UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Regulation of innate immune ontogeny : from preterm neonates to adults Kan, Bernard 2018

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

Item Metadata

Download

Media
24-ubc_2018_may_kan_bernard.pdf [ 5.28MB ]
Metadata
JSON: 24-1.0365287.json
JSON-LD: 24-1.0365287-ld.json
RDF/XML (Pretty): 24-1.0365287-rdf.xml
RDF/JSON: 24-1.0365287-rdf.json
Turtle: 24-1.0365287-turtle.txt
N-Triples: 24-1.0365287-rdf-ntriples.txt
Original Record: 24-1.0365287-source.json
Full Text
24-1.0365287-fulltext.txt
Citation
24-1.0365287.ris

Full Text

  REGULATION OF INNATE IMMUNE ONTOGENY: FROM PRETERM NEONATES TO ADULTS  by Bernard Kan B.Sc., The University of British Columbia, 2009 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Experimental Medicine) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2018 © Bernard Kan, 2018   ii  Abstract  Innate immunity is the first line of defense against infection, and is particularly important in newborns, as they lack immunological memory. During fetal development, innate immunity must be carefully regulated to prevent miscarriage. In contrast, upon reaching term, the innate immune system of the newborn must rapidly become operational to provide protection against exposure to the extra-uterine microbial environment.  Human immune responses are generally highly variable among individuals of all ages. According to current models, immune reactivity is highly influenced by an individual’s genetic make-up. However, studies in my thesis suggest that immune reactivity is also drastically influenced by non-genetic factors. My overall goal during my PhD was to understand the mechanisms regulating innate immune reactivity in humans, across the development spectrum, to better understand why some individuals are more susceptible to significant clinical infections. I hypothesized that examining responses at the systems level would inform me on how innate immune reactivity is regulated throughout life. Candida species (spp) are common neonatal pathogens. Despite the clinical importance of these pathogens, relatively little is known about the maturation of anti-fungal innate immune defenses in newborns. In Chapter 2 of my thesis, I examined innate immune responses to Candida spp. in preterm infants. I discovered that cellular metabolism plays a major role in regulating immune reactivity during fetal life, via regulation of protein translation. In Chapter 3 of my thesis, I applied similar methods to understand the factors driving the variability in innate immune responses in healthy adults. As expected, I found a large iii  diversity in immune responses between individuals. Surprisingly, some of these protein level responses were largely independent of gene transcription events. I provide evidence that metabolic pathways also modulate immune reactivity in healthy adults. Overall, my findings enhance our understanding of the factors regulating immune responses in the highly genetically diverse human population, providing insight into the development of these pathways in the late fetal/early neonatal period, and support a major role for metabolism in regulating immune reactivity in the general population and during ontogeny.    iv  Lay summary   The innate immune system is an important first line of defense against infection. In babies born prematurely, these defenses are immature, resulting in an increased risk of infection. The purpose of my research is to understand why some individuals are more susceptible to infections and how the immune system of preterm babies matures over their development. In the first part of my thesis, I found that the same mechanisms that regulate energy production in cells also play an important role in regulating how strongly immune defenses will react in premature babies. In the second part of my thesis, I show that similar mechanisms also affect the extent to which individuals in the general adult population respond immunologically. Overall, my studies demonstrate that metabolic factors control the susceptibility of babies, but also healthy adults, to infection and suggest some directions to help reduce the burden of infections in preterm babies.   v  Preface   Experiments in this thesis were conceived by B. Kan with supervision by P. M. Lavoie, as detailed:  A version of Chapter 1 was published as follows: B. Kan, H. R. Razzaghian, and P. M. Lavoie, “An Immunological Perspective on Neonatal Sepsis,” Trends Mol. Med., vol. 22, no. 4, pp. 290–302, 2016. A version of Chapter 2 has been submitted to Nature Communications and is currently in the process of peer review. Epidemiological data from figure 2.34 were provided by Dr. Prakesh Shah from the Canadian Neonatal Network. For experiments in figure 2.2 to 2.7, 2.23, 2.25), samples were collected and ELISAs were performed by B. Kan, C. Michalski, and K. Lee. Experiments were designed by B. Kan and C. Michalski. For microarray experiments in figure 2.10 to 2.17, experiments were designed by B. Kan, with input from A. A. Sharma, the Ross Lab and P. M. Lavoie. Samples were collected by B. Kan and K. Lee. RNA libraries were generated by F. Miao (C. Ross Lab), analyses were performed, and figures created by B. Kan and C. Michalski. For the multiplex ELISA experiments in figure 2.16 and 2.3, experiments were designed by B. Kan and K. Lee, and the figure itself was made by B. Kan. Real-time qPCR experiments in figure 2.19 was designed by B. Kan and K. Lee. Polysome experiments in figure 2.18 and 2.20 were designed by B. Kan, C. Michalski, H. Au, with input from R. G. Mirmira and E. Anderson-Baucum, and under supervision by E. Jan and P. M. Lavoie; Samples for this experiment were collected by B. Kan, and C. Michalski. Experiments were performed by C. Michalski with assistance from H. Au. B. vi  Kan and C. Michalski designed the qPCR primers. RT-PCR was done by C. Michalski, qPCR was done by B. Kan and C. Michalski. Figures were generated by B. Kan and C. Michalski. Phagocytosis assays (figure 2.1, 2.8, 2.9, 2.26, 2.31) were designed and run by B. Kan with assistance from H. Fu, samples for these experiments were also collected by B. Kan. Glucose uptake experiments (figure 2.30) were designed and performed by C. Michalski, with input from B. Kan. The figure was made by C. Michalski. Seahorse experiments (figure 2.28) were designed by B. Kan with assistance from M Aharoni-Simon, experiments were performed by B. Kan and C. Michalski. Lactate secretion ELISA experiments (figure 2.29) were performed by C. Michalski with samples collected by B. Kan. The figure was created by C. Michalski. 2-DG blocking experiment (figure 2.27) was designed and performed by B. Kan and C. Michalski. 35S-Met/Cys pulse-labelling experiments (figure 2.21) were designed and performed by B. Kan and C. Michalski, with help from H. Au. Figure A.2 detailing downregulation of electron transport chain component genes in preterm monocytes was prepared by C. Michalski, using data generated from microarray experiments. Chapter 3 includes experiments designed by A. A. Sharma, B. Kan, and K. Lee (as detailed below) with supervision from P.M. Lavoie. Data were analyzed and interpreted by A. A. Sharma, and B. Kan, figures were generated by B. Kan, and A. A. Sharma. Subject recruitment was done by A. A. Sharma, B. Kan, and K. Lee. Processing of blood was performed by A. A. Sharma, B. Kan, and K. Lee. Flow cytometry and ELISA experiments were designed and performed by B. Kan, and K. Lee. Transcriptome experiments (RNA extraction, cDNA synthesis, and microarray experiments) were designed and performed by A. A. Sharma, B. Kan, and K. Lee, with assistance from F. Miao.  vii  Table of contents  ABSTRACT ......................................................................................................................................... ii LAY SUMMARY ................................................................................................................................. iv PREFACE ........................................................................................................................................... v TABLE OF CONTENTS ...................................................................................................................... vii LIST OF TABLES.................................................................................................................................. x LIST OF FIGURES ............................................................................................................................... xi LIST OF ILLUSTRATIONS .................................................................................................................. xiv ACKNOWLEDGEMENTS ................................................................................................................... xv DEDICATION ................................................................................................................................... xvi CHAPTER 1: INTRODUCTION ............................................................................................................ 1 1.1 Clinical impact of infections in newborns ................................................................................... 1 1.2 A brief overview of the immune system ..................................................................................... 1 1.3 Fungal infections in preterm neonates ....................................................................................... 4 1.4 Recognition of Candida spp. by immune cells ............................................................................. 5 1.5 Immune development in humans ............................................................................................... 6 1.6 Development of mucosal immune defenses ............................................................................. 10 1.7 Lessons from neonatal mouse models ...................................................................................... 14 1.8 Considerations unique to the study of immune functions in premature infants ....................... 16 1.9 Immune regulation during ontogeny ........................................................................................ 17 1.10 Metabolic regulation of immune reactivity .............................................................................. 18 viii  1.11 Role of mTOR and HIF-1α ......................................................................................................... 20 1.12 Regulation of protein synthesis ................................................................................................ 21 1.13 Regulation of innate immune function in healthy adults .......................................................... 22 1.14 Hypothesis and objectives ........................................................................................................ 23 CHAPTER 2: CELLULAR METABOLISM BROADLY SUPPRESSES ANTI-FUNGAL INNATE IMMUNE DEFENSES IN HUMAN NEONATES .................................................................................................. 28 2.1 Background .............................................................................................................................. 28 2.2 Methods ................................................................................................................................... 31 2.2.1 Cohorts ................................................................................................................................... 31 2.2.2 Reagents ................................................................................................................................. 31 2.2.3 ELISA and blocking cytokine experiments .............................................................................. 33 2.2.4 Real-time qPCR experiments and data analysis ..................................................................... 33 2.2.5 Pulse-labelling experiments ................................................................................................... 34 2.2.6 Cytokine stimulation .............................................................................................................. 34 2.2.7 Phagocytosis assay ................................................................................................................. 35 2.2.8 Western Blots ......................................................................................................................... 35 2.2.9 Polysome profiling, RT-PCR and pulse labelling ..................................................................... 36 2.2.10 Metabolic assays .................................................................................................................. 36 2.2.11 Microarray analyses ............................................................................................................. 37 2.2.12 ClueGO Analysis ................................................................................................................... 37 2.2.13 Statistical Analysis ................................................................................................................ 38 2.3 Results ...................................................................................................................................... 39 2.3.1 Lack of anti-fungal innate immune recognition in early gestation ........................................ 39 2.3.2 Transcriptome analyses indicate broad metabolic impairments in preterm monocytes ...... 50 2.3.3 Defective translation of key immune response genes in preterm monocytes ...................... 61 2.3.4 MALT1 is required for immune recognition of Candida spp. in human monocytes .............. 71 2.3.5 Altered cellular energy metabolism and protein synthesis in preterm monocytes ............... 74 2.3.6 Developmental regulation of mTOR ...................................................................................... 80 2.3.7 Gestational age-dependent immune recognition and risk of invasive Candida infections ... 85 2.4 Discussion................................................................................................................................. 90 CHAPTER 3: IMMUNE VARIABILITY IN HEALTHY ADULTS ................................................................ 95 3.1  Background .............................................................................................................................. 95 3.2  Methods ................................................................................................................................... 97 3.2.1 Recruitment of human subjects and blood sample collection ............................................... 97 3.2.2 Standard methods to determine variability ........................................................................... 97 3.2.3 Cell purification, stimulation and cytokine detection by ELISA .............................................. 98 3.2.4 Detection of intracellular pro-IL-1 production and caspase-1 activation ............................ 99 ix  3.2.5 Real-time PCR experiments .................................................................................................. 100 3.2.6 Gene expression microarray analysis ................................................................................... 101 3.2.7 Pathway analysis .................................................................................................................. 101 3.2.8 Statistical analyses ............................................................................................................... 102 3.3 Results .................................................................................................................................... 103 3.3.1 Stability of immunological traits among individuals ............................................................ 103 3.3.2 Biological contribution of the variance in individuals’ immunological traits ....................... 108 3.3.3 Co-expressed transcriptomic networks after LPS stimulation ............................................. 111 3.3.4 Cluster of co-expressed genes predictive of IL-1 and IL-6 outcomes................................. 