevelopmental events to be translated across species (Workman, Charvet et al. 2013).

Based on developmental allometry, highly predictive models for timing of neurodevelopmental events were constructed and validated in multiple species. Additionally, their website

(www.translatingtime.net) includes instructions for estimating the event timing in new species which we have done for the pig. For this calculation, we used a gestational length of 114 days and an adult brain weight of 180 grams. Using these parameters, the species constant for the domestic pig equaled 2.984 and the slope 2.809. Using the modeling equation and event scores, estimations of developmental equivalencies can be computed between species. Using the MRI-based whole brain volume estimates for the pig, we compared this to the translating time estimates for whole brain weight for the pig and human. Linear regression of the translating time data estimated that one pig week is roughly equal to 3 human weeks. We estimate through our MRI data, that one pig week is equal to one human month. The authors of translating time indicate that the model is based only on data up to 3 years old in human and thus only applicable to prenatal and early-postnatal development. Similarly, we found their model to fit our data well around the perinatal period with larger deviations as the animal grew towards adulthood.

Additionally, Niblock et al. estimated the comparable age of the piglet to humans based on serotonergic system development (Niblock, Luce et al. 2005). They suggest that a 4, 12, 30, and 60 day old pig is similar to a 1, 4, 6, and 12 month old human. From these data, rough estimations can be made for equivalent ages during the early neonatal period between pigs and humans, but more work is needed to validate these estimations.

2.4 Piglet Models in Neonatal Research

Infection Models

The pig has striking similarities in immune system anatomy, function, and gene expression compared to humans (Freeman, Ivens et al. 2012; Meurens, Summerfield et al.

2012). One study examined immune system similarities between pigs, humans, and rodents and found that the pig was more similar than rodents to humans in 80% of analyzed parameters

(Dawson 2011). As one example in the innate immune response, macrophages stimulated with lipopolysaccharide (LPS) produce nitric oxide in rodents, but this response is not seen in both human and pig macrophages (Pampusch, Bennaars et al. 1998). Because these similarities, the pig presents as important translational animal model that may be more predictive of therapeutic interventions than other models.

The pig has been used extensively for studying pathogens which cause systemic, respiratory, and gastrointestinal tract infections (Meurens, Summerfield et al. 2012). Many of the same pathogens that impact human health are highly prevalent in pigs and are important both from a biomedical research and a pork production standpoint. Staphylococcus aureus, including methicillin-resistant S. aureus, is becoming a worldwide clinical problem and many studies have been conducted in pigs examining wound infection, osteomyelitis, and pneumonia

(Svedman, Ljungh et al. 1989; Luna, Sibila et al. 2009; Jensen, Nielsen et al. 2010). Other human pathogens, such as Bordetella pertussis, are able to infect piglets and present as a very good model for direct translational research (Elahi, Buchanan et al. 2006). Gnotobiotic piglets are the only animal model that is susceptible to human rotovirus, and this model has been used to research pathogenicity and vaccination approaches (Saif, Ward et al. 1996; Meurens,

Summerfield et al. 2012).

Hypoxia-ischemia Models

Due the body weight and size of the piglet, it has a rich history of being used as a model for human neonatal hypoxia ischemia (HI) (Roohey, Raju et al. 1997). Much of the model

12 development was conduct in the 1980‘s including characterizing changes to cerebral blood flow and metabolism. This research, along with fetal sheep studies, described the basic physiology of the response to acute asphyxia including hyper/hypocapnia, hyper/hypoglycemia, and hypovolemic hypertension (Raju 1992).

There are multiple approaches that can be taken to produce hypoxia ischemia encephalopathy (HIE) that model different clinical situations. The first is a whole body approach by reducing the inspired air to hypoxic conditions (i.e. FiO2 = 0.08) (Haaland, Loberg et al. 1997;

Bjorkman, Foster et al. 2006). This produces HI in both the brain as well as the peripheral organs. A second approach uses the whole body HI and then includes an asphyxia component resulting in cardiac arrest and resuscitation (Agnew, Koehler et al. 2003). A review by Alonso-

Spilsbury et al describes in detail pathological changes seen in asphyxia in both a biomedical and pork production view (Alonso-Spilsbury, Mota-Rojas et al. 2005). A third approach is to focus the HI insult just on the brain, excluding peripheral impact. For this, a bilateral carotid ligation technique can be used (LeBlanc, Vig et al. 1991). The physiologic changes and brain damage seen in the piglets are very similar to the human neonate, thus making the piglet an ideal model for therapeutic development.

Magnetic resonance imaging allows for the study of brain injury severity in neonates

(Barkovich, Miller et al. 2006). Similar imaging protocols are able to be used in piglet models of

HIE. HIE causes energy and metabolism disruptions and these can be quantified using magnetic resonance spectroscopy (Thornton, Ordidge et al. 1997; Vial, Serriere et al. 2004). In addition, diffusion-weighted imaging allows for the quantification of changes in water diffusion following HIE insult (Thornton, Ordidge et al. 1997; Cheng, Liu et al. 2005; Munkeby, De Lange et al. 2008). MR angiography can also be used to track perfusion changes during the experimental protocol (Munkeby, Lyng et al. 2004). The direct translation of these imaging methods strengthens the piglet as an animal model for HIE.

An example of how research conducted in the piglet has directly translated to improvements in neonatal care is exemplified by hypothermia treatment for neonatal HIE. HIE, which previously had no treatments, causes high neonatal mortality rates and long-term neurological complications (Higgins, Raju et al. 2011). Research conducted by Thoresen and others have used a piglet model of HIE to determine timing of administration and improvements to survival and brain damage using hypothermia therapy (Thoresen, Haaland et al. 1996;

Haaland, Loberg et al. 1997). Using these time points as guides, clinical studies were conducted which showed improvements to survival and neurodevelopmental disability

(Shankaran, Pappas et al. 2012; Tagin, Woolcott et al. 2012; Thoresen, Tooley et al. 2013).

Based on these studies, the standard clinical practice is to now offer hypothermia therapy for patients with HIE whom meet the inclusion criteria (Zanelli, Naylor et al. 2008; Perlman, Wyllie et al. 2010).

2.5 Microglia and the Developing Brain

Microglial cells are the resident macrophages of the brain and were originally identified by Nissl in 1899 who considered them to be reactive neuroglia with the capacity for migration and phagocytosis (Barron 1995). Rio-Hortega advanced the knowledge of microglia by suggesting that they were of mesenchymal origin and were non-neuronal cells (Barron 1995). It is now known that approximately 10-12% of the cells in the brain are microglia and the regional distribution varies with the highest density in the hypothalamus, basal ganglia, substantia nigra, and hippocampus (Lawson, Perry et al. 1990; Mittelbronn, Dietz et al. 2001). Microglia share many qualities with peripheral macrophage including expression of the markers CD11b, CD14, and F4/80, but they are not of bone marrow origin. Studies in rodents have shown that microglia progenitor cells originate in the extra-embryonic yolk sac and migrate into the CNS via blood circulation between E8.5 and E9.5 (Ginhoux, Greter et al. 2010). The lifespan of

14 microglia is unknown but studies have shown that the microglia population can expand under pathological conditions (Saijo and Glass 2011). There is a sexual dimorphism in microglia number and morphology in the developing brain (Schwarz, Sholar et al. 2012). Males have more microglia early in development and females have an activated phenotype later into development. This dimorphism may be a contributing factor in the sex bias of many neurodevelopmental disorders.

Microglia have been shown to have a multifaceted role in the CNS including maintenance of normal tissue homeostasis, response to inflammation and pathogens, and normal brain development. Microglia have two activation states which can be characterized by cell morphology. During the “resting” or quiescent state, microglia exhibit a small soma with long and thin processes (Hanisch and Kettenmann 2007). Microglia act as surveying cells sampling the CNS microenvironment and display high levels of motility (Nimmerjahn, Kirchhoff et al. 2005). In early development and in response to inflammatory stimuli, microglia enter an active phase, retract their processes, resulting in an amoeboid cell morphology, and exhibit phagocytic actions (Ransohoff and Perry 2009). Pro- and anti-inflammatory signals modulate the microglia response. Microglia express a variety of cytokine and chemokine receptors which sense the inflammatory environment and allow the cells to mount an appropriate response

(Saijo and Glass 2011). Similar to peripheral macrophages, microglia are capable of displaying two different phenotypes, M1 (classical) and M2 (alternative) activation. The M1 phenotype is displayed in response to pro-inflammatory conditions including activation of Toll-like receptors

(TLRs) and interferon-γ (Michelucci, Heurtaux et al. 2009). In response, microglia up-regulate major histocompatibility markers (MHC class I and II) and produce proinflammatory cytokines and free radicals (Perry and Gordon 1988; Gordon 2003). The alternative (M2) phenotype is stimulated with IL-4 and IL-13 and is marked by arginase 1 expression (Gordon 2003). The M2 phenotype is involved with tissue remodeling and wound repair (Song, Ouyang et al. 2000). IL-

10, an anti-inflammatory cytokine, strongly enhances the phagocytic activity of microglia which may contribute to wound repair (Michelucci, Heurtaux et al. 2009). Although microglia are essential for host defense and neuroprotection, over activation and dysregulation during development may be a “vulnerability” factor for later-life pathology (Bilbo and Schwarz 2009).

In the adult brain, the majority of microglia are in the resting state. During development however, microglia are reactive and amoeboid, and have a crucial role in normal brain development (Cuadros and Navascues 1998). Microglia participate in many developmental processes including synaptic remodeling, clearance of dying neurons, and contribute to neuronal proliferation and differentiation (Harry 2013). Microglia participate in synaptic remodeling through release of neurotrophic factors as well as physically modifying the synapses. Microglia produce and release multiple neurotrophic factors including neuronal growth factor (NGF), brain derived neurotrophic factor (BDNF), neurotrophin-3 (NT3), and glial derived neurotrophic factor (GNDF), which induce and modify synaptic plasticity (Kim and de

Vellis 2005; Harry 2013). Microglial processes are found in synaptic clefts where they modulate synaptic strengthening and degradation and are important for shaping the visual cortex during early life (Rochefort, Quenech'du et al. 2002; Tremblay, Lowery et al. 2010). Additionally, microglia participate in synaptic pruning by engulfing and removing dendritic spines (Paolicelli,

Bolasco et al. 2011). Microglia appear in the brain parenchyma at a time when programed cell death is beginning and apoptotic cells start to appear (Wakselman, Bechade et al. 2008; Rigato,

Buckinx et al. 2011). CD11b and DAP12 expressed by microglia control the production of superoxide molecules which promote neuronal cells death (Wakselman, Bechade et al. 2008).

Loss or blockade of either CD11b or DAP12 leads to decreases in developmental apoptosis.

Microglia also contribute to neuronal proliferation through guidance of neuronal progenitor cells and influence differentiation (Aarum, Sandberg et al. 2003). They have been shown to promote

16 differentiation of cholinergic neurons in the basal forebrain and drive a switch from neurogenesis to astrocyte formation (Jonakait, Wen et al. 2000; Antony, Paquin et al. 2011).

2.6 Immune to Brain Communication

Historically, the brain has been considered an immune-privileged organ, but this is now known to be untrue (Dantzer, O'Connor et al. 2008). Within the choroid plexus and meninges, there are multiple immune cells including macrophages and dendritic cells (Dantzer, O'Connor et al. 2008). The brain parenchyma has its own distinct subset of macrophages, microglia cells, which are able to detect and respond to inflammatory signals. In addition to microglia cells, neurons , astrocytes, and oligodendrocytes have cytokine receptors (Sawada, Itoh et al. 1993).

These receptors allow for the sensing of inflammation and contribute to the bi-directional communication between the immune system and central nervous system.

This bi-directional communication allows for a “sixth sense” which is important for modifying physiology and behavior in response to injury or infection. In the periphery, the innate immune system has a variety of mechanism for detecting potential pathogens called pattern- recognition receptors (PRRs) which respond to specific moieties or pathogen-associated molecular patterns (PAMPs) (Dantzer 2009). Upon activation of the PRRs, multiple molecular pathways are initiated to combat the pathogen. The majority of these pathways ultimately lead to the up-regulation nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB)

(Kumar, Kawai et al. 2011). NFκB activation leads to increased synthesis and release of the pro-inflammatory cytokines interleukin (IL)-1β, IL-6, and tumor necrosis factor-α (TNF-α)

(Pugazhenthi, Zhang et al. 2013).

Cytokines that are produced in the periphery are able to act on the brain through both neural and humoral pathways to modify behaviors classically termed “sickness symptoms”

(Dantzer, Konsman et al. 2000; Dantzer and Kelley 2007). Sickness symptom behaviors include anorexia, listlessness, fatigue and malaise, sleep disturbances, and social withdrawal

(Kelley, Bluthe et al. 2003). These behavior modifications are thought to provide protection and reduce energy expenditures allowing the body to focus on fighting the infection. In addition to behavior modifications, physiological changes occur to aid in the clearance of the infection including fever and hypothalamic-pituitary-adrenal axis (HPA) activation (Johnson 2002). Fever induction is caused by pro-inflammatory cytokines activating cells within the preoptic area of the hypothalamus and aids in the clearance of pathogens, especially bacteria (Conti, Tabarean et al. 2004).

In order to induce a behavioral response, peripheral cytokines must pass their signal to the brain. This is accomplished via neural and humoral pathways. In the periphery, the afferent vagus nerve has cytokine receptors and can transmit information about the inflammatory state to the nucleus tractus solitarius and secondary projection areas of parabrachial nucleus, hypothalamic paraventriculum, and supraoptic nuclei (Wan, Janz et al. 1993). Here, the neural signals activate cytokine expression centrally causing sickness behavior (Goehler, Gaykema et al. 1998). Subdiaphragmatic vagotomy attenuates some behavioral responses but leaves the fever response functional indicating multiple messaging pathways for immune-to-brain communication (Kapcala, He et al. 1996; Romanovsky, Simons et al. 1997). Cytokines can also access the brain by humoral pathways through active transport mechanisms and diffusion at circumventricular organs where there is a lack of a blood brain barrier (Banks, Farr et al. 2002).

Endothelial cells that compose the blood brain barrier are capable of detecting pro-inflammatory cytokines and propagate the signal by release of cytokines and prostaglandins into the brain

(Van Dam, De Vries et al. 1996). Once these pro-inflammatory cytokines reach CNS, they can activate microglia cells which will amplify and transmit the signal.

2.7 Pathogenesis of Porcine Reproductive and Respiratory Syndrome Virus

Porcine reproductive and respiratory syndrome virus (PRRS) was first described as

“mystery swine disease” and “blue ear disease” when it emerged in North America and Europe in the 1980’s (Done, Paton et al. 1996). The virus was formally named and characterized in the early 1990’s as belonging to the Arteriviridae family of enveloped viruses (Dokland 2010).

PRRS infection is characterized by marked interstitial pneumonia and reproductive losses and increases the risk for secondary infections (Rossow, Shivers et al. 1999). The economic impact of PRRS on the pork production industry was estimated to be $664 million annually as of 2013

(Holtkamp, Kliebenstein et al. 2013).

Due to the impact of this disease, much research has been conducted into the structure and pathogenicity of the virus. The virus contains a positive-sense (+) RNA genome with a length of 15.1 – 15.5 kb (Dokland 2010). Seven open reading frames (ORF) allow for the coding of 14 non-structural proteins, including proteases, a RNA-dependent RNA polymerase, and a helicase, and multiple structural proteins (Amonsin, Kedkovid et al. 2009). PRRS has a very narrow cell tropism, first infecting alveolar macrophages in the lungs and then spreading to macrophages in other areas (Van Breedam, Delputte et al. 2010). The PRRS virion first attaches to the porcine macrophage by binding to heparin sulfate GAGs on the surface of the macrophage. Next, the M/GP5 glycoprotein on the virion envelope binds to a sialoadhesin receptor and is internalized by clathrin-mediated endocytosis (Van Breedam, Van Gorp et al.

2010). Once internalized, porcine CD163 interacts with GP2 and GP4 on the virion envelope allowing for the release of the viral genome (Calvert, Slade et al. 2007). When the viral genome is within the cell, it replicates, assembles new virion particles, and then exits through a budding process (Zhang, Xue et al. 2010).

The innate immune system has multiple ways to detect and fight PRRS viral infection.

The Toll-like receptors (TLRs) are proteins of the innate immune system that detect distinct pathogen-associated molecular patterns (PAMPs). In response to TLR activation, specific signaling pathways are activated to induce expression of pro- and anti-inflammatory cytokines to combat the infection. TLR3, 7, and 8 are located on the endosome membrane and respond to viral infection. Intracellular double stranded RNA (dsRNA) activates the TLR3 receptor. Upon binding of dsRNA, TLR3 activates interferon regulatory factor 3 (IRF-3) which induces type I interferon and NF-κB. Single stranded RNA (ssRNA) bind to TLR7 and 8 inducing MyD88 and

IRF-7. Similarly to IRF-3, IRF-7 activates type I interferon and NF-κB. The PRRS virus is able to activate TLR3, 7, and 8 as both ssRNA and dsRNA are produced during the viral genome replication (Zhang, Liu et al. 2013).

2.8 PRRSV and Neuroinflammation

The majority of PRRS research has looked at the peripheral aspects of the infection with the goal of developing a vaccine or improving pork production. Upregulation of pro- inflammatory cytokines in the periphery, including IL-6 and IL-1β, is a physiological response by the innate immune system to fight the infection. Through immune-to-brain communication, the brain can sense the inflammatory state of the periphery and reacts with changes to behavior and gene expression in the brain (Dantzer, O'Connor et al. 2008). In addition to neuroinflammation induction through the periphery, some strains of PRRS may have neurovirulence. PRRS has been shown to co-localize with MAC-387 positive cells (macrophage marker) in brain tissue (Rossow, Shivers et al. 1999). It is unknown if these are microglia cells or invading monocytes/macrophages from the periphery.

Evidence of peripheral and central immune activation in neonatal and young piglets has been well characterized (Miguel, Chen et al. 2010; Elmore, Burton et al. 2014). Experimentally

20 inoculating piglets with a highly-virulent strain of PRRS elicits a strong peripheral cytokine response with increased expression of TNF-α, IL-1β, IL-10, IL-6 and IFN-γ in the blood (Miguel,

Chen et al. 2010; Zhang, Liu et al. 2013). Pro-inflammatory cytokines and TLR receptors are also up-regulated in multiple regions of the brain. The source of these pro-inflammatory cytokines may be from activated microglia or reactive astrocytes (Lieberman, Pitha et al. 1989;

Dantzer, O'Connor et al. 2008). Increases in IL-10 and reductions in CD200, nerve growth factor (NGF), and myelin basic protein (MBP) were also evident in the brain of PRRS-infected piglets (Elmore, Burton et al. 2014).

We have previously shown that PRRS infection in piglets leads to drastic microglia activation in the hippocampus with 82% and 43% of isolated microglia being MHC-II positive 13 and 20 days post inoculation (Elmore, Burton et al. 2014). The presence of MHC-II shows that the microglia are activated, but it does not differentiate between the classical M1 (inflammatory) versus alternative M2 (anti-inflammatory) phenotype (Kigerl, Gensel et al. 2009; Olah, Biber et al. 2011). Further characterization of the phenotype of these microglia will help the determine timing of the inflammatory state and whether there is a M1 to M2 switch during disease progression.

The increase in pro-inflammatory cytokines has the potential to impact normal brain development processes and produce deficits in learning and memory. The hippocampus is very important for learning and memory and is susceptible to neuroinflammation (Yirmiya and

Goshen 2011). Using a hippocampal dependent spatial t-maze task, piglets with active PRRS infection show deficits in learning and memory (Elmore, Burton et al. 2014). PRRS infection increased the number of days that the piglets needed to acquire the task. The PRRS piglets also exhibited increased latency to choice and distance moved in the maze showing that they completed the task, but not to the performance level of the controls. There are multiple mechanisms that can explain the learning and memory deficits including changes to synaptic

21 plasticity, neurogenesis, and neuron morphology. The role of inflammation in modulating neurogenesis and neuron morphology are discussed below.

2.9 Inflammation and Neurogenesis

The timelines of development of the human and rodent brain has been well characterized. Early in neurodevelopment, there is a high level of neuronal proliferation and migration that lasts until the early postnatal period (Rice and Barone 2000). The progenitor cells that form these new neurons undergo a switch during mid-gestation and start producing glial cells, including astrocytes and oligodendrocytes. The formation of these cells persists into the postnatal period and even into early adulthood (Rice and Barone 2000). By birth, the majority of neurogenesis has been completed, but there are two areas of the brain which neurogenesis continues throughout life. The subependyma of the lateral ventricles produce new neurons that migrate in the rostral stream into the olfactory bulb and the subgranular zone of the dentate gyrus (DG) produces dentate granule cell neurons (Ekdahl 2012). Much research has been conducted into how inflammation can impact neurogenesis, cell fate, and behavioral consequences in adults. A few studies have looked at prenatal inflammation’s effects on neurogenesis, but there is a void of studies looking at the early postnatal period.

Neural progenitor cells in the DG have a limited-self renewing capacity and produce throughout life (Kempermann, Gast et al. 2003). Of the newly born cells, the majority will die but a few will migrate into the DG granule cell layer and mature into functional neurons (Biebl,

Cooper et al. 2000; Kempermann, Gast et al. 2003). The exact role of these new neurons is unknown, but they may be crucial for hippocampal-dependent forms of spatial memory

(Nakashiba, Cushman et al. 2012). Inflammation has been shown to alter adult hippocampal neurogenesis in a variety of ways including proliferation, survival, differentiation, and incorporation into new networks (Belarbi and Rosi 2013; Kohman and Rhodes 2013). The first

22 studies to show a link between inflammation and changes in neurogenesis were conducted in the early 2000s by Monje et al. and Ekdahl et al. Both studies found that LPS administration, which leads to classical microglia activation, resulted in decreased new neuron survival but no changes in proliferation (Ekdahl, Claasen et al. 2003; Monje, Toda et al. 2003). Minocycline, which inhibits microglia activation, was able to block this decrease in neurogenesis (Ekdahl,

Claasen et al. 2003). Classical microglia activation leads to an upregulation of pro-inflammatory cytokines including TNF-α, IL-1β, and IL-6 (Michelucci, Heurtaux et al. 2009). In vitro and in vivo studies have shown that these pro-inflammatory cytokines are able to neural precursor generation, differentiation, and survival (Vallieres, Campbell et al. 2002; Cacci, Ajmone-Cat et al. 2008; Wu, Hein et al. 2012; Wu, Montgomery et al. 2013). Some studies have also shown that pro-inflammatory cytokines can reduce the proliferation rate (Fujioka and Akema 2010).

Alternative activation of microglia produce the opposite effect. IL-10, IL-4, TGF-β, and IGF-1 promote production of neurons (Aberg, Aberg et al. 2000; Battista, Ferrari et al. 2006; Kiyota,

Okuyama et al. 2010; Kiyota, Ingraham et al. 2012). In addition to affecting proliferation and survival, cytokines also drive changes in cell fate. Pro-inflammatory cytokines causes a switch from neurogenesis to gliogenesis and increase astrocyte formation (Green, Treacy et al. 2012;

Stolp 2013). Inflammation can also impact the electrophysiological properties of newly incorporated neurons (Jakubs, Bonde et al. 2008).

Although the majority of studies looking at the impact of neuroinflammation have been conducted in adults, a few have looked at the impact of prenatal and neonatal inflammation.

Jӓrlestedt et al. administered LPS to mice at P9 and characterized long term survival of neurons and astrocytes (Jarlestedt, Naylor et al. 2013). They found that LPS injection reduced the number of surviving neurons and astrocytes at P41, and this was only true for the dorsal hippocampus. No long term behavioral differences were found. Prenatal immune activation has been shown to both decrease and increase neurogenesis. Prenatal LPS injection leads to decreased cell survival in the offspring, but this is dependent on timing of LPS administration

(Cui, Ashdown et al. 2009). Alternatively, maternal E. coli infection causes increased number of proliferating cells in the offspring and increased expression of BDNF, TrkB, and p-Akt (Jiang,

Zhu et al. 2013). In addition, rats from E. coli infected dams displayed spatial learning and memory deficits.

From these studies, the complexity of the impact of inflammation on neurogenesis can be seen. The phenomena seen in adult studies may not be predictive of what happens during early development. In addition, the timing and duration of inflammatory insult may differentially impact neurogenesis during development.

2.10 Inflammation, Neuron Morphology, and Dendritic Spines

Due to the role of the hippocampus in learning and memory, the structure and function of the cells within this region have been well characterized (Olton, Becker et al. 1979; von Bohlen

Und Halbach 2009). Much of the early work characterized the normal dendritic complexity of the dentate granule cells and CA region pyramidal cells and how this complexity changes in pathological states or in different environments (Green and Juraska 1985). For example, environmental enrichment causes increased complexity within dentate granule cells for female rats, but male rats were unaffected by environmental enrichment (Juraska, Fitch et al. 1985).

Additionally, there is a sexual dimorphism in the isolated environment, where males have more dendritic material per neuron than females. The complexity of dentate granule cells also varies with region where neurons located in the deep 2/3 of the dentate granule cell layer are less complex than those in the superficial 1/3 (Green and Juraska 1985). Dendritic complexity is very plastic and can be influenced by a number of factors, including stress and inflammation, and changes can lead to deficits in learning and memory (McEwen 1999). Stress, via action of adrenal steroids, can lead to atrophy of dendrites in the hippocampus (Woolley, Gould et al.

1990). Peripheral and central immune activation via LPS can also alter dendritic branching and spine density (Milatovic, Zaja-Milatovic et al. 2003; Richwine, Parkin et al. 2008). Using a

24 mouse model of influenza infection, Jurgens et al. showed alterations in CA1 pyramidal neuron morphology and differential effects on morphology in dentate granule cells based on layer location (Jurgens, Amancherla et al. 2012). Much less is known about how early-life inflammatory insult can affect neuron morphology. One study conducted in rats administered

LPS prenatally at E15 (Baharnoori, Brake et al. 2009). The authors found long-term disruptions in the complexity of the dendritic arbor and variable changes in dendritic length. Changes to the normal morphology may be clinically relevant as alterations in neuron morphology and spine density have been found in diseases including schizophrenia and autism (Glantz and Lewis

2000; Kolluri, Sun et al. 2005; Pickett and London 2005)

In addition to dendritic tree complexity, dendritic spines are also sensitive to environmental changes including inflammation (Bitzer-Quintero and Gonzalez-Burgos 2012).

Dendritic spines are small protrusions on the neuronal dendrite that are active or potential sites for synapses. They play a role in synaptic plasticity and have different physiological properties based on size and shape (von Bohlen Und Halbach 2009). Pro-inflammatory cytokines are capable of altering the physiological properties of neurons and dendritic spine development

(Kondo, Kohsaka et al. 2011). For example, a single injection of LPS in young mice is sufficient to destabilize spines, increase spine turnover rate, and reduce spine density after eight weeks

(Kondo, Kohsaka et al. 2011). Additionally, influenza infection is able to reduce the spine density of inner dentate granules cells (Jurgens, Amancherla et al. 2012).

Changes to neuron morphology and spine density may have acute and chronic implications. Alterations have the potential to impact synaptic plasticity and could be an underlying mechanism for deficits in hippocampal-dependent learning and memory (Neves,

Cooke et al. 2008; Kasai, Fukuda et al. 2010). Disruptions during the neonatal period may be of particular importance as this period has the highest rate of synaptogenesis and could lead to lifelong consequences (Rice and Barone 2000).

2.11 Minocycline and Neuroinflammation

Neuroinflammation during the neonatal period can occur with many conditions including hypoxic-ischemic encephalopathy and TORCH infections (Khandaker, Zimbron et al. 2012; Liu and McCullough 2013). Microglia within the CNS have a central role in the inflammatory response, and therapeutics designed to inhibit microglia may be beneficial. Minocycline has been shown to reduce microglia activation and provide anti-inflammatory and anti-apoptotic actions. Minocycline is a second-generation tetracycline antibiotic that is able to cross the blood brain barrier. Clinically, it is routinely used in humans to treat urinary tract infections and acne.

The low doses used for antibiotic treatment is safe for long term use. Because of this, minocycline has been the target of much research into its effectiveness as a treatment for neuroinflammation.

Although a precise mechanism of action has not been found for minocycline in the CNS, multiple actions may collectively provide neuroprotection. Minocycline has been shown to inhibit the proliferation and activation of microglia cells including reductions in inflammatory molecules including matrix metalloproteases, nitric oxide, and pro-inflammatory cytokines (He,

Appel et al. 2001; Stirling, Koochesfahani et al. 2005). Minocycline prevents p38 MAPK signaling in microglia, which is crucial for microglia activation (Walton, DiRocco et al. 1998; Guo and Bhat 2007). Additionally, minocycline provides anti-oxidant and anti-apoptotic actions.

Inhibition of caspase-1, -3, and cytochrome c release all provide anti-apoptotic effects (Chen,

Ona et al. 2000; Kim and Suh 2009). Minocycline also increases expression of Bcl-2 which prevents cell death (Wang, Wei et al. 2004). Collectively, these actions may provide the neuroprotective effects seen.

Because of these actions, minocycline has been researched extensively as a therapeutic for neuroinflammatory disease in both humans and animal models. Animal models have shown

26 both protective and harmful effects in models of ischemia, multiple sclerosis, amyotrophic lateral sclerosis (ALS), and Alzheimer’s as examples (Kim and Suh 2009). High minocycline doses were found to be protective in a rat model of hypoxia ischemia, but exacerbated brain injury in a mouse model showing that minocycline may have species specific responses (Tsuji, Wilson et al. 2004; Fan, Pang et al. 2005). Issues with the proper dosage are problematic. The neuroprotective dose in rodents is very high (45-90 mg/kg/day) compared to the clinically used dose in humans (2-4 mg/kg) and this makes difficult in translating work done in rodent models to human clinical trials (Buller, Carty et al. 2009). Nonetheless, many clinical trials have been completed or are in progress using minocycline for neurodegenerative and psychiatric conditions using low doses. Completed studies have shown mixed results. Minocycline caused faster deterioration in ALS patients but improved negative symptoms in schizophrenia patients

(Gordon, Moore et al. 2007; Chaudhry, Hallak et al. 2012). Differences in the inflammatory mechanism of these diseases may contribute to the differential responses to minocycline.

The use of minocycline in pediatric populations has been limited in humans due to side effects. Tetracyclines act as ion chelators which bind calcium phosphate and can be incorporated into developing teeth and bone leading to stunting of growth (Buller, Carty et al.

2009). Because of this, minocycline is not typically used in neonatal populations. Even though it is not used clinically in pediatric populations, it is still useful in animal models of neonatal neuroinflammation. Benefits seen with minocycline administration may help elucidate the mechanisms behind the disease process and provide more specific targets for future therapeutic development.

2.12 Summary and Objectives

In summary, environmental insults, including infection, during sensitive developmental periods can alter normal brain development. Human epidemiological studies have linked childhood infections with increased risk of developmental and psychiatric disorders, but the exact mechanism behind this is still unknown (Atladottir, Thorsen et al. 2010). Work over the past three decades has shed light on the cause of these disorders including, autism and schizophrenia, and have linked neuroimmune activation and dysregulation as risk factors

(Meyer, Feldon et al. 2011). As shown by the background information, activation of microglia and neuroinflammation have the potential to disrupt normal brain development.

The majority of work looking at the impact of infection and inflammation in early life has been conducted in rodent models using the acute inflammatory stimuli lipopolysaccharide (LPS) or Poly I:C (Meyer and Feldon 2010). While many important discoveries about how early life inflammation affects brain and behavioral have been made in rodent models, additional work in larger animal species and chronic inflammatory stimuli is needed.

The piglet presents as an ideal animal model for testing hypothesis in the neonatal time period. The piglet has very similar brain growth and development to humans and has a rich history of being used as an animal model for neonatal research. There is very little research being conducted on how chronic immune activation, typical of many infectious agents, can affect brain development. Here we use a live virus, PRRS, as a model of respiratory syncytium virus (RSV) in neonates. RSV is a significant clinical burden with over 132,000 children under 5 years of age hospitalized from 1997-2006 (Centers for Disease Control and Prevention 2013).

Neonates are especially prone to severe RSV infection with the median duration of illness of 12 days and 10 percent remaining ill for over four weeks (Steiner 2004). RSV also causes increases in peripheral pro-inflammatory cytokines (Arnold, Konig et al. 1995). Infection with

PRRS provides us a model to test the short- and long-term effects of peripheral inflammation on brain development.

This dissertation was designed with two objectives. The first was to develop noninvasive imaging techniques to measure brain growth and development in the piglet. MRI methods were first developed to measure regional brain volumes and these techniques were then utilized to characterize the normal brain growth of the domestic pig. The creation of an averaged brain and MRI-based atlas for the piglet allowed for improvement on MRI analysis including voxel based morphometry. Methods for additional imaging techniques in the piglet, including DTI and MR-spec, were also produced.