116 3.3.5 Transcriptome events predictive of IL-1β and IL-6 outcomes .............................................. 120 3.4 Discussion............................................................................................................................... 127 CHAPTER 4: INTERPRETATION, SIGNIFICANCE AND FUTURE DIRECTIONS .................................... 132 4.1 Interpretation ......................................................................................................................... 132 4.2  Significance ............................................................................................................................ 136 4.3  Contribution to current literature ........................................................................................... 139 4.4  Limitations and future directions ........................................................................................... 140 BIBLIOGRAPHY ............................................................................................................................. 143 APPENDIX ..................................................................................................................................... 158     x  List of tables  Table 2.1 Gestational Age-Related Prevalence of Infections ...................................................................... 86 Table 2.2 Primer Sequence Used in qPCR Gene Amplification. .................................................................. 89  Table 3.1 Top 25 Statistically Significant Differentially Expressed Genes (LPS vs Unstim) ....................... 113 Table 3.2 Top 20 Pathways Associated with Differentially Expressed Genes (LPS versus Unstim; Gene Set Enrichment Analysis) ................................................................................................................................. 115 Table 3.3 Top 2 Pathways Associated with Genes in Each Main Cluster in Figure 3.8 ............................. 118 Table 3.4 Top 10 Pathways Associated with Genes Predictive of IL-1 or IL-6 outcomes ....................... 123 Table 3.5 Primer Sequences Used for PCR Amplification ......................................................................... 126    xi  List of figures   Figure 2.1 Phagocytosis of C. albicans by CD14+ Monocytes ..................................................................... 41 Figure 2.2 Inflammatory Cytokine Secretion in Response to Various Species of Candida ......................... 42 Figure 2.3 Cytokine Secretion in Response to Various Species of Candida ................................................ 43 Figure 2.4 IL-1β Secretion in Response to Specific PRR Stimulation........................................................... 44 Figure 2.5 IL-6 Production in Response to Specific PAMP Stimulation ....................................................... 45 Figure 2.6 Pro-IL-1β Production in CD-14+ Monocytes ............................................................................... 46 Figure 2.7 IL-1β Production Blocked by Dectin-1 Specific Neutralizing Antibody ...................................... 47 Figure 2.8 Phagocytosis of C. albicans is Not Dependant on Dectin-1 ....................................................... 48 Figure 2.9 Effect of Blocking of Common Phagocytic Receptors on Phagocytosis of Candida ................... 49 Figure 2.10 Pathway Analysis of Differentially Expressed Genes in Preterm Monocytes .......................... 53 Figure 2.11 All Age-Related Differentially Expressed Gene Pathways in Unstimulated Monocytes. ......... 54 Figure 2.12 Heatmap of Gene Components of Major Metabolic Pathways ............................................... 55 Figure 2.13 Heatmap and Unsupervised Clustering of Genes Encoding Ribosomal Proteins .................... 56 Figure 2.14 Principal Component Analysis of Unstimulated Versus LPS Stimulated Monocytes ............... 57 Figure 2.15 Venn Diagram of LPS-Induced Differentially Expressed Genes ............................................... 58 Figure 2.16 Preterm Gene Expression Levels Are Not Correlated with Protein Secretion ......................... 59 Figure 2.17 Cytokine/Chemokine Gene Expression Changes Following LPS-Stimulation in Monocytes ... 60 Figure 2.18 Absorbance Profiles of Polysome Fractions ............................................................................. 64 Figure 2.19 Gene Expression of Signalosome Components ....................................................................... 65 Figure 2.20 Quantification of Signalosome Genes (qPCR) in Total RNA Fractions ..................................... 66 Figure 2.21 Reduced 35S-met/cys Incorporation in Preterm Monocytes.................................................... 67 Figure 2.22 Protein Expression of Dectin-1 ................................................................................................ 68 xii  Figure 2.23 Protein Expression of Bcl-10 And MALT1................................................................................. 69 Figure 2.24 Protein Expression of Syk and Card9 ....................................................................................... 70 Figure 2.25 MALT1 is Essential For Dectin-1 Signaling ............................................................................... 72 Figure 2.26 Phagocytosis is Independent of MALT-1 Paracaspase Activity ................................................ 73 Figure 2.27 Blocking Glycolysis Limits Cytokine Secretion ......................................................................... 75 Figure 2.28 Reduced Glycolytic Activity in Preterm Monocytes ................................................................. 76 Figure 2.29 Reduced Lactate Secretion in Preterm Mononuclear Cells ..................................................... 77 Figure 2.30 Glucose Uptake in CD14+ Monocytes ...................................................................................... 78 Figure 2.31 Effect of Blocking Glycolysis or Translation on Phagocytosis and IL-1β Secretion .................. 79 Figure 2.32 Western Blot of mTOR Phosphorylation and 4EBP1 Expression ............................................. 83 Figure 2.33 Increased Gene Expression of mTOR Regulators in Preterm Monocytes ................................ 84 Figure 2.34 Rates of Candidemia in Preterm Infants .................................................................................. 87 Figure 2.35 Candida Induced IL-1β Production by Gestational Age ........................................................... 88 Figure 3.1 Intra-Individual Correlations Between LPS-Stimulated IL-1β and IL-6 Responses in Adults Over 1 month ..................................................................................................................................................... 105 Figure 3.2 Impact of Age and Sex On IL-1β and IL-6 Outcome Measures ................................................ 106 Figure 3.3 Variability of Caspase-1 Activity by Sex Over One Month ....................................................... 107 Figure 3.4 Estimates of Technical versus Biological Variability in Outcome Measures. ........................... 109 Figure 3.5 Correlation Between IL-6 And IL-1β Outcome Measures Following LPS Stimulation. ............. 110 Figure 3.6 Differentially Expressed Genes Associated with LPS stimulation ............................................ 112 Figure 3.7 Hierarchal Clustering of Chemokine/Cytokine Genes Based on Association with LPS Stimulated Transcriptome ........................................................................................................................................... 114 Figure 3.8 Co-Expression Matrix-Based Network of Unstimulated and LPS Stimulated Genes ............... 117 xiii  Figure 3.9 Significance of Associations Between Expression Levels of Genes Composing Each of the Main Gene Clusters (Figure 3.8) and IL-1 or IL-6 Outcomes. ........................................................................... 119 Figure 3.10 Number of Genes Associated with IL-1β and IL-6 Outcomes (LPS) ....................................... 122 Figure 3.11 Hierarchal Clustering of LPS Stimulated IL-1β and IL-6 Outcomes with Unstimulated (blue) or Stimulated (red) Transcriptome ................................................................................................................ 124 Figure 3.12 Significance of Associations Between Expression Levels of Genes Associated with IL-1 Secretion, and Other IL-1/IL-6 Outcomes ............................................................................................... 125 Figure A.1 Representative Gating Strategy for Pro-IL-1β MFI .................................................................. 158 Figure A.2 Age-Related Changes in Expression of Genes Involved in Mitochondrial Electron-Transport 159    xiv  List of illustrations  Illustration 1.1 Immune Ontogeny .............................................................................................................. 24 Illustration 1.2 Metabolic Adaptation Occurring During Immune Activation. ............................................ 26 Illustration 2.1 Dectin-1 Signalling Via MALT1 Signalosome ...................................................................... 63 Illustration 2.2 mTOR Links Immune Activation to Metabolism and Protein Synthesis ............................. 82 Illustration 3.1 Determining Variability in Functional Immune Responses .............................................. 104   xv  Acknowledgements  Many thanks to the various members of the Lavoie lab for their help over the years. Mihoko Ladd for keeping the lab running efficiently and organized. Ashish Sharma for helping me develop the ideas contained within my thesis, as well as showing me how to think like a scientist. Elizabeth Marchant for all her insights and help throughout the years, and for being a sympathetic ear and sounding board for ideas. Christina Michalski for her assistance in many of the experiments in this thesis, as well as continuing to pursue the ideas developed in this thesis. Kelsey Lee, Helen Fu, Shannon Tang, Hong Li, and Cheryl Wu for their assistance in many of the experiments performed, as well as their assistance in collecting cord blood. A special thanks to the medical staff who aided in sample collection, and to the families of patients who supported this research through the donation of samples. Finally, thank you to Dr. Pascal Lavoie for the years of guidance and mentorship; thank you for always encouraging me to push past my limits.   xvi  Dedication     This thesis is dedicated to my fiancée Kelsey, who has been a constant source of love, strength, and encouragement throughout my graduate studies. And to my parents, who have always supported and guided me. 1  Chapter 1: Introduction   1.1 Clinical impact of infections in newborns   Worldwide, approximately one million neonates (newborns under 28 days of age) die of infection each year. Despite broad, concerted international efforts, infections continue to be a major cause of neonatal mortality [1]. Improving global access to medical care, together with the continued implementation of sanitary measures, could substantially reduce neonatal mortality. However, despite optimal medical care, neonatal infections remain prevalent due to the unique developmental stage and vulnerability of the newborn immune system [2]. To develop therapeutic interventions that can prevent and treat infections in this patient population, we must first gain an understanding of the sources of their vulnerability. Recent technological advances, such as multi-parameter flow cytometry, systems immunology, and next-generation sequencing offer new opportunities to understand and dissect these biological processes in human infants.   1.2 A brief overview of the immune system   The innate immune system is the first line of defence against infection. Cells of the innate system include primarily myeloid cells such as macrophages, neutrophils, basophils, eosinophils, monocytes, and certain subtypes of dendritic cells. The adaptive immune system is primarily comprised of B and T lymphocytes. The hallmark of the adaptive immune system is its 2  ability to develop an “immunological memory” to a wide variety of antigens, thereby resulting in improved immune responses over subsequent pathogen encounters. Like the adaptive immune system that can develop a memory, the innate immune system can also be “trained”, adapting its functional responses upon subsequent pathogen encounters [3].  For the adaptive immune system, the development of an immunological memory proceeds through affinity maturation through hyper-mutation of nucleotide sequences in the hypervariable regions of genes encoding T cell and B cell receptors. For the innate immune system, evidence suggests that epigenetic changes in DNA and chromatin may be important in innate immune training [3]–[9]. Myeloid cells, a main component of the innate system, develop earlier in gestation than lymphoid cells (adaptive). Adaptive immunity provides considerably more advanced immunological protection from infections, although maturation of this system occurs over years of exposure to multiple variants of pathogens. Therefore, in comparison with adults, neonates born at term rely more heavily on innate immunity, while infants born prematurely are entirely dependent upon innate immune protection.  In healthy individuals, recognition of a pathogen by cells of the innate immune system results in production of pro-inflammatory cytokines that signal the recruitment of additional immune cells for clearance of the pathogen at the site of inflammation. Additionally, cytokines also activate the adaptive immune system responsible for the development of an immunological memory against re-infection. Production of pro-inflammatory cytokines occurs following the recognition of microbial pathogens by immune cells via Pattern Recognition Receptors (PRR). PRRs are a series of evolutionarily conserved receptors that are designed to recognize conserved microbial structures. They comprise endosomal, intra-cytoplasmic and 3  extracellular anti-microbial detectors such as Toll-like receptors (TLRs), nucleotide-binding oligomerization domain-containing protein (NOD) and NOD-like receptors (NLRs), C-type lectin receptors (CLRs) and the retinoic acid-inducible gene I (RIG-I) and RIG-I-like receptors (RLRs) [10]. Activation of PRRs by microbes induces the release of additional soluble inflammatory mediators, recruitment of immune cells, increased phagocytosis of microbial particles and oxidative anti-microbial killing by immune cells [11].  Interleukin 1β (IL-1β) and interleukin-6 (IL-6) are two model pro-inflammatory cytokines produced in response to PRR activation. However, their mechanisms of production and action on downstream cells differ in many ways: While production of IL-6 occurs in many different cell types (T cells, neutrophils, dendritic cells, etc.) [12], production of IL-1β is more limited, including monocytes, macrophages and dendritic cells [13]; while production of both cytokines occurs via the NF-κB-signaling pathway, IL-1β requires a second signal for production of its biologically active form. The main reason for this distinction is that while both cytokines are pro-inflammatory, IL-1β is highly pyrogenic, and has the unique ability to potentiate production of other pro-inflammatory cytokines via its own IL-1β receptor [14]. This likely explains the additional regulatory mechanism involved in production of this latter cytokine.  IL-1β production occurs via a two-step mechanism. First, pro-IL-1β is produced following PRR stimulation; this pro-protein remains intracellular and inactive unless a second activation signal occurs. The second signal for IL-1β production is usually associated with cellular damage (extracellular ATP, etc.). This second signal results in the assembly of the inflammasome, and processing of pro- IL-1β either through a canonical caspase-1 dependent pathway, or a non-canonical caspase-8 dependent pathway [15], [16]. Following processing, mature IL-1β is 4  secreted from the cell. The rate-limiting step in IL-1β secretion is cleavage by inflammatory caspases.   1.3 Fungal infections in preterm neonates   Candida spp. are a major cause of neonatal sepsis, and carry a mortality rate of up to 30% in premature infants [17]. In healthy adults, this micro-organism rarely causes invasive disease despite high rates of colonization [18]. However, in the most immature premature newborns, Candida spp. frequently cause invasive disease, although the fundamental mechanisms underlying this increased risk remain unclear. Despite the clinical importance of this pathogen, data on the maturation of anti-fungal immune responses in preterm infants are deeply lacking [19].  Candida albicans and C. parapsilosis account for most cases of fungal sepsis in premature neonates. Although rates of fungal infection vary widely between geographical locations, overall rates of C. parapsilosis are on the rise, particularly in neonates [20]–[23]. Some members of the Candida genus are polymorphic microorganisms that can exist as yeast or hyphal forms [18] (in the case of C. albicans), or yeast and pseudo-hyphal forms (in the case of C. parapsilosis). From an immunological perspective, these two species broadly elicit similar innate immune cytokine responses [24]; however, host phagocytic activity in response to these two species differ, with C. parapsilosis being phagocytosed more efficiently [25], [26], and in a dectin-1 independent manner [27]; however, it should be noted that one study observed higher rates of phagocytosis in response to C. albicans compared to C. parapsilosis [16]. Although both 5  species of Candida are known to cause fungal sepsis, in a model of mixed infection of C. albicans and C. parapsilosis, C. parapsilosis reduced invasion and damage to intestinal epithelial cells caused by C. albicans [28].  1.4 Recognition of Candida spp. by immune cells   Preventing systemic invasion by Candida spp. requires the immune recognition of fungal pathogens via multiple receptors [29]. C-type lectins and Toll-like receptors (TLRs) are the main PRRs involved in the recognition of Candida, which are also species and morphology-dependent [30]. Yeast forms, which predominate in the bloodstream during invasion, are strongly detected by dectin-1, contrary to filamentous forms which are predominantly recognized via TLR2 and TLR4 [31]–[35]. Dectin-1 is the main receptor which binds the fungal cell wall component -1,3-glucan [36], [37], resulting in the production of pro-inflammatory cytokines, including interleukin-1β (IL-1β). The latter is important for immediate innate [38] inflammation, but also for long-lasting adaptive immunity against Candida [39], [40].  At the cellular level, dendritic cells play an important role presenting fungal antigens to T-cells, whereas monocytes (in blood) and macrophages (in tissues) are the main source of IL-1β. Compared to dendritic cells and macrophages, monocytes demonstrate increased reactivity to Candida albicans, thus highlighting their importance in preventing dissemination of this pathogen through blood [41]. Production of IL-1 can occur via two main pathways [42]. In the canonical pathway, activation of PRR results in transcription of the IL1B gene, which is then translated into the pro-IL-1 precursor protein until a second danger-associated signal (e.g. 6  tissue damage) is detected. This second signal leads to proteolytic cleavage of pro-IL-1 and secretion into its mature IL-1 form via the NLRP3/Caspase-1 inflammasome (reviewed in [30]).  Alternatively, in the non-canonical pathway, both the production of pro-IL-1β and its cleavage into mature IL-1β can occur via the MALT1/Bcl10/CARD9 signalosome resulting in activation of caspase-8 [43], [44]. In humans, the major importance of the signalosome is highlighted by data showing increased invasive Candida infections in individuals carrying loss-of-function mutations in CARD9 [45]. However, despite an accumulation of data on the receptor and molecular pathways involved in the recognition of Candida spp. in humans, there is a deep lack of knowledge on the development of these functions in newborns.  1.5 Immune development in humans   Ontogeny defines the period in life where development occurs. Important progress has been made over the last few years in our understanding of hematopoiesis in adults. In comparison, much remains to be studied regarding the development of immune cells and functions during ontogeny. The normal, full-term of gestation in humans is 38-40 weeks (~9 months). Prematurity is defined when an infant is born below 37 weeks of gestation. The current limit of viability in humans is about 22 weeks; unfortunately, infants born below this gestational age cannot survive even with optimal medical care. Premature infants face many short- and long-term health issues that require medical attention. When born below 32 weeks, newborns are considered most immature physiologically and immunologically, with decreased immune functionality correlating with gestational age. Because of incompletely developed 7  organs such as the lungs and bowels, preterm infants require intensive care interventions to survive after birth; these include supplemental oxygen, mechanical ventilation, gavage feeds and intravenous nutrition. Although life-saving, these interventions expose infants to a considerable risk of developing blood stream infections from microbial pathogens present in the natural environment. About one-fourth of all infants born below 32 weeks of gestation will develop a serious infection during the neonatal period [2]. Infants at this stage of development are considerably immunocompromised, because of the incomplete maturation of both innate and adaptive immune defense mechanisms. Infections in premature infants usually occur within the first month of age and involve common organisms such as members of the Candida genus, but also other organisms that are usually not threatening to most healthy adults like Staphylococcus epidermidis and Escherichia coli.  Illustration 1.1 situates premature infants in their immune development. As illustrated in this figure, myelopoiesis begins early in the yolk sac around the 3rd week of gestation, followed by the liver after the 5th week. Until about the 20th to 22nd weeks, the liver assumes most hematopoietic cell production, after which production of blood cells mainly takes place in the bone marrow [46]. The composition of myeloid cells has been studied throughout gestation [47]. Neutrophils are very important in protecting newborns against infections. However, they form a very low proportion of blood cells until about 32 weeks [47]. Consequently, these cells have been the most difficult to study during this period. In preterm newborns, neutrophils have an impaired ability to form neutrophil extracellular traps, which are lattices of extracellular DNA, chromatin and antibacterial proteins that mediate extracellular killing of microorganisms 8  via the generation of potent reactive oxygen species [48]. Recent evidence suggests that neutrophil extracellular traps can form in preterm cells in response to fungal pathogens[49]. Like other immune cell types, monocytes recognize pathogens through innate PRRs. Three main monocyte subsets have been defined in adult blood, according to their expression of the CD14 and CD16 cell surface markers [50]. Before 29 weeks of gestation, a majority of fetal monocytes display an immature phenotype characterized by low CD14 expression [51]. These monocytes are likely to play a more predominant role in tissue remodeling rather than engaging in aggressive immune responses [46]. At this gestational age, both anti-microbial immune recognition and antigen presentation are also significantly impaired even in “classical” high CD14-expressing monocytes because of reduced receptor expression and intracellular signaling [51], [52]. The complement system is also functionally impaired in neonates born prematurely as evidenced by a decrease in complement proteins C3a, Factor H, and Factor I [53], which further limits pathogen recognition. Little activity by extracellular PRR is detectable below 20 to 24 weeks of gestation [54]–[58]. Beyond this gestational age, PRR activity increases to a level comparable to that of a term infant by 33 weeks of gestation, with the earliest activity detected by endosomal (TLR7, 8 and 9), intracytoplasmic (e.g. NLRs), and, subsequently, extracellular PRRs such as TLR1, 2, 4 and 5 and dectin-1 [51], [59]. Between 33 weeks and full term of gestation, only modest changes in PRR activity occur which, for the most part, are within the detection limits and experimental variability of currently available assays. The significance of this “inside-out” hierarchical development of PRR function during fetal ontogeny is unclear [57]. However, it may have important clinical implications. It 9  corroborates well with some of the microbial vulnerabilities clinically observed in early premature infants. For example, activity in TLR2, which plays a predominant role in the recognition of the coagulase-negative Staphylococci (CoNS) epidermidis, develops late in gestation [56]. Correspondingly, infections with this pathogen are most common in infants born below 30 weeks of gestation [60]. Knowledge of the maturation in activity of PRRs across stages of development is also important because these receptors are largely responsible for adjuvant effects during immunizations, and premature infants display significantly reduced antibody responses to routine vaccines [61]. The mechanisms underlying the global dampening of PRR functions in preterm infants at lower gestation are unknown, although it is thought to represent an evolutionarily advantageous functional characteristic aimed at reducing the potential harmful effects of inflammation during the early stages of fetal development, when organs are most vulnerable. Indeed, studies have shown that intra-uterine inflammation can result in major organ damage or a premature birth [62]–[66]. To a large extent, our knowledge of the developing newborn immune system has been derived from cord blood. Nonetheless, recent studies have examined the maturation of immune cells during the neonatal period, at the time infants are most likely to be affected by neonatal pathogens (i.e. around 2-3 weeks of post-natal age). Indeed, using both whole blood and single-cell flow cytometry analyses, Marchant et al. have confirmed that PRR responses (e.g. TLR4, TLR7/8), remain attenuated in post-natal blood obtained from neonates born below 30 weeks of gestation [67]. These results provide important support to the premise that reduced immune activity is responsible for the high vulnerability to infections.    10  1.6 Development of mucosal immune defenses   Colonization of the fetus marks the beginning of an extraordinarily complex lifelong relationship with microbes. To fully understand the vulnerability of the newborn immune system, it is essential to appreciate the distinctive mucosal immune system and the importance mucosal interfaces play in protecting against infections. The intra-uterine environment is no longer considered sterile, and there is a greater appreciation of a “normal” placental and amniotic microbiota in utero. However, the abundance and diversity of this microbial flora probably increases considerably within days of birth, as the infant’s mucosal and epidermal surfaces become colonized. In the case of infants born at full term, a variety of factors have been evolutionarily selected to enhance colonization by beneficial microorganisms. Healthy term infants born vaginally are colonized by the maternal vaginal, intestinal, and skin microbiota shortly after birth [111]. Several factors will influence the diversity of the early microbiome. These include, for example, the way the infant was delivered, nutritional factors, antibiotic exposure and (highly probably) geographical and ethnic factors. In the case of preterm infants, nearly half of them are born via Caesarean section in North America, which likely affects colonization by normal human commensals. Premature infants also often have nutritional deficits and are at risk of prolonged antibiotic use that may further disrupt their normal microbial colonization. Mouse models have been extremely useful in understanding how microorganisms at mucosal interfaces are essential in shaping immune functions during the post-natal period. However, the existence of major functional differences across species has warranted human studies to confirm the relevance of findings from mouse models. Detailed 11  reviews on the development of mucosal immune defenses during ontogeny have been published elsewhere [68]–[70]. However, additional relevant studies performed in the context of neonatal sepsis have been undertaken and are discussed here. The role of the intestine as a functionally distinct immunological organ is also important to point out [71]. Traditionally, our knowledge of the developing human intestinal immune system has been greatly complicated by tenuous immune phenotyping protocols coupled to a lack of accessibility to fetal samples. In recent years, advances in the use of multi-parameter flow cytometry has allowed for more extensive, high-throughput characterization of cell populations using minuscule volumes of biological samples obtainable in a human neonate and has greatly facilitated discoveries in this area. For instance, invariant natural Killer (iNKT) cells are a relatively rare innate-like T cell type particularly important in protecting the host against colonizing microorganisms at mucosal interfaces that can be detected using antigen-loaded tetramerized major histocompatibility complex molecules [72]. Recent data has shown that these intestinal neonatal iNKT cells share unique functional properties; contrary to conventional T cells, intestinal neonatal iNKT cells produce robust Th1 and Th17 responses [73]. In humans, iNKT cells are abundant early in fetal blood [74], although their tissue of origin had remained unclear until recent data showed that these cells accumulate in the small intestine during the second trimester of gestation [73].  Studies in mice have also demonstrated the critical role of the small intestine in regulating both local and systemic Th17 responses [75]. Particularly, IL-17-mediated T helper and  T cell responses are important to prevent invasion of microorganisms at mucosal surfaces [51]. Earlier studies have shown that neonatal CD4+ T cells are intrinsically biased 12  towards a “default” epigenetically-regulated Th2 response with abundant production of IL-4 and IL-10 [76]. Recently, human neonatal CD4+ T cells were also shown to differentiate poorly into IL-17-producing cells upon anti-CD3/CD28 activation in vitro [77], [78].  There are six homologs of IL-17 (IL-17A to F); IL-17A is the best characterized and predominant member of this family, whereas IL-17F also shares partly overlapping functions with IL-17A [79]. IL-17A/F-deficient mice exhibit increased susceptibility to a broad range of bacterial and fungal mucosal infections [79], [80]. In humans, the clinical impact of IL-17 deficiency is more limited, highlighting another important difference across species [81]. Loss of Th17 cells in humans with chronic HIV infection also results in a loss of gut-barrier function and thus leads to subsequent microbial translocation [82]. Adult humans lacking production of IL-17 appear specifically vulnerable to microbes such as Candida spp. and Staphylococcus aureus, which are also predominant neonatal pathogens [83]. To help promote systemic Th17 responses, neonatal antigen-presenting cells produce high levels of IL-1, IL-6, but also high amounts of IL-23, at least in term-born infants [84]. In contrast, monocytes and dendritic cells from preterm neonates below 29 weeks of gestation produce low amounts of IL-1 and IL-23 [51], [58]. This lack of Th17-differentiating cytokine production by neonatal antigen presenting cells of infants born very prematurely may partly explain their increased vulnerability to certain types of mucosal infections by bacterial and fungal pathogens including Staphylococcal and Candida species.  