The second objective was to investigate if peripheral infection with the PRRS virus could lead to changes in brain development. Using the MRI methods developed, changes to regional gray and white matter volumes, white matter track structure, and neurochemical concentration were assessed in piglets inoculated with PRRS. In addition, changes to neurogenesis, cell fate, and hippocampal neuron morphology were characterized. Finally, the therapeutic potential of minocycline for reduction of microglia activation and neuroinflammation was conducted in the

PRRS model.

The use of a highly translatable animal model with peripheral viral infection provides important and clinically relevant information on how peripheral immune activation can change neurodevelopment. Further research into the mechanisms of these changes may provide novel therapeutic targets for treatment of neonatal infections. Long-term studies using this model would also be beneficial to see if changes persist into adulthood.

2.13 References

1. Patten, B.M., The embryology of the pig. 2nd ed1931, Philadelphia: P. Blakiston's Son and Co.

2. Book, S.A. and L.K. Bustad, The Fetal and Neonatal Pig in Biomedical Research. J Anim Sci, 1974. 38(5): p. 997-1002.

3. van Straaten, H.W.M., et al., Neurulation in the pig embryo. Anatomy and Embryology, 2000. 202(2): p. 75-84.

4. Nielsen, K.B., et al., Reelin expression during embryonic development of the pig brain. BMC Neurosci, 2010. 11: p. 75.

5. Dickerson, J.W.T. and J. Dobbing, Prenatal and postnatal growth and development of the central nervous system of the pig. Proc Biol Sci, 1967. 166(1005): p. 384-395.

6. Pond, W.G., et al., Perinatal ontogeny of brain growth in the domestic pig. Proc Soc Exp Biol Med, 2000. 223(1): p. 102-8.

7. Dobbing, J. and J. Sands, Comparative aspects of the brain growth spurt. Early Hum Dev, 1979. 3(1): p. 79-83.

8. Passingham, R.E., Rates of Brain Development in Mammals Including Man. Brain, Behavior and Evolution, 1985. 26(3-4): p. 167-175.

9. Conrad, M.S., R.N. Dilger, and R.W. Johnson, Brain Growth of the Domestic Pig (Sus scrofa) from 2 to 24 Weeks of Age: A Longitudinal MRI Study. Dev Neurosci, 2012. 34(4): p. 291-8.

10. Jelsing, J., et al., The postnatal development of neocortical neurons and glial cells in the Gottingen minipig and the domestic pig brain. J Exp Biol, 2006. 209(Pt 8): p. 1454-62.

11. Lind, N.M., et al., The use of pigs in neuroscience: modeling brain disorders. Neurosci Biobehav Rev, 2007. 31(5): p. 728-51.

12. Niblock, M.M., et al., Comparative anatomical assessment of the piglet as a model for the developing human medullary serotonergic system. Brain Res Brain Res Rev, 2005. 50(1): p. 169-83.

13. Rosa-Neto, P., D.J. Doudet, and P. Cumming, Gradients of dopamine D1- and D2/3- binding sites in the basal ganglia of pig and monkey measured by PET. NeuroImage, 2004. 22(3): p. 1076-83.

14. Larsen, M., et al., The anatomy of the porcine subthalamic nucleus evaluated with immunohistochemistry and design-based stereology. Anat Embryol (Berl), 2004. 208(3): p. 239-47.

15. Agarwal, R.K., et al., Distribution of catecholamines in the central nervous system of the pig. Brain Res Bull, 1993. 32(3): p. 285-91.

16. Anderson, N.J., et al., Characterisation of imidazoline I2 binding sites in pig brain. Eur J Pharmacol, 2005. 519(1-2): p. 68-74.

17. Marcilloux, J.C., et al., Preliminary results of a magnetic resonance imaging (MRI) study of the pig brain placed in stereotaxic conditions. Neurosci Lett, 1993. 156(1-2): p. 113-6.

18. Sørensen, J.C., et al., Oriented sectioning of irregular tissue blocks in relation to computerized scanning modalities:: Results from the domestic pig brain. Journal of Neuroscience Methods, 2000. 104(1): p. 93-98.

19. Felix, B., et al., Stereotaxic atlas of the pig brain. Brain Res Bull, 1999. 49(1-2): p. 1-137.

20. Watanabe, H., et al., MR-based statistical atlas of the Gottingen minipig brain. Neuroimage, 2001. 14(5): p. 1089-96.

21. Rosa-Neto, P., et al., MDMA-evoked changes in [11C]raclopride and [11C]NMSP binding in living pig brain. Synapse, 2004. 53(4): p. 222-233.

22. Saikali, S., et al., A three-dimensional digital segmented and deformable brain atlas of the domestic pig. J Neurosci Meth, 2010. 192(1): p. 102-9.

23. Conrad, M.S., et al., Magnetic resonance imaging of the neonatal piglet brain. Pediatr Res, 2012. 71(2): p. 179-184.

24. Winter, J.D., et al., Noninvasive MRI measures of microstructural and cerebrovascular changes during normal swine brain development. Pediatr Res, 2011. 69(5 Pt 1): p. 418- 24.

25. Alexander, A.L., et al., Diffusion tensor imaging of the brain. Neurotherapeutics, 2007. 4(3): p. 316-29.

26. Thornton, J.S., et al., Anisotropic water diffusion in white and gray matter of the neonatal piglet brain before and after transient hypoxia-ischaemia. Magnetic Resonance Imaging, 1997. 15(4): p. 433-440.

27. Blüml, S., Magnetic Resonance Spectroscopy: Basics, in MR Spectroscopy of Pediatric Brain Disorders, S. Blüml and A. Panigrahy, Editors. 2013, Springer New York. p. 11-23.

28. Smith, D.H., et al., Magnetic resonance spectroscopy of diffuse brain trauma in the pig. J Neurotrauma, 1998. 15(9): p. 665-74.

29. Munkeby, B.H., et al., A piglet model for detection of hypoxic-ischemic brain injury with magnetic resonance imaging. Acta Radiol, 2008. 49(9): p. 1049-57.

30. Fang, M., et al., fMRI mapping of cortical centers following visual stimulation in postnatal pigs of different ages. Life sciences, 2006. 78(11): p. 1197-201.

31. Fang, M., et al., Postnatal changes in functional activities of the pig's brain: a combined functional magnetic resonance imaging and immunohistochemical study. Neurosignals, 2005. 14(5): p. 222-33.

32. Workman, A.D., et al., Modeling transformations of neurodevelopmental sequences across mammalian species. J Neurosci, 2013. 33(17): p. 7368-83.

33. Meurens, F., et al., The pig: a model for human infectious diseases. Trends in microbiology, 2012. 20(1): p. 50-57.

34. Freeman, T., et al., A gene expression atlas of the domestic pig. BMC Biology, 2012. 10(1): p. 1-22.

35. Dawson, H.D., A Comparative Assessment of the Pig, Mouse and Human Genomes, in The Minipig in Biomedical Research2011, CRC Press. p. 323-342.

36. Pampusch, M.S., et al., Inducible nitric oxide synthase expression in porcine immune cells. Vet Immunol Immunopathol, 1998. 61(2-4): p. 279-89.

37. Svedman, P., et al., Staphylococcal wound infection in the pig: Part I. Course. Ann Plast Surg, 1989. 23(3): p. 212-8.

38. Jensen, H.E., et al., A non-traumatic Staphylococcus aureus osteomyelitis model in pigs. In Vivo, 2010. 24(3): p. 257-64.

39. Luna, C.M., et al., Animal models of ventilator-associated pneumonia. Eur Respir J, 2009. 33(1): p. 182-8.

40. Elahi, S., et al., Maternal immunity provides protection against pertussis in newborn piglets. Infect Immun, 2006. 74(5): p. 2619-27.

41. Saif, L.J., et al., The gnotobiotic piglet as a model for studies of disease pathogenesis and immunity to human rotaviruses. Arch Virol Suppl, 1996. 12: p. 153-61.

42. Roohey, T., T.N. Raju, and A.N. Moustogiannis, Animal models for the study of perinatal hypoxic-ischemic encephalopathy: a critical analysis. Early Hum Dev, 1997. 47(2): p. 115-46.

43. Raju, T.N., Some animal models for the study of perinatal asphyxia. Biol Neonate, 1992. 62(4): p. 202-14.

44. Haaland, K., et al., Posthypoxic hypothermia in newborn piglets. Pediatr Res, 1997. 41(4 Pt 1): p. 505-12.

45. Bjorkman, S.T., et al., Hypoxic/Ischemic models in newborn piglet: comparison of constant FiO2 versus variable FiO2 delivery. Brain Res, 2006. 1100(1): p. 110-7.

46. Agnew, D.M., et al., Hypothermia for 24 hours after asphyxic cardiac arrest in piglets provides striatal neuroprotection that is sustained 10 days after rewarming. Pediatr Res, 2003. 54(2): p. 253-62.

47. Alonso-Spilsbury, M., et al., Perinatal asphyxia pathophysiology in pig and human: a review. Anim Reprod Sci, 2005. 90(1-2): p. 1-30.

48. LeBlanc, M.H., et al., MK-801 does not protect against hypoxic-ischemic brain injury in piglets. Stroke, 1991. 22(10): p. 1270-5.

49. Barkovich, A.J., et al., MR imaging, MR spectroscopy, and diffusion tensor imaging of sequential studies in neonates with encephalopathy. AJNR Am J Neuroradiol, 2006. 27(3): p. 533-47.

50. Vial, F., et al., A newborn piglet study of moderate hypoxic-ischemic brain injury by 1H- MRS and MRI. Magn Reson Imaging, 2004. 22(4): p. 457-65.

51. Cheng, Y., et al., Early diffusion weighted imaging and expression of heat shock protein 70 in newborn pigs with hypoxic ischaemic encephalopathy. Postgraduate Medical Journal, 2005. 81(959): p. 589-593.

52. Munkeby, B.H., et al., Morphological and hemodynamic magnetic resonance assessment of early neonatal brain injury in a piglet model. J Magn Reson Imaging, 2004. 20(1): p. 8-15.

53. Higgins, R.D., et al., Hypothermia and Other Treatment Options for Neonatal Encephalopathy: An Executive Summary of the Eunice Kennedy Shriver NICHD Workshop. The Journal of Pediatrics, 2011. 159(5): p. 851-858.e1.

54. Thoresen, M., et al., A piglet survival model of posthypoxic encephalopathy. Pediatr Res, 1996. 40(5): p. 738-48.

55. Thoresen, M., et al., Time is brain: starting therapeutic hypothermia within three hours after birth improves motor outcome in asphyxiated newborns. Neonatology, 2013. 104(3): p. 228-33.

56. Tagin, M.A., et al., Hypothermia for neonatal hypoxic ischemic encephalopathy: an updated systematic review and meta-analysis. Arch Pediatr Adolesc Med, 2012. 166(6): p. 558-66.

57. Shankaran, S., et al., Childhood outcomes after hypothermia for neonatal encephalopathy. N Engl J Med, 2012. 366(22): p. 2085-92.

58. Perlman, J.M., et al., Part 11: Neonatal resuscitation: 2010 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations. Circulation, 2010. 122(16 Suppl 2): p. S516-38.

59. Zanelli, S.A., et al., Implementation of a 'Hypothermia for HIE' program: 2-year experience in a single NICU. J Perinatol, 2008. 28(3): p. 171-5.

60. Barron, K.D., The microglial cell. A historical review. Journal of the Neurological Sciences, 1995. 134, Supplement(0): p. 57-68.

61. Lawson, L.J., et al., Heterogeneity in the distribution and morphology of microglia in the normal adult mouse brain. Neuroscience, 1990. 39(1): p. 151-70.

62. Mittelbronn, M., et al., Local distribution of microglia in the normal adult human central nervous system differs by up to one order of magnitude. Acta Neuropathol, 2001. 101(3): p. 249-55.

63. Ginhoux, F., et al., Fate Mapping Analysis Reveals That Adult Microglia Derive from Primitive Macrophages. Science, 2010. 330(6005): p. 841-845.

64. Saijo, K. and C.K. Glass, Microglial cell origin and phenotypes in health and disease. Nat Rev Immunol, 2011. 11(11): p. 775-787.

65. Schwarz, J.M., P.W. Sholar, and S.D. Bilbo, Sex differences in microglial colonization of the developing rat brain. Journal of Neurochemistry, 2012. 120(6): p. 948-963.

66. Hanisch, U.-K. and H. Kettenmann, Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat Neurosci, 2007. 10(11): p. 1387-1394.

67. Nimmerjahn, A., F. Kirchhoff, and F. Helmchen, Resting Microglial Cells Are Highly Dynamic Surveillants of Brain Parenchyma in Vivo. Science, 2005. 308(5726): p. 1314- 1318.

68. Ransohoff, R.M. and V.H. Perry, Microglial physiology: unique stimuli, specialized responses. Annu Rev Immunol, 2009. 27: p. 119-45.

69. Michelucci, A., et al., Characterization of the microglial phenotype under specific pro- inflammatory and anti-inflammatory conditions: Effects of oligomeric and fibrillar amyloid- beta. J Neuroimmunol, 2009. 210(1-2): p. 3-12.

70. Perry, V.H. and S. Gordon, Macrophages and microglia in the nervous system. Trends Neurosci, 1988. 11(6): p. 273-7.

71. Gordon, S., Alternative activation of macrophages. Nat Rev Immunol, 2003. 3(1): p. 23- 35.

72. Song, E., et al., Influence of alternatively and classically activated macrophages on fibrogenic activities of human fibroblasts. Cell Immunol, 2000. 204(1): p. 19-28.

73. Bilbo, S.D. and J.M. Schwarz, Early-life programming of later-life brain and behavior: a critical role for the immune system. Front Behav Neurosci, 2009. 3: p. 14.

74. Cuadros, M.A. and J. Navascues, The origin and differentiation of microglial cells during development. Prog Neurobiol, 1998. 56(2): p. 173-89.

75. Harry, G.J., Microglia during development and aging. Pharmacol Ther, 2013. 139(3): p. 313-26.

76. Kim, S.U. and J. de Vellis, Microglia in health and disease. J Neurosci Res, 2005. 81(3): p. 302-13.

77. Rochefort, N., et al., Microglia and astrocytes may participate in the shaping of visual callosal projections during postnatal development. J Physiol Paris, 2002. 96(3-4): p. 183- 92.

78. Tremblay, M.-È., R.L. Lowery, and A.K. Majewska, Microglial Interactions with Synapses Are Modulated by Visual Experience. PLoS Biol, 2010. 8(11): p. e1000527.

79. Paolicelli, R.C., et al., Synaptic Pruning by Microglia Is Necessary for Normal Brain Development. Science, 2011.

80. Wakselman, S., et al., Developmental neuronal death in hippocampus requires the microglial CD11b integrin and DAP12 immunoreceptor. J Neurosci, 2008. 28(32): p. 8138-43.

81. Rigato, C., et al., Pattern of invasion of the embryonic mouse spinal cord by microglial cells at the time of the onset of functional neuronal networks. Glia, 2011. 59(4): p. 675- 95.

82. Aarum, J., et al., Migration and differentiation of neural precursor cells can be directed by microglia. Proc Natl Acad Sci U S A, 2003. 100(26): p. 15983-8.

83. Jonakait, G.M., et al., Macrophage cell-conditioned medium promotes cholinergic differentiation of undifferentiated progenitors and synergizes with nerve growth factor action in the developing basal forebrain. Exp Neurol, 2000. 161(1): p. 285-96.

84. Antony, J.M., et al., Endogenous microglia regulate development of embryonic cortical precursor cells. J Neurosci Res, 2011. 89(3): p. 286-98.

85. Dantzer, R., et al., From inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci, 2008. 9(1): p. 46-56.

86. Sawada, M., et al., Expression of cytokine receptors in cultured neuronal and glial cells. Neurosci Lett, 1993. 160(2): p. 131-4.

87. Dantzer, R., Cytokine, sickness behavior, and depression. Immunol Allergy Clin North Am, 2009. 29(2): p. 247-64.

88. Kumar, H., T. Kawai, and S. Akira, Pathogen recognition by the innate immune system. Int Rev Immunol, 2011. 30(1): p. 16-34.

89. Pugazhenthi, S., et al., Induction of an Inflammatory Loop by Interleukin-1β and Tumor Necrosis Factor-α Involves NF-kB and STAT-1 in Differentiated Human Neuroprogenitor Cells. PLoS One, 2013. 8(7): p. e69585.

90. Dantzer, R., et al., Neural and humoral pathways of communication from the immune system to the brain: parallel or convergent? Auton Neurosci, 2000. 85(1-3): p. 60-5.

91. Dantzer, R. and K.W. Kelley, Twenty years of research on cytokine-induced sickness behavior. Brain, Behavior, and Immunity, 2007. 21(2): p. 153-160.

92. Kelley, K.W., et al., Cytokine-induced sickness behavior. Brain Behav Immun, 2003. 17 Suppl 1: p. S112-8.

93. Johnson, R.W., The concept of sickness behavior: a brief chronological account of four key discoveries. Veterinary Immunology and Immunopathology, 2002. 87(3-4): p. 443- 450.

94. Conti, B., et al., Cytokines and fever. Front Biosci, 2004. 9: p. 1433-49.

95. Wan, W., et al., Differential induction of c-Fos immunoreactivity in hypothalamus and brain stem nuclei following central and peripheral administration of endotoxin. Brain Res Bull, 1993. 32(6): p. 581-7.

96. Goehler, L.E., et al., Interleukin-1 induces c-Fos immunoreactivity in primary afferent neurons of the vagus nerve. Brain Res, 1998. 804(2): p. 306-10.

97. Kapcala, L.P., et al., Subdiaphragmatic vagotomy inhibits intra-abdominal interleukin-1 beta stimulation of adrenocorticotropin secretion. Brain Res, 1996. 728(2): p. 247-54.

98. Romanovsky, A.A., et al., The vagus nerve in the thermoregulatory response to systemic inflammation. Am J Physiol, 1997. 273(1 Pt 2): p. R407-13.

99. Banks, W.A., S.A. Farr, and J.E. Morley, Entry of blood-borne cytokines into the central nervous system: effects on cognitive processes. Neuroimmunomodulation, 2002. 10(6): p. 319-27.

100. Van Dam, A.M., et al., Interleukin-1 receptors on rat brain endothelial cells: a role in neuroimmune interaction? FASEB J, 1996. 10(2): p. 351-6.

101. Done, S.H., D.J. Paton, and M.E. White, Porcine reproductive and respiratory syndrome (PRRS): a review, with emphasis on pathological, virological and diagnostic aspects. Br Vet J, 1996. 152(2): p. 153-74.

102. Dokland, T., The structural biology of PRRSV. Virus Research, 2010. 154(1-2): p. 86-97.

103. Rossow, K.D., et al., Porcine reproductive and respiratory syndrome virus infection in neonatal pigs characterised by marked neurovirulence. Vet Rec., 1999. 144(16): p. 444- 448.

104. Holtkamp, D.J., J.B. Kliebenstein, and E.J. Neumann, Assessment of the economic impact of porcine reproductive and respiratory syndrome virus on United States pork producers. Journal of Swine Health and Production, 2013. 21(2): p. 72-84.

105. Amonsin, A., et al., Comparative analysis of complete nucleotide sequence of porcine reproductive and respiratory syndrome virus (PRRSV) isolates in Thailand (US and EU genotypes). Virol J, 2009. 6: p. 143.

106. Van Breedam, W., et al., Porcine reproductive and respiratory syndrome virus entry into the porcine macrophage. Journal of General Virology, 2010. 91(7): p. 1659-1667.

107. Van Breedam, W., et al., The M/GP(5) glycoprotein complex of porcine reproductive and respiratory syndrome virus binds the sialoadhesin receptor in a sialic acid-dependent manner. PLoS Pathog, 2010. 6(1): p. e1000730.

108. Calvert, J.G., et al., CD163 expression confers susceptibility to porcine reproductive and respiratory syndrome viruses. Journal of virology, 2007. 81(14): p. 7371-9.

109. Zhang, C., et al., Profiling of cellular proteins in porcine reproductive and respiratory syndrome virus virions by proteomics analysis. Virol J, 2010. 7: p. 242.

110. Zhang, L., et al., Comparative expression of Toll-like receptors and inflammatory cytokines in pigs infected with different virulent porcine reproductive and respiratory syndrome virus isolates. Virol J, 2013. 10(1): p. 135.

111. Miguel, J.C., et al., Expression of inflammatory cytokines and Toll-like receptors in the brain and respiratory tract of pigs infected with porcine reproductive and respiratory syndrome virus. Veterinary Immunology and Immunopathology, 2010. 135(3-4): p. 314- 319.

112. Elmore, M.R., et al., Respiratory viral infection in neonatal piglets causes marked microglia activation in the hippocampus and deficits in spatial learning. J Neurosci, 2014. 34(6): p. 2120-9.

113. Lieberman, A.P., et al., Production of tumor necrosis factor and other cytokines by astrocytes stimulated with lipopolysaccharide or a neurotropic virus. Proc Natl Acad Sci U S A, 1989. 86(16): p. 6348-52.

114. Kigerl, K.A., et al., Identification of two distinct macrophage subsets with divergent effects causing either neurotoxicity or regeneration in the injured mouse spinal cord. J Neurosci, 2009. 29(43): p. 13435-44.

115. Olah, M., et al., Microglia phenotype diversity. CNS Neurol Disord Drug Targets, 2011. 10(1): p. 108-18.

116. Yirmiya, R. and I. Goshen, Immune modulation of learning, memory, neural plasticity and neurogenesis. Brain, Behavior, and Immunity, 2011. 25(2): p. 181-213.

117. Rice, D. and S. Barone, Jr., Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models. Environ Health Perspect, 2000. 108 Suppl 3: p. 511-33.

118. Ekdahl, C.T., Microglial activation - tuning and pruning adult neurogenesis. Frontiers in Pharmacology, 2012. 3.

119. Kempermann, G., et al., Early determination and long-term persistence of adult- generated new neurons in the hippocampus of mice. Development, 2003. 130(2): p. 391-9.

120. Biebl, M., et al., Analysis of neurogenesis and programmed cell death reveals a self- renewing capacity in the adult rat brain. Neurosci Lett, 2000. 291(1): p. 17-20.

121. Nakashiba, T., et al., Young dentate granule cells mediate pattern separation, whereas old granule cells facilitate pattern completion. Cell, 2012. 149(1): p. 188-201.

122. Belarbi, K. and S. Rosi, Modulation of adult-born neurons in the inflamed hippocampus. Front Cell Neurosci, 2013. 7: p. 145.

123. Kohman, R.A. and J.S. Rhodes, Neurogenesis, inflammation and behavior. Brain Behav Immun, 2013. 27C: p. 22-32.

124. Monje, M.L., H. Toda, and T.D. Palmer, Inflammatory Blockade Restores Adult Hippocampal Neurogenesis. Science, 2003. 302(5651): p. 1760-1765.

125. Ekdahl, C.T., et al., Inflammation is detrimental for neurogenesis in adult brain. Proceedings of the National Academy of Sciences, 2003. 100(23): p. 13632-13637.

126. Cacci, E., et al., In vitro neuronal and glial differentiation from embryonic or adult neural precursor cells are differently affected by chronic or acute activation of microglia. Glia, 2008. 56(4): p. 412-425.

127. Wu, M.D., et al., Adult murine hippocampal neurogenesis is inhibited by sustained IL- 1beta and not rescued by voluntary running. Brain Behav Immun, 2012. 26(2): p. 292- 300.

128. Wu, M.D., et al., Sustained IL-1β expression impairs adult hippocampal neurogenesis independent of IL-1 signaling in nestin+ neural precursor cells. Brain, Behavior, and Immunity, 2013. 32(0): p. 9-18.

129. Vallieres, L., et al., Reduced hippocampal neurogenesis in adult transgenic mice with chronic astrocytic production of interleukin-6. J Neurosci, 2002. 22(2): p. 486-92.

130. Fujioka, H. and T. Akema, Lipopolysaccharide acutely inhibits proliferation of neural precursor cells in the dentate gyrus in adult rats. Brain Research, 2010. 1352(0): p. 35- 42.

131. Kiyota, T., et al., AAV serotype 2/1-mediated gene delivery of anti-inflammatory interleukin-10 enhances neurogenesis and cognitive function in APP+PS1 mice. Gene Ther, 2012. 19(7): p. 724-33.

132. Kiyota, T., et al., CNS expression of anti-inflammatory cytokine interleukin-4 attenuates Alzheimer's disease-like pathogenesis in APP+PS1 bigenic mice. FASEB J, 2010. 24(8): p. 3093-102.

133. Battista, D., et al., Neurogenic niche modulation by activated microglia: transforming growth factor β increases neurogenesis in the adult dentate gyrus. European Journal of Neuroscience, 2006. 23(1): p. 83-93.

134. Aberg, M.A., et al., Peripheral infusion of IGF-I selectively induces neurogenesis in the adult rat hippocampus. J Neurosci, 2000. 20(8): p. 2896-903.

135. Stolp, H.B., Neuropoietic cytokines in normal brain development and neurodevelopmental disorders. Molecular and Cellular Neuroscience, 2013. 53(0): p. 63- 68.

136. Green, H.F., et al., A role for interleukin-1beta in determining the lineage fate of embryonic rat hippocampal neural precursor cells. Mol Cell Neurosci, 2012. 49(3): p. 311-21.

137. Jakubs, K., et al., Inflammation regulates functional integration of neurons born in adult brain. J Neurosci, 2008. 28(47): p. 12477-88.

138. Jarlestedt, K., et al., Decreased survival of newborn neurons in the dorsal hippocampus after neonatal LPS exposure in mice. Neuroscience, 2013. 253(0): p. 21-8.

139. Cui, K., et al., Effects of prenatal immune activation on hippocampal neurogenesis in the rat. Schizophrenia Research, 2009. 113(2-3): p. 288-297.

140. Jiang, P., et al., The persistent effects of maternal infection on the offspring's cognitive performance and rates of hippocampal neurogenesis. Prog Neuropsychopharmacol Biol Psychiatry, 2013. 44(0): p. 279-89.

141. von Bohlen Und Halbach, O., Structure and function of dendritic spines within the hippocampus. Ann Anat, 2009. 191(6): p. 518-31.

142. Olton, D.S., J.T. Becker, and G.E. Handelmann, Hippocampus, space, and memory. Behavioral and Brain Sciences, 1979. 2(03): p. 313-322.

143. Green, E.J. and J.M. Juraska, The dendritic morphology of hippocampal dentate granule cells varies with their position in the granule cell layer: a quantitative Golgi study. Experimental Brain Research, 1985. 59(3): p. 582-586.

144. Juraska, J.M., et al., Sex differences in the dendritic branching of dentate granule cells following differential experience. Brain Research, 1985. 333(1): p. 73-80.

145. McEwen, B.S., Stress and hippocampal plasticity. Annu Rev Neurosci, 1999. 22: p. 105- 22.

146. Woolley, C.S., E. Gould, and B.S. McEwen, Exposure to excess glucocorticoids alters dendritic morphology of adult hippocampal pyramidal neurons. Brain Res, 1990. 531(1- 2): p. 225-31.

147. Richwine, A.F., et al., Architectural changes to CA1 pyramidal neurons in adult and aged mice after peripheral immune stimulation. Psychoneuroendocrinology, 2008. 33(10): p. 1369-77.

148. Milatovic, D., et al., Pharmacologic suppression of neuronal oxidative damage and dendritic degeneration following direct activation of glial innate immunity in mouse cerebrum. J Neurochem, 2003. 87(6): p. 1518-26.

149. Jurgens, H.A., K. Amancherla, and R.W. Johnson, Influenza Infection Induces Neuroinflammation, Alters Hippocampal Neuron Morphology, and Impairs Cognition in Adult Mice. The Journal of Neuroscience, 2012. 32(12): p. 3958-3968.

150. Baharnoori, M., W.G. Brake, and L.K. Srivastava, Prenatal immune challenge induces developmental changes in the morphology of pyramidal neurons of the prefrontal cortex and hippocampus in rats. Schizophr Res, 2009. 107(1): p. 99-109.

151. Pickett, J. and E. London, The neuropathology of autism: a review. J Neuropathol Exp Neurol, 2005. 64(11): p. 925-35.

152. Kolluri, N., et al., Lamina-specific reductions in dendritic spine density in the prefrontal cortex of subjects with schizophrenia. Am J Psychiatry, 2005. 162(6): p. 1200-2.

153. Glantz, L.A. and D.A. Lewis, Decreased dendritic spine density on prefrontal cortical pyramidal neurons in schizophrenia. Arch Gen Psychiatry, 2000. 57(1): p. 65-73.

154. Bitzer-Quintero, O.K. and I. Gonzalez-Burgos, Immune system in the brain: a modulatory role on dendritic spine morphophysiology? Neural Plast, 2012. 2012: p. 348642.

155. Kondo, S., S. Kohsaka, and S. Okabe, Long-term changes of spine dynamics and microglia after transient peripheral immune response triggered by LPS in vivo. Mol Brain, 2011. 4: p. 27.

156. Neves, G., S.F. Cooke, and T.V. Bliss, Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. Nat Rev Neurosci, 2008. 9(1): p. 65-75.

157. Kasai, H., et al., Structural dynamics of dendritic spines in memory and cognition. Trends Neurosci, 2010. 33(3): p. 121-9.

158. Khandaker, G.M., et al., Childhood infection and adult schizophrenia: a meta-analysis of population-based studies. Schizophr Res, 2012. 139(1-3): p. 161-8.

159. Liu, F. and L.D. McCullough, Inflammatory responses in hypoxic ischemic encephalopathy. Acta Pharmacol Sin, 2013. 34(9): p. 1121-30.

160. Stirling, D.P., et al., Minocycline as a neuroprotective agent. Neuroscientist, 2005. 11(4): p. 308-22.

161. He, Y., S. Appel, and W. Le, Minocycline inhibits microglial activation and protects nigral cells after 6-hydroxydopamine injection into mouse striatum. Brain Res, 2001. 909(1-2): p. 187-93.

162. Guo, G. and N.R. Bhat, p38alpha MAP kinase mediates hypoxia-induced motor neuron cell death: a potential target of minocycline's neuroprotective action. Neurochem Res, 2007. 32(12): p. 2160-6.

163. Walton, K.M., et al., Activation of p38MAPK in Microglia After Ischemia. Journal of Neurochemistry, 1998. 70(4): p. 1764-1767.

164. Kim, H.S. and Y.H. Suh, Minocycline and neurodegenerative diseases. Behav Brain Res, 2009. 196(2): p. 168-79.

165. Chen, M., et al., Minocycline inhibits caspase-1 and caspase-3 expression and delays mortality in a transgenic mouse model of Huntington disease. Nature medicine, 2000. 6(7): p. 797-801.

166. Wang, J., et al., Minocycline up-regulates Bcl-2 and protects against cell death in mitochondria. J Biol Chem, 2004. 279(19): p. 19948-54.

167. Fan, L.W., et al., Minocycline attenuates lipopolysaccharide-induced white matter injury in the neonatal rat brain. Neuroscience, 2005. 133(1): p. 159-68.

168. Tsuji, M., et al., Minocycline worsens hypoxic-ischemic brain injury in a neonatal mouse model. Exp Neurol, 2004. 189(1): p. 58-65.

169. Buller, K.M., et al., Minocycline: A neuroprotective agent for hypoxic-ischemic brain injury in the neonate? Journal of Neuroscience Research, 2009. 87(3): p. 599-608.

170. Gordon, P.H., et al., Efficacy of minocycline in patients with amyotrophic lateral sclerosis: a phase III randomised trial. The Lancet Neurology, 2007. 6(12): p. 1045- 1053.

171. Chaudhry, I.B., et al., Minocycline benefits negative symptoms in early schizophrenia: a randomised double-blind placebo-controlled clinical trial in patients on standard treatment. Journal of Psychopharmacology, 2012. 26(9): p. 1185-1193.

172. Atladottir, H.O., et al., Association of Hospitalization for Infection in Childhood With Diagnosis of Autism Spectrum Disorders: A Danish Cohort Study. Arch Pediatr Adolesc Med, 2010. 164(5): p. 470-477.

173. Meyer, U., J. Feldon, and O. Dammann, Schizophrenia and autism: both shared and disorder-specific pathogenesis via perinatal inflammation? Pediatr Res, 2011. 69(5 Pt 2): p. 26R-33R.

174. Meyer, U. and J. Feldon, Epidemiology-driven neurodevelopmental animal models of schizophrenia. Prog Neurobiol, 2010. 90(3): p. 285-326.

175. Centers for Disease Control and Prevention, Respiratory Syncytial Virus Activity - United States, July 2011- January 2013. MMWR, 2013. 62(08): p. 141-144.

176. Steiner, R.W., Treating acute bronchiolitis associated with RSV. Am Fam Physician, 2004. 69(2): p. 325-30.

177. Arnold, R., et al., Cytokine (IL-8, IL-6, TNF-alpha) and soluble TNF receptor-I release from human peripheral blood mononuclear cells after respiratory syncytial virus infection. Immunology, 1995. 85(3): p. 364-72.