Increasing evidence suggests an important influence of the neonatal microbiome on health outcomes. In healthy term newborns, a diverse intestinal microbial flora promotes the development of mucosal barriers and prevents the outgrowth of pathogenic microorganisms 13  [85], [86]. Early transient intestinal dysbiosis was associated with an increased risk of asthma during early childhood [87]. Anomalies in the placental microbiome have been associated with premature birth (reviewed in [88]). Also, accumulating evidence suggests that the composition of the microbiome plays an important role in protecting the newborn against infections. For example, disruption of the normal flora with broad-spectrum antibiotics increases the risk of neonatal sepsis. [89]–[91]. In preterm neonates, distinct microbial colonization of the gut linked to increased CoNS and decreased colonization by protective commensals such as Bacteroides, Bifidobacterium, and Lactobacillus has also been associated with increased risks of neonatal sepsis [92], [93] and necrotizing enterocolitis [94], [95]. Data also suggest that abnormal PRR signaling at mucosal surfaces in preterm infants may favor the development of limited microbial diversity favoring the colonization of pathogenic micro-organisms [91], [93], [96]. These findings have direct clinical relevance since administration of probiotics has been shown to reduce the risk of neonatal sepsis and necrotizing enterocolitis in premature infants [97]. More recently, important observations were published regarding the cellular composition of breast milk [98], including anti-microbial peptides [99], [100], and how the immune properties of human milk vary across lactation [98], [99]. Factors such as breast milk composition, antibiotic exposure, and method of birth all play a role in the colonization status of infants. Large clinical trials are also ongoing in developing countries; results so far indicate that administration of probiotics may substantially reduce neonatal mortality due to sepsis [101].   14  1.7 Lessons from neonatal mouse models   Despite recent progress, the study of the developing immune system in human newborns remains extremely challenging, both practically and ethically. The miniaturization of immunological methods has greatly helped the development of neonatal mouse models of sepsis. These models have confirmed the limited myelopoietic capacity of the newborn and its impact on survival of newborn mice during sepsis [102]. Using a poly-microbial infection model, Wynn and Moldawer demonstrated that rodent neonates rely heavily on innate immune defense during sepsis [103]. Moreover, the brain of a developing newborn is also particularly vulnerable to inflammation, and neonatal infections carry an important risk of major long-term neurological disability in premature infants [104]. To study mechanisms of brain injury during neonatal infections, Levy and Mallard developed an interesting mouse model [105] in which they demonstrated that inflammation due to Staphylococcus epidermidis caused brain injury even in the absence of bacterial entry into the CSF. These results suggest that low-grade inflammation produced during transient bacteremia could be sufficient to cause damage to a developing neonatal brain. These findings have important clinical implications, since earlier studies have indicated that transient bacteremia commonly occurs in infants with sepsis [106], [107]. Another study has demonstrated persistent sub-clinical inflammation in the absence of overt clinical signs of infection in these infants throughout the neonatal period, possibly due to an ongoing exposure to intensive care therapies [108]. In light of these findings, better strategies are required to limit the sustained exposure of premature infants to any potential triggers of inflammation or sources of transient bacteremia, and to prevent infections [109]. 15  The coupling of neonatal mouse models to genome-wide transcriptome studies has also pointed to important functional differences between neonatal and adult responses [110]. Indeed, the nature of innate immune PRR pathways involved during sepsis differs between neonates and adults, highlighting the importance of age-appropriate models [102]. However, major differences in the nature of such transcriptome responses depend on the model used [111], and well-recognized differences between humans and mice may limit the application of findings across species, justifying direct observations in humans [112]. Most recently, two important genome-wide gene expression studies have provided insights into the molecular pathways activated in newborns with severe infections [113], [114]. For instance, marked transcriptional alterations in metabolic pathways associated with glucose transport, glycolysis, and cholesterol homeostasis, transport, and metabolism were detected in newborns during active sepsis [113]. On the other hand, transcriptional pathways related to antigen presentation processes via MHC II were down-regulated in septic infants, although the lower gestational age of infants from the “septic” group may have confounded these results. The specific up-regulation of immune suppressor pathways, such as TNFAIP3, CD71, and SOCS1/3 in infants with sepsis may help explain why some infants fail to clear infections [113]. While these data provide important insights into how immaturity of the immune system can increase the risk of infection in newborns, important differences across species make it difficult to extrapolate some of these data to humans.    16  1.8 Considerations unique to the study of immune functions in premature infants   Premature infants are often born suffering pathological conditions such as following chorioamnionitis, which may potentially impact functional measurements of immune responses in cord blood. Chorioamnionitis has been strongly associated with preterm labor. For example, evidence of chorioamnionitis can be detected in about 40% of mothers delivering an infant prematurely below 29 weeks of gestation. It is also a major risk factor for poor neonatal health outcomes, including increased risk of neonatal sepsis. Infections of the fetal membranes expose the fetus to an inflammatory environment, which can be damaging to the developing brain and may also progress to neonatal sepsis. Based on animal models, chorioamnionitis can also modulate fetal immune responses, an effect that has been termed “immune-paralysis” [115], [116]. However, we lack direct evidence of this phenomenon in human infants. Many of these infants are also exposed to corticosteroids (an immunosuppressive drug), which is often administered to pregnant mothers to promote fetal lung maturation. However, multivariable statistical analyses of immune responses in these infants following rigorous clinical definitions of antenatal exposures suggest that the functional immunological differences measured from cord blood are probably developmentally related, rather than due to clinical exogenous perinatal factors linked to their premature birth [57].   17  1.9 Immune regulation during ontogeny   Although pro-inflammatory responses are essential to protect against infection, inappropriate activation of inflammatory cascades can also result in organ damage, preterm labor (as mentioned above) and in some cases, death. As such, regulation of inflammation is a fine balancing act; failure to induce inflammation may result in a the lack of an appropriate adaptive immune response or continued infection; in contrast, over activation of inflammation can result in tissue damage and death [63], [64], [117]. In the context of preterm birth, a suppressed innate immune response results in the neonate being highly vulnerable to infection. In contrast, as the fetus reaches full term, innate immune functions must rapidly become operational to cope with the large-scale exposure to various microbes. The mechanisms of regulation of the activity of the innate immune system are unclear. Previous observations show that these functions are globally downregulated, which make it difficult to ascribe them to a common/shared set of receptors or signaling molecules [54], [55], [57], [118]–[120]. In healthy adults, innate immune responses show high variability between individuals, likely due to a combination of genetic and environmental factors, collectively referred to as the “exposome” (reviewed in [121], [122]).    18  1.10 Metabolic regulation of immune reactivity   Despite major progress made in the characterization of the neonatal immune system at various developmental stages, the fundamental mechanisms regulating the maturation of immune functions in humans during ontogeny remain completely unknown. Recent data has revealed an important immune regulatory role for erythrocytes during the neonatal period. These cells have been shown to actively suppress both innate and adaptive immune responses through the expression of arginase, which inhibits nitric oxide synthesis [123]. However, these mechanisms are unable to account for the broad suppression of innate immune responses when cells are examined singularly ex vivo.  A recent major breakthrough in our understanding of the regulation of immune responses is the discovery that a major metabolic switch occurs in cells, and that this switch is required to accommodate the massive upregulation of genes following an immune activation [124]–[127]. The essence of these changes is depicted in Illustration 1.2. At rest, immune cells largely utilize oxidative phosphorylation to generate energy. Upon encountering a pathogen, cells turn off oxidative phosphorylation to rely almost exclusively on glycolysis for energy production (reviewed in [128]); these metabolic changes are followed by activation of inflammatory immune pathways. Mitochondria play a central role in operating this metabolic switch, by controlling the metabolic fate of glucose, as well as the utilization of fatty acids. In contrast, upon encountering an anti-inflammatory stimulus, enhanced lipid metabolism results in increase production of anti-inflammatory cytokines (IL-10, TGF-β), as well as pro-resolving lipid mediators (e.g. resolvins and protectins) [129].  19  In contrast to decades of dogma, it has recently become appreciated that innate immune cells do display a form of memory [6], [124], [130]–[136]. Depending on the nature of an initial stimulus, innate immune cells (such as monocytes), can become non-responsive, anti-inflammatory, or hyper-inflammatory to subsequent stimuli; these phenotypes are known as immune-paralysis, immune-tolerance, and immune training respectively. The basis of this memory is largely controlled by metabolic mechanisms [6], [124], [131], [132], [135]. As in the case of activated immune cells, trained monocytes primarily utilize glycolysis, whereas tolerant cells skew towards fatty acid oxidation [137], [138]; however, cells that are immune-paralyzed downregulate all major metabolic pathways [139], [140].  Metabolic control over the pro- or anti-inflammatory nature of an immune response appears to be a general requirement amongst immune cells, as suggested by similar changes observed during T cell [141], monocyte/macrophage [142], and dendritic cell activation [143]. Corresponding metabolic shifts to glycolysis in pro-inflammatory vs fatty acid oxidation in anti-inflammatory responses were observed in both human and mouse cells [125], [126], [129], [144]. As another example, activation of effector CD4+ T cells requires a shift from oxidative phosphorylation to glycolysis; in contrast, activation of regulatory T cells results in a shift in cellular energy metabolism toward fatty acid oxidation [141]. The broad and conserved extent of these changes and the unique metabolic state of the fetus suggests that a lack of similar metabolic changes may contribute to immune impairments in preterm neonates. For instance, during fetal life, mitochondria have a fragmented appearance [145], have reduced mass [146], and display a more limited energy production capacity [147], [148]. In addition, preterm neonates display increased lipid metabolism relative to their glycolytic activity [149]; this is 20  linked to an altered metabolic response during sepsis [113]. Major primary metabolic differences have been described between neonates born prematurely, at term, and adults [147], [150]. Specifically, oxidative phosphorylation pathways are down-regulated in preterm neonates [148], [151], [152], which may be compensated for by increasing fatty acid oxidation [153], [154].   1.11 Role of mTOR and HIF-1α   Two main regulators, Hypoxia-Inducible Factor 1-alpha (HIF-1α), and mechanistic Target of Rapamycin (mTOR) have been shown to play an important role in regulating this metabolic switch [124] (reviewed in [143]). HIF-1α is a transcription factor largely responsible for the switch from oxidative phosphorylation to glycolytic pathways; additionally, HIF-1α activity can be induced by inflammatory stimuli through mTOR signaling [124]. Little is known about the regulation of HIF-1α and mTOR during ontogeny. However, due to the unique maternal-fetal blood circulation system, the fetus naturally evolves in a relatively hypoxic environment that is due to its unique blood circulation. Regulation through HIF-1α is influenced by a variety of stimuli, including inflammatory signaling [155] and oxygen levels [156], thereby providing a potential link between low oxygen concentration in utero and decreased immune responsiveness [127]. mTOR is known to play a role in signal transduction; stimuli such as stress, nutrient availability, and immune signalling, result in assembly of the mTOR complex 1 or 2 (mTORC1/mTORC2), and induction of transcription, protein synthesis, autophagy, metabolic regulation, and growth [157], [158].  21  1.12 Regulation of protein synthesis   Transcriptomic studies provide interesting insight into gene expression, from which a functional hypothesis can be generated and tested. However, it is important to note that there is a level of regulation between the transcriptome and proteome; in fact the correlation between transcript and protein abundance is surprisingly low [159]. This discrepancy between transcript and protein abundance demonstrates that protein expression is also regulated post-transcriptionally. In the context of innate immune responses, cells must balance between rapid reactivity and inappropriate activation. As such, by regulating protein synthesis/translation, immune cells are able to achieve further control over the scale of their cytokine secretion [160]. Whether this level of control is broadly applied or specific to certain transcripts remains to be seen. Regulation of protein synthesis during an immune response can occur via several mechanisms (reviewed in [160]–[164]); initiation of translation can be regulated by activating or deactivating eIF2α [165]. Alternatively, translation can be stopped by sequestering gene transcripts. Transcripts can be sequestered in stress granules [166]; additionally, GAPDH has been shown to bind and sequester the TNFα transcript [139]. Interestingly, many of the mechanisms regulating protein synthesis are linked to cell stress and mTOR signaling. Another potential link between metabolic function and regulation of synthesis of specific mRNAs is the discover of 5’ terminal oligopyrimidine (TOP) mRNAs; transcripts containing this motif are repressed in an mTOR dependent fashion in response to cellular stress, and are largely related to protein synthesis and ribosome function [167]. Like TOP mRNAs, TISU (translation initiator of short 5′ UTR) mRNAs are also sensitive to mTOR regulation; however, mRNAs containing TISU 22  motifs are related to metabolic function [168], whereas TOP mRNAs are mostly related to ribosome function.  