CHAPTER 3

EARLY POSTNATAL RESPIRATORY VIRAL INFECTION INDUCES STRUCTURAL AND NEUROCHEMICAL CHANGES IN THE NEONATAL PIGLET BRAIN

3.1 Abstract

Respiratory infections during the neonatal period are common, yet little is known about their impact on human brain development because studies in infants are either unethical or extremely difficult. Furthermore, results from rodent models are difficult to translate to human infants due to the differences in brain development and morphology. To overcome some of these barriers, in the present study domestic piglets that have brain growth and morphology similar to human infants were inoculated with porcine reproductive and respiratory syndrome virus (PRRSV) on postnatal day (PD) 7 and magnetic resonance imaging (MRI) was used to assess brain macrostructure (voxel-based morphometry), microstructure (diffusion tensor imaging) and biochemistry (MR spectroscopy) at PD 29 or 30. PRRSV piglets exhibited signs of infection throughout the post inoculation period and had elevated plasma levels of TNFα at the end of the study, indicating PRRSV caused a sustained inflammatory response. PRRSV infection increased the volume of several components of the ventricular system including the cerebral aqueduct, fourth ventricle, and the lateral ventricles. Group comparisons between control and

PRRSV piglets defined 8 areas (≥20 clusters) where PRRSV piglets had less gray matter volume; 5 areas where PRRSV piglets had less white matter volume; and 4 relatively small areas where PRRSV piglets had more white matter. Of particular interest was a bilateral reduction in gray and white matter in the primary visual cortex. PRRSV piglets tended to have reduced fractional anisotropy (an indicator of white matter development) in the corpus callosum.

Additionally, N-acetylaspartate, creatine, and myo-inositol were decreased in the hippocampus of PRRSV piglets suggesting disrupted neuronal and glial health and energy imbalances.

These finding show that early-life infection can affect brain growth and development.

3.2 Introduction

Infectious disease is the most common cause of illness in children, with acute respiratory infection constituting the most prevalent reason for hospitalization in children under one year of age (Hall, Weinberg et al. 2009). This is a concern for neurodevelopment because the phase of rapid brain growth that begins in the late prenatal period continues into the early postnatal period due to dendritic growth, synaptogenesis, and intense glial cell proliferation (Huttenlocher

1979; Dietrich, Bradley et al. 1988; Rice and Barone 2000). During infection, immune-to-brain signaling pathways activate microglia, causing neuroinflammation (Dantzer and Kelley 2007).

Developing and mature neurons, as well as glia, have numerous pro-inflammatory cytokine receptors(Sawada, Itoh et al. 1993), and untimely inflammation in rodent models, especially in the prenatal period, is known to affect neural development and increase risk for behavioral disorders that manifest later (Meyer, Feldon et al. 2011). Little is known, however, about peripheral infection in the neonatal period, especially in humans and other animals whose brain is gyrencephalic and experience major perinatal growth.

Quantitative magnetic resonance imaging (MRI) can provide important information on brain development in early childhood and adolescence but because of potential health concerns for infants, few studies have focused on the period from birth to 2 years of age when dramatic brain development occurs and progress on understanding the influence of infection on brain development in infants has been slow. Furthermore, results from rodent models commonly used to investigate neurodevelopment are difficult to translate to human infants due to the substantial differences in brain development and morphology. In this regard, the domestic piglet (Sus scrofa domestica) may be an excellent translational model. Similar to humans, the major brain growth spurt in pigs extends from the late prenatal to the postnatal period (Dobbing and Sands

1979). Gross anatomical features, including gyral pattern and distribution of gray and white matter of the neonatal piglet brain are similar to that of human infants (Dickerson and Dobbing

1967; Thibault and Margulies 1998). Moreover, their physical size allows neuroimaging instruments designed for humans to be used with piglets. Indeed, structural MRI, functional

MRI, and positron emission tomography have all been conducted in pigs (Fang, Lorke et al.

2005; Jakobsen, Pedersen et al. 2006; Conrad, Dilger et al. 2012). Additionally, piglets can undergo cognitive testing at a young age and show far more overlap with humans in genes involved in immunity compared to rodents (Dilger and Johnson 2010; Dawson 2011; Meurens,

Summerfield et al. 2012). Thus, piglets represent a gyrencephalic species with brain growth similar to humans that can be used in highly controlled experiments to explore how infection affects brain structure and function.

In the present study, we used MRI to assess brain structure (voxel-based morphometry), connectivity (diffusion-tensor imaging) and metabolites (spectroscopy) to test the hypothesis that infection with porcine reproductive and respiratory syndrome virus (PRRSV) in the early postnatal period affects brain development in piglets. PRRSV infects mononuclear myeloid cells in lungs inducing interstitial pneumonia in piglets (Done, Paton et al. 1996). Furthermore, postnatal PRRSV infection in piglets was recently reported to activate brain microglial cells, cause neuroinflammation, and reduce performance in a spatial learning and memory task

(Elmore, Burton et al. 2014).

3.3 Materials and Methods

Animals, Housing, and Feeding

Naturally farrowed crossbred piglets from three separate litters (7 males and 7 females) were obtained from the University of Illinois swine herd. Piglets were brought to the biomedical animal facility on PD 2 (to allow for colostrum consumption from the sow), where upon arrival, they were assigned to either the control group (3 males and 3 females) or the PRRSV infection group (4 males and 4 females) based on sex, litter of origin, and body weight and housed

44 individually in cages (0.76 m L x 0.58 m W x 0.47 m H) designed for neonatal piglets (Elmore,

Burton et al. 2014). Two of the PRRSV females developed severe diarrhea and were removed from the study. Each cage was positioned in a rack, with stainless steel perforated side walls and clear acrylic front and rear doors within one of two separate but identical disease containment chambers that have been described (Elmore, Burton et al. 2014). Each cage was fitted with flooring designed for neonatal animals (Tenderfoot/NSR, Tandem Products, Inc.,

Minneapolis, MN, USA). A toy (plastic Jingle BallTM, Bio-Serv, Frenchtown, NJ, USA) was provided to each piglet. Room temperature was maintained at 27ºC and each cage was equipped with an electric heat pad (K&H Lectro-KennelTM Heat Pad, K&H Manufacturing, LLC,

Colorado Springs, CO, USA). Piglets were maintained on a 12-h light/dark cycle; however, during the dark cycle minimal lighting was provided.

Piglets were fed a commercial sow milk replacer (Advance Liqui-Wean, Milk Specialties

Co., Dundee, IL, USA). Milk was reconstituted daily to a final concentration of 206 g/L using tap water and supplied at a rate of 285 mL/kg BW (based on daily recorded weights) to a stainless steel bowl via a peristaltic pump (Control Company, Friendswood, TX). Using this automated feeding system (similar to that described previously (Dilger and Johnson 2010)), piglets received their daily allotted milk over 18 meals (once per hour from 0600 h to 2400 h). All animal experiments were in accordance with the National Institute of Health Guidelines for the

Care and Use of Laboratory Animals and approved by the University of Illinois at Urbana-

Champaign Institutional Animal Care and Use Committee.

Experimental Design and Assessment of Sickness

At PD 7, piglets were inoculated intranasal with either 1mL of 1x10^5 50% tissue culture infected dose (TCID 50) of live PRRSV (strain P129-BV, obtained from the School of Veterinary

Medicine at Purdue University, West Lafayette, IN, USA) or sterile phosphate buffered saline

(PBS). Daily body weights (kg) were obtained to monitor piglets’ growth. In addition, daily

45 rectal temperatures were obtained PD 7 through PD 28. The willingness of the piglets to consume their first daily meal was determined from PD 7 to PD 28 using a feeding score (1 = no attempt to consume the milk; 2 = attempted to consume the milk, but did not finish within 1 min;

3 = consumed all of the milk within 1 min). Body weight, rectal temperature, and feeding score data were analyzed as a two-way (treatment x day) repeated measures ANOVA using the

MIXED procedure in SAS (SAS Institute Inc, Cary, NC, USA). There were no significant differences due to sex, therefore it was not included in final analysis of body weight, rectal temperature, and feeding score data. Significance was accepted at p<0.05.

The presence of PRRSV antibodies in the serum of all piglets at the end of the study was analyzed by the Veterinary Diagnostic Laboratory (University of Illinois, Urbana, Illinois) using a PRRSV-specific ELISA kit (IDEXX Laboratories). This assay has 98.8% sensitivity and

99.9% specificity, with an S/P ratio of >0.4 indicating a positive sample. Serum tumor necrosis factor alpha (TNF-α) levels at the end of the study were determined using porcine-specific sandwich enzyme immunoassays (R&D Systems). Data analysis was conducted using the

MIXED procedure in SAS. Serum TNF-α concentrations were analyzed as a two-way

(treatment x sex) ANOVA. Sex was not significant so it was removed from the model and data were reanalyzed as a one-way ANOVA to determine the effects of treatment. Significance was accepted at p<0.05.

At PD 29 and PD30 control and PRRSV piglets were transported to the Biomedical

Imaging Center at the Beckman Institute and anesthetized using a telazol:ketamine:xylazine

(100/50/50 mg/kg/BW; Fort Dodge Animal Health). Anesthesia was maintained by inhalation of isoflurane (98% oxygen/2% isoflurane). A MRI compatible pulse oximeter was used to monitor heart rate and oxygen saturation throughout the scanning procedure. After piglets entered a sustainable plane of deep anesthesia, they were placed in the MRI machine. Piglets were restrained and remained immobilized throughout the MRI scans to prevent motion artifact. After

46 conclusion of the scans, the piglets were euthanized via intracardiac injection of sodium pentobarbital (Fatal Plus, 72 mg/kg BW).

All MRI was conducted using a Siemens MAGNETOM Trio 3T imager and a 32-channel head coil (Siemens, Erlangen, Germany). Descriptions of the MRI sequences have been previously described, and are briefly explained below (Radlowski et al, 2013). The total scan time per pig was roughly 50 minutes.

Structural MRI Acquisition and Analysis

For brain structure analysis, anatomic images were acquired using a 3D T1-weighted magnetization-prepared rapid gradient-echo (MPRAGE) sequence. Three repetitions were acquired using the following parameters: repetition time = 1,900 ms; echo time = 2.49 ms; inversion time = 900 ms; flip angle = 9°; matrix = 256 x 256, 224 slices with thickness = 0.7 mm.

The final voxel size was 0.7 mm isotropic across the entire head from the tip of the snout to the cervical/thoracic spinal cord junction as described previously (Conrad, Dilger et al. 2012).

The three MPRAGE image sets were averaged and the brains were manually extracted using FSL (Analysis Group, FMRIB, Oxford, UK) (Jenkinson, Beckmann et al. 2012). The extracted brains were brought in to a common stereotactic space by 12 parameter affine registration to a digital piglet brain atlas (http://pigmri.illinois.edu). The remainder of the voxel- based morphometry technique was conducted using SPM8 (Wellcome Department of Clinical

Neurology, London; http://www.fil.ion.cul.ac.uk) in MATLAB version 8.0.0.783 (Mathworks). The brains were segmented into gray matter, white matter, and CSF classifications using the prior probability maps available in the piglet brain atlas dataset (http://pigmri.illinois.edu). Next, the

Diffeomorphic Anatomical Registration using Exponentiated Lie Algebra (DARTEL) package of

SPM8 was used to create study specific templates (Ashburner 2007). Changes from default include a bounding box of -30.1 to 30.1, -35 to 44.8, -28 to 31.5; and a voxel size of 0.7 mm3. A linear bend energy was used as the cost function. Flow fields generated from the DARTEL

47 procedure were converted to warp fields which represent the expansion/shrinkage from the subject to atlas space. The modulated data were smoothed with a 4 mm full-width half maximum (FWHM).

Due to the small sample size, the statistical non-parametric methods (SnPM) toolbox was used to compare the modulated gray and white matter data between the control and

PRRSV groups (http://go.warwick.ac.ui/tenichols/snpm). SnPM uses non-parametric permutation tests which is advantageous for small numbers per treatment (Nichols and Holmes

2002). A maximum number of permutations were used with a FWHW of 7 mm for variance smoothing. ANCOVA was used for global normalization and a mean voxel value was used for

Global Calculation. All other options were default. No covariates were used.

Two-sample permutation t tests were performed voxel-wise for both gray and white matter to detect differences between the control and PRRSV treatments. Due to the small numbers per treatment, an uncorrected p < 0.01 was used. A threshold of at least 20 edge- connected voxels (clusters) was set and clusters that appeared on the edges of the brain were excluded. The differences between control and PRRSV piglets are displayed in the pseudo-t statistic images (Figures 2 and 3). The location of the significant voxel clusters were estimated using an MRI-based atlas for adult domestic pigs (Saikali, Meurice et al. 2010).

In addition to the VBM, brain region volume estimations were calculated. Using the warp fields from the DARTEL procedure, inverse warps were created for each subject. The inverse warp represents the expansion/shrinkage from the atlas to the subject space. The inverse warps were then applied to the regions of interest (ROI) maps found in the piglet brain atlas

(http://pigmri.illinois.edu). The volumes were then estimated for each ROI for each subject using FSLstats. Data analysis was conducted using the MIXED procedure in SAS. ROI volumes were analyzed as a two-way (treatment x sex) ANOVA. Significance was accepted at p<0.05. . Unless otherwise stated, data are presented as Least Square Means (LSM) ±

Standard Error of the Means (SEM).

Diffusion Tensor Imaging and Analysis

Analysis of neuronal connectivity was conducted using a diffusion tensor imaging (DTI) sequence to visualize fiber tracts non-invasively. A diffusion-weighted echo-planar-imaging

(DW-EPI) sequence was used with the following parameters: repetition time = 5000 ms; echo time = 91 ms; averages = 3; diffusion weightings = 2; b-value of 1000 s/mm2 across 30 directions; two images with a b-value of 0 s/mm2. Forty slices with a 2.0 mm thickness were collected with a matrix size of 100 x 100 for a final voxel size of 2.0 mm isotropic.

The DW-EPI images were then analyzed using the diffusion toolbox in the FSL software package to create fractional anisotropy (FA) and mean diffusivity (MD) images. The brain was manually extracted. Masks of the corpus callosum and white matter from the piglet brain atlas were nonlinearly transformed to the subject’s MPRAGE space and then linearly transformed into the DTI space. The corpus callosum had a threshold of 0.15 applied to compensate for expansion caused by interpolation. The white matter ROI had a threshold of 0.5 applied and was dilated twice. Averages were taken in the regions of interest with a FA>0.2 mask to select only white matter tracts. Averages were taken over the whole brain with a FA>0.2 mask to assess global FA values. MD values were averaged over the same region of interest. Data analysis was conducted using the T-TEST procedure of SAS software. A two sample two-sided t-test design was used to compare control versus PRRSV animals. Significance was accepted at p<0.05, statistical trends at p<0.10.

MR Spectroscopy Parameters and Analysis

Magnetic resonance spectroscopy (MRS) was performed using a spin echo chemical shift sequence with the following parameters: repetition time = 3000 ms; echo time = 30 ms; 128 averages. The voxel size was 12 x 25 x 12 mm and was placed over the left and right dorsal hippocampus. Six presaturation bands were placed around the volume of interest. Automatic

49 and/or higher-order shims were conducted before each spectroscopic scan to ensure a FWHM of less than 18 Hz. Both water-suppressed and non-water suppressed data were collected.

All MRS data were subsequently analyzed by LC Model 6.3 fitting program

(Provencher). Absolute levels, in institutional units, were obtained by using eddy current correction procedure and water scaling. Water-suppressed time domain data were analyzed between 0.2 and 4.0 ppm without further T1 or T2 correction. There was no post-suppression of water while using the LC Model. The basis-set of metabolite phantom spectra was used as prior knowledge; the impact of lack of post-suppression was minimal. Absolute concentrations were calculated for each metabolite. Cramer-Rao lower bounds (S.D.%) were calculated by the

LC Model to evaluate the quantitative result of absolute levels. Only those levels that were less than 20% were considered reliable and were included for further statistical analysis. Data analysis was conducted using the MIXED procedure in SAS. Metabolite concentrations were analyzed as a two-way (treatment x sex) ANOVA. Sex was not a significant main effect for any metabolite so it was removed from the model and data were reanalyzed as a one-way ANOVA to determine the effects of treatment. Significance was accepted at p<0.05.

3.4 Results

PRRSV Infection and Measures of Sickness

Serum ELISA results indicated all control piglets were negative for PRRSV at the end of the study, whereas all piglets inoculated were positive for PRRSV. Body weight, rectal temperature, and the willingness to consume the first daily meal (feeding score) were determined to provide an indication of the sickness response of piglets infected with PRRSV

(Figure 3.1). Analysis of body weight data showed a significant effect of day (F (28, 275) =

31.51, p < 0.001), but neither treatment (F (1, 10) = 0.53) nor the day × treatment interaction (F

(27, 275) = 1.25) were significant indicating control and PRRSV piglets gained similar weight over the course of the study. Nonetheless, analysis of feeding score data revealed a significant effect of treatment (F (1, 10) = 10.21, p < 0.01), day (F (22, 214) = 2804.45, p < 0.001), and a trend toward a treatment × day interaction (F (21, 214) = 1.58, p = 0.056), indicating PRRSV piglets’ motivation to consume the first meal of the day was reduced towards the last half of the study. It is noteworthy that this is the timeframe when body weight of control and PRRSV piglets began to separate. Analysis of rectal temperature data showed a significant effect of treatment (F (1, 10) = 43.19), day (F (22, 215) = 3.56), and a treatment × day interaction (F (21,

215) = 2.88) (all, p < 0.0001). PRRSV piglets became febrile 3 d after inoculation and remained so during most of the experimental period. The proinflammatory cytokine TNF-α was significantly increased in PRRSV piglets (F (1, 20) = 98.87 p < 0.001) with an average serum concentration of 135.6 ± 13.0 pg/mL compared to 17.1 ± 1.7 pg/mL in controls. Collectively, these data indicate infection with PRRSV in the neonatal period induced clinical signs of illness.

Brain Region Volume Estimation

Using the deformation maps created during the DARTEL procedure, the regions of interest (ROI) from the piglet brain atlas (http://pigmri.illinois.edu) were warped to the individual subject’s brain. Volume estimations were made for 19 regions of interest (Table 3.1). In addition, gray and white matter volumes were estimated by the segmented data from the

DARTEL procedure and whole brain volume was estimated using the extracted brain. Overall, there were very few differences in brain region volumes between control and PRRSV piglets or between male and female piglets (Table 3.1). An exception was that PRRSV infection increased the volume of several components of the ventricular system including the cerebral

51 aqueduct, fourth ventricle, and the lateral ventricles. In the case of the cerebral aqueduct and fourth ventricle, the effects of PRRSV were greater in females compared to males.

Voxel-based Morphometry

Although there were few effects of PRRSV on brain region volumes, the VBM analysis

(Table 3.2) revealed eight clusters where PRRSV piglets had significantly less gray matter compared to control piglets. The local maxima coordinates were mapped to the piglet brain atlas and histological atlas to estimate the anatomic region (http://pigmri.illinois.edu) (Felix,

Leger et al. 1999). Of particular interest are two clusters located in the left and right primary visual cortex which had the highest pseudo-t value. Pseudo-t maps for areas that have reduced gray matter volume in PRRSV piglets are shown in Figure 3.2. There were no areas where the

PRRSV piglets had increased gray matter volumes compared to controls.

Analysis of white matter volumes revealed four regions where PRRSV piglets had increased white matter compared to controls (Table 3.2). These regions had smaller clusters and lower pseudo-t values compared to clusters where controls had more white matter volume.

There were five clusters where PRRSV piglets had less white matter volume than controls

(Table 3.2). Pseudo-t maps for areas that have reduced white matter volume in PRRSV piglets are shown in Figure 3.3.

Diffusion Tensor Imaging

Fractional anisotropy maps were compared between control and PRRSV piglets to evaluate white matter tract differences. An example FA map is shown in Figure 3.4. Three areas of interest were evaluated for FA and MD differences including whole brain, corpus

52 callosum, and white matter (Figure 3.4). There were no differences in whole brain and white matter FA (p > 0.10 in both cases). However, when the region of interest was restricted to the corpus callosum, control piglets tended to have a larger FA than PRRSV piglets (p = 0.06).

There were no significant differences in whole brain, corpus callosum, or white matter for MD (p

> 0.10 in each case).

Magnetic Resonance Spectroscopy

Differences in absolute metabolite concentrations between control and PRRSV piglets were evaluated in the hippocampus using MRS. The hippocampus was chosen because

PRRSV infection (a) activates microglia in this brain region; and (b) inhibits performance in a hippocampal-dependent spatial T-maze task (Elmore, Burton et al. 2014). A single voxel was placed over the left and right dorsal hippocampus. This location and a representative MR spectrum are shown in Figure 3.5. Nine metabolites/metabolite combinations reached Cramer-

Rao lower bounds of SD < 20% in all subjects. These metabolites and their absolute concentrations are shown in Table 3.3. Four of the metabolites were significantly reduced in the

PRRSV piglets including creatine/phosphocreatine (Cr+PCr), N-Acetylasparate (NAA), NAA/N-

Acetylaspartylglutamate (NAAG), and Myo-inositol (Ins).

3.5 Discussion

Respiratory infections during the neonatal period are common, yet little is known about their impact on brain development. Progress in this area has been slow because studies in human infants are either impossible, due to obvious ethical considerations, or extremely difficult.

Furthermore, results from rodent models commonly used to investigate neurodevelopment are difficult to translate to human infants due to the substantial differences in brain development and

53 morphology. To overcome some of these barriers, the present study was conducted using a highly tractable and translational piglet model.

Previously, we used MRI and a longitudinal study design and showed that brain growth in the domestic piglet is similar in many regards to the human neonate (Conrad, Dilger et al.

2012). Logistical modeling suggested that total brain volume from birth to 4 wks of age increased 45%. Therefore, in the present study inoculation with PRRSV and the subsequent infection coincided with a period of rapid brain growth. PRRSV is an enveloped, positive-sense, single-stranded RNA virus that belongs to the Arteriviridae family within the order Nidovirales

(Done, Paton et al. 1996). In young pigs PRRSV primarily infects and replicates in cells of the monocyte/macrophage lineage (Duan, Nauwynck et al. 1997). Once infected, macrophages can migrate to lymphatic tissue and enter circulation allowing for the spread of PRRSV (Shin and

Molitor 2002). Either by eliciting a pro-inflammatory response that activates immune-to-brain signaling pathways or by entering the CNS, PRRSV activates microglia and causes neuroinflammation. In a recent study, several pro-inflammatory cytokines were elevated in serum 20 d after inoculation with PRRSV, and a number of pro-inflammatory genes, including interferon-γ, TNFα, and IL-1β, were up regulated in several brain regions at the same time post inoculation (Elmore, Burton et al. 2014). Furthermore, the percentage of activated microglia, as indicated by expression of MHC class II, was markedly increased in piglets with PRRSV infection (Elmore, Burton et al. 2014). Microglial activation was positively correlated with fever, and negatively correlated with food motivation and learning and memory (Elmore, Burton et al.

2014). In the present study, piglets were febrile throughout the study and circulating TNFα was elevated in samples collected after MRI (i.e., 21 d after inoculation). Thus, it is reasonable to conclude that PRRSV infected piglets experienced a sustained neuroinflammatory response throughout the study period.

We employed multiple MRI-based techniques to assess the effects of PRRSV infection on piglet brain macrostructure (VBM), microstructure (DTI), and biochemistry (MRS). Although

54 this may be the first study to use a MRI-based approach to investigate if early-life infection affects brain development, we recently used a similar strategy to explore brain development in piglets born small-for-gestational-age (i.e., a model of intrauterine growth restriction; Radlowski et al.) and piglets provided formula supplemented with phospholipids and gangliosides

(Radlowski et al.). In the present study, VBM revealed multiple brain areas in PRRSV piglets with less gray matter volume. The two most significant clusters were located bilaterally in the primary visual cortex. White matter volume was also reduced in these regions. The primary visual cortex develops quickly with a high rate of synaptogenesis after birth and, therefore, may be sensitive to environmental insults at this time (Johnson 1990; Bourgeois and Rakic 1993;

Gogtay, Giedd et al. 2004). Microglia are vital for synaptogenesis in the primary visual cortex and activation may disrupt normal development (Rochefort, Quenech'du et al. 2002). While we wish to be cautious not to overstate the significance of these findings in neonatal piglet model, it is worth noting that a neuroimmune-based developmental hypothesis has been proposed for several psychiatric diseases in humans that are associated with deficiencies in the primary visual cortex (Meyer, Feldon et al. 2011). First, the primary visual cortex in autistic patients appears normal, but other areas, such as the fusiform gyrus, change (van Kooten, Palmen et al.

2008). It is thought that the visual processing abnormalities in autism may be due to top-down processing differences (Hadjikhani, Chabris et al. 2004). Second, some schizophrenia patients present with visual processing problems which are thought to be due to a disruption in the early development in the visual pathway (Butler, Schechter et al. 2001). Postmortem histological analysis has shown that schizophrenia patients have roughly a 25% reduction in neuron number and volume in the primary visual cortex compared to controls (Dorph-Petersen, Pierri et al.

2007). The reduced volumes seen in the PRRSV piglets maybe indicative of the sensitivity to inflammation and may result in long term consequences. Either delays in synaptogenesis or degeneration of established connections could explain the volume differences. Additional

55 studies to characterize the histological changes in the primary visual cortex of infected piglets are warranted.

In addition to voxel-based morphometry, region of interest analysis was used to estimate volumes of 19 brain regions. Changes were found in the ventricular system, including the cerebral aqueduct, lateral, and fourth ventricle, where PRRSV piglets had an expanded volume.

Increased lateral ventricle volumes have been shown in both autism and schizophrenia (Chua and McKenna 1995; Piven, Arndt et al. 1995; Palmen, Hulshoff Pol et al. 2005; Rosa,

Schaufelberger et al. 2010). We also found a sexual dimorphism in the corpus callosum where males had a smaller corpus callosum than females, similar to humans (Ardekani, Figarsky et al.

2012).

Myelination begins in late gestation, has its highest rate of formation in early-life, and is susceptible to inflammation (Nakagawa, Iwasaki et al. 1998; Lenroot and Giedd 2006; Abraham,

Vincze et al. 2010; Favrais, van de Looij et al. 2011). Disruptions to white matter tracts, including the corpus callosum, have been found in psychiatric disease (Alexander, Lee et al.

2007; Cheung, Cheung et al. 2008; Cheung, Chua et al. 2009). Here, we found no differences in the FA values across the whole brain and white matter. There was a trend in the corpus callosum where the PRRSV piglets had decreased FA values which may indicate a disruption in myelination. Previous studies in rodents have found that inflammatory cytokines block oligodendrocytes maturation and cause axonopathy including reduced myelinated axon diameter (Favrais, van de Looij et al. 2011). The inflammatory response that is seen in the

PRRSV piglets includes an increase in pro-inflammatory cytokines and this may be a mechanism for the changes seen (Elmore, Burton et al. 2014). Additionally, we did not find any differences in MD suggesting that there were no changes in cellular integrity. Further analysis is needed in this model to determine if other white matter tracts are affected by the infection.

MR-spectroscopy is a powerful tool to quantify chemical metabolite concentrations non- invasively (Soares and Law 2009). Nine metabolites were quantified in the hippocampus with

56 four being significantly lower in the PRRSV piglets including creatine/phosphocreatine, myo- inositol, NAA, and NAAG. Reductions in creatine and myo-inositol suggest an energy imbalance and decrease in glial cells or delayed gliogenesis compared to controls. NAA is a marker of the density and viability of neurons and is reduced in many neuroinflammatory conditions (Soares and Law 2009). The reduced NAA in the PRRSV piglets may be a sign of neuronal loss or developmental impairment. Similar findings have been shown in hypoxia ischemia models using piglets suggesting that energy-dysregulation, including secondary energy failure, may be present after early-life infection (Munkeby, De Lange et al. 2008).

In conclusion, the present study found that PRRSV infection causes structural changes in gray and white matter, as well as neurochemical changes. Due to the small number of animals used for this study, we were limited in the analysis of covariates in VBM including sex.

Further histological analysis of the primary visual cortex and corpus callosum would be beneficial to further characterize the changes seen. Regardless, early-life infection and inflammation have a significant impact on normal brain development. In future studies, it will be important to examine long-term endpoints to determine the lasting impact of early-life infection.

3.6 Figures and Tables

A 10

4 Weight (kg)

2 Control PRRSV 0

2 3 4 5 6 7 8 9 2 3 4 5 6 9 0 1 2 6 7 8 9 0 10 11 1 1 1 1 1 17 18 1 2 2 2 23 24 25 2 2 2 2 3 B Day of Age 41

# # 40 # C) ## #  # * # # * * # * # * * 39

Temperature ( Temperature 38 * = <0.05 # = <0.01 37 7 8 9 1 2 3 4 5 6 0 1 2 3 4 5 6 7 10 1 1 1 1 1 1 17 18 19 2 2 2 2 2 2 2 2 28 Day of Age C 3.2

3.0

2.8

2.6

2.4 Feeding ScoreFeeding

2.2

2.0 7 8 9 0 8 9 8 1 11 12 13 14 15 16 17 1 1 20 21 22 23 24 25 26 27 2 Day of Age

Figure 3.1. PRRSV piglets show a febrile response and reduced feeding score (A to C). There were no differences in weight between the two treatments (A). PRRSV piglets had significantly higher temperatures starting on the third day that lasted throughout the study. PRRSV piglets also had a significant reduction in feeding score. Data are shown as the mean ± SEM (*p<0.05, #p<0.01; n=6 per treatment).

Figure 3.2. VBM analysis showing clusters where PRRSV pigs had significantly less gray matter than controls. Axial slices are shown with a 1 mm slice thickness. The color bar indicates significance level (pseudo t-statistic).

Figure 3.3. VBM analysis showing areas where PRRSV pigs had significantly less white matter than controls. Axial slices are shown with a 1 mm slice thickness. The color bar indicates significance level (pseudo t-statistic).

ABC

0.36 0.32 0.34 Control PRRSV 0.30 # 0.33 0.34 0.32 0.28 0.32 0.31 0.26 Fractional Anisotropy Fractional Fractional Anisotropy Fractional Anisotropy Fractional 0.30 DEF Control 1.5 1.5 1.5 PRRSV

1.0 1.0 1.0

0.5 0.5 0.5 Mean Diffusivity Mean Diffusivity Mean Diffusivity 0.0 0.0 0.0 G

Figure 3.4. PRRSV piglets have a trend towards reduced FA in the corpus callosum (A to G). There were no differences in FA found in whole brain (A) and white matter (C). There was a trend that the controls had a higher FA than PRRSV in the corpus callosum (B; p=0.0637). There were no differences in MD in whole brain (D), corpus callosum (E), or white matter (F). A representative FA map in a axial, sagittal, and coronal slice from left to right is shown (G). Data are shown as the mean ± SEM (n=6 per treatment). Units for mean diffusivity are 10-3 mm2/s.

Figure 3.5. Representative proton MR spectroscopy spectrum and location of the voxel placement. The single voxel was placed across the left and right dorsal hippocampus for each animal.

Table 3.1. Brain region volume estimations. Volume averages for 22 regions are reported by sex and treatment. Significance due to sex, treatment, and the interaction are shown.

Cluster Local Maxima Cluster level Coordinates*

Comparison Anatomic Region (voxels) (P-value) x y z Pseudo-t Left Primary Visual Ctx 616 .0032 -9 2 16 4.82

Right Primary Visual Ctx 530 .0022 10 1 17 4.44 Left Insular Ctx 118 .0032 -8 25 4 3.32

Prefrontal Ctx 169 .0043 0 22 -2 3.04 Control>PRRSV Left Auditory Ctx 30 .0087 -18 6 8 2.58

Gray Matter Right Fusiform Gyrus 22 .0087 15 -11 1 2.54

Left Fusiform Gyrus 175 .0054 -13 -10 3 2.51

Right Primary Visual Ctx 69 .0065 6 -3 20 2.39

PRRSV>Control

No significant findings

Left Primary Visual Ctx 208 .0022 -6 1 17 4.60

Right Primary Visual Ctx 615 .0032 8 -1 17 4.51

Right Primary Somatosensory Ctx 57 .0097 12 15 15 3.94

Left Primary Somatosensory Ctx 436 .0032 -11 9 17 3.61 Control>PRRSV

.0032 -13 2 14 3.48

White Left Primary Somatosensory Ctx 33 .0087 -8 22 15 2.35 Matter

Prefrontal Cortex 311 .0097 3 31 1 2.59 PRRSV>Control

Right Internal Capsule 28 .0065 8 9 10 1.75

Left Hippocampus 50 .0097 -6 -6 2 1.73

Left Internal Capsule 59 .0097 -7 9 10 1.73

Table 3.2. Voxel-based morphometry analysis of gray and white matter differences. Significant clusters of at least 20 voxels are shown with the cluster size (voxels), cluster significance, location, and pseudo-t value.

Metabolite Control (SEM) PRRSV (SEM) P-value Creatine + Phosphocreatine (Cr + PCr) 4.3395 (0.0915) 3.6992 (0.1163) 0.0015 N-Acetylaspartate (NAA) 5.3473 (0.0579) 4.6953 (0.1977) 0.0101 NAA + N-Acetylaspartylglutamate (NAAG) 5.7728 (0.0286) 5.0563 (0.1845) 0.0033 Myoinositol (Ins) 9.5028 (0.4302) 8.2737 (0.3395) 0.0488 Phosphocholine + Glycerophosphocholine (GPC + PCh) 1.6775 (0.0506) 1.6253 (0.0561) 0.5056 Glutamate + Glutamine (Glu + Gln) 6.6178 (0.3099) 7.4558 (0.6084) 0.2478 MM20 8.4418 (0.5506) 9.0770 (0.4523) 0.3937 MM09+Lip09 4.4567 (0.0708) 4.5043 (0.0941) 0.6941 MM20+Lip20 9.2922 (0.4385) 9.7607 (0.4048) 0.4506

Table 3.3. Absolute metabolite concentration in the hippocampus. Single voxel MRS for nine metabolites that met criteria for analysis (%S.D.<20). Data are shown as the mean ± SEM (n=6 per treatment).

3.7 References

Abraham, H., A. Vincze, et al. (2010). "Myelination in the human hippocampal formation from midgestation to adulthood." Int J Dev Neurosci 28(5): 401-410.

Alexander, A. L., J. E. Lee, et al. (2007). "Diffusion tensor imaging of the corpus callosum in Autism." NeuroImage 34(1): 61-73.

Ardekani, B. A., K. Figarsky, et al. (2012). "Sexual Dimorphism in the Human Corpus Callosum: An MRI Study Using the OASIS Brain Database." Cerebral Cortex.

Ashburner, J. (2007). "A fast diffeomorphic image registration algorithm." NeuroImage 38(1): 95- 113.

Bourgeois, J. and P. Rakic (1993). "Changes of synaptic density in the primary visual cortex of the macaque monkey from fetal to adult stage." The Journal of Neuroscience 13(7): 2801-2820.

Butler, P. D., I. Schechter, et al. (2001). "Dysfunction of early-stage visual processing in schizophrenia." Am J Psychiatry 158(7): 1126-1133.

Cheung, C., S. E. Chua, et al. (2009). "White matter fractional anisotrophy differences and correlates of diagnostic symptoms in autism." J Child Psychol Psychiatry 50(9): 1102- 1112.

Cheung, V., C. Cheung, et al. (2008). "A diffusion tensor imaging study of structural dysconnectivity in never-medicated, first-episode schizophrenia." Psychol Med 38(6): 877-885.

Chua, S. E. and P. J. McKenna (1995). "Schizophrenia--a brain disease? A critical review of structural and functional cerebral abnormality in the disorder." Br J Psychiatry 166(5): 563-582.

Conrad, M. S., R. N. Dilger, et al. (2012). "Brain Growth of the Domestic Pig (Sus scrofa) from 2 to 24 Weeks of Age: A Longitudinal MRI Study." Dev Neurosci 34(4): 291-298.

Conrad, M. S., R. N. Dilger, et al. (2012). "Magnetic resonance imaging of the neonatal piglet brain." Pediatr Res 71(2): 179-184.

Dantzer, R. and K. W. Kelley (2007). "Twenty years of research on cytokine-induced sickness behavior." Brain, Behavior, and Immunity 21(2): 153-160.

Dawson, H. D. (2011). A Comparative Assessment of the Pig, Mouse and Human Genomes. The Minipig in Biomedical Research, CRC Press: 323-342.

Dickerson, J. W. T. and J. Dobbing (1967). "Prenatal and postnatal growth and development of the central nervous system of the pig." Proc Biol Sci 166(1005): 384-395.

Dietrich, R. B., W. G. Bradley, et al. (1988). "MR evaluation of early myelination patterns in normal and developmentally delayed infants." American Journal of Roentgenology 150(4): 889-896.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Dobbing, J. and J. Sands (1979). "Comparative aspects of the brain growth spurt." Early Hum Dev 3(1): 79-83.

Done, S. H., D. J. Paton, et al. (1996). "Porcine reproductive and respiratory syndrome (PRRS): a review, with emphasis on pathological, virological and diagnostic aspects." Br Vet J 152(2): 153-174.

Dorph-Petersen, K.-A., J. N. Pierri, et al. (2007). "Primary visual cortex volume and total neuron number are reduced in schizophrenia." The Journal of Comparative Neurology 501(2): 290-301.

Duan, X., H. J. Nauwynck, et al. (1997). "Virus quantification and identification of cellular targets in the lungs and lymphoid tissues of pigs at different time intervals after inoculation with porcine reproductive and respiratory syndrome virus (PRRSV)." Vet Microbiol 56(1-2): 9- 19.

Elmore, M. R., M. D. Burton, et al. (2014). "Respiratory viral infection in neonatal piglets causes marked microglia activation in the hippocampus and deficits in spatial learning." J Neurosci 34(6): 2120-2129.

Fang, M., D. E. Lorke, et al. (2005). "Postnatal changes in functional activities of the pig’s brain: a combined functional magnetic resonance imaging and immunohistochemical study." Neurosignals 14(5): 222-233.

Favrais, G., Y. van de Looij, et al. (2011). "Systemic inflammation disrupts the developmental program of white matter." Ann Neurol 70(4): 550-565.

Felix, B., M. E. Leger, et al. (1999). "Stereotaxic atlas of the pig brain." Brain Res Bull 49(1-2): 1-137.

Gogtay, N., J. N. Giedd, et al. (2004). "Dynamic mapping of human cortical development during childhood through early adulthood." Proc Natl Acad Sci U S A 101(21): 8174-8179.

Hadjikhani, N., C. F. Chabris, et al. (2004). "Early visual cortex organization in autism: an fMRI study." Neuroreport 15(2): 267-270.

Hall, C. B., G. A. Weinberg, et al. (2009). "The burden of respiratory syncytial virus infection in young children." N Engl J Med 360(6): 588-598.

Huttenlocher, P. R. (1979). "Synaptic density in human frontal cortex - developmental changes and effects of aging." Brain Res 163(2): 195-205.

Jakobsen, S., K. Pedersen, et al. (2006). "Detection of α2-Adrenergic Receptors in Brain of Living Pig with 11C-Yohimbine." J Nucl Med 47(12): 2008-2015.

Jenkinson, M., C. F. Beckmann, et al. (2012). "FSL." NeuroImage 62(2): 782-790.

Johnson, M. H. (1990). "Cortical maturation and the development of visual attention in early infancy." J. Cognitive Neuroscience 2(2): 81-95.

Lenroot, R. K. and J. N. Giedd (2006). "Brain development in children and adolescents: insights from anatomical magnetic resonance imaging." Neurosci Biobehav Rev 30(6): 718-729.

Meurens, F., A. Summerfield, et al. (2012). "The pig: a model for human infectious diseases." Trends in microbiology 20(1): 50-57.

Meyer, U., J. Feldon, et al. (2011). "Schizophrenia and autism: both shared and disorder- specific pathogenesis via perinatal inflammation?" Pediatr Res 69(5 Pt 2): 26R-33R.

Munkeby, B. H., C. De Lange, et al. (2008). "A piglet model for detection of hypoxic-ischemic brain injury with magnetic resonance imaging." Acta Radiol 49(9): 1049-1057.

Nakagawa, H., S. Iwasaki, et al. (1998). "Normal myelination of anatomic nerve fiber bundles: MR analysis." American Journal of Neuroradiology 19(6): 1129-1136.

Nichols, T. E. and A. P. Holmes (2002). "Nonparametric permutation tests for functional neuroimaging: a primer with examples." Hum Brain Mapp 15(1): 1-25.

Palmen, S. J., H. E. Hulshoff Pol, et al. (2005). "Increased gray-matter volume in medication- naive high-functioning children with autism spectrum disorder." Psychol Med 35(4): 561- 570.

Piven, J., S. Arndt, et al. (1995). "An MRI study of brain size in autism." Am J Psychiatry 152(8): 1145-1149.

Provencher, S. (September 14, 2013). "LCModel and LCMgui User's Manual." from http://www.s-provencher.com/pub/LCModel/manual/manual.pdf.

Rice, D. and S. Barone, Jr. (2000). "Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models." Environ Health Perspect 108 Suppl 3: 511-533.

Rochefort, N., N. Quenech'du, et al. (2002). "Microglia and astrocytes may participate in the shaping of visual callosal projections during postnatal development." J Physiol Paris 96(3-4): 183-192.

Rosa, P. G., M. S. Schaufelberger, et al. (2010). "Lateral ventricle differences between first- episode schizophrenia and first-episode psychotic bipolar disorder: A population-based morphometric MRI study." World J Biol Psychiatry 11(7): 873-887.

Saikali, S., P. Meurice, et al. (2010). "A three-dimensional digital segmented and deformable brain atlas of the domestic pig." J Neurosci Meth 192(1): 102-109.

Sawada, M., Y. Itoh, et al. (1993). "Expression of cytokine receptors in cultured neuronal and glial cells." Neurosci Lett 160(2): 131-134.

Shin, J. H. and T. W. Molitor (2002). "Localization of porcine reproductive and respiratory syndrome virus infection in boars by in situ riboprobe hybridization." J Vet Sci 3(2): 87- 96.

Soares, D. P. and M. Law (2009). "Magnetic resonance spectroscopy of the brain: review of metabolites and clinical applications." Clinical Radiology 64(1): 12-21.

Thibault, K. L. and S. S. Margulies (1998). "Age-dependent material properties of the porcine cerebrum: effect on pediatric inertial head injury criteria." J Biomech 31(12): 1119-1126. van Kooten, I. A., S. J. Palmen, et al. (2008). "Neurons in the fusiform gyrus are fewer and smaller in autism." Brain 131(Pt 4): 987-999.

CHAPTER 4

EARLY POSTNATAL RESPIRATORY VIRAL INFECTION ALTERS HIPPOCAMPAL NEUROGENESIS, CELL FATE, AND NEURON MORPHOLOGY IN THE NEONATAL PIGLET

4.1 Abstract

Respiratory viral infections are common during the neonatal period in humans, but little is known about how early-life infection impacts brain development. The current study used a neonatal piglet model with porcine reproductive and respiratory syndrome virus (PRRSV) to evaluate how chronic neuroinflammation affects hippocampal neurogenesis and neuron morphology. Piglets in the neurogenesis study received one bromodeoxyuridine injection on postnatal day (PD) 7 and then were inoculated with PRRSV. Piglets were sacrificed at PD 28 and the number of number of BrdU+ cells and cell fate were quantified in the dentate gyrus.

PRRSV piglets showed a 24% reduction in the number of newly divided cells forming neurons.

Approximately 15% of newly divided cells formed microglia, but this was not affect by sex or

PRRSV. Additionally, there was a sexual dimorphism of new cell survival in the dentate gyrus where males had more cells than females, and PRRSV infection caused a decreased survival in males only. Golgi impregnation was used to characterize dentate granule cell morphology.

Sholl analysis revealed that PRRSV causes a change in inner granule cell morphology where the first branch point is extended further away from the cell body. Males had more complex dendritic arbors than females in the outer granule cell layer, but this was not affected by

PRRSV. There were no changes to dendritic spine density or morphology distribution. These findings suggest that early-life infection can impact brain development.

4.2 Introduction

Respiratory infections are common in neonates, but little is known about their impact on short- and long-term brain development (Hall, Weinberg et al. 2009). This is important, however, because brain development in the neonatal period is characterized by extensive dendritic growth, synaptogenesis, gliogenesis, and myelination (Huttenlocher 1979; Dietrich,

Bradley et al. 1988; Rice and Barone 2000). Furthermore, although neurogenesis occurs primarily during the prenatal period, the subependyma of the lateral ventricles and granule cell layer of the hippocampal dentate gyrus are two regions where neurogenesis continues postnatal and into adulthood (Ekdahl 2012). During infection, immune-to-brain signaling pathways activate brain microglia, causing increased production of pro-inflammatory cytokines (Dantzer and Kelley 2007). Developing and mature neurons and glia have numerous pro-inflammatory cytokine receptors and are sensitive to inflammatory conditions (Dantzer and Kelley 2007).

Therefore, understanding the impact of early-life infection on brain development is crucial. A number of psychiatric disorders are associated with neuroimmune alterations and are thought to have developmental origins (Boksa 2010; Meyer, Feldon et al. 2011).

Studies in adult animals with a fully mature brain suggest infection-related neuroinflammation inhibits neurogenesis and alters neuron morphology. One brain area that is especially vulnerable to inflammatory insults and that is important for spatial learning and memory is the hippocampus (Elmore, Dilger et al. 2012). Peripheral immune activation with bacterial lipopolysaccharide (LPS) increased the expression of pro-inflammatory cytokines like interleukin (IL)-1β, IL-6, and tumor necrosis factor-alpha (TNF-α) in the brain (Kelley, Bluthé et al. 2003) and inhibited the survival of newborn neurons in the dentate gyrus without impacting cell proliferation (Ekdahl, Claasen et al. 2003). In vitro and in vivo studies also showed that pro- inflammatory cytokines affect neural precursor generation, differentiation, and survival

(Vallieres, Campbell et al. 2002; Cacci, Ajmone-Cat et al. 2008; Wu, Hein et al. 2012; Wu,

Montgomery et al. 2013).

Pro-inflammatory cytokines can further impair synaptic plasticity by inhibiting production of neurotrophins like BDNF, inhibiting long-term potentiation, and altering the architecture of dendrites (Lynch 2002; Milatovic, Zaja-Milatovic et al. 2003; Richwine, Parkin et al. 2008; Tong,

Balazs et al. 2008; Jurgens, Amancherla et al. 2012). Although these and many other studies show that infection and neuroinflammation inhibits neurogenesis and alters neuron morphology in the adult brain, there has been little research on how infection affects neurogenesis or neuron morphology in the critical early postnatal period (Green and Nolan 2014).

Therefore, the goal of this study was to determine the impact of respiratory viral infection in the early postnatal period on hippocampal neurogenesis, cell fate, and neuron morphology in a domestic piglet model. The piglet is an ideal model for this type of investigation because it has a gyrencephalic brain that grows and develops similar to human infants. (Conrad, Dilger et al. 2012). In the present study, piglets were experimentally infected on postnatal day (PD) 7 with porcine reproductive and respiratory syndrome virus (PRRSV) and neurogenesis and neuron morphology were determined with brain tissue collected PD 28. PRRSV activates microglia in the hippocampus of piglets and causes increased pro-inflammatory cytokine production and deficits in hippocampal-dependent learning and memory (Elmore, Burton et al.

2014). Here we show that PRRSV infection impacts new cell survival, cell fate, and granule cell morphology. These findings are the first to show that a live viral infection in the neonatal period alters neurogenesis and neuron morphology.

4.3 Materials and Methods

Animals, Housing, and Feeding

Naturally farrowed crossbred piglets from six separate litters (20 males and 20 females) were obtained from the University of Illinois swine herd. Piglets were brought to the biomedical animal facility on PD 2 (to allow for colostrum consumption from the sow) and placed in individual cages (0.76 m L x 0.58 m W x 0.47 m H) designed for neonatal piglets (Elmore,

Burton et al. 2014). Each cage was positioned in a rack, with stainless steel perforated side walls and clear acrylic front and rear doors within one of two separate but identical disease containment chambers that have been described (Elmore, Burton et al. 2014). Each cage was fitted with flooring designed for neonatal animals (Tenderfoot/NSR, Tandem Products, Inc.,

Colorado Springs, CO, USA). Piglets were maintained on a 12-h light/dark cycle; however, during the dark cycle minimal lighting was provided.

Piglets were fed a commercial sow milk replacer (Advance Liqui-Wean, Milk Specialties

Co., Dundee, IL, USA). Milk was reconstituted daily to a final concentration of 206 g/L using tap water and supplied at a rate of 285 mL/kg BW (based on daily recorded weights) to a stainless steel bowl via a peristaltic pump (Control Company, Friendswood, TX). Using this automated feeding system (similar to that described previously (Dilger and Johnson 2010)), piglets received their daily allotted milk over 18 meals (once per hour). All animal experiments were in accordance with the National Institute of Health Guidelines for the Care and Use of Laboratory

Animals and approved by the University of Illinois at Urbana-Champaign Institutional Animal

Care and Use Committee.

Experimental Design and Treatments

Upon arrival, piglets were assigned to either the control group or the PRRSV infection group based on sex, litter of origin, and body weight. To determine the effect of early-life viral infection on neurogenesis and cell fate, on PD 7, 24 piglets (12 males and 12 females) were injected i.p. with BrdU (50 mg/kg, Sigma, St. Louis, MO, USA) and then inoculated intranasal with either 1mL of 1x10^5 50% tissue culture infected dose (TCID 50) of live PRRSV (strain

P129-BV, obtained from the School of Veterinary Medicine at Purdue University, West

Lafayette, IN, USA) or sterile phosphate buffered saline (PBS). Two PRRSV piglets developed diarrhea and were unable to complete the study. All remaining piglets (control male, n=6; control female, n=6; PRRSV male, n=5; PRRSV female, n=5) were sacrificed at PD 28. To determine the effect of early-life viral infection on hippocampal neuron morphology, a total of 16 piglets (8 males and 8 females) piglets were inoculated with PRRSV or sterile PBS at PD 7.

One PRRSV piglet developed diarrhea and was unable to complete the study. All remaining piglets were sacrificed at PD 31 (control male, n=4; control, female n=4; PRRSV male, n=4;

PRRSV female, n=3).

Assessment of Infection

Daily body weights (kg) were obtained to monitor piglets’ growth. In addition, daily rectal temperatures were obtained starting at PD 7. The willingness of the piglets to consume their first daily meal was determined starting at PD 7 using a feeding score (1 = no attempt to consume the milk; 2 = attempted to consume the milk, but did not finish within 1 min; 3 = consumed all of the milk within 1 min).

98.8% sensitivity and 99.9% specificity, with an S/P ratio of >0.4 indicating a positive sample.

Serum TNF-α levels at the end of the study were determined using porcine-specific sandwich enzyme immunoassays (R&D Systems, Minneapolis, MN, USA).

Perfusions and Tissue Processing

For the neurogenesis and cell fate study, all animals were sacrificed and perfused at PD

28. A telazol:ketamine:xylazine solution was administered intramuscularly at 4.4 mg/kg BW

(50.0 mg of tiletamine, plus 50.0 mg of zolazepam, reconstituted with 2.50 mL ketamine [100 g/L] and 2.50 mL xylazine [100 g/L]; Fort Dodge Animal Health, Fort Dodge, IA, USA). Eyeblink and pain reflexes were tested to confirm deep anesthesia before piglets were perfused transcardially with a phosphate buffer solution (PBS) and a 4% paraformaldehyde/PBS solution.

Brains were extracted and postfixed overnight. The hippocampus was removed and transferred to PBS with 30% sucrose until the tissue sank (~2 days). The entire hippocampus was sectioned using a cryostat into 40 µm sections in an axial plane from dorsal to ventral and collected into a 12-well plate. Six sections were collected into each well with a distance of 480

µm separating each well. Tissues from dorsal region of the hippocampus were used for staining.

BrdU-DAB

The DAB staining was adapted from previously described work (Kohman, DeYoung et al. 2012). Briefly, free floating sections were washed in Tris-buffering solution (TBS) and treated with 0.6% hydrogen peroxide solution for 30 min. Next, sections were placed in 50% de-ionized formamide for 90 min to denature DNA. Sections were then placed in a 10% 20x saline sodium citrate buffer for 15 min, 2N hydrochloric acid for 30 min at 37°C, and then 0.1 M boric acid (pH 8.5) for 10 min. After rinsing, sections were blocked with a solution of 0.3%

Triton-X and 3% goat serum in TBS (TBS-X) for 30 min. The sections were then incubated with the primary antibody rat anti-BrdU (1:200; AbD Serotec, Raleigh, NC, USA) in TBS-X for 72 h at

4°C. Sections were then rinsed with TBS, blocked with TBS-X for 30 min, and then incubated

75 with a biotinylated goat anti-rat secondary antibody (1:250, Jackson ImmunoResearch

Laboratories, West Grove, PA, USA) for 100 min. The ABC system (Vector, Burlingame, CA,

USA) and diaminobenzidine kit (DAB; Sigma, St. Louis, MO, USA) were used for the chromogen.

Immunofluorescence

For the BrdU/NeuN double staining and the BrdU/GFAP/IBA-1 triple staining, a similar procedure was used as with DAB staining. The primary antibodies consisted of rat anti-BrdU

(1:100; AbD Serotec, Raleigh, NC, USA), mouse anti-NeuN (1:50, Millipore, Temecula, CA,

USA), mouse anti-GFAP (1:50, Santa Cruz Biotechnology, Santa Cruz, CA, USA), and rabbit anti-IBA1 (1:1000, Wako Chemicals, Richmond, VA, USA). All secondary fluorescent antibodies (All 1:250, Alexa-488, Cy3, and Alexa-647) were made in goat and incubated for 3 h at room temperature.

BrdU-DAB Image Analysis

Slides containing the DAB-stained tissue were digitized using a Nanozoomer digital pathology system at 20x lens power (Hamamatsu Photonics, Hamamatsu, Japan). The hippocampus of each tissue section was then exported at 10x resolution. Using ImageJ, regions of interest (ROI) were drawn for the suprapyramidal blade, infrapyramidal blade, and hilus of the dentate gyrus and ROI volumes measured. For the suprapyramidal and infrapyramidal blade, the granule cell layer and subgranular zone were included. An overview of the hippocampus with a highlighted view of the suprapyramidal blade can be seen in Figure

4.1. ImageJ was used to automatically count the number of positively labeled cells in all three

ROIs. The data are expressed as BrdU-positive cells per cubic micrometer. Thirty-two sections

(~2400 positive cells) were used to validate the automated methods versus manual hand counting. Correlation analysis validated the automatic method with a slope no different from a

1-to-1 line and a Pearson r value of 0.99. Unbiased estimation was used to correct for cells that could be intersecting with either the top or bottom of the tissue section. An average BrdU- positive nucleus was 6.5 μm in diameter (320 sampled) which is 16.25% of 40 μm BrdU-positive cell counts were multiplied by 0.8375 for unbiased estimation correction.

Immunofluorescence Analysis

A Zeiss LSM 700 confocal microscope (20x objective) was used to acquire z-stack images with a 0.5 micron slice thickness in the dentate gyrus including the granule cell layer and subgranular zone. Images were deconvoluted using Autoquant (Media Cybernetics, Rockville,

MD, USA). For the BrdU/NeuN co-localization, cells from both the suprapyramidal and infrapyramidal blades were acquired but later combined for analysis due to no differences between the two regions. Only cells from the suprapyramidal blade were sampled for the

BrdU/GFAP/IBA-1 triple staining.

Raters were blinded to the treatments and manually counted the BrdU-positive and the number of either NeuN, GFAP, or IBA-1-positive cells that co-localized with BrdU-positive nuclei. Data is expressed as the proportion of BrdU-positive cells that co-localized with another cell marker.

Hippocampal Neuronal Architecture Staining

To determine the effect of early-life viral infection on hippocampal neuron architecture, animals were euthanized at PD 31. A similar dose of TKX (see above) was used for induction of anesthesia and the animals were euthanized by intracardiac injection of sodium pentobarbital

(86.0 mg/kg of B.W.; Fatal Plus®, Vortech Pharmaceuticals, Dearborn, MI, USA). The brain was removed and the left hippocampus was extracted and processed for Golgi-Cox staining as previously described (Richwine, Parkin et al. 2008; Jurgens, Amancherla et al. 2012). Briefly, the left hippocampus was submerged in Golgi-Cox solution for two weeks at which point daily test slices were taken to track neuron filling. Tissues were removed after 4-wks, dehydrated, and embedded in 12% celloiden. The dorsal hippocampus was sectioned at 140 µm and mounted on glass slides. Experimenters responsible for neuron tracing and spine density measurement were blinded to the conditions.

Neuron Selection and tracing

Hippocampal neuron morphology was quantified using a Zeiss Axio Imager A.1 microscope and a computer-based system (Neurolucida; MBF Bioscience, Williston, VT, USA).

NeuroExplorer (MBF Bioscience) was used for visualization and analysis of neuron tracings. An example of a stained dentate granule cell neuron and tracing is shown in Figure 4.2. Dentate granule cell neurons were selected from the suprapyramidal blade and were distinct from other neurons, not truncated, and were well filled. The granule cell neurons were traced at 40X magnification and then segments of the dendrites were captured at 100X magnification for dendritic spine analysis. Previous research has shown that the complexity of granule cells in the inner 2/3 of the granule cell layer (towards hilus) are different than cells in the outer 1/3

(towards molecular layer) (Green and Juraska 1985). Therefore, 5 granule cell neurons from

78 each region were traced and analyzed per pig. After tracing, an estimation of dendritic complexity was determined by calculating the total dendritic length and intersections. Dendritic tree morphology was analyzed using Sholl ring analysis. For the Sholl analysis, 3D concentric spheres with an increasing radius (20 μm increments) were placed around the cell body. The number of intersections of the dendrites and the concentric rings per radial distance from the soma were quantified.

Quantification of Spine Density and Morphology

Spine density measurements were conducted on the same cells quantified for Sholl analysis. For each dentate gyrus granule cell, three dendritic segments were traced. Only 2-5° order branches and dendrites that were 20 µm or greater in length were selected. Each segment was at least 50 µm away from the cell soma. After neuron tracing was completed, dendritic spines were counted using Neurolucida. Spines were counted on both sides of the dendritic segment and classified according to their shapes; either thin, stubby, mushroom, filopodium, or branched (Tashiro and Yuste 2003). The density of spines is expressed as number of spines per micron of dendrite.

Statistical Analysis

Data analysis was conducted using the MIXED procedure of SAS (SAS Institute Inc,

Cary, NC, USA). Sickness measures were analyzed as a two-way (treatment × day) repeated measures ANOVA with trial as a blocking factor. There were no significant differences due to sex, therefore it was not included in the final analysis of sickness symptoms. Neurogenesis data were analyzed as a two-way (treatment × sex) ANOVA. Total dendritic length and total

79 intersections in the inner and outer layer were analyzed separately as a two-way (treatment × sex) ANOVA. Data from the Sholl analysis were analyzed as a three-way (treatment × sex × distance) repeated measures ANOVA with pig as a random effect and nested within treatment and sex. When a main effect or interaction was significant, post hoc Student’s t tests using

Fisher’s least significant differences were used. Statistical significance was set at p < 0.05.

Data are presented as means ± SEM.

4.4 Results

PRRSV Infection and Measures of Sickness

Serum ELISA results indicated all control piglets were negative for PRRSV at the end of the study, whereas all piglets inoculated with PRRSV were positive for PRRSV. Body weight, rectal temperature, and the willingness to consume the first daily meal (feeding score) were determined to provide an indication of the sickness response of piglets infected with PRRSV

(Figure 4.3). Analysis of body weight data showed a significant effect of day (F (30, 962) =

61.98, p < 0.001) and a day × treatment interaction (F (30, 962) = 2.25, p = 0.001), where controls had higher weight gain toward the conclusion of the study. Overall, the effect of treatment on body weight was not significant (F (1, 34) = 1.21). Analysis of feeding score data revealed a significant effect of treatment (F (1, 34) = 34.67, p < 0.001), day (F (24, 734) = 1.99, p = 0.003), and a treatment × day interaction (F (23, 734) = 2.08, p = .002), indicating PRRSV piglets’ motivation to consume the first meal of the day was reduced. Analysis of rectal temperature data showed a significant effect of treatment (F (1, 34) = 80.15), day (F (24, 766) =

7.34), and a treatment × day interaction (F (24, 766) = 3.49 (all, p < 0.001). PRRSV piglets became febrile 3 d after inoculation and remained so during most of the experimental period. At the conclusion of the study, plasma cytokine TNF-α concentration was significantly higher in

PRRSV piglets (209.2 ± 36.3 pg/mL) compared to controls (21.5 ± 4.3 pg/mL) (F (1, 34) = 32.97

80 p < 0.001). Collectively, these data indicate infection with PRRSV in the neonatal period induced clinical signs of illness that persisted throughout the study period.

Hippocampal Cell Proliferation and Survival

BrdU was injected at PD 7, just before PRRSV inoculation, and brain tissue was collected 3-wks later. Therefore, the number of BrdU+ cells represents the basal level of cell division at PD 7, subsequent division of cells labeled at PD 7, and the effects of viral infection on the survival of these labeled cells (Figure 4.4). Analysis of BrdU+ cell density in the suprapyramidal blade revealed a significant effect of sex (F (1, 18) = 8.32, p = 0.01) and a sex × treatment interaction (F (1, 18) = 5.50, p = 0.03), whereby control males had higher density of

BrdU+ cells than control females (24,274 ± 2604 and 11,612 ± 2063 cells/mm3, respectively, p =

0.001) and PRRSV caused a reduction of BrdU+ cells in males only (14,445 ± 1582 cells/mm3, p = 0.01). In the infrapyramidal blade, only the effect of sex was significant (F (1, 17) = 7.68, p

= 0.01). The density of BrdU+ cells in the hilus was also quantified. A significant effect of sex

(F (1, 18) = 8.04, p = 0.01) and a sex × treatment interaction (F (1, 18) = 6.8, p = 0.02) was found, whereby female controls had a lower density of BrdU+ cells than male controls (3,624 ±

678 and 10,691 ± 1786 cells/mm3, respectively, p = 0.001) and whereas PRRSV numerically reduced the number of BrdU+ cells in males, it numerically increased the number of BrdU+ cells in females.

Cell Fate

Immunofluorescence was used to determine the fate of BrdU+ cells in the granule cell layer and subgranular zone (Figure 4.5). To estimate the percentage of Brdu+ cells that

81 developed into mature neurons, the number of BrdU+ cells that co-labeled with NeuN was determined. Two-way ANOVA of the percentage of double-labeled cells revealed a significant effect of treatment (F (1, 18) = 34.42 p < 0.001) where in control piglets more than 80% of the cells labeled with BrdU at PD 7 differentiated into mature neurons by PD 28 but in PRRSV piglets only 57% of the cells labeled with BrdU at PD 7 differentiated into mature neurons. The marked reduction in neurogenesis caused by PRRSV was similar in both males and females.

To estimate the percentage of BrdU+ cells that developed into microglia, the number of BrdU+ cells that co-labeled with IBA-1 was determined. Roughly 15% of the cells labeled with BrdU at

PD 7 differentiated into microglia by PD 28 and this was not influenced by sex (F (1, 18) = 1.05), treatment (F (1, 18) = 0.15), or the sex × treatment interaction (F (1, 18) = 0.21). Tissue sections were also stained with GFAP in an attempt to quantify the number of BrdU+ cells that developed into astrocytes. However, due to the high density of astrocytes within the region analyzed and problems staining two intracellular markers, we were unable to clearly identify double labeled cells (data not shown).

Dendritic Arborization

Total dendritic length and total intersection in the Sholl analysis were used to quantify the overall complexity of the inner and outer granule cell neurons (Figure 4.6). A significant difference between the inner and outer granule cell neurons was evident (F (1, 22) = 9.86, p =

0.0048), therefore the two regions were analyzed separately for the overall complexity. Two- way ANOVA of the total dendritic length and the number of intersections of the inner dentate granule cell neurons showed neither an effect of sex, treatment or a sex × treatment interaction

(all p > 0.05). However, there was a significant effect of sex on dendritic length (F (1, 11) =

4.95, p = 0.048) and number of intersections (F (1, 11) = 5.99, p = 0.0328) for neurons in the

82 outer layer of the dentate gyrus where males had longer more complex dendrites than females.

A similar trend was found when analyzing dendritic length and intersections within each Sholl interval. A repeated-measures ANOVA showed a significant effect of sex (F (1, 11) = 5.95, p =

0.0328) and distance from the soma (F (16, 176) = 55.02, p < 0.0001) on intersections for the outer dentate granule cell neurons. Males (2.11 ± 0.11) had more intersections than females

(1.72 ± 0.12). Interestingly, analysis of the inner dentate granule cell neurons revealed a significant effect of distance (17, 187) = 25.29, p < 0.0001) and a distance × treatment interaction (F (17, 187) = 2.26, p = 0.0041). The altered dendritic arborization was due to shifting of the initial branching points away from the cell soma (within 40-140 μm) in the PRRSV piglets (p < 0.05).

Spine Density

For the inner and outer portion of the granule cell layer spines were counted on both sides of the dendritic segment and classified according to their shape: thin, stubby, mushroom, filopodium, or branched (Figure 4.7 and 4.8) Neither spine density or classification were affected in either the inner or outer granule cell layers by sex, treatment, or the sex × treatment interaction (all p > 0.05).

4.5 Discussion

Respiratory infections during early-life are common, but knowledge of their impact on brain development is lacking. Due to ethical considerations and the difficulty of conducting experiments in human neonates, progress in this area has been slow. Furthermore, the translation of data from rodent neurodevelopmental models to human infants is difficult due to

83 substantial differences in brain morphology and development. To overcome several obstacles, the current study was conducted using a highly tractable and translational piglet model.