1.13 Regulation of innate immune function in healthy adults   Among healthy adult controls, our lab has repeatedly observed that there is a large degree of variability between individuals. In twins, it has previously been shown that a large degree of variability in human immune responses is non-heritable [169]; this finding suggests that at least for specific immune functions, such as IL-1β secretion, environmental factors are largely responsible for inter-individual variability. It has been proposed that cumulative environmental/ecological factors, known as the exposome [122], may play a role in determining the diversity in human immune responses; this hypothesis is based on the fact that there is an increase in variability in innate immune gene expression in older individuals [170]. Although many studies examining immune variability use transcript levels as a stand-in for immune reactivity, this approach is confounded by low correlation between transcript and protein levels [159]. As it has previously been demonstrated that innate immune variability is affected by individual differences in the exposome, we hypothesize that transcriptomic events do not fully capture the mechanisms that determine immune diversity in healthy adults. As such, we aimed to determine to what degree gene expression correlates with specific functional outcomes. Additionally, I propose that the simplistic model of examining mRNA levels to determine immune outcomes fails to consider complex gene interactions that may play a large role in 23  regulating immune diversity. An example of more distal gene clusters that affect immune function would be metabolic genes, as previously mentioned.  1.14 Hypothesis and objectives   Although premature birth presents many medical issues, it also provides insights into the development of the human immune system. The overall goal of this thesis is to identify the mechanisms that regulate inflammation in the premature neonate and in the healthy adult. Due to the potential adverse functional consequence of an inappropriate activation of the innate immune system, I hypothesized that innate immune responses in fetuses and newborns are broadly regulated and that examining immune responses at the systems level would inform on how this reactivity of the innate immune system is regulated throughout life. Objective 1 (Chapter 2): To examine anti-fungal immune responses in preterm infants Objective 2 (Chapter 2): Identify the molecular mechanisms that are responsible for immune suppression in preterms. Objective 3 (Chapter 3): Determine mechanisms that are responsible for immune variability observed in healthy adults.   24   Illustration 1.1 Immune Ontogeny  The development of hematopoiesis begins early in embryonic life, in the yolk sac followed by the liver, bone marrow and other secondary hematopoietic organs. The blue arrows illustrate hematopoietic cells and their progenitors migrating between hematopoietic organs at different stages of embryonic and fetal life. The diagram also represents the aorta-gonad-mesonephric (AGM) region which is an important structure supporting the production of hematopoietic cells and their progenitors early on, before the development of other more definitive hematopoietic organs. Lymphopoiesis begins after erythropoiesis and myelopoiesis. Infants born very prematurely, between 22 (defining the absolute gestational limit of viability in humans) and 32 weeks are at high risk of infections. This period also corresponds to the maturation of anti-microbial Pattern Recognition Receptors (PRR), beginning with endosomal/cytoplasmic 25  followed by extracellular PRR, and the development of mature, adult-like fetal T cells from an earlier wave of fetal T cells. Concomitant to this is the passive immunization of fetuses through trans-placental maternal antibody transfer.   26   Illustration 1.2 Metabolic Adaptation Occurring During Immune Activation.   Resting immune cells, such as monocytes or T cells primarily produce energy (ATP) from breaking down glucose into pyruvate, and through mitochondrial oxidative phosphorylation. The activation of Pattern Recognition Receptors (PRR) by microbes leads to the activation of the master transcription regulator NF-κB, which in turn results in inflammatory cytokine gene expression and production (e.g. IL-1β, TNF-α). Concurrently, cells also undergo a metabolic shift to aerobic glycolysis, a phenomenon also known as the Warburg effect, whereby energy production occurring through glycolysis breaks down glucose into succinate. Although less energy is produced for each glucose molecule, this pathway does provide large quantities of 27  anabolic substrates required to feed into the high gene transcriptional activity occurring during immune activation. Succinate and NF-κB also stabilize the transcription factor HIF-1α, resulting in further up-regulation of inflammatory cytokine gene expression. The shift away from oxidative phosphorylation also allows these cells to produce reactive oxygen species (ROS) more efficiently. In contrast, anti-inflammatory signaling (via STAT6) causes an increase in fatty acid uptake, mitochondrial biogenesis and increased mitochondrial function. This metabolic shift towards fatty acid oxidation leads to up-regulation of anti-inflammatory cytokines (e.g. IL-10, TGF-β) and pro-resolving lipid mediators such as resolvins and protectins.    28  Chapter 2: Cellular metabolism broadly suppresses anti-fungal innate immune defenses in human neonates   2.1 Background   Candida spp. are a major cause of neonatal sepsis, with reported mortality rates from fungal sepsis ranging up to 30% in severely premature infants [17]. In healthy adults, this micro-organism rarely causes invasive disease, despite high rates of colonization [18]. However, in the most immature preterm newborns, Candida spp. frequently cause invasive disease. The fundamental mechanisms underlying this increased risk of infection remain unclear. Despite the clinical importance of this pathogen, data on the maturation of anti-fungal immune responses in preterm infants are deeply lacking [19].  Candida albicans and C. parapsilosis account for most cases of fungal sepsis in premature neonates. Although rates of fungal infection vary widely between geographical locations, overall rates of C. parapsilosis are on the rise, particularly in neonates [20]–[23]. Candida spp. are polymorphic microorganisms that can exist as yeast or hyphal forms [18] (in the case of C. albicans), or yeast and pseudo-hyphal forms (in the case of C. parapsilosis). From an immunological perspective, these two species broadly elicit similar innate immune cytokine responses [24]; however, host phagocytic activity in response to these two species differ, with C. parapsilosis being phagocytosed more efficiently [25], [26], and in a dectin-1 independent manner [27]; however, it should be noted that one study observed higher rates of phagocytosis in response to C. albicans compared to C. parapsilosis [16].  29  Preventing systemic invasion by Candida spp. requires immune recognition, phagocytosis, and killing of fungi via multiple receptors [29]. C-type lectins and Toll-like receptors (TLRs) are the main PRRs involved in the recognition of Candida, which is also species and morphology-dependent [30]. Yeast forms, which predominate in the bloodstream during invasion, are strongly detected by dectin-1; in contrast, filamentous forms are predominantly recognized via TLR2 and TLR4 [31]–[35]. Dectin-1 is the main receptor which binds the fungal cell wall component -1,3-glucan [36], [37], resulting in the production of pro-inflammatory cytokines such as interleukin-6 (IL-6), and interleukin-1β (IL-1β). The latter is important for immediate innate [38] inflammation, but also for long-lasting mucosal adaptive immunity against Candida [39], [40].  At the cellular level, monocytes (in blood) and macrophages (in tissues) are the main sources of IL-1β. Compared to macrophages, monocytes demonstrate increased reactivity to Candida albicans, thus highlighting their importance in preventing dissemination of this pathogen through blood [41]. Production of IL-1 can occur via two main pathways [42]. In the canonical pathway, activation of PRR results in expression of the IL1B gene, which is then translated into the pro-IL-1 precursor protein until a second danger-associated signal (e.g. tissue damage) is detected. This leads to proteolytic cleavage of pro-IL-1 and secretion into its mature IL-1 form via the NLRP3/Caspase-1 inflammasome (reviewed in [30]). Alternatively, in the non-canonical pathway, both the production of pro-IL-1β and its cleavage into mature IL-1β can occur via the MALT1/Bcl10/CARD9 signalosome resulting in activation of caspase-8 [43], [44]. In humans, the major importance of the signalosome is highlighted by data showing increased invasive Candida infections in individuals carrying loss-of-function mutations in 30  CARD9 [45]. However, despite an accumulation of data on the receptor and molecular pathways involved in the recognition of Candida spp. in humans, there is a deep lack of knowledge on the development of these functions in newborns. Multiple studies have shown that mononuclear cells from preterm infants display profoundly reduced PRR responses and secretion of inflammatory cytokines (see [54], [55], [57], [118]–[120] for examples).  Innate immune cells importantly dictate the outcome of an infection especially in newborns that lack an immunological memory against most antigens. During ontogeny, the innate immune system is deployed before the adaptive immune system. In humans, myeloid cells are first produced in the fetal liver (<16 to 24 weeks of gestation) and later on during gestation (>22 weeks) in the bone marrow [171]. Between the 22nd-23rd gestational weeks, the earliest premature infants can survive outside the womb with intensive care, anti-microbial PRR responses are profoundly attenuated and gradually emerge in a hierarchical fashion until the term of gestation [57], [119]. The broad extent of attenuation in anti-microbial functions at this stage, involving multiple PRR families has been difficult to assign to a common, shared set of regulatory molecules [172], [173]. Despite an accumulation of these studies, the fundamental mechanisms regulating innate immune reactivity at this critical developmental stage remain unknown [119]. To address this question, we used an unbiased systems approach combining transcriptomic, metabolic and polysome profiling.   31  2.2 Methods   2.2.1 Cohorts   Rates of invasive Candida infections (all species) were obtained from prospectively collected data from the Canadian Neonatal Network database of all infants admitted to a neonatal intensive care unit in Canada between 2003 and 2013. Since invasive Candida infections are rare in marginally preterm newborns, we focused our analysis on those born below 33 weeks. Blood samples were obtained from preterm infants (<33 weeks), term infants (cord blood) and healthy adult volunteers (peripheral blood), in sodium heparin vacutainers, and following written informed consent. To ensure that the lack of dectin-1 response was not due to an enrichment in dectin-1 SNP rs16910526 among preterm births, genotyping for the common Y238X mutations [174] was performed in a previously published cohort of 177 preterm infants below 31 weeks of gestation [175] and only 2 were homozygous for the rare variant, similar to the adult population [176]. Protocols were approved by the UBC Children’s & Women’s Research Ethics Board.  2.2.2 Reagents    Pediatric clinical isolates of Candida albicans and Candida parapsilosis, identified by mass spectrometry (Bruker Daltonics, Billerica, MA) were obtained from the BC Children’s Hospital Microbiology Lab (Vancouver, Canada). We used fixed Candida spp. yeast particles 32  which are closer antigenically to yeast forms [18]. Fungi were grown over 4 days at 30°C in BHI (Brain-Heart Infusion) broth (Oxoid, Nepean, ON). Yeast in exponential growth phase were harvested after 4 days, centrifuged and fixed in 10% paraformaldehyde prior to counting. To ensure reproducibility, batches of Candida particles were prepared and used throughout the study. Human dectin-1 and dectin-1 neutralizing antibodies were purchased from R&D Systems (#MAB1859) and Invivogen (#Mabg-hdect2), respectively, and used at 5 µg/ml for cytokine studies, and 10 µg/ml for phagocytosis assays. Human CD-206 neutralizing antibody was purchased from Biolegend (#321111) and used at 10 µg/ml for phagocytosis assays. Human DC-SIGN neutralizing antibody was purchased from Invivogen (#mab-hdcsg) and used at 10 µg/ml for phagocytosis assays. Curdlan (β-1, 3-glucan from Alcaligenes faecalis), a dectin-1 specific agonist, was obtained from Wako (#032-09902) and used at 10 µg/ml in cytokine studies. Lipopolysaccharide (LPS), a TLR-4 agonist purified from Escherichia coli, was obtained from InvivoGen (#tlrl-eblps) and used at 10 ng/ml. Zymosan, a TLR-2/dectin-1 agonist was obtained from Invivogen (#tlrl-zyn) and used at 10 µg/ml. Mannan was purchased from Sigma-Aldrich (#M3640-1G) and used at 3 mg/ml in phagocytosis assays. Laminarin, a soluble β-glucan dectin-1 blocking molecule was purchased from Sigma-Aldrich (#L9634-500mg) and used at 5 µg/ml in phagocytosis assays. The following anti-target antibodies were used for Western blots (source): Syk (Abcam, #Ab155187), MALT1 (Cell Signaling Tech., #2494S), BCL10 (Cell Signaling Tech., #4237S), CARD9 (Abcam, #Ab133560), phospho-mTOR (Cell Signaling Tech., #5536P), mTOR (Cell Signaling Tech., #4517S), 4EBP1 (Cell Signaling Tech., #9644P), β-actin (Abcam, #Ab75186), and goat anti-mouse (LI-COR Odyssey, #926-68070) and anti-rabbit antibodies (#926-32213).  33   2.2.3 ELISA and blocking cytokine experiments   For enzyme-linked immunosorbent assays, supernatants stored at -80oC were analyzed in batches for IL-1β and IL-6 (eBioscience, #88-7261-76), or using a multiplex ELISA assay for IL-1α, IL-1β, IL-1RA, IL-6, TNF-α and IL-10 (R&D Systems) as per manufacturer’s instructions. For blocking experiments, MNCs were pre-incubated for 1 hour at 37oC with neutralizing anti-dectin-1 or anti-dectin-2 antibodies, mepazine hydrochloride (10µM, EMD Millipore #5005000001), cyclohexamide (100 ug/ml, Sigma Aldrich #C4859), ST 2825 (a MyD88 inhibitor, 10µM, AdooQ Bioscience, #A15248-1), GW5074 (a Raf-1 inhibitor, 1 µM, Cayman Chemical #10010368-1), Piceatannol (a Syk inhibitor, 20 µM, Cayman Chemical, #10009355-5). For flow cytometry analyses, MNCs were washed in PBS (3x) and stained for surface expression of CD14 (eBioscience, #25-0149-41), dectin-1 (AbD Serotec, #MCA4661A488). Intracellular cytokine staining was performed fixing/permeabilizing cells with Foxp3 Staining Buffer (eBioscience, #005523-00), and anti-IL-1β antibody (BioLegend, #508208). Data were acquired on a LSRII flow cytometer (Becton Dickenson) and analyzed using flowJo (flowJo, LLC).  2.2.4 Real-time qPCR experiments and data analysis   Total RNA was isolated using TRIzol LS (Thermo Fisher Scientific) followed by chloroform extraction and cleaned using RNeasy Mini spin columns (Qiagen) followed by ethanol precipitation. mRNA was reverse transcribed using SuperScript® III First-Strand Synthesis 34  SuperMix (Thermo Fisher Scientific, #18080400), and qPCR experiments were carried out in triplicates on a ViiA 7 system (Applied Biosystems) using both Power SYBR® Green (Thermo Fisher Scientific, #4368706) or TAQ-Man. Data were normalized on dsRED (ΔCt method) as detailed in [177]. Briefly, ΔCt was calculated for each fraction against dsRED. 2^ΔCt was then calculated, and the sum of all fractions determined. mRNA was expressed as a percentage of the total mRNA content represented by the sum of mRNA content in every fraction ((2^ΔCt x 100)/sum).   2.2.5 Pulse-labelling experiments   Monocytes were stimulated for 2.5 hours and pulsed with 35S-methionine (22.6 ug/ml) for the last 30 minutes of the experiment. Cells were washed with PBS and lysed in RIPA buffer (Santa Cruz Biotechnology); equivalent amounts of protein were loaded on polyacrylamide gel. Blots were imaged on film.  2.2.6 Cytokine stimulation   Mononuclear cells (MNC) were isolated by density centrifugation of blood diluted using LymphoprepTM (STEMCELL Technologies). Monocytes were purified from MNCs using an EasySepTM positive CD14 enrichment kit (STEMCELL Technologies, #18058) using 1 mM EDTA to prevent cell clumping. MNC or monocytes were cultured in RPMI 1640 (Roswell Park Memorial Institute medium) supplemented with 2 mM L-glutamine (Gibco, #25030), 10% human AB 35  serum (Gibco #11875-093) and Pen-strep (Gibco, #15140-122). MNCs were stimulated in round-bottomed 96-well plate (Corning Life Sciences, cat.3799) with fixed Candida (multiplicity of infection, MOI=5), LPS (10 ng/ml), Curdlan (10 µg/ml), Zymosan (10 µg/ml) for 24 hours unless otherwise specified.   