Previously, we have characterized the normal brain growth of the domestic pig using magnetic resonance imaging (Conrad, Dilger et al. 2012). This study showed that piglet brain growth is similar to human neonates and that from birth to 4-wks of age, total brain volume increased 45% (Conrad, Dilger et al. 2012). Therefore, in the present study inoculation with

PRRSV and subsequent infection occurred in a period of substantial brain growth. PRRSV is an enveloped, positive-sense, single-stranded RNA virus that belongs to the Arteriviridae family within the order Nidovirales (Done, Paton et al. 1996). In young pigs PRRSV primarily infects and replicates in cells of the monocyte/macrophage lineage (Duan, Nauwynck et al. 1997).

Once infected, macrophages can migrate to lymphatic tissue and enter circulation allowing for the spread of PRRSV (Shin and Molitor 2002). PRRSV can activate microglia and cause production of pro-inflammatory cytokines in the brain either through immune-to-brain signaling pathways or by entering the CNS. In a recent study, several pro-inflammatory cytokines were elevated in serum 20 d after inoculation with PRRSV, and a number of pro-inflammatory genes, including interferon-γ, TNF-α, and IL-1β, were up regulated in several brain regions at the same time post inoculation (Elmore, Burton et al. 2014). Furthermore, the percentage of activated microglia, as indicated by expression of MHC class II, was markedly increased in piglets with

PRRSV infection (Elmore, Burton et al. 2014). Microglial activation was positively correlated with fever, and negatively correlated with food motivation and learning and memory (Elmore,

Burton et al. 2014). In the present study, piglets exhibited a sustained febrile response and increased circulating TNF-α at the conclusion of the experiment. This strongly suggests that

PRRSV piglets had a sustained neuroinflammatory response throughout the study period.

Both in vivo and in vitro studies have shown that inflammation can lead to altered cell fate for newly divided cells. IL-1β causes a switch from neuronal to glial differentiation in vitro

(Green, Treacy et al. 2012). IL-1β can also inhibit the proliferation of neural progenitor cells and proliferation of newly born neurons (Green and Nolan 2014). Similar to the in vitro data, prenatal and early postnatal immune activation in rodents causes a reduction in neuronal differentiation (Bland, Beckley et al. 2010; Graciarena, Depino et al. 2010). Both increases and decreases in gliogenesis have been reported and may be time and insult dependent

(Ratnayake, Quinn et al. 2012; Jarlestedt, Naylor et al. 2013). Here PRRSV infection was found to reduce the number of new cells differentiating into neurons by approximately 25% in both males and females. New neurons in the dentate have been shown to be necessary for hippocampal function including pattern separation and learning and memory, and disruptions could lead to cognitive deficits (Deng, Aimone et al. 2010; Villeda, Luo et al. 2011). The number of newly divided microglia cells was consistent across both sex and treatment at 15% of BrdU+ cells. There was a very high density of microglia in the subgranular zone and within the granule cell layer, but the absolute numbers were not quantified. We also stained for astrocytes using

GFAP, but were unable to clearly identify BrdU+/GFAP+ cells. Astrocytes densely populated the subgranular zone and only had projections into the granule cell layer, but no cell bodies.

The astrocytes were so dense that even with capturing z-stack image sets it was difficult to positively identify a BrdU+/GFAP+ cell and not rule out that two cells were in close proximity.

Nonetheless, , the data show that in healthy control piglets 80% of BrdU+ cells develope into neurons, and 15% into microglia. This leaves only 5% undetermined. In piglets infected with

PRRSV, however, only 55% of the BrdU+ cells develop into neurons, and 15% into microglia.

This leaves 30% of the BrdU+ cells unidentified. The undetermined cells could be astrocytes, oligodendrocytes, or undifferentiated cells. Further, characterization is needed.

Research with adult rodents has shown that peripheral immune activation can reduce the survival of new neurons in the dentate gyrus (Ekdahl, Claasen et al. 2003). The majority of neurogenesis studies in adult rodents have used males only. As many developmental disorders have a higher incidence in one sex or the other, it is important to study both males and females.

Here we find that there is a sexual dimorphism in the number of surviving BrdU+ cells in the dentate gyrus with males having more surviving cells than females. This dimorphism is also seen in the dentate gyrus and CA1 region of the neonatal rat (Zhang, Konkle et al. 2008;

Bowers, Waddell et al. 2010). PRRSV infection causes a significant reduction of surviving cells in males, but does not affect females. This suggests that either females are not as susceptible to inflammation or there is a protective mechanism to combat these signals. Alternatively, at postnatal day 4, male rats have significantly more microglia in the dentate gyrus than females which could lead to sex differences in the inflammatory response (Schwarz, Sholar et al. 2012).

Studies in rodents have used both acute inflammatory stimuli, such as LPS, or systemic infection with Escherichia coli in male mice during the early postnatal period. (Bland, Beckley et al. 2010; Jarlestedt, Naylor et al. 2013) These studies show reductions in new cell survival in the dentate gyrus, consistent with what we found in the male piglets.

In addition to neurogenesis, changes in hippocampal neuron morphology were assessed. Dentate granule cells were characterized as they have been shown to be sensitive to inflammatory insult (Jurgens, Amancherla et al. 2012). The soma location within the granule cell layer may be associated with “age” of the neuron as new cells are born in the subgranular zone and migrate into the inner granule cell layer (Mongiat and Schinder 2011). Although we didn’t specifically test for cell age, the microenvironment may be different for the inner granule cell neurons and inflammation may affect them differently (Mongiat and Schinder 2011).

PRRSV infection caused a shifting in the shape of the inner granule cell dendritic tree, extending the primary dendrite length before the first branching points. These results are dissimilar to adult influenza studies which found that inflammation in adulthood caused a retraction of dendrites (Jurgens, Amancherla et al. 2012). The reason for the extension of the primary dendrite is not clear. The highest density of microglia was found in the subgranular zone, so it may be that the inner granule layer cells are exposed to a higher pro-inflammatory cytokine load. Additionally, studies have found that size of the dentate increases with prenatal

86 inflammation, so it is possible that the primary dendrite must travel through more cells before starting to branch (Golan, Lev et al. 2005).

Additionally, results showed that there was a sexual dimorphism in the complexity of outer dentate granule cell neurons, but the shape of the dendritic tree was similar among males and females. This difference is similar to rodents (Juraska, Fitch et al. 1985). We also found that the outer dentate granule cell neurons were more complex than the inner neurons, also similar to rodents (Green and Juraska 1985). Although no differences were found in dendritic spine density, the distribution of spine morphology is noteworthy. During early development, stubby spines are dominant with some filipodia (Nimchinsky, Sabatini et al. 2002). As synaptic connections are made and strengthened, the spine takes on a more mature mushroom morphology (Bourne and Harris 2008). Our data illustrate that the most abundant morphology was stubby and mushroom spines, a similar trend to the postnatal distribution in rodents (Harris,

Jensen et al. 1992).

In conclusion, inflammation during early-life has the potential to cause short- and long- term disruptions in brain development. Our data indicate that neonatal respiratory infection reduces cell survival, changes cell fate, and alters hippocampal cell morphology. The conclusions of this study are limited as changes were only quantified at one end point. The

PRRSV piglets were almost symptomatically recovered at this time. Tracking changes during peak sickness and at long-term endpoints would be useful. Additional work is needed to characterize the unidentified BrdU+ cells. Regardless of these limitations, we show that early- life respiratory infection can impact brain development. Further work with this model will allow for testing of therapeutic strategies to modulate the neuroimmune response with aims of preventing adverse developmental outcomes.

4.6 Figures

Figure 4.1. The representative section shows an overview of the hippocampus (top) with a magnified view of the suprapyramidal blade of the dentate gyrus (bottom). BrdU+ cells are indicated by DAB staining. Magnification = 2.5X (top), 20X (bottom); scale bar, 100 μm.

Figure 4.2. A-B, Representative example of a Golgi-stained dentate gyrus neuron (A) with corresponding tracing and Sholl analysis (B). Magnification = 40X; scale bar 100 μm.

A B 10 41 * * * 8 * 40 * C) * *

 * * * * * * * * * ** 6 * * * * * * * * * 39 *

4 Weight (kg)

Temperature( 38 2 Control PRRSV 0 37

2 3 4 5 6 7 8 9 0 1 4 5 7 8 1 4 5 7 8 1 2 7 8 9 2 1 1 12 13 1 1 16 1 1 19 20 2 22 23 2 2 26 2 2 29 30 3 3 10 11 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 Day of Age Day of Age C 3.5

3.0

2.5 * * * * * * * * * * Feeding Score Feeding 2.0 *

1.5 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Day of Age

Figure 4.3. A-C, Daily recorded sickness measures. Body weight was recorded from the start of the experiment (A) and temperature (B) and feeding score (C) were measured starting prior to inoculation at 7 days of age. Data presented are means ± SEM (*p < 0.05 compared to controls).

A B 30 30 a ) )

3 a Control 3 Control 25 PRRSV 25 PRRSV (x10 (x10 3 3 a 20 20 b b b b 15 b 15

10 10

BrdU+ / mm cells 5 BrdU+ / mm cells 5 Male Female Male Female C 15 ) 3 a Control 12 PRRSV (x10 3 9 ab ab

6 b

BrdU+ cells / mm 0 Male Female

Figure 4.4. A-C, Density of newly divided (BrdU+) cells in the suprapyramidal blade (A), infrapyramidal blade (B), and hilus (C) of the denate gyrus. Groups are separated by sex and treatment. Letters indicate the groups that were significantly different (p < 0.05).

Control Control C PRRSV D PRRSV 100 25 a a 80 20 b b 60 15

40 10

20 5 % BrdU-IBA1/BrdU % BrdU-NeuN/BrdU 0 0 Male Female Male Female

Figure 4.5. A-D, Representative sections showing double labeling (A) of antibodies against BrdU (new cell; red) and NeuN (mature neuron; green). Representative section showing triple labeling (B) with antibodies against BrdU (new cell; blue), Iba-1 (macrophage/microglia; red), and GFAP (astrocyte; green). The granule cell layer is located in the upper half of each photomicrograph with the hilar region located in the bottom half. The proportion of newly divided cells that express NeuN (C) and IBA-1 (D) are plotted by sex and treatment. Means ± SEM are plotted with letters indicating differences between groups (p < 0.05). Magnification = 20X; scale bar = 50 μm.

A Male B Male Female Female 1000 50 m)  800 * 40 600 30 *

400 20

200 10 Total Intersections

0 0 Total Dendritic Length ( Inner DG Outer DG Inner DG Outer DG C D 6 Male 5 Control Female 4 * * PRRSV 4 * 3 * Outer DG Inner DG 2 2 # of Intersections of #

# of Intersections of # 1

0 0 0 0 2 20 100 200 300 100 20 300 Distance From the Soma ( m) Distance From the Soma ( m)

Figure 4.6. A-D, Females had reduced total dendritic length (A) and total intersections (B) in the outer dentate granule (DG) cell layer. Sholl analysis revealed that females had fewer intersections in the outer DG (C) and that PRRSV causes shifting of the first branching points away from the soma in the inner DG (D). Data are presented as means ± SEM (*p < 0.05 compared to controls).

1.0 Control 0.8 PRRSV m

 0.6

0.4 Spines/ 0.2

0.0 Inner Outer

Figure 4.7. PRRSV infection did not alter spine density of dendritic granule cells located in the inner or outer DG. Data are represented as means ± SEM.

0.6

m 0.4 

Spines/ 0.2

0.0 Thin Stubby Mushroom Filopodium Branched B 0.6

m 0.4 

Control

Spines/ 0.2 PRRSV

0.0 Thin Stubby Mushroom Filopodium Branched

Figure 4.8. A-B. PRRSV infection did not alter the distribution of spine morphology in the inner DG (A) or outer DG (B). Data are represented as means ± SEM.

4.7 References

Bland, S. T., J. T. Beckley, et al. (2010). "Enduring consequences of early-life infection on glial and neural cell genesis within cognitive regions of the brain." Brain Behav Immun 24(3): 329-338.

Boksa, P. (2010). "Effects of prenatal infection on brain development and behavior: A review of findings from animal models." Brain, Behavior, and Immunity 24(6): 881-897.

Bourne, J. N. and K. M. Harris (2008). "Balancing structure and function at hippocampal dendritic spines." Annu Rev Neurosci 31: 47-67.

Bowers, J. M., J. Waddell, et al. (2010). "A developmental sex difference in hippocampal neurogenesis is mediated by endogenous oestradiol." Biol Sex Differ 1(1): 8.

Cacci, E., M. A. Ajmone-Cat, et al. (2008). "In vitro neuronal and glial differentiation from embryonic or adult neural precursor cells are differently affected by chronic or acute activation of microglia." Glia 56(4): 412-425.

Conrad, M. S., R. N. Dilger, et al. (2012). "Brain Growth of the Domestic Pig (Sus scrofa) from 2 to 24 Weeks of Age: A Longitudinal MRI Study." Dev Neurosci 34(4): 291-298.

Dantzer, R. and K. W. Kelley (2007). "Twenty years of research on cytokine-induced sickness behavior." Brain, Behavior, and Immunity 21(2): 153-160.

Deng, W., J. B. Aimone, et al. (2010). "New neurons and new memories: how does adult hippocampal neurogenesis affect learning and memory?" Nat Rev Neurosci 11(5): 339- 350.

Dietrich, R. B., W. G. Bradley, et al. (1988). "MR evaluation of early myelination patterns in normal and developmentally delayed infants." American Journal of Roentgenology 150(4): 889-896.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Done, S. H., D. J. Paton, et al. (1996). "Porcine reproductive and respiratory syndrome (PRRS): a review, with emphasis on pathological, virological and diagnostic aspects." Br Vet J 152(2): 153-174.

Ekdahl, C. T. (2012). "Microglial activation - tuning and pruning adult neurogenesis." Frontiers in Pharmacology 3.

Ekdahl, C. T., J.-H. Claasen, et al. (2003). "Inflammation is detrimental for neurogenesis in adult brain." Proceedings of the National Academy of Sciences 100(23): 13632-13637.

Elmore, M. R., R. N. Dilger, et al. (2012). "Place and direction learning in a spatial T-maze task by neonatal piglets." Anim Cogn 15(4): 667-676.

Golan, H. M., V. Lev, et al. (2005). "Specific neurodevelopmental damage in mice offspring following maternal inflammation during pregnancy." Neuropharmacology 48(6): 903-917.

Graciarena, M., A. M. Depino, et al. (2010). "Prenatal inflammation impairs adult neurogenesis and memory related behavior through persistent hippocampal TGFβ1 downregulation." Brain, Behavior, and Immunity 24(8): 1301-1309.

Green, E. J. and J. M. Juraska (1985). "The dendritic morphology of hippocampal dentate granule cells varies with their position in the granule cell layer: a quantitative Golgi study." Experimental Brain Research 59(3): 582-586.

Green, H. F. and Y. M. Nolan (2014). "Inflammation and the developing brain: consequences for hippocampal neurogenesis and behavior." Neurosci Biobehav Rev.

Green, H. F., E. Treacy, et al. (2012). "A role for interleukin-1beta in determining the lineage fate of embryonic rat hippocampal neural precursor cells." Mol Cell Neurosci 49(3): 311-321.

Hall, C. B., G. A. Weinberg, et al. (2009). "The Burden of Respiratory Syncytial Virus Infection in Young Children." New England Journal of Medicine 360(6): 588-598.

Harris, K. M., F. E. Jensen, et al. (1992). "Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic physiology and long-term potentiation." J Neurosci 12(7): 2685-2705.

Huttenlocher, P. R. (1979). "Synaptic density in human frontal cortex - developmental changes and effects of aging." Brain Res 163(2): 195-205.

Jarlestedt, K., A. S. Naylor, et al. (2013). "Decreased survival of newborn neurons in the dorsal hippocampus after neonatal LPS exposure in mice." Neuroscience 253(0): 21-28.

Juraska, J. M., J. M. Fitch, et al. (1985). "Sex differences in the dendritic branching of dentate granule cells following differential experience." Brain Research 333(1): 73-80.

Jurgens, H. A., K. Amancherla, et al. (2012). "Influenza Infection Induces Neuroinflammation, Alters Hippocampal Neuron Morphology, and Impairs Cognition in Adult Mice." The Journal of Neuroscience 32(12): 3958-3968.

Kelley, K. W., R.-M. Bluthé, et al. (2003). "Cytokine-induced sickness behavior." Brain, Behavior, and Immunity 17(1, Supplement 1): 112-118.

Kohman, R. A., E. K. DeYoung, et al. (2012). "Wheel running attenuates microglia proliferation and increases expression of a proneurogenic phenotype in the hippocampus of aged mice." Brain, Behavior, and Immunity(0).

Lynch, M. A. (2002). "Interleukin-1 beta exerts a myriad of effects in the brain and in particular in the hippocampus: analysis of some of these actions." Vitam Horm 64: 185-219.

Meyer, U., J. Feldon, et al. (2011). "Schizophrenia and autism: both shared and disorder- specific pathogenesis via perinatal inflammation?" Pediatr Res 69(5 Pt 2): 26R-33R.

Milatovic, D., S. Zaja-Milatovic, et al. (2003). "Pharmacologic suppression of neuronal oxidative damage and dendritic degeneration following direct activation of glial innate immunity in mouse cerebrum." J Neurochem 87(6): 1518-1526.

Mongiat, L. A. and A. F. Schinder (2011). "Adult neurogenesis and the plasticity of the dentate gyrus network." Eur J Neurosci 33(6): 1055-1061.

Nimchinsky, E. A., B. L. Sabatini, et al. (2002). "STRUCTURE AND FUNCTION OF DENDRITIC SPINES." Annual Review of Physiology 64(1): 313-353.

Ratnayake, U., T. A. Quinn, et al. (2012). "Behaviour and hippocampus-specific changes in spiny mouse neonates after treatment of the mother with the viral-mimetic Poly I:C at mid-pregnancy." Brain Behav Immun 26(8): 1288-1299.

Rice, D. and S. Barone, Jr. (2000). "Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models." Environ Health Perspect 108 Suppl 3: 511-533.

Richwine, A. F., A. O. Parkin, et al. (2008). "Architectural changes to CA1 pyramidal neurons in adult and aged mice after peripheral immune stimulation." Psychoneuroendocrinology 33(10): 1369-1377.

Schwarz, J. M., P. W. Sholar, et al. (2012). "Sex differences in microglial colonization of the developing rat brain." Journal of Neurochemistry 120(6): 948-963.

Shin, J. H. and T. W. Molitor (2002). "Localization of porcine reproductive and respiratory syndrome virus infection in boars by in situ riboprobe hybridization." J Vet Sci 3(2): 87- 96.

Tashiro, A. and R. Yuste (2003). "Structure and molecular organization of dendritic spines." Histol Histopathol 18(2): 617-634.

Tong, L., R. Balazs, et al. (2008). "Interleukin-1 beta impairs brain derived neurotrophic factor- induced signal transduction." Neurobiol Aging 29(9): 1380-1393.

Vallieres, L., I. L. Campbell, et al. (2002). "Reduced hippocampal neurogenesis in adult transgenic mice with chronic astrocytic production of interleukin-6." J Neurosci 22(2): 486-492.

Villeda, S. A., J. Luo, et al. (2011). "The ageing systemic milieu negatively regulates neurogenesis and cognitive function." Nature 477(7362): 90-94.

Wu, M. D., A. M. Hein, et al. (2012). "Adult murine hippocampal neurogenesis is inhibited by sustained IL-1beta and not rescued by voluntary running." Brain Behav Immun 26(2): 292-300.

Wu, M. D., S. L. Montgomery, et al. (2013). "Sustained IL-1β expression impairs adult hippocampal neurogenesis independent of IL-1 signaling in nestin+ neural precursor cells." Brain, Behavior, and Immunity 32(0): 9-18.

Zhang, J. M., A. T. Konkle, et al. (2008). "Impact of sex and hormones on new cells in the developing rat hippocampus: a novel source of sex dimorphism?" Eur J Neurosci 27(4): 791-800.

CHAPTER 5

MINOCYCLINE ADMINISTRATION DOES NOT ATTENUATE MICROGLIA ACTIVATION AND NEUROINFLAMMATION INDUCED BY PORCINE REPRODUCTIVE AND RESPIRATORY SYNDROME VIRUS

5.1 Abstract

Microglia cells are the resident macrophages of the brain and are important for their immunological response as well as maintaining normal tissue homeostasis. Through immune- to-brain communication, microglia provide a “sixth sense,” monitoring the immunological state of the periphery and allow for the propagation and amplification of inflammatory signals to the central nervous system. Activation of microglia during the neonatal period has the potential to disrupt many developmental processes. Porcine reproductive and respiratory syndrome virus

(PRRSV) is a respiratory viral infection which causes microglia activation and neuroinflammation in piglets. Using a neonatal piglet model with PRRSV infection, we tested the efficacy of using a chronically administered, high dose of minocycline to prevent neuroinflammation. Minocycline is a second generation tricyclic antibiotic that has been shown to prevent microglia activation and reduce neuroinflammation in vitro and in rodent models.

Minocycline did not rescue weight gain changes or prevent febrile response in the PRRSV piglets and caused a reduction in feeding score in the PRRSV animals. Minocycline treatment failed to reduce microglia activation in the hippocampus of PRRSV piglets and induced activation in control piglets. Additionally, minocycline treatment caused a significant increase in

IL-1β and a trend towards increased TNF-α in the hippocampus. PRRSV piglets also had a significant reduction in BDNF. Overall, these data indicate that high dose minocycline does not attenuate neuroinflammation due to PRRSV. The high dose, commonly used in rodent models, may itself lead to neuroinflammation in the neonatal period possibly due to increased intracranial hypertension or bilirubin-induced brain damage.

100

5.2 Introduction

Microglia have been shown to have a multifaceted role in the central nervous system including maintenance of normal tissue homeostasis, response to inflammation and pathogens, and normal brain development (Harry 2013). Microglia are able to “sense” the inflammatory state of the periphery through immune-to-brain communication and become activated in response to peripheral inflammation (Dantzer, Konsman et al. 2000). This bi-directional communication allows modification of physiology and behavior in response to injury or infection.

Upon activation, microglia can create a neuroinflammatory environment through production of prostaglandins, free radicals, and pro-inflammatory cytokines (Perry and Gordon 1988; Gordon

2003). Much work has been done to characterize the impact of inflammation on brain function and behavior in adults, but much less is known about how it can affect brain growth and development. Immune activation during the neonatal time period may disrupt synaptogenesis, dendritic growth, gliogenesis, and myelination, which are all sensitive to inflammation (Rice and

Barone 2000; Richwine, Parkin et al. 2008; Schmitz and Chew 2008; Bitzer-Quintero and

Gonzalez-Burgos 2012).

Porcine reproductive and respiratory syndrome virus (PRRSV) is an enveloped, positive- sense, single-stranded RNA virus that belongs to the Arteriviridae family within the order

Nidovirales (Done, Paton et al. 1996). PRRSV causes interstitial pneumonia in piglets by infecting the mononuclear myeloid cells in the lungs and activates microglia either through immune-to-brain signaling pathways or by entering the brain (Done, Paton et al. 1996). In a recent study, PRRSV infection caused elevated expression of pro-inflammatory genes, including interferon-γ, TNF-α, and IL-1β, in several brain regions and caused elevated peripheral pro- inflammatory cytokines 20 d after inoculation with PRRSV (Elmore, Burton et al. 2014).

Furthermore, 82% and 43% of microglia cells were activated (MHC II+) 13 and 20 d after inoculation (Elmore, Burton et al. 2014). Microglial activation was positively correlated with

101 fever, and negatively correlated with learning and memory and food motivation (Elmore, Burton et al. 2014).

We previously found PRRSV infection in the neonatal piglet to have profound effects on brain development. Using an MRI-based approach, PRRSV infection caused reductions in grey and white matter in multiple brain regions including the primary visual cortex (Chapter 2).

PRRSV infection may have caused white matter disruptions in the corpus callosum and also reduced metabolite concentration in the hippocampus. In a separate but similar study, we found

PRRSV infection to inhibit neurogenesis by ~25% and alter dendritic architecture of granule cell neurons. Collectively, these findings suggest a close relationship between infection, microglia activation, neuroinflammation, and neurodevelopment. As alterations in neuroimmune activity have been proposed to underlie psychiatric disorders including autism and schizophrenia, it might be useful to inhibit microglial cells during infection at critical stages of development

(Meyer, Feldon et al. 2011).

Minocycline is a second-generation tetracycline antibiotic that has been shown to have neuroprotective effects (Kim and Suh 2009). Minocycline is able to cross the blood-brain barrier where it exerts both an anti-inflammatory and anti-apoptotic action. These effects are mediated by reductions in microglia activation through a p38 MAPK-dependent mechanism

(Tikka and Koistinaho 2001). Minocycline has shown both protective and deleterious results as a therapeutic in rodent models of hypoxia-ischemia, multiple sclerosis, Parkinson’s disease, and

Alzheimer’s disease (Yrjanheikki, Keinanen et al. 1998; Yang, Sugama et al. 2003; Casarejos,

Menendez et al. 2006; Choi, Kim et al. 2007; Chen, Pi et al. 2010). Multiple human clinical trials have also been conducted with minocycline, with the results showing concern about the translatability of rodent studies to humans (Blum, Chtarto et al. 2004; Kim and Suh 2009). To date, there is a significant gap of evaluation of minocycline treatment in larger animal models which may be more predictive of human outcomes.

102

Therefore, the goal of the current study was to evaluate chronic minocycline administration as a potential therapeutic for neuroinflammatory changes due to PRRSV infections. Changes in sickness symptoms, microglia activation, and hippocampal gene expression are evaluated.

5.3 Materials and Methods

Animals, Housing, and Feeding

Naturally farrowed, crossbred piglets from three separate litters (12 females and 12 males) were obtained from the University of Illinois swine herd. Piglets were brought to the biomedical animal facility on PD 2 (to allow for colostrum consumption from the sow) and placed in individual cages (0.76 m L × 0.58 m W × 0.47 m H) designed for neonatal piglets (Elmore,

Colorado Springs, CO, USA). Piglets were maintained on a 12-h light/dark cycle; however, during the dark cycle minimal lighting was provided.

Piglets were fed a commercial sow milk replacer (Advance Liqui-Wean, Milk Specialties

103 feeding system (similar to that described previously (Dilger and Johnson 2010)), piglets received their daily allotted milk over 18 meals (once per hour). All animal experiments were in accordance with the National Institute of Health Guidelines for the Care and Use of Laboratory

Animals and approved by the University of Illinois at Urbana-Champaign Institutional Animal

Care and Use Committee.

Experimental Design and Treatments

Upon arrival, piglets were assigned to groups based on sex, litter of origin, and body weight. A 2 × 2 factorial arrangement with the variables of ± PRRSV and ± minocycline was used with three males and three females per group. At PD 7, 24 piglets (12 males and 12 females) were inoculated intranasal with either 1mL of 1x10^5 50% tissue culture infected dose

(TCID 50) of live PRRSV (strain P129-BV, obtained from the School of Veterinary Medicine at

Purdue University, West Lafayette, IN, USA) or sterile phosphate buffered saline (PBS).

Piglets in the minocycline group were administered one 100 mg capsule of minocycline hydrochloride orally every day starting at PD 7 (Ranbaxy Pharmaceuticals, Jacksonville, FL,

USA). Minocycline was administered 2 hours prior to the first feeding using a cat piller. The cat piller allows for non-invasive administration of capsules in larger animals. Controls were subjected to similar stress of capsule administration using the cat piller. Six male piglets were removed from the study due to bacterial infection or weight loss due to diarrhea. All remaining piglets were sacrificed at PD 28 [CON/CON: n = 5 (2M, 3F); CON/MINO: n = 5 (2M, 3F);

PRRSV/CON: n = 3 (0M, 3F); PRRSV/MINO: n = 5 (2M, 3F)].

104

Assessment of Infection

Daily body weights (kg) were obtained to monitor piglets’ growth. In addition, daily rectal temperatures were obtained PD 7 through PD 28. The willingness of the piglets to consume their first daily meal was determined from PD 7 to PD 28 using a feeding score (1 = no attempt to consume the milk; 2 = attempted to consume the milk, but did not finish within 1 min; 3 = consumed all of the milk within 1 min).

99.9% specificity, with an S/P ratio of >0.4 indicating a positive sample.

Euthanasia and Tissue Collection

All piglets were euthanized at PD 28 by first anesthetizing using a telazol:ketamine:xylazine solution (50.0 mg of tiletamine, plus 50.0 mg of zolazepam, reconstituted with 2.50 mL ketamine [100 g/L] and 2.50 mL xylazine [100 g/L]; Fort Dodge

Animal Health, Fort Dodge, IA, USA) administered at 4.4mg/kg BW. Piglets were euthanized via intracardiac (i.c.) administration of sodium pentobarbital (86.0 mg/kg of BW; Fatal Plus®,

Vortech Pharmaceuticals, Dearborn, MI, USA). Brains were extracted, weighed, and dissected for tissue collection. Brain tissues from the left hemisphere were collected, rinsed in PBS, and stored at -80ºC in RNA later (Ambion, Carlsbad, CA, USA) until processing. The right hippocampus was used for microglia isolation.

105

Microglia Isolation and Flow Cytometry

To assess the activational state, microglia were isolated and then characterized using flow cytometry as previously reported (Elmore, Burton et al. 2014). Briefly, the right hippocampus was collected from each piglet and rinsed in PBS. Tissues were homogenized in

DPBS (Sigma, St. Louis, MO, USA) with 0.2% glucose by pushing them through a 70 μm nylon mesh strainer (BD Biosciences, San Jose, CA, USA). The cells were spun at 600 x g for 6 min and the supernatant removed and resuspended with 70% isotonic Percoll (GE Healthcare,

Waukesha, WI, USA). A discontinuous gradient with 50%, 35%, and 0% isotonic Percoll was overlaid and spun for 20 min at 2000 x g. The microglia were collected from the interface of the

50% and 70% layers and spun down to form a pellet. The pellet was resuspended in flow buffer

(PBS with 1% bovine serum albumin, Fisher, Waltham, MA, USA), 0.1% sodium azide (Sigma), and 20 mM glucose (Sigma). Fc receptors were blocked with purified CD16/CD32 antibodies

(eBiosciences, CA, USA). The microglia were incubated with CD11b (Biolegend, San Diego,

Ca, USA), CD45 (AbD Serotec, Raleigh, NC, USA), and MHC II (Antibodies Online, Atlanta, GA,

USA). Using a LSR II Flow Cytometer (BD Biosciences), microglia were gated based on size, granularity, and side scatter, as well as being CD11b+/CD45Int. Activated microglia were characterized as being MCH II-positive.

Real Time PCR

RNA was extracted from approximately 50 mg of the left hippocampus of each piglet using the Tri Reagent protocol (Sigma). cDNA synthesis was conducted with a QuanTect

Reverse Transcription Kit (Qiagen, Valencia, CA, USA) with removal of genomic DNA. Total

RNA quantity and purity was assessed with a NanoDrop system (Thermo Scientific, Wilmington,

DE, USA). Sequences for porcine BDNF, IL-10, IL-1β, IL-6, TGF-β, and TNF-α have been

106 previously reported and ribosomal protein L 19 (RPL19) was used as the housekeeping gene

(Elmore, Burton et al. 2014). Real-time PCR was conducted according to the Life Technologies

(Grand Island, NY, USA) Taqman Assay on Demand Gene Expression protocol. All PCRs were plated in triplicate on a 384-well RT-PCR plate. An ABI PRISM 7900HT system (Life

Technologies) was used with standard cycle parameters. Data was analyzed using the comparative threshold cycle (Ct) method (Livak and Schmittgen 2001) and are expressed as relative changes in mRNA expression compared to controls.

Statistical Analysis

Data were analyzed using the MIXED procedure of SAS (SAS Institute, Cary, NC, USA).

For the sickness symptom measures, a three-way (PRRSV × Minocycline × Day) repeated measures ANOVA with pig as the random effect was used. All other measures were analyzed using a two-way (PRRSV × Minocycline) ANOVA. Means were compared using the least significant difference (LSD) option in SAS and are presented as means ± SEM.

5.4 Results

Sickness Measures

Serum ELISA results indicated all control piglets were negative for PRRSV at the end of the study, whereas all piglets inoculated were positive for PRRSV. Sickness responses for the piglets with PRRSV and minocycline were tracked by measuring body weight gain, temperature, and feeding scores (Figure 5.1). There was a significant main effect of day for body weight (F

(26, 364) = 20.29, p < 0.001) indicating that all pigs increased weight of the study period. There was also a significant day × PRRSV interaction (F (26, 364) = 2.24, p < 0.001) showing that

107

PRRSV pigs had reduced weight gain compared to controls. There was no impact of minocycline treatment on weight gain (F (1, 14) = 0). Rectal temperatures showed a significant main effect of PRRSV (F (1, 14) = 11.46), p = 0.004) where PRRSV piglets had increased temperatures (39.2 ± 0.2 °C) compared to controls (38.3 ± 0.2 °C). Feeding scores had a main effect of day (F (20, 280) = 1.76, p = 0.02), minocycline (F (1, 14) = 9.05, p = 0.009), PRRSV (F

(1, 14) = 45.81, p < 0.001). In addition there were significant day × minocycline (F (20, 280) =

1.62, p = 0.05), day × PRRSV (F (20, 280) = 1.76, p = 0.02), minocycline × PRRSV (F (1, 14) =

9.05, p = 0.009), and a day × minocycline × PRRSV interaction (F (20, 280) = 1.62, p = 0.05).

The PRRSV animals had reduced feeding scores compared to controls, but this differed in timing based on minocycline treatment. The PRRSV/CON piglets had reduced feeding scores during the first week post-inoculation while the PRRSV/MINO piglets displayed a reduced feeding score during the last two weeks of the study. Collectively, these data indicate infection with PRRSV in the neonatal period induced clinical signs of illness.

Brain Weights

The whole excised brain weights were measured (Figure 5.2). There was a main effect of PRRSV (F (1, 14) = 9.22, p = 0.009) where the PRRSV piglets had reduced brain weight

(42.1 ± 1.5 g) compared to controls (48.2 ± 1.3 g). There was also a trend for minocycline treatment (F (1, 14) = 4.35, p = 0.06). When the raw brain weights were standardized to body weight (brain weight/body weight), these differences disappeared (all p > 0.05) indicating that these difference can be contributed to body size.

108

Microglia Activation

Cells were isolated from the hippocampus, identified as microglia (CD11b+ and CD45Int), and activation was measured by MHC II using flow cytometry (Figure 5.3). The proportion of activated microglia cells showed a significant effect of minocycline only (F (1, 14) = 11.86, p =

0.004) where minocycline piglets had increased activation (62%) compared to controls (19%).

Gene Expression

Gene expression changes in pro-inflammatory, anti-inflammatory, and neurotrophic factors were measured (Figure 5.4). PRRSV piglets showed a decrease in BDNF expression

(0.57 ± 0.11) compared to controls (0.93 ± 0.10, F (1, 13) = 5.67, p = 0.03). For IL-1β, one outlier was removed in PRRSV/MINO group. There was a significant effect of PRRSV (F (1, 12)

= 5.27, p = 0.04) where PRRSV piglets had increased expression (3.01 ± 0.46) compared to controls (1.60 ± 0.41). There was also a significant effect of minocycline (F (1, 12) = 12.64, p =

0.004) where minocycline piglets (3.40 ± 0.41) had higher expression than controls (1.21 ±

0.46). There was a trend towards increased TNF-α in the minocycline groups (F (1,13) = 3.80, p = 0.07). There were no significant differences found in IL-10, IL-6, and TGF-β expression (all p > 0.05).

5.5 Discussion

Minocycline has potential for being a neuroprotective therapeutic for inflammatory conditions in the central nervous system. Many studies have shown protective rolls of minocycline in in vitro and in vivo acute inflammatory insults, but chronic administration paradigms and testing in large animal models are lacking (Suk 2004; Yong, Wells et al. 2004).

109

Minocycline is thought to reduce neuroinflammation through a variety of mechanisms, but the specific targets in the brain are being still being elucidated. Studies have shown that minocycline has potent anti-inflammatory, anti-apoptotic, and antioxidant properties driven by inhibition of p38 MAPK and inhibition of microglia activation (Plane, Shen et al. 2010). Here, we tested whether chronic high dose minocycline could reduce microglia activation and neuroinflammation caused by PRRSV in a neonatal piglet model. We found that minocycline potentiated microglia activation and inflammatory cytokines in the hippocampus, contrary to previous studies.

There is a substantial difference in dosage used in rodents and what has been deemed safe for humans. Clinically, minocycline is routinely used to treat urinary tract infections and acne with a dose of 2-4 mg/kg/day. Oral administration is most common as it provides 95-100% bioavailability (Saivin and Houin 1988). The majority of rodent studies use intraperitoneal injections and higher doses. In rodent models of hypoxia ischemia, 45-90 mg/kg was found to be the neuroprotective dosage range, but this appears to be species specific where it is protective in rats, but exacerbates damage in mice (Buller, Carty et al. 2009). For this study, we chose a higher minocycline dose than clinically used in humans, similar to the rodent neuroprotective range. One 100 mg tablet was administered daily starting at 7 days of age.

The effective dosing was on a sliding scale based on body weight grain where the average dose at 7 days of age was 46 mg/kg/day and dropped to 17 mg/kg/day by the conclusion of the study.

Monitoring tolerance of this dose was important as many side effects have been reported with high doses in humans. The most common adverse side effects include gastrointestinal irritability, vestibular toxicity, and central nervous system disturbances (Buller,

Carty et al. 2009). Intracranial hypertension can be caused by minocycline and can lead to multiple clinical manifestations (Lander 1989). We did not observe any gastrointestinal disturbances in the control minocycline group. The PRRSV minocycline group did develop

110 diarrhea, but this is a common manifestation of PRRSV infection. The diarrhea was not qualitatively different than the PRRSV control group. We did observe gait ataxia and hypertonic limbs in two control minocycline and one PRRSV minocycline animal. The ataxia lasted for roughly 3 days in each animal and occurred roughly 2-weeks after the start of minocycline administration. The ataxia may be a sign of intracranial hypertension as it has been documented before in humans (Round and Keane 1988).

Minocycline is not typically used during pregnancy and for infants as it is an ion chelator that binds to calcium phosphate and is incorporated into developing teeth and bone (Buller,

Carty et al. 2009). We observed this in our minocycline piglets during brain extraction as the skull had marked yellow coloration from minocycline (Figure 5.5). Additionally, tetracyclines can disrupt binding of bilirubin to albumin leading to kernicterus, or bilirubin-induced brain damage

(Buller, Carty et al. 2009). Clinical features of kernicterus include reduced feeding, seizures, and hypertonicity (Dennery, Seidman et al. 2001). Kernicterus may be a contributing factor in this study as we observed hypertonic limbs in piglets (see above) and the PRRSV minocycline group had reduced feeding scores beginning after one week of minocycline administration.

Kernicterus has also been shown to cause neuroinflammation and damage neurons, astrocytes, oligodendrocytes, and microglia (Watchko and Tiribelli 2013). Unconjugated bilirubin activates microglia through a p38 MAPK and ERK1/2 pathway which leads to increased expression of TNF-α, IL-1β, and IL-6 (Silva, Vaz et al. 2010). This may explain the results found in this study. Minocycline treatment significantly increased microglia activation in both the control and PRRSV animals. Pro-inflammatory gene expression, including IL-1β and a trend in

TNF-α, were observed in the minocycline piglets. Although we did not observe jaundice in the minocycline piglets, chronic exposure to even low levels of unconjugated bilirubin can lead to neuroinflammation.

111

Use of minocycline in other large animal models of neuroinflammation have not shown protective effects of minocycline, indicating differences between large animal and rodent studies. Kay et al. used CLN6 South Hampshire sheep as a model for neuronal ceroid lipofuscinoses (NCLs, or Batten disease) which is a fatal neurodegenerative disease (Kay and

Palmer 2013). They found that chronic high dose oral minocycline was successful in delivering the drug to the CNS, but it did not inhibit microglia activation, astrocytosis, or neuronal loss.

Limitations to this study are the limited animal number, especially in the PRRSV/CON group. Only 3 females were left in this group at the conclusion of the study impacting the ability to include sex in statistical analysis. Nonetheless, the results of this study show the complexity of minocycline use. Results obtained in rodent studies may not translate to large animal models and may not be predictive of outcomes in human patients. More precise knowledge of the mechanisms of minocycline in the brain of large animals would be beneficial for study design.

Finding the balance between having a high enough dose for neuroprotection, but low enough to prevent side effects is key. High dose minocycline is not appropriate for neuroprotection in a piglet PRRSV model, but additional research on other dosing regimens are warranted.

112

5.6 Figures

A 8 * * * * * * 6

4 Weight (kg) Control Control 2 Control Minocycline PRRSV Control PRRSV Minocycline

0 2 3 4 5 6 7 8 9 9 3 6 10 11 12 13 14 15 16 17 18 1 20 21 22 2 24 25 2 27 28 Day of Age B 41

40 C)  39

Temperature ( Control Control 37 Control Minocycline PRRSV Control PRRSV Minocycline 36 7 8 9 1 2 4 5 6 7 0 1 2 3 6 7 8 10 1 1 13 1 1 1 1 18 19 2 2 2 2 24 25 2 2 2 Day of Age C 3.0

2.5

2.0 bc bc bc ac ab bc bc bc bc

bc bc

First Feeding Score Feeding First bc 1.5 ac Control Control Control Minocycline PRRSV Control PRRSV Minocycline 1.0 7 8 9 1 2 4 6 9 2 4 6 10 1 1 13 1 15 1 17 18 1 20 21 2 23 2 25 2 27 Day of Age

Figure 5.1. Sickness symptom measures. PRRSV piglets displayed reduced weight gain (A) and higher temperatures (B) than controls. The PRRSV piglets also showed decreases in the first feeding score (C). Data are represented as means ± SEM. (*<0.05 for PRRSV groups compared to CON; a = PRRSV/CON is significantly less than the control groups; b =

113

PRRSV/MINO is significantly less than the control groups; c = the PRRSV/CON and PRRSV/MINO are significantly different).

A * B

60 15

55 10 50

45 5 40 Brain Weight (g)

35 0 Brain Weight / Body Weight e l l e l e n o o n o n li li ntrol cline tr o y yc n yc C c o c o l Contr oc C inocycli o n no trol M tr Mi Mi n l n RSV o o o C PRRSV Contr SV Min C PR R ntrol RSV Contr PR Co PR

Figure 5.2. Brain weights. The PRRSV piglets had a significantly lower brain weight than controls (A). When standardized by brain weight, this difference disappeared (B). Data are represented as means ± SEM. (* < 0.05).

114

1.0

0.8

0.6

0.4

0.2 MHC II % Cells, + MHC

0.0

trol rol n cline nt y ocycline Co n Mi RSV Control Co PR ntrol Co PRRSV Minoc

Figure 5.3. Microglia activation. The minocycline treatment groups had significantly higher MHC-II expression than controls (p = 0.002). Data are represented as means ± SEM.

115

6 Control Control Control Minocycline PRRSV Control 4 PRRSV Minocycline

2 Relative ExpressionRelative

F - -6 - - N 1 IL F F D IL-10 IL G N B T T

Figure 5.4. Relative abundance of mRNA in the hippocampus. The PRRSV piglets had significantly lower expression of BDNF (p = 0.03). There were also significant main effects of PRRSV (p = 0.04) and minocycline (p = 0.004) for IL-1β. No significant differences were found in IL-10, IL-6, TGF-β, or TNF-α. Data are represented as means ± SEM.

116

Figure 5.5. Minocycline incorporation into the skull. Minocycline incorporation into the skull during ossification can be seen by comparing a control piglet skull (A) to a yellow tinted minocycline piglet skull (B). An additional view with the cuts through the skull show the extent of coloring caused by minocycline (C).

117

5.7 References

Bitzer-Quintero, O. K. and I. Gonzalez-Burgos (2012). "Immune system in the brain: a modulatory role on dendritic spine morphophysiology?" Neural Plast 2012: 348642.

Blum, D., A. Chtarto, et al. (2004). "Clinical potential of minocycline for neurodegenerative disorders." Neurobiol Dis 17(3): 359-366.

Buller, K. M., M. L. Carty, et al. (2009). "Minocycline: A neuroprotective agent for hypoxic- ischemic brain injury in the neonate?" Journal of Neuroscience Research 87(3): 599- 608.

Casarejos, M. J., J. Menendez, et al. (2006). "Susceptibility to rotenone is increased in neurons from parkin null mice and is reduced by minocycline." J Neurochem 97(4): 934-946.

Chen, X., R. Pi, et al. (2010). "Combination of methylprednisolone and minocycline synergistically improves experimental autoimmune encephalomyelitis in C57 BL/6 mice." J Neuroimmunol 226(1-2): 104-109.

Choi, Y., H. S. Kim, et al. (2007). "Minocycline attenuates neuronal cell death and improves cognitive impairment in Alzheimer's disease models." Neuropsychopharmacology 32(11): 2393-2404.

Dantzer, R., J. P. Konsman, et al. (2000). "Neural and humoral pathways of communication from the immune system to the brain: parallel or convergent?" Auton Neurosci 85(1-3): 60-65.

Dennery, P. A., D. S. Seidman, et al. (2001). "Neonatal hyperbilirubinemia." N Engl J Med 344(8): 581-590.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Done, S. H., D. J. Paton, et al. (1996). "Porcine reproductive and respiratory syndrome (PRRS): a review, with emphasis on pathological, virological and diagnostic aspects." Br Vet J 152(2): 153-174.

Gordon, S. (2003). "Alternative activation of macrophages." Nat Rev Immunol 3(1): 23-35.

Harry, G. J. (2013). "Microglia during development and aging." Pharmacol Ther 139(3): 313- 326.

Kay, G. W. and D. N. Palmer (2013). "Chronic oral administration of minocycline to sheep with ovine CLN6 neuronal ceroid lipofuscinosis maintains pharmacological concentrations in the brain but does not suppress neuroinflammation or disease progression." J Neuroinflammation 10(1): 97.

118

Kim, H. S. and Y. H. Suh (2009). "Minocycline and neurodegenerative diseases." Behav Brain Res 196(2): 168-179.

Lander, C. M. (1989). "Minocycline-induced benign intracranial hypertension." Clin Exp Neurol 26: 161-167.

Livak, K. J. and T. D. Schmittgen (2001). "Analysis of relative gene expression data using real- time quantitative PCR and the 2(-Delta Delta C(T)) Method." Methods 25(4): 402-408.

Meyer, U., J. Feldon, et al. (2011). "Schizophrenia and autism: both shared and disorder- specific pathogenesis via perinatal inflammation?" Pediatr Res 69(5 Pt 2): 26R-33R.

Perry, V. H. and S. Gordon (1988). "Macrophages and microglia in the nervous system." Trends Neurosci 11(6): 273-277.

Plane, J. M., Y. Shen, et al. (2010). "Prospects for minocycline neuroprotection." Archives of Neurology 67(12): 1442-1448.

Rice, D. and S. Barone, Jr. (2000). "Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models." Environ Health Perspect 108 Suppl 3: 511-533.

Richwine, A. F., A. O. Parkin, et al. (2008). "Architectural changes to CA1 pyramidal neurons in adult and aged mice after peripheral immune stimulation." Psychoneuroendocrinology 33(10): 1369-1377.

Round, R. and J. R. Keane (1988). "The minor symptoms of increased intracranial pressure: 101 patients with benign intracranial hypertension." Neurology 38(9): 1461-1464. Saivin, S. and G. Houin (1988). "Clinical pharmacokinetics of doxycycline and minocycline." Clin Pharmacokinet 15(6): 355-366.

Schmitz, T. and L. J. Chew (2008). "Cytokines and myelination in the central nervous system." ScientificWorldJournal 8: 1119-1147.

Silva, S. L., A. R. Vaz, et al. (2010). "Features of bilirubin-induced reactive microglia: from phagocytosis to inflammation." Neurobiol Dis 40(3): 663-675.

Suk, K. (2004). "Minocycline suppresses hypoxic activation of rodent microglia in culture." Neuroscience Letters 366(2): 167-171.

Tikka, T. M. and J. E. Koistinaho (2001). "Minocycline provides neuroprotection against N- methyl-D-aspartate neurotoxicity by inhibiting microglia." J Immunol 166(12): 7527-7533.

Watchko, J. F. and C. Tiribelli (2013). "Bilirubin-Induced Neurologic Damage — Mechanisms and Management Approaches." New England Journal of Medicine 369(21): 2021-2030.

Yang, L., S. Sugama, et al. (2003). "Minocycline enhances MPTP toxicity to dopaminergic neurons." J Neurosci Res 74(2): 278-285.

Yong, V. W., J. Wells, et al. (2004). "The promise of minocycline in neurology." Lancet Neurol 3(12): 744-751.

119

Yrjanheikki, J., R. Keinanen, et al. (1998). "Tetracyclines inhibit microglial activation and are neuroprotective in global brain ischemia." Proc Natl Acad Sci U S A 95(26): 15769- 15774.

120

CHAPTER 6

SUMMARY AND SIGNIFICANCE

The aim of this research was to determine if neonatal respiratory viral infection induces changes in the developing brain. Previously, we have shown that porcine reproductive and respiratory viral infection (PRRS) causes marked microglia activation, increased pro- inflammatory cytokines, and deficits in hippocampal dependent learning and memory. The present work further expands upon this research to look at possible mechanisms for the deficits in learning and memory and possible therapeutics.

Initial studies were conducted to develop magnetic resonance imaging techniques for the neonatal piglet. Three works are presented in the appendices that allowed for studies in chapter 3 to be conducted. First, basic methods were developed to measure regional brain volumes in the piglet using structural MRI techniques, and these data are presented in appendix

A. Using this technique, a longitudinal study was conducted that characterized the regional brain growth curves in the domestic pig (Appendix B). This study was significant as it allowed for the comparison of the regional brain growth to humans and this confirmed the usefulness of the piglet as an animal model for human brain development. Finally, an averaged brain and

MRI-based atlas was created for the neonatal piglet (Appendix C). This allowed for further refinement of the structural MRI analysis including the ability to conduct voxel-based morphometry studies. Additional method optimization and protocols were developed for the piglet including diffusion tensor imaging and magnetic resonance spectroscopy.

Using these new methods, the results from Chapter 3 show that neonatal PRRS infection does impact the macro development of the CNS. There were significant reductions in gray and white matter, most notably in the primary visual cortex. This was the first time that voxel-based morphometry has been used in a piglet model. There was also a trend for reductions in fractional anisotropy in the corpus callosum, which suggest either delay or disruption of myelination. Additionally, there were differences found in metabolite

121 concentrations within the hippocampus. Importantly, this research shows that MRI is an important research tool for tracking brain growth and development, and is sensitive to insults including infection. Further research is needed in the brain areas which had reductions in gray and white matter volume in the PRRS piglets. Histological confirmation of this volume decrease and characterization of the cellular changes are needed.

In Chapter 4, cellular changes due to PRRS infection were characterized in the hippocampus. Focus was placed on two developmental processes within the dentate gyrus which have been implicated in learning and memory. First, the dentate gyrus is one of two brain areas which has postnatal neurogenesis. Bromodeoxyuridine administration, concurrent with

PRRS infection, allowed for the characterization of the impact of PRRS on new cell survival and cell fate. A sexual dimorphism in new cell survival was found in which males had more surviving cells than females. Additionally, only males had reduced survival of new cells with

PRRS infection. This shows that there may be vulnerability differences between the sexes, and using male and female animals in research is important. Cell fate was also disrupted with

PRRS infection causing a decrease in the number of newly divided cells that form neurons. The number of newly formed microglia were not affected by PRRS. Secondly, we found differences due to sex and PRRS in dentate granule cell morphology. There were many similarities found between rodents and pigs including males having more complex dendritic trees than females and complexity differences due to location in the granule cell layer. PRRS infection caused a morphological change in the inner granule cell neurons, with the initial sprouting point further away from the cell soma. There were no changes in spine density or the distribution of the morphology of the spines. These data show important changes at the cellular level caused by early-life respiratory infection which may have implications for learning and memory.

Lastly, the studies in Chapter 5 were meant to study the mechanism of the changes in the previous two chapters and evaluate minocycline as a possible therapeutic for neuroinflammation. Minocycline is a second generation tetracycline antibiotic which prevents

122 the activation of microglia. Using minocycline for the duration of PRRS infection, this study sought to prevent microglia activation in the PRRS piglets. The results indicate that minocycline administration increased microglia activation, contrary to the majority of studies published in other species. Minocycline may have caused a worsening of feeding scores in the PRRS animals and also caused increased levels of pro-inflammatory gene expression in the hippocampus. The results from this study may be due to a high dose, albeit neuroprotective in rodents, that was used. High doses of minocycline have the potential to increase intracranial pressure and cause bilirubin-induced brain damage. Thus, this shows that high dose minocycline is not appropriate in the neonatal time period. Further research into finding a dose that is high enough to exhibit neuroprotective effects, yet low enough to prevent side effects is warranted. Collectively, this dissertation shows the early-life respiratory viral infection can effect brain development. Finding better therapeutics to modulate the neuroimmune response could prevent these changes and lead to better short- and long-term developmental outcomes.

123

APPENDIX A

MAGNETIC RESONANCE IMAGING OF THE NEONATAL PIGLET BRAIN

A.1 Abstract

Appeal for the domestic pig as a preclinical model for neurodevelopmental research is increasing. One limitation, however, is lack of MRI methods for brain volume quantification in the neonatal piglet. To address this void, anatomic MRI data (non-longitudinal) were acquired from 2-week (n=6) and 5-week (n=6) old pigs using a three-dimensional T1-weighted magnetization prepared gradient echo (MPRAGE) sequence on a MAGNETOM Trio 3T imager.

Manual segmentation was performed for volume estimations of total brain, cortical, diencephalon, brainstem, cerebellar, and hippocampal regions. Hippocampal volumes were also estimated by postmortem histological analysis using the Cavalieri method with planimetric analysis. Strong correlations were found between the hippocampal volume estimates using

MRI and the histological section estimates of hippocampal volume, indicating accurate volumetric measurements can be obtained using this MRI protocol. In addition to developing and validating MRI methods for estimating brain volume, the present study also provides evidence of substantial brain growth during this early period. The results indicate that MRI can provide accurate estimates of brain region volume changes over time during the neonatal period in piglets. The availability of a piglet model for use in longitudinal studies may be useful to understanding effects of experimental factors on brain growth and development.

This material has been previously published. The copyright owner has provided permission to reprint. Conrad Matthew S, Dilger Ryan N, Nickolls Alec, Johnson Rodney W (2012) Magnetic resonance imaging of the neonatal piglet brain. Pediatr Res 71:179-84

124

A.2 Introduction

The domestic pig (Sus scrofa), due to its anatomic and physiologic similarities to humans, is a common well accepted preclinical model in cardiovascular, metabolic, and pediatric nutrition research (Miller and Ullrey 1987; Bellinger, Merricks et al. 2006; Dixon and

Spinale 2009). For the reason that pigs and humans also share similar patterns in brain growth and development, appeal for the pig as a model for neurodevelopmental research has recently increased (Pond, Boleman et al. 2000; Lind, Moustgaard et al. 2007). In both the pig and human, the major brain growth spurt extends from late prenatal to early postnatal, which is different from other common animal models (Dobbing and Sands 1979). Both pigs and humans are born with a brain weight that is roughly 25% of the final adult weight (Dobbing and Sands

1979). Additionally, gross anatomical features including gyral pattern and distribution of grey and white matter of the neonatal pig brain are similar to that of human infants (Dickerson and

Dobbing 1967; Thibault and Margulies 1998). Furthermore, due to the fact that the pig is a precocial species with fairly well developed sensory and motor systems at birth, it is amenable to behavioral testing to assess cognitive development as early as one to two weeks of age

(Dilger and Johnson 2010). Thus, the neonatal piglet is a potential preclinical translational model for investigating the effects of early-life experiential factors (e.g., infection, nutrition) on brain and cognitive development. As the piglet is already used to study parenteral and enteral nutrition in preterm infants (Ganessunker, Gaskins et al. 1999), and preterm birth is associated with a high rate of developmental delay and cognitive disability (Wood, Marlow et al. 2000), the availability of the piglet for investigating brain and cognitive development would be particularly welcome. One limitation, however, has been the ability to make use of quantitative magnetic resonance imaging (MRI) for in vivo assessment of brain growth and development in neonatal piglets.

Quantitative MRI is providing important information on brain development in early childhood and adolescence (Pfefferbaum, Mathalon et al. 1994; Giedd, Vaituzis et al. 1996;

125

Giedd, Blumenthal et al. 1999; Gilmore, Lin et al. 2007; Knickmeyer, Gouttard et al. 2008) but because of potential health concerns for infants, few studies have focused on the period from birth to 4 years of age when dramatic brain development occurs. Most MRI studies of infants have been with those born prematurely or with significant health complications (Beauchamp,

Thompson et al. 2008; Dubois, Benders et al. 2008; Thompson, Wood et al. 2008). Only recently was MRI used to assess structural brain development of healthy full term infants from birth to 2 years of age (Knickmeyer, Gouttard et al. 2008). The first year after birth, total brain volume more than doubled and hemispheric grey and white matter increased 149% and 11%, respectively. Although this application of MRI in human infants is impressive, the ability to address mechanistic questions related to experiential influences on brain growth and development cannot be readily done due to practical or ethical concerns.

In older pigs, it is possible to obtain reliable estimates of brain volumes using MRI

(Jelsing, Rostrup et al. 2005), and MRI has been used in experimental neurosurgical procedures and to investigate several neuropathological conditions (Munkeby, De Lange et al.

2008; Rosendal, Pedersen et al. 2009; Björkman, Miller et al. 2010; Kuluz, Samdani et al. 2010;

Sandberg, Crandall et al. 2010). To our best knowledge, however, MRI has not been used to obtain brain volumetric data for in vivo quantitative analysis in neonatal piglets. As a first step toward addressing this void, here we report MRI scanning and analysis protocols for estimating total brain, cortex, diencephalon, cerebellum, hippocampus, and brainstem volumes in neonatal piglets at 2- and 5-weeks of age. The reliability of the MRI-based volume estimates was assessed for the hippocampus by the Cavalieri method and planimetric analysis with physical histological sections obtained postmortem. The results indicate that MRI can provide an accurate in vivo estimate of brain region volume changes over time during the neonatal period.

Furthermore, substantial increases in volumes for all brain regions examined were evident during the three-week time period. The potential to quantify brain growth and development in

126 vivo using MRI in a longitudinal study design makes the piglet an attractive preclinical animal model.

A.3 Material and Methods

Subjects

A total of twelve pigs, six male and six female (Sus scrofa domestica, York breed), from three litters, were obtained from the University of Illinois swine herd at two days of age and placed into an artificial rearing system that has been previously described (Dilger and Johnson

2010). Briefly, each piglet was housed individually in an acrylic-sided cage with a 12 hour light/dark cycle. Ambient temperature was maintained at 27˚C with intracage temperatures maintained at 32˚C using radiant heaters located directly over the piglets. Piglets received a commercially available milk replacer (Milk Specialties Global Animal Nutrition, Carpentersville,

IL) throughout the 5-week study. Milk replacer powder was reconstituted fresh each day and 50 mL of liquid diet was delivered to individual feeding bowls once an hour from 8:00 a.m. to midnight using a custom, automated system (Dilger and Johnson 2010). The pigs had an average weight of 1.86 kg ± 0.05 at birth and 3.08 kg ± 0.05 and 9.13 kg ± 0.34 at 2- and 5- weeks of age, respectively. All procedures were in accordance with the National Institute of

Health Guidelines for the Care and Use of Laboratory Animals and approved by the University of Illinois Institutional Animal Care and Use Committee.

Experimental Procedures

At 2-weeks of age, six of the pigs, three male and three female, were subjected to MRI procedures. Immediately before scanning, pigs were anesthetized by intramuscular (i.m.) injection of a telazol:ketamine:xylazine solution (TKX; 4.4 mg/kg body weight), and placed in a prone position in the MRI scanner. The pigs remained anesthetized during the MRI scanning procedure (approximately 20 minutes) and afterwards were euthanized by intracardiac

127 administration of an overdose of sodium pentobarbital (390 mg/ml Fatal Plus - administered at 1 ml/5 kg body weight). Brains were removed and a 2.5 cm3 tissue block encompassing the left hippocampus was excised and placed in 4% paraformaldyhyde. Tissue blocks were stored in fixative at 4°C for 3 weeks before processing. Tissue blocks from male and female littermates were selected and processed for histological volume determination. The remaining six pigs were scanned at 5-weeks of age after which they were euthanized and their brain tissue was collected and processed as described above.

Imaging

Magnetic resonance scanning was conducted using a Siemens MAGNETOM Trio 3-T imager by employing a Siemens 32-channel head coil (Erlangen, Germany). Anatomic images were acquired using a three-dimensional T1-weighted magnetization prepared gradient echo

(MPRAGE) sequence with the following parameters: repetition time, TR = 1900 ms; echo time,

TE = 2.48ms; inversion time, TI = 900 ms, flip angle = 9˚, matrix = 256 x 256 (interpolated to

512 x 512), slice thickness = 1.0 mm. The final voxel size was 0.35 mm x 0.35 mm x 1.0 mm across the entire head from the tip of the snout to the cervical/thoracic spinal cord junction.

Images were exported in the Digital Imaging and Communications in Medicine (DICOM) format and analyzed using three-dimensional visualization software (AMIRA®, Visage Imaging, Inc.,

San Diego, CA).

In vivo Quantification of Brain Region Volume Using MRI

The volumes for total brain, cortex, diencephalon, cerebellum, hippocampus, and brainstem were quantified using AMIRA® 3-D visualization software. Unsharp masking post- processing was performed to improve grayscale inhomogeneities (Brinkmann, Manduca et al.

1998). Manual segmentation of the brain regions was facilitated by a Wacom Cintiq 21UX graphic input screen and pen (Wacom, Vancouver, WA). The anatomical criterion for the

128 regions were based upon the stereotaxic atlas of Félix et al (1999). The boundaries for segmentation are shown in Figure 1. For the hippocampus, the medial border was defined by the choroidal fissure and the lateral border was the inferior horn of the lateral ventricle. The superior border was the body of the lateral ventricle and the inferior border was parahippocampal gyrus. The total brain volume was defined as the sum of all cortical, subcortical, cerebellar, brainstem, and partial spinal cord volumes. The spinal cord was segmented to the posterior boundary of the cerebellum. The volumes were calculated using the material statistics function of AMIRA® which estimates the three-dimensional volume of the manually segmented regions given the aforementioned voxel size. To determine the reliability of these measures, segmentation was conducted by two independent observers and the intra- and interobserver agreements calculated.

Quantification of Hippocampal Volume from Histological Sections

Paraformaldehyde-fixed hippocampal tissue was dehydrated and paraffin-embedded using a Leica ASP300 tissue processor (Leica, Wetzlar, Germany). The tissue block was sectioned in the axial plane at 5 µm intervals and every 10th section was collected. The serial sections were hematoxylin and eosin (H&E) stained using a Sakura Tissue-Tek DRS H & E stainer system and coverslipped with a Sakura Tissue-Tek Glas coverslipper (Sakura, Tokyo,

Japan). The slides were digitized at 20x magnification using a Nanozoomer Digital Pathology

System (Hamamatsu, Hamamatsu City, Japan), creating high resolution virtual slide images.

Digital images of the hippocampal region were exported from the virtual slides as JPEGs for subsequent segmentation. Digital images of the serial sections were imported into AMIRA® and the z-plane depth was adjusted to account for the 50 µm distance between collected slices.

Images were aligned and the hippocampus was manually segmented for each section. The hippocampal region was defined as the dentate gyrus, CA1, and CA3 as shown in a

129 representative slide in Figure 2. After all tissue sections were segmented, the material statistics function of AMIRA® was used to estimate hippocampal volume.

Statistics

For analysis of brain region volumes in 2- to 5-week old piglets, data were subjected to two-way

ANOVA (Age × Sex) using SAS software (SAS, Cary, NC). It was determined that there was unequal variance between the data obtained for 2- and 5-week old piglets, and to correct for this, data were log transformed before ANOVA. There was no significant effect of sex, therefore the male and female data were combined for calculation of the means. Significance level was set at p<0.05. Validation of MRI volume estimation of the hippocampus was based on the technique employed in human studies by Bobinski et al. (Bobinski, de Leon et al. 2000). The correlative relationship between hippocampal volumes obtained by MRI and histological measurement was determined using GraphPad Prism 5 software (GraphPad Software, La Jolla,

CA). The Pearson product-moment correlation coefficient was used to determine the strength of linear dependence between the two volume estimations.

The intraobserver and interobserver agreements for manual segmentation were calculated based on statistical methods by Bland and Altman (Bland and Altman 1986). The

intraobserver agreement was calculated by the following formula where x1st and x2nd were repeat measures of regional volumes of 6 pigs using manual segmentation by one rater

(M.S.C). This rater was blinded to the pig brain images being segmented and the order pig brain images were segmented was randomized.

 x1st  x2nd  Intraobserver agreement index = 100   100 x1st  x2nd / 2 

130

The interobserver agreement was calculated by the following formula where xa and xb are measures of regional volumes of each pig using manual segmentation by two different raters

(M.S.C and A.N.).

 xa  xb  Interobserver agreement index = 100   100 xa  xb / 2 

The intraobserver coefficient of variation (COV) was defined as 2 SD of the following equation by the British Standards Institution (Institution 1998).

x  x 1st 2nd x1st  x2nd / 2

The interobserver COV was defined as 2 SD of the following equation

x  x a b xa  xb / 2

A.4 Results

Total brain and subregion volumes for 2- and 5-week old piglets (n = 6) are shown in

Table 1. As expected, there was a main effect of age on total brain, cortex, diencephalon, cerebellum, hippocampus, and brainstem volumes, demonstrating considerable volume increases from the 2- to 5-week period. There was no main effect of sex for any brain region studied, indicating that the volumes were similar in males and females at this age (p > 0.05 for all measures). Because of this, data for males and females (n = 6) were combined. Volume of the left and right hippocampal formation was similar within each animal irrespective of sex or age. Since no asymmetries were observed, the left and right hippocampal volumes were combined to create a total hippocampal volume measure at each age. A three-dimensional segmented brain including all of the aforementioned brain regions can be seen in Figure 3.