2.2.7 Phagocytosis assay   MNC were incubated in 96-well plates in presence of Candida pre-stained with DAPI (Biolegend, #422801) for 15 min at room temperature. After 1-hour incubation, cells were centrifuged, washed in Trypan Blue/PBS for quenching of any residual fluorescence of non-phagocytized Candida. The cells were then washed once more in PBS before staining using a CD14 Pe-Cy7-conjugated antibody (eBioscience, #25-0149-42). Cells treated with cytochalasin D (10 µg/ml, Sigma-Aldrich) were used as negative control. Samples were analyzed on an Amnis Image Stream Imager. Data was analyzed using Amnis IDEAS software (v.5.0.343.0)  2.2.8 Western Blots   After stimulation (LPS or curdlan), monocytes were washed, lysed in RIPA buffer in presence of phosphatase/protease inhibitors (Santa Cruz Biotechnology, #sc-24948). Protein quantification was performed using a Pierce 660 nm protein assay (Thermo Fisher Scientific). Lysates boiled in 4X Laemeli buffer (Bio-Rad) supplemented with 2-mercaptoethanol were run on 4-20% mini-PROTEAN TGX gradient gels (Bio-Rad), transferred to PVDF membranes, and 36  incubated with primary and secondary antibodies. Blots were imaged on a LI-COR Odyssey 9120.  2.2.9 Polysome profiling, RT-PCR and pulse labelling   For polysome profiling experiments, monocytes were pre-cultured with cycloheximide, then washed once in PBS/cycloheximide and lysed in polysome lysis buffer (Mammalian ARTseq Ribosome Profiling Kit, Illumina, #RPHMR12126). Lysates were layered onto sucrose gradients in the presence of heparin (1 mg/mL), and spun in an ultracentrifuge (SW41, Beckman Coulter) at 4°C. After fractionation, purified, in vitro transcribed dsRED RNA was proportionally spiked into pooled monosome, disome, heavy or light polysome fractions as internal control. The primer sequences used are presented in Table 2.2.   2.2.10 Metabolic assays   Glucose uptake in monocytes was measured by flow cytometry using 2NBDG-flow. In brief, previously frozen MNCs were rested for 1h at 37°C in cRPMI before counting, and stimulated for 2h in presence of LPS (10ng/ml), with 2-(N-[7-Nitrobenz-2-oxa-1,3-diazol-4-yl]Amino)-2-Deoxyglucose (2-NBDG, 15uM) added for the last 40 min. Cells were washed twice before reading on a LSR or Fortessa flow cytometer (Becton Dickenson) and analyzed using flowJo (flowJo, LLC). For extracellular flux analyses, monocytes were plated on Cell-tak (Corning, #354240) coated XF96 well plates. Cells were rested in a non-CO2 incubator, and glycolysis 37  stress test was performed as per manufacturer protocol (Agilent Seahorse, #103020-100). After the assay, cells were lysed, and protein assay (Pierce 660nm) performed. ECAR was normalized to protein content. L-lactate was measured by colorimetry (Abcam, # ab65331).  2.2.11 Microarray analyses   Total RNA from purified CD14+ monocytes was extracted using the MagJET RNA purification kit (ThermoFisher Scientific). RNA was quantified using NanoDrop 1000 spectrophotometer, and its integrity was determined using the Agilent 2100 Bioanalyzer. Samples were hybridized onto an Illumina HumanHT-12 v4 Expression BeadChips and scanned with an Illumina iScan System (Illumina, San Diego, CA, USA). Data were analyzed using R. Data was preprocessed by quantile normalization and log2 transformation. Probes that were non-expressed (based on their detection p-values) were removed. For differential gene analyses, linear models were fit using limma [178]. Probes with FDR <= 0.05 were considered differentially expressed. Gene ontology analysis and Network images were generated using the ClueGO 2.3.3 plugin [179] for Cytoscape 3.5.1 [180]. Microarray data are deposited in GEO under accession code GSE104510.   2.2.12 ClueGO Analysis   Genes that were differentially expressed by two-fold or more compared to adult or term samples were included in pathway analysis using the ClueGO plugin (v2.3.3) in Cytoscape 38  (v3.5.1). GO biological-processes ontology was used. Degree of functional enrichment was determined by sorting enriched terms based on p-value, with a p-value cut off of 0.001. GO tree intervals between levels 3 and 8 and Kappa Score Threshold of 0.04 were used.  2.2.13 Statistical Analysis   GraphPad Prism was used for statistical analyses. To avoid excessive statistical comparisons, we used inferential statistics to test for differences between groups only when indicated from the data, or by experimental design. Differences between groups were analyzed using 2-tailed t tests or ANOVA as specified in the figure legends. Statistical significance is considered at p values of <0.05.    39  2.3 Results   2.3.1 Lack of anti-fungal innate immune recognition in early gestation   To examine responses to Candida species, we first compared the cells’ ability to phagocytose clinical strains of Candida between preterm, term neonatal, and adult mononuclear cells. Notably, preterm cells phagocytosed Candida albicans equally well as adults (Figure 2.1). Given the role of C-type lectins and TLRs in the recognition of invasive Candida yeast forms in humans, we also compared whether these PRRs were able to elicit a cytokine response between age groups. Surprisingly, mononuclear cells from preterm neonates were unable to respond to Candida albicans or C. parapsilosis, as demonstrated by a lack of production of IL-1, IL-6 ( Figure 2.2), but also other cytokines (Figure 2.3). Preterm cells also did not respond to stimulation of dectin-1 (using curdlan) or TLR4 (using LPS) (Figure 2.4, Figure 2.5, Figure 2.6) despite a strong response detected in adult and term neonatal cells. Preterm production of IL-1β in response to TLR-2 stimulation via zymosan was also diminished compared to term and adult, but higher than LPS and curdlan. This finding is consistent with a broad functional impairment involving multiple PRRs [51]. The importance of dectin-1 in our system was confirmed by efficient blocking of cytokine responses using a neutralizing receptor antibody (Figure 2.7). Dectin-1 blocking using the same antibody did not block phagocytosis of Candida particles (Figure 2.8). Phagocytosis of Candida was unaffected by blocking other receptors known to be important for the uptake of 40  this micro-organism (Figure 2.9), consistent with reports of functional redundancy among phagocytic receptors [30]. Altogether, our data suggest a primary involvement for dectin-1 in mediating responses to our clinical strains of Candida, in human mononuclear cells. Additionally, our initial observations suggest broad functional impairments in PRRs in preterm newborns, limiting cytokine responses, but not phagocytosis of Candida.   41    Figure 2.1 Phagocytosis of C. albicans by CD14+ Monocytes Phagocytosis of Candida, as measured as a percentage of monocytes that have incorporated particles (bars and whiskers), including a representative flow microscopy diagram (bar = 10 µm). Data pooled from multiple experiments over 14 months (9 to 17 subjects per age group; preterm samples range between 24 and 32 weeks gestation)   42    Figure 2.2 Inflammatory Cytokine Secretion in Response to Various Species of Candida Cytokine response (blood mononuclear cells) to C. albicans or C. parapsilosis (24h stimulation; bars and whiskers; n=10 preterm, n=12 term, n=18 adult C. albicans; n=6 preterm, n=8 term, n=12 adult C. parapsilosis). One-way ANOVA with post-hoc testing between each age groups for each given stimulus condition, * p<0.05, ** p<0.01, *** p<0.001   43   Figure 2.3 Cytokine Secretion in Response to Various Species of Candida Cytokine response (blood mononuclear cells) to C. albicans or C. parapsilosis (24h stimulation; bars and whiskers; n=4 per age group). IL-1α and TNFα are pro-inflammatory cytokines, while IL-1RA, and IL-10 are important for resolution of inflammation.      44   Figure 2.4 IL-1β Secretion in Response to Specific PRR Stimulation Production of IL-1β (24h; n=21 preterm, n=13 term, n=11 adult) in blood mononuclear cells in response to specific TLR-4 (LPS), or TLR-2 (zymosan) stimulation, Dectin-1 (curdlan). One-way ANOVA comparing age groups within each stimulus condition, only significant results shown.  * p<0.05, ** p<0.01, and *** p<0.001.    45   Figure 2.5 IL-6 Production in Response to Specific PAMP Stimulation Production of IL-6 (24h; n=10 preterm, n=7 term, n=5 adult) in blood mononuclear cells in response to specific TLR-4 (LPS), or Dectin-1 (curdlan) stimulation (24 hours). One-way ANOVA comparing age groups within each stimulus condition, only significant results shown.  * p<0.05, ** p<0.01, and *** p<0.001.    46   Figure 2.6 Pro-IL-1β Production in CD-14+ Monocytes Production of pro-IL-1β (5h, n=7 <27 weeks GA preterm, n=15 >27 weeks GA preterm, n=11 term, n=20 adult) in response to LPS, zymosan or curdlan in mononuclear cells (bars and whiskers; n=12 preterm, n=9 term, n=5 adult) as measured by flow cytometry (gated on CD14+ monocytes) after 24 hours of stimulation. Representative gating strategy and dot plots shown in Figure A.1. One-way ANOVA comparing age groups within each stimulus condition, only significant results shown. * p<0.05, ** p<0.01, and *** p<0.001.     47     Figure 2.7 IL-1β Production Blocked by Dectin-1 Specific Neutralizing Antibody Antibody blocking of dectin-1 reduces cytokine production in adult mononuclear cells (one-sided t-test with Welch’s correction for unequal variance; 6 to 12 subjects per condition). Ca=C. albicans   48   Figure 2.8 Phagocytosis of C. albicans is Not Dependant on Dectin-1 Phagocytosis of C. albicans (% monocytes) upon blocking with anti-dectin-1 receptor antibody, same-isotype control or Cytochalasin-D (Cyto-D). Cyto-D is an inhibitor of cytoskeletal rearrangement and used as a negative control (bars and whiskers; n=14 preterm, n=11 term, n=12 adult). Except for Cyto-D, none of the treatment conditions resulted in statistically significant blocking. Y-axis: percent of CD14+ monocytes that had phagocytosed at least one Candida particle. * p<0.05, ** p<0.01, and *** p<0.001 by paired 2-sided student t-test.    49   Figure 2.9 Effect of Blocking of Common Phagocytic Receptors on Phagocytosis of Candida Blocking of C. albicans (Ca) or C. parapsilosis (Cp) phagocytosis (% monocytes) using laminarin (dectin-1), mannan (dectin-2/mannose receptor), anti-DC-SIGN, anti-CD206 or a combination (bars = mean ± SD; 2 to 3 subjects/age group). Cytochalasin D (Cyto-D), an inhibitor of actin polymerization was used as a negative control of phagocytosis. Note that despite the low number of subjects, none of the blockings appear complete except for Cyto-D (control). Y-axis: percent of CD14+ monocytes that had phagocytosed at least one labelled Candida particle.     50  2.3.2 Transcriptome analyses indicate broad metabolic impairments in preterm monocytes   The lack of cytokine response to Candida in preterm cells was concerning given the major risk of invasive disease in these infants. However, the molecular basis for these impairments is unknown. To investigate this, we used an unbiased systems analysis approach comparing adult, term and preterm monocytes through genome-wide transcriptomic profiling. A gene ontology analysis comparing unstimulated monocytes from preterm, term neonates and adults, revealed major differences located mainly in metabolic pathways including glycolysis, oxidative phosphorylation and beta-oxidation (Figure 2.10, Figure 2.11). Notably, differences were seen through all three main energetic pathways (Figure 2.12), but also in a profound 51  down-regulation of ribosomal genes ( Figure 2.13). A more detailed schematic showing specific downregulation of electron transport chain component genes is shown in Figure A.2. 52  Upon stimulation with LPS, a large proportion of genes (40%) were upregulated after 5 hours in all 3 age groups (Figure 2.14, Figure 2.15). In contrast upon curdlan stimulation, we detected little functional response in preterm monocytes, suggesting a defect in signalling through dectin-1 (not shown). In independent qPCR experiments, expression of the IL1b and IL6 cytokine genes was detectable after 30 min, peaking between 5 and 8 hours in all three age groups. Most notably, we found that the transcriptional response of cytokine/chemokine genes was comparatively strong in all three (preterm, term and adult) age groups, including expression of Il1b, Tnfa, and Il6 (Figure 2.16), as well as other cytokine/chemokine genes ( Figure 2.17). This strong response at the gene expression level contrasted with the lack of response at the protein level (Figure 2.4, Figure 2.5, Figure 2.16), particularly for IL-1β, and IL-6. These data suggested to us a defect in protein synthesis in preterm immune cells.   53   Figure 2.10 Pathway Analysis of Differentially Expressed Genes in Preterm Monocytes Pathway analysis of differentially expressed genes in resting preterm monocytes reveal large scale differences in metabolic and protein synthesis pathways. Preterm gene expression compared to mean expression values from all age groups (monocytes; n = 8 to 12 subjects/age group; preterm ranged from 24 – 28 weeks gestation)   54   Figure 2.11 All Age-Related Differentially Expressed Gene Pathways in Unstimulated Monocytes. Gene-set enrichment analysis of clusters (colors, with pathway names) of differentially expressed genes across age groups (unstimulated preterm, term and adult monocytes; data from same samples as in Figure 2.10).   55   Figure 2.12 Heatmap of Gene Components of Major Metabolic Pathways Detailed examination of genes important to major metabolic pathways reveal large-scale downregulation of metabolic pathways in preterm monocytes. (Colour scale represents Log2(ratio) signal against mean).  56   Figure 2.13 Heatmap and Unsupervised Clustering of Genes Encoding Ribosomal Proteins Detailed examination of ribosomal genes reveals large-scale downregulation of ribosomal genes in preterm monocytes. (Colour scale represents Log2(ratio) signal against mean)  57  -1 0 0 0 1 0 0-1 0 001 0 0AdultTermPretermP C 1  (2 8 % )PC2 (9.3%)L P SN o  L P S Figure 2.14 Principal Component Analysis of Unstimulated Versus LPS Stimulated Monocytes Principal component analysis of unstimulated versus LPS stimulated monocytes (n=6 to 12 subjects/age group; preterm infants in this experiment ranged from 24 to 28 weeks gestational age). First principal component is identified as LPS stimulation and is responsible for 28% of observed variance. Second principal component is identified as age.   58     Figure 2.15 Venn Diagram of LPS-Induced Differentially Expressed Genes  Differentially expressed genes defined as +/-2-fold difference against unstimulated condition (FDR 5%).     59    Figure 2.16 Preterm Gene Expression Levels Are Not Correlated with Protein Secretion Heatmap of cytokine gene expression after LPS stimulation (5h) in monocytes compared to cytokine production (24h stimulation, effect of age by one-way ANOVA; mean ± SD; n = 3 to 6 subjects per age group) in mononuclear cells by Luminex ELISA (mean ± SD; n=3 preterm, n=4 term, n=3 adult), as measured by Luminex ELISA. (Colour scale represents Log2(ratio) signal against mean)   60    Figure 2.17 Cytokine/Chemokine Gene Expression Changes Following LPS-Stimulation in Monocytes Detailed examination of gene expression changes following LPS-stimulation shows that preterm monocytes differentially express cytokine/chemokine genes to a similar degree as other age groups    61  2.3.3 Defective translation of key immune response genes in preterm monocytes   To assess whether preterm monocytes lacked translation of immune response genes, we performed polysome analysis. Given the complete lack of dectin-1 response, we focused our experiments on this pathway to address whether key immune signaling proteins downstream of this receptor were expressed (Illustration 2.1). Lysates from adult, term and preterm monocytes were subjected to sucrose gradient fractionation; the concentration of specific mRNAs in monosome, disome, and light and heavy polysome fractions were measured by RT-qPCR.  