131

To determine the reliability of the MRI-based volume estimates, the volume of the hippocampus was also determined from physical sections obtained postmortem following MRI.

The hippocampus was chosen because it was smallest of all areas examined and most difficult to manually segment. Furthermore, technical challenges prevented whole brain slicing at 5µm intervals. Using the histological sections, segmentation of individual hippocampal subregions

(dentate, CA1, and CA3) was possible, however, due to the small size of the hippocampus in the young pig, it was not possible to identify these hippocampal subregions by MRI. Thus, only whole hippocampal volume was compared between the two methods. To reduce variance due to genetic variability, piglets from the same litter were used for the histological validation. A total of 4 piglets were used, two that were 2-weeks of age and two that were 5-weeks of age.

The average histological-determined volumes at 2-weeks and 5-weeks of age were 165.29 mm3

(± 9.75) and 246.70 mm3 (± 11.74), respectively. The average MRI-determined volumes at 2- and 5-weeks of age were 487.83 mm3 (± 38.28) and 658.87 mm3 (± 47.25), respectively. While tissue blocks were not measured prior to processing, the difference in the MRI and histological- determined volumes represents a reduction in the total volume (i.e. shrinkage) of approximately

64%. Finally, a strong correlation between hippocampal volumes determined by MRI and histology was evident (r = 0.9878, P = 0.0122).

The median intraobserver and interobserver agreement for two independent raters for

MRI-based volume estimates is shown in table 2. Intraobserver agreement was consistent across brain regions with the median values ranging from 94.4 to 96.6%. The coefficients of variation ranged 2.1 to 6.4% across brain regions. The interobserver agreement medians ranged from 96.6 to 98.8% with coefficients of variations ranging from 2.0 to 5.2%.

A.5 Discussion

MRI has been used in older pigs to quantify whole brain and sub-structure volumes, but to our best knowledge, no studies have used MRI to characterize normal brain development

132 during the neonatal time period (Jelsing, Rostrup et al. 2005). Therefore, the current study sought to determine if MRI could be used to estimate volumes of different brain regions in the neonatal period. The important results showed that MRI and manual segmentation reliably estimated volume changes over time of different brain regions. Furthermore, substantial increases in volumes for all brain regions examined were evident during the three-week time period.

Several challenges are presented when using MRI during the neonatal time period including poor spatial resolution and low tissue contrast which causes difficulties discerning between grey matter, white matter, and cerebrospinal fluid (Shi, Fan et al. 2010). Because of this, using automated segmentation procedures in the neonate has been difficult. An additional obstacle of structural analysis in the neonatal pig brain is the lack of a brain atlas specific to this age period. Currently, the only MRI-based atlas for the pig is in adult animals, which is not suitable for neonatal brain analysis (Saikali, Meurice et al. 2010). Because there is currently no atlas available for pigs during the neonatal time period, we chose to use manual segmentation for volume analysis. This technique requires significantly more time than semi- and fully automated segmentation methods, but provides accurate volume estimations. Additional imaging data sets are needed for creation of structural MRI atlases, which will allow for more complex, automated segmentation methods. A future goal is to utilize more sophisticated techniques used for human studies, such as neonatal cortical mapping, once aged-based MRI- atlases have been created in the piglet (Dubois, Benders et al. 2008). Nonetheless, the current procedure can be used in a longitudinal experimental design to assess the effects of experiential factors (e.g., nutrition, infection) on brain growth and development. This is important because similar to humans (Dobbing and Sands 1979), overall brain weight of piglets in the neonatal period is about 25-35% of normal adults.

Studies employing MRI for volumetric analysis of specific brain regions typically validated this modality using classic histological techniques and the Cavalieri method (Jelsing,

133

Rostrup et al. 2005). Here, we used a digital planimetry method with high-resolution segment image sets from both MRI and histological sections. The planimetry method is a reliable method for volume estimation in which areas of interest are outlined on a computer screen and the included pixels in series of images is used for volume calculation (Cotter, Miszkiel et al.

1999). Using this method, we demonstrated a high significant correlation between hippocampal volume determined from the histological sections and hippocampal volume determined using

MRI, showing MRI can provide accurate relative differences over time.

In addition to developing MRI methods for estimating brain volume, by examining 2- week and 5-week old piglets, the present study also provides evidence of substantial brain growth during this early three-week period. A limitation of the present study was that it was not a longitudinal design due to the need to obtain brain sections for estimation of hippocampal volume using the Cavalieri method. Nonetheless, the volumes for cortex and subcortical regions including the hippocampus, diencephalon, cerebellum, and brainstem were 30-43% greater in 5-week old piglets compared to 2-week old piglets. No main effect of sex on absolute volumes or differences in volumes between 2- and 5-week olds was evident for these brain regions. These data are similar to a study of humans aged 3 months to 30 years which found no effects of sex on grey and white matter volumes after accounting for head size (Pfefferbaum,

Mathalon et al. 1994).

Additional limitations of this study include the simplicity of the volumetric analysis technique and the histological methods for validation of the MRI volumes. Due to lack of age specific MRI-based atlases in the neonatal piglet, more sophisticated regional volume estimation methods could not be used. Our lab is collecting MRI data to construct these averaged brain MRI-atlases in order to utilize techniques such as voxel-based morphology and deformation-based morphology which have been shown to be powerful methods for volumetric analysis in the neonatal brain (Boardman, Counsell et al. 2006). In addition, the automated scull stripping features of many semi- and fully-automated procedures that were developed for

134 humans do not currently work with the pig due to anatomical differences. For this study, only two independent observers were evaluated to construct the coefficients of variation (COV).

Although it would be ideal to evaluate additional observers, we obtained low COVs for our methods, and the use of only two observers is similar to other manual segmentation protocols for human brain region analysis (Bokde, Teipel et al. 2005; Yu, Zhang et al. 2010). For the histological validation, there were differences in the absolute volumes between the MRI and histologically-determined hippocampal volumes. These differences were due to the dehydration and paraffin embedding process of the hippocampal tissue. We chose this approach in order to be able to slice the tissue at the small intervals used in order to increase the resolution of our histological method. Because of this, the correlation analysis is based on relative differences in volume rather than absolute volumes. In addition, another limitation is that tissues from only four brains were used for the correlation analysis, although all of these brains were from littermate pigs allowing for reduction in variability. Despite the limitations of this study, the data supports this as a reliable method for quantification of brain region volumes in the neonatal piglet and future work will allow for more semi- and fully-automated segmentation of the piglet brain.

In summary, reliable estimates of brain volume changes over time in neonatal piglets can be obtained using MRI and manual segmentation procedures. Because the pig is a gyrencephalic species thought to have brain growth and development similar to humans, it may serve as a preclinical translational model for studying the developmental origins of neuropsychiatric diseases. Using the methods introduced in this paper, ongoing studies will characterize normal brain development in the pig from the neonatal period to adulthood in a longitudinal design.

135

A.6 Figures and Tables

Figure A.1. MRI segmentation boundaries. The left (yellow) and right (red) hippocampus are segmented in the coronal, axial, and sagittal planes of view. The cortex (green), diencephalon (teal), brainstem (dark blue), and cerebellum (pink) are also shown.

136

Figure A.2. Hippocampus segmentation criteria for histological tissue sections. The outlined area includes the dentate gyrus (DG), CA1, and CA3 regions of a horizontal cross-sectional slice of hippocampal tissue from the dorsobasal portion of the hippocampus. The subiculum (Sub) was not included in volume analysis. Image was taken at 12.5X magnification.

137

Figure A.3. Three-dimensional brain reconstruction. The 3-D reconstruction shows the anatomical position of the left hippocampus (yellow) and right hippocampus (red) in relation to the cortex (green), diencephalon (teal), brainstem (dark blue), and cerebellum (pink).

138

2 Week 5 Week % Region Age Old Old Increase

mm3 (SE) mm3 (SE) p <

Total Brain 46162 64958 41% 0.0001 Volume (642) (1341)

32636 46779 Cortex 43% 0.0001 (736) (1224)

Diencephalon 5118 (85) 6892 (189) 35% 0.0001

Total 1066 (42) 1464 (58) 37% 0.0003 Hippocampus

Cerebellum 3919 (69) 5380 (93) 37% 0.0001

Brainstem 3422 (62) 4443 (144) 30% 0.0001

Table A.1. Brain growth from 2- to 5-weeks of age. All volumes are mean (SE) in mm3. Three animals of each sex were used at 2- and 5-weeks of age.

139

Region Intraobserver Interobserver

Median Agreement, 94.7 98.1 Total Brain % Volume COV, % 4.5 4.0

Median Agreement, 94.4 97.2 % Cortex COV, % 6.4 5.2

Median Agreement, 95.6 98.0 % Diencephalon COV, % 5.3 4.3

Median Agreement, 95.2 96.6 % Cerebellum COV, % 3.0 2.0

Median Agreement, 94.8 98.8 Total % Hippocampus COV, % 2.1 2.5

Median Agreement, 94.9 97.4 % Brainstem COV, % 4.9 3.5

Table A.2. Intra- and interobserver agreements and coefficients of variation for segmented regions. All values are percentages.

140

A.7 References

Beauchamp, M. H., D. K. Thompson, et al. (2008). "Preterm infant hippocampal volumes correlate with later working memory deficits." Brain 131(Pt 11): 2986-2994.

Bellinger, D. A., E. P. Merricks, et al. (2006). "Swine models of type 2 diabetes mellitus: insulin resistance, glucose tolerance, and cardiovascular complications." ILAR J 47(3): 243-258.

Björkman, S. T., S. M. Miller, et al. (2010). "Seizures are associated with brain injury severity in a neonatal model of hypoxia-ischemia." Neuroscience 166(1): 157-167.

Bland, J. M. and D. G. Altman (1986). "Statistical methods for assessing agreement between two methods of clinical measurement." Lancet 1(8476): 307-310.

Boardman, J. P., S. J. Counsell, et al. (2006). "Abnormal deep grey matter development following preterm birth detected using deformation-based morphometry." Neuroimage 32(1): 70-78.

Bobinski, M., M. J. de Leon, et al. (2000). "The histological validation of post mortem magnetic resonance imaging-determined hippocampal volume in Alzheimer's disease." Neuroscience 95(3): 721-725.

Bokde, A. L. W., S. J. Teipel, et al. (2005). "Reliable manual segmentation of the frontal, parietal, temporal, and occipital lobes on magnetic resonance images of healthy subjects." Brain Res Brain Res Protoc 14(3): 135-145.

Brinkmann, B. H., A. Manduca, et al. (1998). "Optimized homomorphic unsharp masking for MR grayscale inhomogeneity correction." IEEE Trans Med Imaging 17(2): 161-171.

Cotter, D., K. Miszkiel, et al. (1999). "The assessment of postmortem brain volume; a comparison of stereological and planimetric methodologies." Neuroradiology 41(7): 493- 496.

Dickerson, J. W. T. and J. Dobbing (1967). "Prenatal and postnatal growth and development of the central nervous system of the pig." Proc Biol Sci 166(1005): 384-395.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Dixon, J. A. and F. G. Spinale (2009). "Large animal models of heart failure: a critical link in the translation of basic science to clinical practice." Circ Heart Fail 2(3): 262-271.

Dobbing, J. and J. Sands (1979). "Comparative aspects of the brain growth spurt." Early Hum Dev 3(1): 79-83.

Dubois, J., M. Benders, et al. (2008). "Mapping the early cortical folding process in the preterm newborn brain." Cereb Cortex 18(6): 1444-1454.

Felix, B., M. E. Leger, et al. (1999). "Stereotaxic atlas of the pig brain." Brain Res Bull 49(1-2): 1-137.

141

Ganessunker, D., H. R. Gaskins, et al. (1999). "Total parenteral nutrition alters molecular and cellular indices of intestinal inflammation in neonatal piglets." JPEN J Parenter Enteral Nutr 23(6): 337-344.

Giedd, J. N., J. Blumenthal, et al. (1999). "Brain development during childhood and adolescence: a longitudinal MRI study." Nat Neurosci 2(10): 861-863.

Giedd, J. N., A. C. Vaituzis, et al. (1996). "Quantitative MRI of the temporal lobe, amygdala, and hippocampus in normal human development: ages 4-18 years." J Comp Neurol 366(2): 223-230.

Gilmore, J. H., W. Lin, et al. (2007). "Regional gray matter growth, sexual dimorphism, and cerebral asymmetry in the neonatal brain." J Neurosci 27(6): 1255-1260.

Institution, B. S. (1998). Accuracy (trueness and precision) of measurement methods and results. Alternative methods for the determination of the precision of a standard measurement method. BS 5725, British Standards Institution. 5725-5.

Jelsing, J., E. Rostrup, et al. (2005). "Assessment of in vivo MR imaging compared to physical sections in vitro--a quantitative study of brain volumes using stereology." Neuroimage 26(1): 57-65.

Knickmeyer, R. C., S. Gouttard, et al. (2008). "A structural MRI study of human brain development from birth to 2 years." J Neurosci 28(47): 12176-12182.

Kuluz, J., A. Samdani, et al. (2010). "Pediatric spinal cord injury in infant piglets: description of a new large animal model and review of the literature." J Spinal Cord Med 33(1): 43-57.

Lind, N. M., A. Moustgaard, et al. (2007). "The use of pigs in neuroscience: modeling brain disorders." Neurosci Biobehav Rev 31(5): 728-751.

Miller, E. R. and D. E. Ullrey (1987). "The pig as a model for human nutrition." Annu Rev Nutr 7: 361-382.

Munkeby, B. H., C. De Lange, et al. (2008). "A piglet model for detection of hypoxic-ischemic brain injury with magnetic resonance imaging." Acta Radiol 49(9): 1049-1057.

Pfefferbaum, A., D. H. Mathalon, et al. (1994). "A quantitative magnetic resonance imaging study of changes in brain morphology from infancy to late adulthood." Arch Neurol 51(9): 874-887.

Pond, W. G., S. L. Boleman, et al. (2000). "Perinatal ontogeny of brain growth in the domestic pig." Proc Soc Exp Biol Med 223(1): 102-108.

Rosendal, F., M. Pedersen, et al. (2009). "MRI protocol for in vivo visualization of the Gottingen minipig brain improves targeting in experimental functional neurosurgery." Brain Res Bull 79(1): 41-45.

Saikali, S., P. Meurice, et al. (2010). "A three-dimensional digital segmented and deformable brain atlas of the domestic pig." J Neurosci Meth 192(1): 102-109.

142

Sandberg, D. I., K. M. Crandall, et al. (2010). "Pharmacokinetic analysis of etoposide distribution after administration directly into the fourth ventricle in a piglet model." J Neurooncol 97(1): 25-32.

Shi, F., Y. Fan, et al. (2010). "Neonatal brain image segmentation in longitudinal MRI studies." Neuroimage 49(1): 391-400.

Thibault, K. L. and S. S. Margulies (1998). "Age-dependent material properties of the porcine cerebrum: effect on pediatric inertial head injury criteria." J Biomech 31(12): 1119-1126.

Thompson, D. K., S. J. Wood, et al. (2008). "Neonate hippocampal volumes: prematurity, perinatal predictors, and 2-year outcome." Ann Neurol 63(5): 642-651.

Wood, N. S., N. Marlow, et al. (2000). "Neurologic and developmental disability after extremely preterm birth." N Engl J Med 343(6): 378-384.

Yu, X., Y. Zhang, et al. (2010). "Comprehensive brain MRI segmentation in high risk preterm newborns." PLoS One 5(11): e13874.

143

APPENDIX B

BRAIN GROWTH OF THE DOMESTIC PIG (SUS SCROFA) FROM 2 TO 24 WEEKS OF AGE: A LONGITUDINAL MRI STUDY

B.1 Abstract

An animal model with brain growth similar to humans, that can be used in MRI studies to investigate brain development, would be valuable. Our lab has developed and validated MRI methods for regional brain volume quantification in the neonatal piglet. The aim of this study was to utilize the MRI-based volume quantification technique in a longitudinal study to determine brain growth in domestic pigs from 2 to 24 wks of age. MRI data were acquired from pigs 2 to 24 wks of age using a three-dimensional MPRAGE sequence on a MAGNETOM Trio

3T imager. Manual segmentation was performed for volume estimates of total brain, cortical, diencephalon, brainstem, cerebellar, and hippocampal regions. Logistic modeling procedures were used to characterize brain growth. Total brain volume increased 130% (± 12%) and 121%

(± 7%) from 2 to 24 wks in males and females, respectively. The maximum increase in total brain volume occurred about age 4 wks and 95% of whole brain growth occurred by age 21-23 wks. Logistical modeling suggests there are sexually dimorphic effects on brain growth. For example, in females, cortex was smaller (P = 0.04). Furthermore, the maximum growth of the hippocampus occurred about 5 wks earlier in females than males, and the window for hippocampal growth was significantly shorter in females than males (P = 0.02, P = 0.002 respectively) . These sexual dimorphisms are similar to what is seen in humans. In addition to providing important data on brain growth for pigs, this study shows pigs can be used to obtain longitudinal MRI data. The large increase in brain volume in the postnatal period is similar to human neonates and suggests pigs can be used to investigate brain development.

This material has been previously published. The copyright owner has provided permission to reprint. Conrad MS, Dilger RN, Johnson RW (2012) Brain growth of the domestic pig (Sus scrofa) from 2 to 24 weeks of age: A longitudinal MRI study. Dev Neuroscience 34(4) 291-8

144

B.2 Introduction

At birth the human infant brain is just 25% of adult size. It undergoes massive postnatal growth so by age 2 overall brain size reaches about 85% of adult volume (Knickmeyer, Gouttard et al. 2008). This period of accelerated brain growth may be the most important phase of postnatal brain development in humans, as it is the result of synaptogenesis, gliosis, and myelination (Rice and Barone 2000). The rapid growth and development of the brain is accompanied by an equally rapid development of cognitive and motor function (Forssberg 1999;

Rovee-Collier and Barr 2007). Thus, this early postnatal phase is recognized as a period of increased vulnerability to injury, and disruption of neurodevelopment by environmental insults may have long-lasting or permanent effects on brain structure and function. The idea that environmental insults in an early sensitive period affect behavior later is supported by studies showing that early-life stress increases the likelihood for several neuropsychological disturbances later in life (Hackman, Farah et al. 2010); and that iron deficiency between age 6 months and 2 years leads to long-term deficits in learning and memory (Graham, Heim et al.

1999; McCann and Ames 2007; Thomas, Grant et al. 2009). Nonetheless, very little is known about how human brain growth and development proceeds during this period let alone how insults interfere.

Quantitative magnetic resonance imaging (MRI) has emerged as a powerful tool for assessing human brain development. However, most large scale MRI studies of brain development have been done with older children (>4 years of age); and use of MRI in infants and younger children has mostly been restricted to those born preterm (Giedd, Snell et al. 1996;

Peterson, Anderson et al. 2003). Only recently, was MRI used to assess brain growth in healthy full term infants from birth to age 2 years (Knickmeyer, Gouttard et al. 2008). This study has provided an impressive view of the massive brain growth and development that occurs postnatal in healthy full term infants and underscores this as being an influential period. Although this application of MRI in human infants is impressive, the ability to address mechanistic questions

145 related to early-life insults cannot be readily done due to practical or ethical concerns.

Therefore, a pre-clinical animal model with brain growth and development similar to humans that can be used in MRI studies to investigate how different insults at critical periods of rapid brain growth affect development and function would be valuable.

Due to its anatomic and physiologic similarities with humans, the domestic pig (Sus scrofa) is a preferred pre-clinical model in several areas including pediatric nutrition and organ transplant surgery (Miller and Ullrey 1987; Dixon and Spinale 2009). Because pigs and humans are thought to share similar brain growth and development patterns, appeal for the pig as a model for human infant brain development has increased. The major brain growth spurt extends from late prenatal to early postnatal in both the pig and human, which is different from other common animal models (Dobbing and Sands 1979). Additionally, gross anatomical features including gyral pattern and distribution of grey and white matter of the neonatal porcine brain are similar to that of human infants (Dickerson and Dobbing 1967; Pond, Boleman et al.

2000; Lind, Moustgaard et al. 2007). The physical size of piglets also allows for quantitative structural MRI of the brain using clinical scanners. Structural MRI scanning and analysis protocols have been developed for neonatal and adult domestic pigs (Saikali, Meurice et al.

2010; Conrad, Dilger et al. 2012) and for adult Göttingen pigs (Jelsing, Rostrup et al. 2005).

More advanced techniques including, functional MRI and diffusion tensor imaging have also been used (Fang, Lorke et al. 2005; Munkeby, De Lange et al. 2008). However, we are unaware of previous studies where brain growth trajectory was assessed in pigs using MRI in a longitudinal study. Previously, our lab has developed MRI methods for quantifying brain region volumes in the neonatal piglet (Conrad, Dilger et al. 2012). The purpose here was to use these techniques to determine total brain and brain region volumes in a cohort of male and female domestic pigs on 7 occasions from age 2 wks to near sexual maturity at age 24 wks. The results using this noninvasive longitudinal approach confirm substantial postnatal brain growth in domestic pigs. Furthermore, using logistic modeling procedures to determine the maximum

146 volume and the age of maximum growth rate, we revealed a number of attributes of pig brain growth that are similar to humans, including evidence for sexual dimorphic effects.

B.3 Material and Methods

Subjects

A total of fifteen pigs, six intact male and nine female (Sus scrofa domestica, York breed), from seven litters, were obtained from the University of Illinois swine herd at two days of age and placed into an artificial rearing system that was previously described (Dilger and

Johnson 2010). Briefly, each piglet was housed individually in an acrylic-sided caging unit with a 12 hour light/dark cycle. Ambient temperature was maintained at 27˚C with intra-cage temperatures maintained at 32˚C using overhead radiant heaters. Piglets were provided a commercially available liquid milk replacer (Milk Specialties Global Animal Nutrition,

Carpentersville, IL) until 3 wks of age at which time they were moved to individual floor pens

(4.65 m2) and provided ad libitum access to water and a corn-soybean meal based diet formulated to provide recommended levels of all essential nutrients (NRC 1998). Pigs were weighed weekly to determine weight gain. All procedures were in accordance with the National

Institute of Health Guidelines for the Care and Use of Laboratory Animals and approved by the

University of Illinois Institutional Animal Care and Use Committee.

Image Acquisition

At 2 wks of age pigs were subjected to MRI scanning procedures to obtain brain images for in vivo estimation of volume. Before scanning, pigs were anesthetized by intramuscular injection of a telazol:ketamine:xylazine solution (TKX; 4.4 mg/kg body weight), and placed in a prone position in the MRI scanner. Pigs remained anesthetized during the MRI scanning procedure (approximately 20 min) and were returned to their pen afterwards where they were monitored until recovering from anesthesia. All pigs underwent additional MRI scans at 4, 8, 12,

147

16, 20, and 24 wks of age. At the conclusion of the 24 wk MRI scan, pigs were euthanized by intracardiac administration of an overdose of sodium pentobarbital (390 mg/ml Fatal Plus - administered at 1 ml/5 kg body weight).

Magnetic resonance scanning was conducted using a three-dimensional T1-weighted magnetization prepared gradient echo (MPRAGE) sequence. A 32-channel head coil was used for pigs up to 8 wks of age, and a flexible 6-channel coil was used thereafter. The parameters used were: repetition time, TR = 1900 ms; echo time, TE = 2.48 ms; inversion time, TI = 900 ms, flip angle = 9˚, matrix = 256 x 256 (interpolated to 512 x 512), slice thickness = 1.0 mm.

The final voxel size was 0.35 mm x 0.35 mm x 1.0 mm.

Brain Region Volume Estimation

Digital Imaging and Communications in Medicine (DICOM) images were imported and analyzed using three-dimensional visualization software (AMIRA®, Visage Imaging, Inc., San

Diego, CA). The whole brain, cortex, diencephalon, cerebellum, hippocampus, and brainstem were manually segmented using a Wacom Cintiq 21UX graphic input screen and pen (Wacom,

Vancouver, WA) and criteria previously described (Conrad, Dilger et al. 2012). Each region was manually segmented in the three anatomical planes using the atlas by Felix et al. (Félix, Léger et al. 1999), and regional volume estimates were calculated using the AMIRA© software. The reliability of this technique in pigs has been validated (Conrad, Dilger et al. 2012).

Statistics

All data analysis was done using SAS software (SAS, Cary, NC). Weight gain data were subjected to a repeated two-way mixed model ANOVA (Age × Sex). Brain region growth curves

148 were constructed for each brain area in each individual pig using the logistic growth model shown below (Grossman and Koops 2003) and the proc nlin function of SAS:

A brain region volume  1 e(ageC) / B

Parameter estimations were computed for maximum brain region volume (A), “duration” of growth (B), and age of maximum growth rate (C) for each pig. The “duration” of growth term

(B) is defined as B   3 /   where sigma is the standard deviation of the logistic function

(Gupta and Gnanadesikan 1966). Individual parameter estimations were then analyzed using a mixed model including sex, litter, and the interaction. For analysis of hippocampal asymmetries, hippocampal data were subjected to a two-way mixed model ANOVA (Sex × Hemisphere).

There were no significant differences due to hemisphere (data not shown); therefore left and right hippocampal volumes were combined to create a total hippocampal volume measure. The logistic growth model was further used to estimate the age when total brain and brain region volume was 50%, 75%, and 95% of maximum volume. Significance level was set at p < 0.05.

B.4 Results

Body Weight

Whole body weight for male and female pigs from 2 to 24 wks of age is shown in Figure

1. There were differences in body weight due to age (F(6,78) = 702.88, p = <0.0001), sex

(F(1, 13) = 14.25, p = 0.0023), and an Age x Sex interaction (F(6,78) = 2.51, p = 0.0283), with males weighing more than females beginning at 20 wks of age.

Total Brain and Brain Region Volumes

A representative set of DICOM images from an individual female pig across all ages is shown in Figure 2a, and DICOM images in Figure 2b show the brain of a 12 wk old female pig in three planes. Total brain volume and volume of cortex, hippocampus, diencephalon,

149 cerebellum, and brain stem were calculated after manual segmentation in three planes and the changes from 2 to 24 wks of age are shown in Figure 3 as are the growth curves estimated by logistic modeling. The estimations for maximum volume, “duration” of growth, and the age when maximum growth rate occurred are presented for both males and females in Table 1.

Total brain volume increased 130% (± 12%) and 121% (± 7%) from 2 to 24 wks in males and females, respectively. The maximum increase in total brain volume occurred when pigs were about 4 wks old and 95% of whole brain growth occurred by 21-23 wks of age. Cortical volumes changed similarly although the estimated maximum volume was greater for males than females (F(1,2) = 21.21, P = 0.04). In 2 wk old piglets, hippocampal volume (right and left hemispheres combined) was similar in males and females. However, the growth trajectory of the hippocampus was sex dependent and at 24 wks, the hippocampus was larger in males than females (F(1,2) = 141.60, P = 0.007). The maximum increase in hippocampal volume occurred earlier in females than males (3 wks of age vs. 8 wks of age; F(1,2) = 51.68, P = 0.02). Based on logistic modeling, 95% of hippocampal growth occurred in females by 24 wks of age whereas in males this milestone was not expected until after 39 wks of age (F(1,2) = 551.99, P = 0.002).

In the diencephalon, there were no significant differences in maximum volume, duration of growth, and age of maximum growth rate due to sex. For cerebellar growth, males showed a significantly larger maximum volume than females (F(1,2) = 35.98, P = 0.03), but there were no sex differences in duration of growth. In males the age of maximum growth rate occurred roughly one wk later than females (F(1,2) = 40.14, P = 0.02). The maximum volume of the brainstem was similar for males and females, but the age of maximum growth rate occurred about a wk later in males (F(1,2) = 75.53, P = 0.01).

From the logistic growth model we further estimated the age when total brain and brain region volumes were 50%, 75%, and 95% of maximum volumes (Table 2). For example, total brain volume of males was estimated to be 50% and 75% of maximum volume at 4 and 11 wks

150 of age, respectively. On the other hand, volume of the hippocampus in males was estimated to be 50% and 75% of maximum volume at about 8 and 20 wks of age, respectively.

B.5 Discussion

In the present study, each piglet was subjected to an MRI scan on seven occasions from

2 to 24 wks of age to determine brain growth throughout the neonatal period to near sexual maturity. The results using this noninvasive longitudinal approach confirm substantial postnatal brain growth, with the most rapid growth occurring at about age 4 wks when the brain is ~50% of maximum volume. The period of rapid brain growth continued to about age 12 wks, when growth rate began to slow. Results from two recent cross-sectional studies where MRI was used to estimate brain volume of pigs at different ages and weights, also suggest rapid brain growth in the postnatal period (Winter, Dorner et al. 2011; Conrad, Dilger et al. 2012). The large increase in total brain volume in the postnatal period is similar to human neonates and suggests this is a critical period where disruption of developmental processes by environmental insults can have long-lasting or permanent effects on brain structure and function. Therefore, the domestic pig may serve as a pre-clinical model for postnatal human brain development.

Quantitative MRI has yielded important information on brain development in early childhood and adolescence (Giedd, Snell et al. 1996), but because of potential health concerns for infants, few studies have focused on the period from birth to 4 years of age when dramatic brain development occurs. Most MRI studies of infants have been with those born prematurely or with significant health complications. Only recently was MRI used to assess structural brain development of healthy full term infants from birth to 2 years of age (Knickmeyer, Gouttard et al.

2008). The first year after birth, total brain volume more than doubled and hemispheric grey and white matter increased 149% and 11%, respectively. With total brain volume estimates for healthy infants and healthy adults (Lieberman, Tollefson et al. 2005) it was determined that total brain volume at 2-4 wks of age is ~36% of adult volume; and at 1 year and 2 years of age, total

151 brain volume is ~72% and 83% of adult volume, respectively (Knickmeyer, Gouttard et al.

2008). Thus, in healthy human infants there is enormous brain growth the first year after birth, making this a period of increased vulnerability to injury. Investigating questions related to early- life environmental insults and brain development and function is difficult in human infants due to obvious practical and ethical concerns. Thus, in addition to providing important data on brain growth trajectory for pigs, the present study shows pigs can be used to obtain longitudinal data in MRI studies. This indicates pigs can be used in MRI studies to investigate how different insults at critical periods of rapid brain growth affect development and function. This is particularly useful because unlike some common animal models, piglets can be tested in learning and memory tasks at an early age when brain growth is maximal (Dilger and Johnson

2010; Elmore, Dilger et al. 2012)

Despite substantial brain growth the first 2 years after birth, in humans total brain volume continues to increase until about puberty, with the total brain volume peaking at 10.5 years of age in females and 14.5 years of age in males (Lenroot, Gogtay et al. 2007); thereafter, total brain volume diminishes with age so total cerebral volume follows an inverted U-shaped trajectory (Giedd and Rapoport 2010). Maximum total brain volume in human males is 8-10% larger than females (Goldstein, Seidman et al. 2001). In the present study, total brain volume data obtained from pigs age 2 to 24 wks were best described using a logistic (sigmoidal) model.

Consistent with what has been reported for humans, the logistic model suggested maximum total brain volume for male pigs was ~8% larger than for females. The final MRI scan occurred near puberty and how total brain volume for pigs might change further into adulthood is not known.

From the present study, we cannot determine composition of brain growth. In humans, most neurogenesis and migration occurs early in the prenatal period (starting in the first month of gestation) with astrocyte and oligigodendrocyte proliferation extending from the late prenatal period into the postnatal period (Rice and Barone 2000). The increase in grey matter the first

152 few years of life is likely due to glial cell proliferation, dendritic and axonal arborization, and synapse formation, while white matter increases are due to myelination (Mrzljak, Uylings et al.

1991; Miller and Ono 1998). Similar to humans, the domestic pig does not have substantial neocortical neurogenesis in the postnatal period (Jelsing, Nielsen et al. 2006). This is different than the Göttingen minipig, which has significant neuronal and glia cell development in the postnatal period (Jelsing, Nielsen et al. 2006). Therefore, the substantial increase in brain volume seen in the current study is likely due to changes in the neuropil (dendrites, axons, glia) and myelination, but not neurogenesis. A limitation to our study is that grey and white matter volumes could not be determined due to the use of manual segmentation. However, a recent cross sectional study of domestic pigs suggests grey and white matter volume continues to increase at least through the first 12 wks of age (Winter, Dorner et al. 2011).