This experiment was extremely challenging technically. Despite the limited number of monocytes obtainable from preterm infants, we were able to perform polysome profiling in all three age groups (Figure 2.18). Expression of the Malt1, Bcl10 and Card9 genes were detected comparably in total mRNA fractions between all three age groups (Figure 2.19). In general, the overall distribution of mRNAs across the polysome gradient was also similar between preterm, term and adult samples, but distinct between genes, suggesting that each mRNA is translated with different intensity (Figure 2.20). Actb, Clec7a and Card9 mRNAs, which respectively encode β-actin, the dectin-1 receptor and Caspase recruitment domain-containing protein 9, were associated with heavy polysome in preterm monocytes. The presence of these mRNAs in the preterm heavy polysome fraction indicate that these mRNAs are translated even at low gestational age. In contrast, the majority of Bcl10 and Malt1 mRNAs was associated with monosomes and disomes in preterm and adult monocytes, and only Malt1 mRNA was detected in heavy polysomes, in term monocytes (Figure 2.20). These results indicate that dectin-1 62  signaling genes are differentially translated across the gestational age spectrum. Pulse-labelling experiments showed blunted 35S-methionine/cysteine incorporation in preterm monocytes (Figure 2.21), indicating reduced translation at rest, as well as in response to LPS. Next, we asked whether these proteins were expressed. The dectin-1 protein was comparable expressed between all 3 age groups (Figure 2.22). In contrast, expression of the MALT1, but also the Syk and CARD9 proteins were severely reduced in preterm monocytes (Figure 2.23, Figure 2.24). For Bcl10, we were unable to detect significant protein expression in preterm monocytes by Western blotting (Figure 2.23). Therefore, the lack of protein expression for some of the essential signalosome proteins, despite corresponding mRNA expression, reinforced our suspicion for a translation defect in preterm immune cells.      63   Illustration 2.1 Dectin-1 Signalling Via MALT1 Signalosome Dectin-1 signalling is dependent upon assembly and activation of the Bcl10/Card9/MALT1 signalosome   64   Figure 2.18 Absorbance Profiles of Polysome Fractions Ribosome bound mRNA is not enriched in any age group (n=4/age group) 65   Figure 2.19 Gene Expression of Signalosome Components Quantification in each of signalosome genes (qPCR) in in monocytes (4 to 6 subjects/age group).    66   Figure 2.20 Quantification of Signalosome Genes (qPCR) in Total RNA Fractions Quantification of signalosome genes (qPCR) in total RNA fractions (top right panel) in monocytes (n= 4 subjects/age group; preterm subjects ranged between 24 and 27 weeks of gestation).   67    Figure 2.21 Reduced 35S-met/cys Incorporation in Preterm Monocytes Representative 35S-met/cys pulse labelling experiments using purified CD14-expressing monocytes. Cumulative quantification (whole lane) from 3 independent experiments (mean ± SD). Monocyte viability was confirmed using trypan blue exclusion to be >95% before pulse-labelling. Results are from equal loading of protein in each lane as determined by Bradford method. (n=3/group)      68    Figure 2.22 Protein Expression of Dectin-1 Surface expression of Dectin-1 (flow cytometry, gated on CD14-expressing cells; bars and whiskers; n=11 preterm, n=14 term, n=23 adult). 69    Figure 2.23 Protein Expression of Bcl-10 And MALT1 Representative western blots of MALT1 and Bcl-10 protein expression in monocytes after 0, 30 or 60 min LPS stimulation; cumulative Western blot quantification of proteins from all independent experiments performed in the same conditions (4 adult and preterm, and 2 term subjects; mean ± SD   70    Figure 2.24 Protein Expression of Syk and Card9 Representative western blots of Syk and Card9 protein expression cumulative quantification from 3 independent experiments (mean ± SD).   71  2.3.4 MALT1 is required for immune recognition of Candida spp. in human monocytes   The role of the MALT1 signalosome in dectin-1 responses in myeloid cells has been mainly studied in mice macrophages and dendritic cells, and has not been formally established in human monocytes [43], [181]. Moreover, due to the major role of MALT1 in T and B cells activation, it is also unclear whether signaling through this molecule is also essential for myeloid responses to Candida in humans [182].  To confirm this, we examined the effect of blocking MALT1 on both cytokine responses and phagocytosis of Candida species. Complete loss of cytokine response to curdlan was observed with MALT1 inhibition (Figure 2.25 A). When testing responses to clinical strains of Candida, MALT1 blocking also completely abrogated cytokine responses to C. albicans or C. parapsilosis (Figure 2.25 B). Blocking of Syk also partially abrogated responses to these two micro-organisms, indicating that the response to Candida primarily involves signalling through dectin-1 (Figure 2.25 B). On the other hand, neither blocking of MyD88 nor Raf-1 abrogated responses to this micro-organism (Figure 2.25 B).  Given that preterm monocytes are fully able to phagocytose Candida, we also examined the role of MALT1 in this function. MALT1 inhibition did not block uptake of Candida spp., indicating a non-essential role in phagocytosis (Figure 2.26). Altogether, these results confirm an essential role of the MALT1 signalosome in the recognition, but not the phagocytosis of whole Candida in human monocytes.    72   Figure 2.25 MALT1 is Essential For Dectin-1 Signaling  (A) IL-1β production (24h) in response to curdlan or LPS, or (B) in response to C. albicans or C. parapsilosis , in presence of a MALT1 inhibitor (mepazine hydrochloride), upon blocking of dectin-1 or dectin-2, or upon inhibition of MyD88, Raf or Syk (n = 6 experiments; bars = mean ± SD). Two-sided paired T-tests. 73   Figure 2.26 Phagocytosis is Independent of MALT-1 Paracaspase Activity Effect of MALT1 inhibitor (using mepazine hydrochloride), dectin-1 blocking antibody or Cytochalasin-D on the phagocytosis of C. albicans by monocytes (bars and whiskers; n=6 adult samples).    74  2.3.5 Altered cellular energy metabolism and protein synthesis in preterm monocytes   Recent data have shown that innate immune cells require increased glycolysis to provide the energy and metabolic intermediates necessary for fueling anabolic processes, such as the massive protein synthesis required during activation [183]. While these studies have been mainly conducted in macrophages and dendritic cells, we confirmed the importance of this mechanism also in human monocytes by showing inhibition of cytokine production with 2-DG, a non-functional glucose analog that blocks glycolysis (Figure 2.27).  Considering the reduced translation in preterm monocytes, we asked whether glycolysis could be impaired. To this end, we compared the glycolytic capacity of preterm, term and adult monocytes. Notably, glycolysis was severely diminished in preterm monocytes (Figure 2.28 A, B). The reduced glycolytic activity in preterm monocytes was also reflected in reduced lactate production (Figure 2.29) and reduced glucose uptake (Figure 2.30), at lower gestation, upon LPS but also at rest (in unstimulated cells). In term monocytes, glycolysis was variably affected, suggesting a transitional functional state. Given that phagocytosis is preserved in preterm cells, these data raise an important mechanistic question: do these metabolic constraints also affect phagocytosis? Indeed, the requirement for glycolysis and/or translation has been recently studied for cytokine responses, but only sparsely for other types of immune responses [183]. Consequently, we showed that phagocytosis of Candida was unaffected by blocking glycolysis, or even by blocking of de novo protein synthesis (Figure 2.31). These results indicate a specific requirement of glycolysis for PRR-mediated cytokine responses, but not for phagocytosis.  75   Figure 2.27 Blocking Glycolysis Limits Cytokine Secretion Effect of blocking glycolysis using 2-DG on cytokine responses (24h stimulation, mononuclear cells; 2-sided paired t- tests; bars and whiskers n=10 IL-1β, n=7 IL-6).     76   Figure 2.28 Reduced Glycolytic Activity in Preterm Monocytes (A) Extracellular acidification rates (monocytes), under baseline glucose-free conditions, after addition of glucose, oligomycin and 2-deoxy-d-glucose (2-DG, glycolytic inhibitor; representative of 3 independent experiments). (B) Glycolytic capacity (cumulative data pooled from 3 independent experiments with preterm samples ranging from 26 to 29 weeks gestation; 2-sided paired t-test; mean ± SD)   77     Figure 2.29 Reduced Lactate Secretion in Preterm Mononuclear Cells Lactate secretion in mononuclear cells after LPS stimulation (24h, p value = effect of age by 2-way ANOVA with matching for subjects between stimulation conditions; n=10 preterm, n=9 term, n=10 adult). Note that increase in lactate upon LPS was only statistically different between adults and preterm (p = 0.002) when tested by multiple t-tests correcting for multiple comparisons using the Holm-Sidak method and assuming equal variance.   - +02468Lactate (mM)A g e : p  =  0 .0 0 1L P SAdultTermPreterm78   Figure 2.30 Glucose Uptake in CD14+ Monocytes Fluorescent 2-(N-[7-Nitrobenz-2-oxa-1,3-diazol-4-yl]Amino)-2-Deoxyglucose (2NBDG) uptake measured by flow cytometry comparing preterm, term and adult monocytes, at rest and following LPS stimulation (n = 7 adult, 7 term and 5 preterm samples). P value represents difference between adult versus preterm (2-sided unpaired t-test between adults and preterm, on LPS condition only)    79   Figure 2.31 Effect of Blocking Glycolysis or Translation on Phagocytosis and IL-1β Secretion Effect of blocking glycolysis (using 2-DG) or translation (using cycloheximide, or CHX) on the phagocytosis of C. albicans (Ca) and IL-1β secretion n=3.     80  2.3.6 Developmental regulation of mTOR   The data thus far raise an additional important question: how are glycolysis and translation regulated in preterm monocytes? Mechanistic target of Rapamycin (mTOR) is a key regulator of the metabolic switch towards glycolysis during immune activation (Illustration 2.2) [184]. Therefore, we examined the mTOR regulator node developmentally. Interestingly, phosphorylation of mTOR was generally reduced following LPS stimulation in preterm cells (Figure 2.32). Preterm cells also displayed reduced expression of its main downstream target 4EBP1, which is important in mediating the effect of mTOR on translation downstream of PRR activation (Figure 2.32).  Next, we more closely examined regulatory gene expression within the mTOR pathway. Again, due to the absence of dectin-1/MALT1 signaling, we focused our analyses on LPS-stimulated cells. Major differences were observed in neonatal monocytes in the mTOR pathway, including a decreased expression of the upstream mTOR activator RAC-alpha serine/threonine-protein kinase (encoded by the Akt1 gene). Most notably, expression of the negative mTOR regulator NAD-dependent deacetylase sirtuin-1 (Sirt1) and the DNA damage inducible transcript-4-like (Ddit4l) molecules were profoundly upregulated in preterm monocytes. Ddit4l is a paralog of the DNA damage inducible transcript-4 (Ddit4) that has also been shown to inhibit mTOR [185], [186] (Figure 2.33).  Expression of the insulin receptor substrate 2 (Irs2) gene, which is an upstream regulator of mTOR, the genes encoding the adenosine monophosphate-activated protein kinase (prkaa2) and the hypoxia-inducible factor 1-alpha (Hif1a) were also upregulated in preterm 81  monocytes after LPS stimulation comparing age groups, which may represent compensatory mechanisms. Furthermore, mitochondrial electron transport chain transcripts were severely reduced in neonatal monocytes overall, even more so at lower gestation, suggesting a gestational age-dependent reduction in mitochondrial energy capacity (Figure A.2).    82   Illustration 2.2 mTOR Links Immune Activation to Metabolism and Protein Synthesis Depiction of signaling events between Toll-like receptor and Raf-1-mediated dectin-1 activation, mTOR phosphorylation, and increased glycolysis and protein synthesis   83   Figure 2.32 Western Blot of mTOR Phosphorylation and 4EBP1 Expression Western blot of mTOR/4EBP1 expression and mTOR phosphorylation (representative experiment in monocytes).   84   Figure 2.33 Increased Gene Expression of mTOR Regulators in Preterm Monocytes Expression of mTOR regulators in unstimulated and LPS-stimulated monocytes from microarray data (only P values significant across age groups are shown, by 2-way ANOVA 6-12 subjects/age group, not corrected for multiple comparisons).   85  2.3.7 Gestational age-dependent immune recognition and risk of invasive Candida infections   Finally, we sought to determine how reduced cytokine responses in preterm cells may impact infants’ risk of invasive infections. The relationship between immune functions and preterm newborns’ increased risk for infection is an area that is very much understudied. This requires large cohorts and systematic biological outcomes which are difficult to achieve together in the same study. However, attempting to provide evidence for a relationship between low Candida responses and preterms’ risk of Candidemia, we reviewed data from 39,336 infants born below 33 weeks of gestation in Canada over 10 years. We found that rates of invasive infections exponentially increased with decreasing gestational age (Table 2.1, Figure 2.34). When specifically looking at Candida infections, rates of candidemia also increased exponentially at lower gestation, confirming previous studies [23].  Moreover, rates of invasive Candida infections inversely correlated with in vitro responses to this micro-organism (Figure 2.35). Altogether, this association supports a role in these immune deficits in increasing the risk of invasive infections in these infants, at least in part. However, this requires more definitive evidence to be able to causally link immune functions with preterm infants’ vulnerability to this pathogen.     86    Gestational age (weeks)   24 25 26 27 28 29 30 31 and 32 Total number of infants 1822 2618 2975 3493 4124 4492 5560 14252 Total number of infants who had infection(s) 702 981 861 867  772 655 521 714 Total infants with Candida infections 67 64 35 22 17 12 15 10  Table 2.1 Gestational Age-Related Prevalence of Infections  Prevalence of infections in infants born below 33 weeks of gestation in Canada between 2003 and 2013. Data was provided by the Canadian Neonatal Network (www.canadianneonatalnetwork.org).       87      Figure 2.34 Rates of Candidemia in Preterm Infants Incidence of Candidemia in infants born below 33 weeks of gestation in Canada between 2003 and 2013 (n = 39,336 infants; bars = 95%CI).     2 4 2 6 2 8 3 0 3 205G A   (w e e k s )Candidemia/100 infants88    Figure 2.35 Candida Induced IL-1β Production by Gestational Age Candida-induced production of IL-1β by gestational age from mononuclear cells (n = 31 preterm infants), compared to term neonates (n=12) or adults (n=18); pink area and error bars = 95%CI.     89   Primer name Sequence Dectin-1-forward 5’- TCGACTCTCAAAGCAATACCAG -3’ Dectin-1-reverse 5’- CCACAGCTATCACCAGTATTACC -3’ dsRed-forward 5’- TGAAGCTGAAGGTGACCAAG-3’ dsRed-reverse 5’- GGACAGCTTCTTGTAGTCG-3’ MALT1-forward 5’-AGGCTATGGAACACACTGAAG-3’ MALT1-reverse 5’-ACCACTGATATTGAACAAAAGGATG-3’ CARD9-forward 5’-CACCCAGCTCTCAGACAAAG-3’ CARD9-reverse 5’-CTTAACAAACGGCCCCAATG-3’ BCL10-forward 5’-GAAATATAAAACTAGAACATCTGAAAGGAC-3’ BCL10-reverse 5’-TGGTACATGACAGTGGATGC-3’ ACTB-forward 5’-CCTTGCACATGCCGGAG-3’ ACTB-reverse 5’-ACAGAGCCTCGCCTTTG-3’  Table 2.2 Primer Sequence Used in qPCR Gene Amplification.   90  2.4 Discussion    In this study, we provide novel insights into the fundamental mechanisms regulating immune activation during ontogeny. Innate immune cells play a cornerstone role in the early and late phase of immune activation during infections, initially providing alarmin signals to instruct adaptive immune cells in the identification of dangerous stimuli, but also later in allowing proper healing and recovery following tissue injury. Our data revealing the importance of cellular energy metabolism in regulating the activity of innate immune cells during development offers a mechanism to reconcile how their functions can be broadly constrained during this period.  