In humans, the cerebellum is 8-13% larger in males than females (Giedd, Snell et al.

1996; Tiemeier, Lenroot et al. 2010). Here we show that the cerebellum in adult male pigs is roughly 10% larger than in females. Also in humans, males show peak cerebellar volume later in life at 15.6 years of age compared to 11.8 in females (Tiemeier, Lenroot et al. 2010). We did not find any difference in the age range of cerebellar growth in pigs, but we did find that the peak growth rate occurred earlier in life for females. Developmental growth patterns of subcortical structures depend on the age range being studied. During early postnatal development, newborn through 2 years of age, subcortical structures develop similar in males and females (Knickmeyer, Gouttard et al. 2008). Later in development, from 5 to 18 years of age, subcortical structures including the putamen and globus pallidus have been shown to be larger in males (Giedd, Snell et al. 1996). In this study, we found that the diencephalon and brainstem have similar growth patterns in male and female pigs.

Studies on human hippocampal development have shown significant growth from birth through 2 years of age (Pfluger, Weil et al. 1999). At age 2 years, the hippocampus is about

85% of adult size but data for males and females have not been reported separately so if sexual

153 dimorphic effects are present early on is not clear. Sexual dimorphic effects are, however, present later. In females, e.g., maximum volume is smaller and achieved at a younger age compared to males (Giedd, Vaituzis et al. 1996; Pfluger, Weil et al. 1999). Our data suggest a similar sex-dependent growth trajectory for the pig hippocampus. For example, the period of maximum growth occurred about 5 wks earlier in females than males, and the window for hippocampal growth was significantly shorter in females than males. One feature of hippocampal development that is different between pigs and humans relates to left/right asymmetries. Multiple studies of human subjects have determined that the right hippocampus is significantly larger in both males and females, and this difference is present from birth (Giedd,

Vaituzis et al. 1996; Pfluger, Weil et al. 1999; Utsunomiya, Takano et al. 1999; Thompson,

Wood et al. 2009). In our study, no difference was found between the left and right hippocampal formation. The importance of hippocampal asymmetry is unknown, but asymmetries can be affected by environmental insults, and changes in asymmetries have been seen in patients with psychiatric illness (Stefanis, Frangou et al. 1999; Wang, Joshi et al. 2001).

Although we did not find differences, it may be an important measure to include in future studies.

The present study has several other limitations. First, the study included a small sample size, meaning the sexual dimorphisms in brain growth and development observed should be cautiously interpreted. However, our confidence is bolstered because variation was minimized by: (1) using a longitudinal design where each pig served as its own control, (2) including littermates of each sex when possible; and (3) maintaining pigs individually in an identical environment throughout the study. A second limitation was the use of manual segmentation for quantifying brain region volume. Technical challenges arise when imaging the neonatal brain including poor spatial resolution and low tissue contrast (Shi, Fan et al. 2010). In addition, there is no MRI-based atlas available for the neonatal piglet. Construction of an atlas set across multiple ages in the pig would allow for more complex automated segmentation protocols.

154

Nonetheless, this manual segmentation protocol has been used previously and is a reliable method for determining brain region volume changes in the pig (Conrad, Dilger et al. 2012). A third limitation is that pigs were only scanned through 24 wks of age, which is near sexual maturity (Lind, Moustgaard et al. 2007). Beyond this age, most domestic pigs are too large to fit in a standard MRI machine bore. This limitation precluded us from extending the study. Thus, use of an open bore magnet may be useful to track changes past this age in order to determine if there are brain volume loses after puberty as seen in many brain areas in the human (Lenroot,

Gogtay et al. 2007). Fourth, the data in this study suggests there are a few differences in brain growth in pigs compared to humans. For example, we did not find any sexual dimorphisms in maximum volumes of the diencephalon and brainstem regions; nor did we observe asymmetries in the hippocampus. These differences present a limitation of the pig as an animal model for human neurodevelopment.

In summary, the present study shows the normal brain growth pattern in the domestic pig from 2 to 24 wks of age. The brain undergoes significant growth during this period and several sexual dimorphisms affecting growth, including maximum volume, appear to be similar for humans and pigs. The study shows that domestic pigs can be used in longitudinal MRI- based studies to investigate brain growth and development, and suggests the pig can serve as a preclinical model for human neurodevelopment.

155

B.6 Figures and Tables

Figure B.1. Body weight of pigs from 2 to 24 wks of age. Body weight of male pigs higher than female pigs starting at 20 wks of age. The error bars indicate SEM. * = p < 0.01 ** = p < 0.001

156

Figure B.2. Representative MRI images displaying brain growth from 2 to 24 wks of age. The top images are T1-weighted axial sections (skull stripped) from a female pig at each age (A). The bottom images show the coronal, axial, and sagittal views left to right of a 12 wk old female pig (B). Segmentation is shown for cortex (white), diencephalon (green), hippocampi (blue), cerebellum (red), and brainstem (yellow).

157

Figure B.3. Brain region growth curves and logistic models. The graphs show total brain and brain region volumes for male and female pigs from 2 to 24 wks of age, as well as logistic growth models for male and female pigs. Error bars indicated SEM.

158

Table B.1. Average logistic growth parameter estimates by sex for each brain region. The averages by sex for maximum volume (A), “duration” of growth (B), and age of maximum growth rate (C) are indicated. Mixed model analysis was used to determine the main effect of sex on each parameter.

159

Percent of 50% 75% 95% Adult Volume Age Age Age SEM SEM SEM (weeks) (weeks) (weeks)

4.08 0.12 11.12 0.30 22.94 0.61 Total Brain Male Volume Female 3.81 0.14 10.30 0.30 21.19 0.57

Male 4.01 0.22 11.34 0.53 23.66 1.07 Cortex Female 3.70 0.19 10.22 0.39 21.19 0.74

Male 8.27 1.24 19.82 2.37 39.21 4.36 Whole Hippocampus Female 3.27 0.16 11.35 0.25 24.92 0.43

Male 4.36 0.22 10.74 0.45 21.47 0.87 Diencephalon Female 4.42 0.07 11.26 0.12 22.75 0.26

Male 5.54 0.16 11.97 0.43 22.76 0.89 Cerebellum Female 4.89 0.05 10.68 0.13 20.41 0.29

Male 5.95 0.46 14.44 0.92 28.69 1.72 Brainstem Female 4.61 0.18 11.92 0.46 24.20 0.93

Table B.2. Age when the brain volumes reach 50, 75, and 95% of maximum volume. Using the logistic models, estimates, with standard errors, were computed for the ages when different brain regions of each sex reach 50, 75, and 95% of adult brain volume for each measure. These data can be compared to human data to determine comparable growth timelines for designing experimental time windows.

160

B.7 References

Conrad, M. S., R. N. Dilger, et al. (2012). "Magnetic resonance imaging of the neonatal piglet brain." Pediatr Res 71(2): 179-184.

Dickerson, J. W. T. and J. Dobbing (1967). "Prenatal and postnatal growth and development of the central nervous system of the pig." Proc Biol Sci 166(1005): 384-395.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Dixon, J. A. and F. G. Spinale (2009). "Large animal models of heart failure: a critical link in the translation of basic science to clinical practice." Circ Heart Fail 2(3): 262-271.

Dobbing, J. and J. Sands (1979). "Comparative aspects of the brain growth spurt." Early Human Development 3(1): 79-83.

Elmore, M. R., R. N. Dilger, et al. (2012). "Place and direction learning in a spatial T-maze task by neonatal piglets." Anim Cogn 15(4): 667-676.

Fang, M., D. E. Lorke, et al. (2005). "Postnatal changes in functional activities of the pig's brain: a combined functional magnetic resonance imaging and immunohistochemical study." Neurosignals 14(5): 222-233.

Félix, B., M.-E. Léger, et al. (1999). "Stereotaxic atlas of the pig brain." Brain Res Bull 49(1-2): 1-137.

Forssberg, H. (1999). "Neural control of human motor development." Curr Opin Neurobiol 9(6): 676-682.

Giedd, J. N. and J. L. Rapoport (2010). "Structural MRI of Pediatric Brain Development: What Have We Learned and Where Are We Going?" Neuron 67(5): 728-734.

Giedd, J. N., J. W. Snell, et al. (1996). "Quantitative magnetic resonance imaging of human brain development: ages 4-18." Cereb Cortex 6(4): 551-560.

Giedd, J. N., A. C. Vaituzis, et al. (1996). "Quantitative MRI of the temporal lobe, amygdala, and hippocampus in normal human development: ages 4-18 years." J Comp Neurol 366(2): 223-230.

Goldstein, J. M., L. J. Seidman, et al. (2001). "Normal sexual dimorphism of the adult human brain assessed by in vivo magnetic resonance imaging." Cereb Cortex 11(6): 490-497.

Graham, Y. P., C. Heim, et al. (1999). "The effects of neonatal stress on brain development: Implications for psychopathology." Dev Psychopathol 11(03): 545-565.

Grossman, M. and W. J. Koops (2003). "Modeling extended lactation curves of dairy cattle: a biological basis for the multiphasic approach." J Dairy Sci 86(3): 988-998.

161

Gupta, S. S. and M. Gnanadesikan (1966). "Estimation of the parameters of the logistic distribution." Biometrika 53(3-4): 565-570.

Hackman, D. A., M. J. Farah, et al. (2010). "Socioeconomic status and the brain: mechanistic insights from human and animal research." Nat Rev Neurosci 11(9): 651-659.

Jelsing, J., R. Nielsen, et al. (2006). "The postnatal development of neocortical neurons and glial cells in the Gottingen minipig and the domestic pig brain." J Exp Biol 209(Pt 8): 1454-1462.

Jelsing, J., E. Rostrup, et al. (2005). "Assessment of in vivo MR imaging compared to physical sections in vitro--a quantitative study of brain volumes using stereology." Neuroimage 26(1): 57-65.

Knickmeyer, R. C., S. Gouttard, et al. (2008). "A structural MRI study of human brain development from birth to 2 years." J Neurosci 28(47): 12176-12182.

Lenroot, R. K., N. Gogtay, et al. (2007). "Sexual dimorphism of brain developmental trajectories during childhood and adolescence." NeuroImage 36(4): 1065-1073.

Lieberman, J. A., G. D. Tollefson, et al. (2005). "Antipsychotic Drug Effects on Brain Morphology in First-Episode Psychosis." Arch Gen Psychiatry 62(4): 361-370.

Lind, N. M., A. Moustgaard, et al. (2007). "The use of pigs in neuroscience: modeling brain disorders." Neurosci Biobehav Rev 31(5): 728-751.

McCann, J. C. and B. N. Ames (2007). "An overview of evidence for a causal relation between iron deficiency during development and deficits in cognitive or behavioral function." Am J Clin Nutr 85(4): 931-945.

Miller, E. R. and D. E. Ullrey (1987). "The pig as a model for human nutrition." Annu Rev Nutr 7: 361-382.

Miller, R. H. and K. Ono (1998). "Morphological analysis of the early stages of oligodendrocyte development in the vertebrate central nervous system." Microsc Res Techniq 41(5): 441- 453.

Mrzljak, L., H. B. M. Uylings, et al. (1991). Chapter 9 Neuronal development in human prefrontal cortex in prenatal and postnatal stages. Progress in Brain Research. C. G. V. E. J. P. C. D. B. M. A. C. H.B.M. Uylings and M. G. P. Feenstra, Elsevier. Volume 85: 185-222.

Munkeby, B. H., C. De Lange, et al. (2008). "A piglet model for detection of hypoxic-ischemic brain injury with magnetic resonance imaging." Acta Radiol 49(9): 1049-1057.

NRC (1998). Nutrient Requirements of Swine. Washington, DC, National Academic Press.

Peterson, B. S., A. W. Anderson, et al. (2003). "Regional Brain Volumes and Their Later Neurodevelopmental Correlates in Term and Preterm Infants." Pediatrics 111(5): 939- 948.

162

Pfluger, T., S. Weil, et al. (1999). "Normative volumetric data of the developing hippocampus in children based on magnetic resonance imaging." Epilepsia 40(4): 414-423.

Pond, W. G., S. L. Boleman, et al. (2000). "Perinatal ontogeny of brain growth in the domestic pig." Proc Soc Exp Biol Med 223(1): 102-108.

Rice, D. and S. Barone, Jr. (2000). "Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models." Environ Health Perspect 108 Suppl 3: 511-533.

Rovee-Collier, C. and R. Barr (2007). Infant Learning and Memory. Blackwell Handbook of Infant Development, Blackwell Publishing Ltd: 139-168.

Saikali, S., P. Meurice, et al. (2010). "A three-dimensional digital segmented and deformable brain atlas of the domestic pig." J Neurosci Meth 192(1): 102-109.

Shi, F., Y. Fan, et al. (2010). "Neonatal brain image segmentation in longitudinal MRI studies." Neuroimage 49(1): 391-400.

Stefanis, N., S. Frangou, et al. (1999). "Hippocampal volume reduction in schizophrenia: effects of genetic risk and pregnancy and birth complications." Biological Psychiatry 46(5): 697- 702.

Thomas, D. G., S. L. Grant, et al. (2009). "The role of iron in neurocognitive development." Dev Neuropsychol 34(2): 196-222.

Thompson, D. K., S. J. Wood, et al. (2009). "MR-determined hippocampal asymmetry in full- term and preterm neonates." Hippocampus 19(2): 118-123.

Tiemeier, H., R. K. Lenroot, et al. (2010). "Cerebellum development during childhood and adolescence: A longitudinal morphometric MRI study." NeuroImage 49(1): 63-70.

Utsunomiya, H., K. Takano, et al. (1999). "Development of the temporal lobe in infants and children: analysis by MR-based volumetry." AJNR Am J Neuroradiol 20(4): 717-723.

Wang, L., S. C. Joshi, et al. (2001). "Statistical Analysis of Hippocampal Asymmetry in Schizophrenia." NeuroImage 14(3): 531-545.

Winter, J. D., S. Dorner, et al. (2011). "Noninvasive MRI measures of microstructural and cerebrovascular changes during normal swine brain development." Pediatr Res 69(5 Pt 1): 418-424.

163

APPENDIX C

AN IN VIVO THREE-DIMENSIONAL MAGNETIC RESONANCE IMAGING-BASED AVERAGED BRAIN AND ATLAS OF THE NEONATAL PIGLET (SUS SCROFA)

C.1 Abstract

Due to the fact that morphology and perinatal growth of the piglet brain is similar to humans, use of the piglet as a translational animal model for neurodevelopmental studies is increasing. Magnetic resonance imaging (MRI) can be a powerful tool to study neurodevelopment in piglets, but many of the MRI resources have been produced for adult humans. Here, we present an average in vivo MRI-based atlas specific for the

4-week-old piglet. In addition, we have developed probabilistic tissue classification maps. These tools can be used with brain mapping software packages (e.g. SPM and

FSL) to aid in voxel-based morphometry and image analysis techniques. The atlas enables efficient study of neurodevelopment in a highly tractable translational animal with brain growth and development similar to humans.

164

C.2 Introduction

Use of the domestic pig (Sus scrofa) as a translational animal model for neuroscience research is increasing (Lind, Moustgaard et al. 2007). The pig is an attractive model because like humans, the major brain growth spurt extends from the late prenatal to the postnatal period (Dobbing and Sands 1979). Gross anatomical features, including gyral pattern and distribution of gray and white matter of the neonatal piglet brain are similar to that of human infants (Dickerson and Dobbing 1967; Thibault and Margulies 1998). Moreover, their physical size allows neuroimaging instruments designed for humans to be used with piglets. Indeed, structural magnetic resonance imaging (MRI), functional MRI, and positron emission tomography have all been conducted in pigs (Ishizu, Smith et al. 2000; Watanabe, Andersen et al. 2001; Fang,

Lorke et al. 2005; Jakobsen, Pedersen et al. 2006). Finally, due to their precocial nature, piglets can be weaned at birth or after caesarian delivery, maintained with relative ease, and used in behavioral testing paradigms to assess learning at an early age (Dilger and Johnson 2010; Elmore, Dilger et al. 2012). Thus, piglets represent a gyrencephalic species with brain growth similar to humans that can be used in highly controlled experiments to explore how environmental insults such as nutrient deficiencies, infection, or social stress affect brain structure and function.

Previously, we reported MRI techniques for quantifying brain region volumes in the neonatal piglet (Conrad, Dilger et al. 2012). These techniques were used to quantify normal brain growth of the domestic pig from 2- to 24-weeks of age (Conrad,

Dilger et al. 2012). During this period, brain volume increased 121-130% and several sexual dimorphisms were identified. For example, the maximum growth rate of the

165

hippocampus occurred 5 weeks earlier in females than males but the adult hippocampal

volume was greater in males, similar to patterns observed in humans (Giedd, Vaituzis

et al. 1996; Pfluger, Weil et al. 1999). In that study, labor-intensive manual

segmentation was used to estimate whole brain and brain region volume.

More advanced semi- and fully-automated structural analysis techniques, such

as voxel-based morphometry, have been used in human studies for over a decade

(Ashburner and Friston 2000). These techniques have also been used in rodents and

non-human primates (Sawiak, Wood et al. 2009; McLaren, Kosmatka et al. 2010). A

prerequisite to using these advanced methods in piglets is the availability of a

standardized atlas that serves as a template for registration and spatial normalization.

An example of a standardized template for humans is the MNI152 atlas (Evans, Collins

et al. 1993; Collins, Neelin et al. 1994). There are also atlases available for rodents and

non-human primates (Ma, Hof et al. 2005; McLaren, Kosmatka et al. 2009). A three-

dimensional digital brain atlas has been made for a 6-month old domestic pig using

high-resolution MRI and histological slices (Saikali, Meurice et al. 2010). Although this is a very good atlas for localizing specific brain areas in the adult pig, it is not appropriate for the young piglet because of the significant difference in brain size

(Conrad, Dilger et al. 2012). An atlas has also been created for the adult Göttingen minipig, but due to size differences between breeds (15-18 kg for an adult Göttingen minipig vs. 5-7 kg for a 4-week-old domestic pig), it also is not appropriate for use with domestic neonatal piglets (Watanabe, Andersen et al. 2001). In addition, the Göttingen minipig has significant neuronal and glial cell development in the postnatal period,

166

something not seen in humans and the domestic pig (Jelsing, Nielsen et al. 2006). This

suggests the domestic pig may be a better model for human neurodevelopment.

Therefore, our goal was to create an in vivo MRI-based atlas specifically for the 4-week-

old domestic piglet.

Here, we report the creation of a T1-weighted population-averaged template for

4-week-old piglets. Probabilistic tissue classification maps for gray matter, white matter,

and cerebrospinal fluid were also generated. Using the average template, we manually

segmented nineteen regions of interest to create the neonatal piglet atlas. These

packages, which are now publically available (http://pigmri.illinois.edu), will allow for

more sophisticated brain structure analysis, including voxel-based morphometry, in the

neonatal piglet.

C.3 Materials and Methods

Ethics Statement

All animal experiments were in accordance with the National Institute of Health

Guidelines for the Care and Use of Laboratory Animals and approved by the University

of Illinois at Urbana-Champaign Institutional Animal Care and Use Committee.

Subjects

Fifteen pigs, nine female and six male (Sus scrofa domestica, York breed), were

obtained from the University of Illinois swine heard. The pigs were placed into an

artificial rearing system 48-hours after birth (previously described by Dilger and Johnson

(Dilger and Johnson 2010)). Briefly, each pig was placed in an acrylic-sided cage (0.76

167

mL x 0.58 mW x 0.47 mH) and provided an enrichment toy and towel. Overhead

radiant heaters maintained the temperature at 27°C. Piglets were maintained on a 12-

hour light/dark cycle. Piglets were fed a commercially available piglet milk replacer

(Advance Liqui-wean, Milk Specialties Co., Dundee, IL) at 285 mL/kg body weight. The piglets received this milk in 18 meals during the day using an automated liquid feeding

system (Dilger and Johnson 2010). The average body weight of the pigs at 2 and 28

days of age was 1.67 (± 0.08) kg and 5.35 (± 0.39) kg, respectively. The animals were

part of a longitudinal study characterizing the normal brain development of the pig

(Conrad, Dilger et al. 2012).

Image Acquisition

The image acquisition procedures have been described previously (Conrad,

Dilger et al. 2012). Briefly, piglets were transported to the MRI facility where they were

anesthetized with an intramuscular injection of a telazol:ketamine:xylazine (TKX; 4.4

mg/kg body weight), and placed in a prone position in a Siemens Trio 3 T imager

(Siemens, Erlangen, Germany). A three-dimensional T1-weighted magnetization

prepared gradient echo (MPRAGE) sequence was used with a 32-channel coil

(Siemens 32-channel head coil). The sequence parameters were: repetition time, TR =

1900 ms; echo time, TE = 2.48 ms; inversion time, TI = 900 ms, flip angle = 9°, matrix =

256 x 256 (interpolated to 512 x 512), slice thickness 1.00 mm; this produced a voxel

size of 0.35 mm x 0.35 mm x 1.0 mm. A total of 192 slices were acquired. The images

used for creating the atlas were acquired when pigs were 4-weeks of age (average

body weight, 5.35 kg ±0.39).

168

Averaged Brain Creation

First, the Digital Imaging and Communication in Medicine (DICOM) images were reconstructed into 3D volumes for each pig. Using the 3D volumes, a binary mask was manually drawn over the brain tissue for each pig using the FSLView function of FSL

(Analysis Group, FMRIB, Oxford, UK) (Jenkinson, Beckmann et al. 2012). The mask was then used to extract the brain from the original data set. Next, the “fast” function of

FSL was used to create segmentation maps while also bias correcting for field inhomogeneity and intensity normalization. An extracted brain from one pig was chosen as the “template” data set and the other fourteen data sets were linearly registered with a rigid-body transformation using the coregister module of SPM (University College

London, London, UK). The aligned images were averaged using “SPM Image

Calculator” to create the “averaged template.” An additional iteration was conducted to linearly realign all of the original data sets to the averaged template to improve alignment.

This template represents the average of all of the data sets, but does not factor in the average shape or morphology of the brain. To compensate for this, we non-linearly registered all of the original brains using the Normalization module in SPM to this template to generate individual deformation fields for each animal. Many atlases, including human infant atlases, use different variations of non-linear registration to improve the averaged brain (Altaye, Holland et al. 2008; Shi, Fan et al. 2010). The deformation fields for all pigs were then averaged. The inverse of this average deformation field was then applied to the template in order to compensate for the

169

average brain shape. This resulted in an averaged brain with compensation for average

brain shape.

Tissue Probability Maps

The “fast” function in FSL does not require prior tissue maps as it uses a hidden

Markov random field model and an expectation-maximization algorithm (Zhang, Brady

et al. 2001). Using the “fast” function, we determined the partial volume tissue

estimation for each individual brain which had been linearly registered during the second iteration above. The partial volume estimates were used as they approximate the proportion of each tissue type in each voxel. Individual gray matter, white matter, and CSF partial volume estimates (PVE) maps were created for each pig. The deformation fields created for each pig during the normalization procedure during atlas construction were then applied to the PVEs to bring them into the averaged brain space.

The PVE were then averaged by tissue type to create the prior probability maps. These maps are provided as raw files and smoothed with a 1 mm full width at half maximum

(FWHM) Gaussian smoothing kernel (Evans, Kamber et al. 1994).

Manual Segmentation of Brain Regions of Interest

After completion of the averaged brain and tissue probability maps, both were reoriented along the anterior and posterior commissure (ac-pc) and cropped to reduce file size. The origin (zero reference point) was set to be the anterior limit of the posterior commissure in the midsagittal plane to be consistent with previously published MRI and histological atlases on adult pigs (Felix, Leger et al. 1999; Watanabe, Andersen et al.

170

2001; Saikali, Meurice et al. 2010). The origin can be seen in Figure 1. Using the

histological atlas by Felix et al. (Felix, Leger et al. 1999) as reference, nineteen regions

of interest (ROI) were identified and manually segmented on the averaged brain using

FSLView. Since no histological atlas is available for the piglet, these regions are larger, generalized regions that could be easily identified and segmented on the averaged brain. The ROIs were segmented in all three orthogonal views to increase accuracy. A three-dimensional reconstruction of the regions of interest was created using AMIRA®

(Visage Imaging, Inc., San Diego, CA.)

C.4 Results and Discussion

Averaged Brain and Tissue Probability Maps

Our objective was to develop a T1-weighted average brain specific for the neonatal piglet. This atlas was designed with features including high spatial resolution, a standard coordinate space, easy access and visualization, and the ability to be linked to current and future atlases, which are necessary for proper brain atlas construction

(Van Essen and Dierker 2007; McLaren, Kosmatka et al. 2009; Saikali, Meurice et al.

2010). Using MRI data from fifteen animals, a series of linear and non-linear transformations were used to create the averaged brain. Axial slices through an individual case and the averaged brain are shown in Figure 2. In addition, tissue probability maps were created for gray matter, white matter and cerebrospinal fluid. A coronal comparison of the three tissue types is shown in Figure 3. The tissue probability maps give the probability that a voxel is classified as one of the three tissue types.

171

Brain Regions

Using the neonatal piglet averaged brain, we manually segmented nineteen regions. A list of these structures can be found in Table 1. These regions are provided in a single NIfTI file containing corresponding region labels, as well as individual binary files. Figure 4 shows a cross section of the brain highlighting a few regions of interest.

The combined image file can be deformed to new data sets to allow for region identification using automated procedures. In addition, the individual region of interest masks can be deformed to allow for quantitative volume estimation for each brain region.

Applications

The goal of this study was to create an averaged brain for the neonatal piglet that will allow for more sophisticated analysis including voxel-based morphometry. The averaged brain serves as a starting template for voxel-based morphometry and can be used with any software that uses the NIfTI file type. In addition, the averaged brain can serve as a standard template for multimodal registration, including diffusion tensor imaging. In data sets including diffusion tensor imaging, the atlas can also serve to create predefined regions of interest for seeding or interrogation of imaging metrics.

Tissue probability maps serve as a priori inputs for tissue class segmentation.

We used the “fast” function from FSL to create tissue classifications for individual piglets as this does not require priors. The resultant probability maps generated can be used with the “Segment” function of SPM, as this software uses the priors in its segmentation algorithm, and FSL which can also incorporate priors if they are available (Ashburner and Friston 2000; Ashburner and Friston 2005).

172

Atlas Availability

The T1-weighted averaged brain, tissue probability maps, and ROI labels are freely available at http://pigmri.illinois.edu. The files are freely distributed and fall under the

Illinois Open Source License.

173

C.5 Figures and Tables

Figure C.1. Location of origin for the atlas. Sagittal, coronal, and axial views of the origin location for the atlas.

174

Figure C.2. Comparison of an individual MRI data set and the atlas. The axial slices are from -18 mm to 18 mm in 6 mm increments. The top row is a representative individual case and the bottom row is the atlas.

175

Figure C.3. Tissue probability maps. Coronal slices through the origin showing the atlas T1 image, gray matter, white matter, and cerebrospinal fluid (CSF) tissue probability maps. The maps have been smoothed with a 1 mm FWHM Gaussian filter.

176

Figure C.4. Regions of interest. Sagittal, coronal, and axial slices through the origin showing manually segmented regions of interest.

177

Label Number Brain Region 1 Caudate 2 Cerebellum 3 Left Cortex 4 Right Cortex 5 Lateral Ventricle 6 Third Ventricle 7 Cerebral Aqueduct 8 Fourth Ventricle 9 Left Hippocampus 10 Right Hippocampus 11 Medulla 12 Midbrain 13 Pons 14 Putamen and Globus Pallidus 15 Hypothalamus 16 Thalamus 17 Olfactory Bulb 18 Corpus Callosum 19 Internal Capsule

Table C.1. The regions of interest found in the atlas and their respective label numbers.

178

C.6 References

Altaye, M., S. K. Holland, et al. (2008). "Infant brain probability templates for MRI segmentation and normalization." NeuroImage 43(4): 721-730.

Ashburner, J. and K. J. Friston (2000). "Voxel-Based Morphometry—The Methods." NeuroImage 11(6): 805-821.

Ashburner, J. and K. J. Friston (2005). "Unified segmentation." NeuroImage 26(3): 839-851.

Collins, D. L., P. Neelin, et al. (1994). "Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space." J Comput Assist Tomogr 18(2): 192-205.

Conrad, M. S., R. N. Dilger, et al. (2012). "Brain Growth of the Domestic Pig (Sus scrofa) from 2 to 24 Weeks of Age: A Longitudinal MRI Study." Dev Neurosci 34(4): 291-298.

Conrad, M. S., R. N. Dilger, et al. (2012). "Magnetic resonance imaging of the neonatal piglet brain." Pediatr Res 71(2): 179-184.

Dickerson, J. W. T. and J. Dobbing (1967). "Prenatal and postnatal growth and development of the central nervous system of the pig." Proc Biol Sci 166(1005): 384-395.

Dilger, R. N. and R. W. Johnson (2010). "Behavioral assessment of cognitive function using a translational neonatal piglet model." Brain Behav Immun 24(7): 1156-1165.

Dobbing, J. and J. Sands (1979). "Comparative aspects of the brain growth spurt." Early Hum Dev 3(1): 79-83.

Elmore, M. R., R. N. Dilger, et al. (2012). "Place and direction learning in a spatial T-maze task by neonatal piglets." Anim Cogn 15(4): 667-676.

Evans, A. C., D. L. Collins, et al. (1993). 3D statistical neuroanatomical models from 305 MRI volumes. Nuclear Science Symposium and Medical Imaging Conference, 1993., 1993 IEEE Conference Record.

Evans, A. C., M. Kamber, et al. (1994). An MRI-Based Probabilistic Atlas of Neuroanatomy. Magnetic Resonance Scanning and Epilepsy. S. D. Shorvon, D. R. Fish, F. Andermann, G. M. Bydder and H. Stefan, Springer US. 264: 263-274.

Felix, B., M. E. Leger, et al. (1999). "Stereotaxic atlas of the pig brain." Brain Res Bull 49(1-2): 1-137.

Giedd, J. N., A. C. Vaituzis, et al. (1996). "Quantitative MRI of the temporal lobe, amygdala, and hippocampus in normal human development: ages 4-18 years." J Comp Neurol 366(2): 223-230.

179

Ishizu, K., D. F. Smith, et al. (2000). "Positron emission tomography of radioligand binding in porcine striatum in vivo: Haloperidol inhibition linked to endogenous ligand release." Synapse 38(1): 87-101.

Jakobsen, S., K. Pedersen, et al. (2006). "Detection of α2-Adrenergic Receptors in Brain of Living Pig with 11C-Yohimbine." J Nucl Med 47(12): 2008-2015.

Jelsing, J., R. Nielsen, et al. (2006). "The postnatal development of neocortical neurons and glial cells in the Gottingen minipig and the domestic pig brain." J Exp Biol 209(Pt 8): 1454-1462.

Jenkinson, M., C. F. Beckmann, et al. (2012). "FSL." NeuroImage 62(2): 782-790.

Lind, N. M., A. Moustgaard, et al. (2007). "The use of pigs in neuroscience: modeling brain disorders." Neurosci Biobehav Rev 31(5): 728-751.

Ma, Y., P. R. Hof, et al. (2005). "A three-dimensional digital atlas database of the adult C57BL/6J mouse brain by magnetic resonance microscopy." Neuroscience 135(4): 1203-1215.

McLaren, D. G., K. J. Kosmatka, et al. (2010). "Rhesus macaque brain morphometry: a methodological comparison of voxel-wise approaches." Methods 50(3): 157-165.

McLaren, D. G., K. J. Kosmatka, et al. (2009). "A population-average MRI-based atlas collection of the rhesus macaque." NeuroImage 45(1): 52-59.

Pfluger, T., S. Weil, et al. (1999). "Normative volumetric data of the developing hippocampus in children based on magnetic resonance imaging." Epilepsia 40(4): 414-423.

Saikali, S., P. Meurice, et al. (2010). "A three-dimensional digital segmented and deformable brain atlas of the domestic pig." J Neurosci Meth 192(1): 102-109.

Sawiak, S. J., N. I. Wood, et al. (2009). "Voxel-based morphometry in the R6/2 transgenic mouse reveals differences between genotypes not seen with manual 2D morphometry." Neurobiol Dis 33(1): 20-27.

Shi, F., Y. Fan, et al. (2010). "Neonatal brain image segmentation in longitudinal MRI studies." Neuroimage 49(1): 391-400.

Thibault, K. L. and S. S. Margulies (1998). "Age-dependent material properties of the porcine cerebrum: effect on pediatric inertial head injury criteria." J Biomech 31(12): 1119-1126.

Van Essen, D. C. and D. L. Dierker (2007). "Surface-Based and Probabilistic Atlases of Primate Cerebral Cortex." Neuron 56(2): 209-225.

Watanabe, H., F. Andersen, et al. (2001). "MR-Based Statistical Atlas of the Göttingen Minipig Brain." NeuroImage 14(5): 1089-1096.

Zhang, Y., M. Brady, et al. (2001). "Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm." IEEE T Med Imaging 20(1): 45-57.

180