Protein synthesis is an energetically expensive process, consuming ~45% of cellular ATP in the resting cell [187]. Upon immune activation, glycolysis has been shown to be essential to provide rapid energy, to fuel the major energy requirement but also to supply metabolic intermediates necessary for protein synthesis [125]. Considering our data, we posit that limiting innate immune reactivity in utero is not only necessary during the early establishment of self-tolerance but serves to obviate the energetic cost of dispensable immune responses to an infection before fetuses reach a viable stage. Our observations that metabolic requirements differ for other immune functions such as phagocytosis is in keeping with previous studies [128], and may stand from the need of innate immune cells to retain some functions, namely the ability to uptake particles or cells for tissue remodelling throughout embryonic and fetal development [188].  91  Our data contributes to the understanding of the immunological basis underlying premature infants’ vulnerability to fungal infections and provide important insight on the role of dectin-1 and MALT1 in the immune responses to Candida in humans. Based on a large population-based Canadian cohort over 10 years, we demonstrate that the prevalence of invasive Candida infection sharply increases at lower gestational ages, and inversely correlates with immune response to Candida. The prevalence and genera of candidemia in preterm infants vary widely across the world [189], with highest rates reported amongst infants up to 33 weeks of gestation [190], [191] whereas in other geographical areas invasive infections are observed only in extremely premature infants [192], [193]. The reasons for these differences are not entirely clear, but may be due to variations in clinical practices, including the use of broad spectrum antibiotics, and in-hospital sanitary measures, which have been shown to affect rates of colonization [194]. Our large study population and our data adds to our understanding of the epidemiology of Candida infections in newborns across the world.  Previous in vitro studies of antifungal innate immune responses have been largely conducted using murine dendritic cells and macrophages. In dendritic cells, blocking MALT1 using siRNA strategies abrogated cytokine responses to multiple strains of Candida [43], [181]. On the other hand, the role of MALT1 in anti-fungal responses in humans has been obscured by significant T cell and B cell activation defects in humans with inherited loss-of-function mutations, in addition to the potential impact on innate immune functions [51]. Here, we confirm that MALT1 plays a central role in the immune recognition of Candida in human monocytes. Moreover, the observation that blocking of Syk and RAF only partially abrogated 92  responses to Candida suggests that other, yet unidentified Syk-independent receptors can activate human myeloid cells via MALT1.  In contrast, in vivo studies indicate a partial role for dectin-1 in the recognition of Candida in humans [195]. In keeping with these previous observations, blocking of the dectin-1/2 receptors in our studies only partially affected the recognition of Candida. Moreover, uptake of Candida was also only marginal reduced by blocking of dectin-1/2. These observations are consistent with multiple overlapping functions of PRRs in the recognition of Candida [34], [195], [196], and with clinical observations that humans with functional dectin-1 mutations are susceptible only to mild mucocutaneous, but not invasive Candida infections [174]. Overall, our results add to the role of MALT1 and dectin-1 in the immune recognition of Candida in human monocytes and further support a need for a broad attenuation of immune functions in preterm infants across multiple PRR families in order to explain their vulnerability to this pathogen.  The transcriptome and metabolic profile of preterm monocytes ex vivo is reminiscent of the immune-paralyzed phenotype that was recently described in human adult subjects following an endotoxin challenge [140]. Immune-paralysis is characterized by a large-scale downregulation of major metabolic pathways (i.e. glycolysis, fatty acid oxidation, and oxidative phosphorylation), and a generally low metabolic activity, in contrast to that of an activated innate immune phenotype, in which glycolytic pathways are upregulated [135]. Preterm birth is associated with intra-uterine infections, raising the possibility that the signature observed in preterm monocytes is a metabolic consequence of a chronic exposure to chorioamnionitis. This remains to be explored further. However, our previous observations that immune responses 93  are generally suppressed in infants regardless of any detectable signs of histological chorioamnionitis argue against this possibility. Rather, we propose that these processes are developmentally regulated potentially through epigenetic mechanisms. Consistent with a role for epigenetics in regulating immune activation  in a cell-heritable manner during ontogeny, differences in DNA methylation [9] and histone structure [197] have been reported between preterm and term monocytes.  Our data raise important questions that merit further investigations. For instance, how translation is affected by the metabolic constraints in preterm monocytes is unclear. Intriguingly, transcripts that were relatively more abundant also showed less detectable protein. These data suggest that there is a reduction in capacity for protein synthesis in preterm cells that becomes limiting upon activation of immune cells. 5’TOP-containing mRNAs such as ribosomal proteins and translation factors, 5'non-TOP mRNAs and mRNAs with short 5'UTRs have been shown to be more sensitive to mTOR activity and thus in such case the availability of eIF4F complexes (via regulating 4E-BP) may limit the synthesis of those proteins [198]–[200]. In preterm cells, it is possible that under activation conditions where mTOR activity is reduced, the rate of translation of these proteins may be less than the rate of turn-over, due to cis-acting elements in their 5'UTRs. Consistent with our data, inhibition of mTOR using PP242 reduced translation of MALT1 [200]. In summary, our data reveal that preterm infants show profoundly reduced immune responses to Candida is associated with a lack of downstream intracellular signalling proteins involved in the recognition of this micro-organism including MALT1, and Bcl10. Additionally, we show that preterm neonatal monocytes have impaired glycolytic and translation capacity, at 94  the same time providing a mechanism to explain the globally suppressed immune reactivity in myeloid cells before the third trimester of gestation. Future studies are required to understand how glycolysis and translation are selectively impaired during this developmental stage. However, at the same time our data suggests an important role for mTOR inhibition, potentially through DDIT4L [186], and SIRT1. The availability of pre-clinical studies making use of these inhibitors [201] may offer unexplored potential therapeutic avenues to prevent infections in this high-risk age group.  95  Chapter 3: Immune variability in healthy adults   3.1  Background   There is considerable variability in immune responses in humans. The source of this variability is incompletely understood though it likely affects the outcome of conditions such as autoimmunity and malignancies. Moreover, understanding how this diversity is regulated at the molecular level in newborns may shed light on the regulation of these responses in adults.  Genome-wide association studies (GWAS) examine relationships between genetic variants and biological traits or diseases. The related concept of expression Quantitative Trait Loci (eQTLs), are genetic loci that contribute to changes in gene expression. eQTLs have been traditionally used as a proxy for assessing changes in immune responses [202]. For example over a thousand eQTLs modulate gene responses to LPS stimulation in human monocytes [202]–[204]. A potential limitation of this approach is that the expression of a gene often does not always correlate with function, or even with the expression of the corresponding protein [159]. Furthermore, environmental influences have been shown to modulate an individual’s ability to respond immunologically, and may dilute the effect of genetic variants on these immune responses [205]–[209]. Data provided in Chapter 2 of my thesis suggest that energy metabolism can profoundly impact the reactivity of immune cells in newborns developmentally [3], [6], [124], [210]. In a twin study, heritability accounted for a relatively low portion of the overall variance in immune responses within the general (adult) population [169]. In conjunction with cellular metabolic and environmental influences, there are examples of other 96  modulatory factors, such as our own microbiome through production of specific microbial metabolites [209]. Age, ethnicity, sex, geographical location, microbial interactions and genetic polymorphisms are but a few of the many factors that contribute to inter-individual variability [169], [205], [207], [211]. Finally, changes in immune gene expression have been reported over time in older individuals [170], suggesting a cumulative effect from environmental exposures. Altogether these observations suggest that an individual’s genetic make-up may not directly predict complex immunological phenotypes. A corollary to this hypothesis is that immune phenotypes may only be marginally represented by gene expression events, cautioning against the overemphasis of traditional eQTL or GWAS approaches to examine the impact of genetic diversity on immune functions.  In Chapter 3 of my thesis, I examined how variability in transcriptome responses is associated with production of two specific model cytokines: IL-1β and IL-6. I hypothesized that the transcriptome would predict production of these two cytokines. However, I found that variability in the transcriptome inconsistently predicted outcomes along these two cytokine response pathways. These results suggest cumulative stochastic effects modulating the impact of gene expression on immune function along these pathways. This has important implications in our understanding of how genetic variants impact immune diversity in human populations.    97  3.2  Methods   3.2.1 Recruitment of human subjects and blood sample collection   Peripheral blood was collected from healthy adult volunteers following written informed consent, in sodium heparin-anticoagulated Vacutainers (BD Biosciences) and processed within 2 hours of collection. Our study was approved by the C&W Research Ethics Board (protocol# H05-70519 and H07-02681). Ethnicity of participants was self-assessed.  3.2.2 Standard methods to determine variability   Initial (pilot) experiments were performed to establish the reproducibility of the different cytokine outcomes measures used in my study, based on 24 adults (12 male and 12 female). Briefly, peripheral blood mononuclear cells were extracted, cultured, and stimulated with LPS +/- ATP for five hours. Following stimulation, supernatants were collected for IL-1β and IL-6 cytokine ELISAs, flow cytometry was performed to determine monocyte composition, caspase-1 activity levels, and pro-IL-1β levels. These measures were repeated with the same individuals one month later to determine the intra-individual variability over time. The technical variability of outcomes was assessed in the following ways: 1) the replicate technical variability was assessed by comparing simple duplicate measures within each experiment; 2) The temporal technical variability was determined by analyzing the same aliquots of samples from 98  each experiment stored at -80°C, re-analyzed (e.g. ELISA, and RT-qPCR) 3 to 6 months apart, as a measure of stability of reagents/samples over the period of subject enrollment; 3) the total variability of the measures, which include both the biological and technical variability, was assessed using independent samples (i.e. fresh peripheral blood collected 1 month apart). Finally, the biological variability in each outcome measure was estimated by subtracting the technical variability from the total variability. After these pilot experiments, another series of measures were conducted on a group of 100 adult Caucasian males aged 18-65 years (we ended up recruiting fewer, as detailed in results). For this second phase study, we also measured expression of the IL1B and IL6 mRNA, pro-IL-1β, IL-6 and IL-1β proteins, and caspase-1 activity using the same rigorous standard protocols used in the pilot phase, at one single time point. Added to these measures, both unstimulated and stimulated (5h LPS) monocytes were subjected to parallel genome-wide transcriptome microarrays for gene expression analyses (see below).  Throughout this entire study, matched lots of reagents (antibodies, ligands, and kits) were used to minimize experimental variability (Illustration 3.1). Transcriptome studies were also conducted in a single batch, at the end of the enrollment period.  3.2.3 Cell purification, stimulation and cytokine detection by ELISA   Peripheral blood mononuclear cells (PBMC) were obtained as described in Chapter 2. Monocytes were isolated by positive selection (>95% purity, assessed by flow cytometry using a CD14-PECy7-conjugated fluorescent antibody) from freshly isolated PBMC, using EasySep TM for 99  Human Monocyte enrichment kit (STEMCELL Technologies, Vancouver Canada). PBMC (5.0 x 105 cells/well) or monocytes (5.0 x 105 cells/well) were stimulated with lipopolysaccharides (LPS, 10 ng/ml, unless specified otherwise), for 5 hours with or without ATP (5mM, MP Biomedical) added in the last hour of culture. These stimulation conditions were based on standardized duration of stimulation period and LPS dosage corresponding to peak responses in previous experiments [51]. LPS (TLR4 agonist) was obtained from InvivoGen (catalog# tlrl-pelps). Following LPS stimulation, mononuclear cells were immediately analyzed by flow cytometry (caspase-1 activity, pro-IL-1β expression). Supernatants from stimulated/unstimulated mononuclear cells (for IL-1 and IL-6) and monocyte cell pellets (for qPCR and transcriptome arrays) were immediately frozen at -80°C for batch analyses. IL-1 and IL-6 in culture supernatants were quantified in batches by Enzyme-Linked ImmunoSorbent Assay (ELISA, eBioscience).  3.2.4 Detection of intracellular pro-IL-1 production and caspase-1 activation   For detection of intracellular IL-1, cells were stained using an anti-CD14-PE-Cy7-conjugated and anti-IL-1-FITC-conjugated antibodies (eBioscience) in the Foxp3/Transcription Factor Staining Buffer Set (Cat# 00-5523-00, eBioscience). For caspase-1 activation (expressed as percentage activated cells), cells were stained using the FITC-conjugated Fluorescent-Labeled Inhibitor of Caspase-1 Z-YVAD (FLICA, ABD Serotec) as described [212]. Data were acquired immediately after staining on a LSR-IITM flow cytometer (Becton Dickenson). During data acquisition, standard voltage settings were set using single antibody staining cell controls. 100  Gatings were set using fluorescence-minus-one (FMO) cell controls. Compensations were determined using single color positive and negative control CompBeads (BD Biosciences). Flow cytometry data were analyzed with FlowJo vX.0.7 (TreeStar Inc., OR) for Windows.  3.2.5 Real-time PCR experiments    Messenger RNA was extracted using the RNeasy column extraction kit (Qiagen). DNAse (Qiagen) I-treated RNA was reverse transcribed into cDNA using Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Scientific). Real-time qPCR was performed on a ViiA 7 System (Applied Biosystems). The PCR conditions included a 10-minute pre-incubation step at 95°C, followed by 45 cycles of a sequence at 95°C for 5 seconds, 60°C for 10 seconds and 72°C for 5 seconds, and a heat dissociation melting curve to identify products. Sample cDNA copy numbers were calculated based on the standard curve generated by serial dilutions of DNA plasmids containing each gene studied, using the second derivative amplification threshold values (Ct). IL1B gene expression levels were normalized to that of -actin (IL1B:ACTB). Similar results were obtained whether normalizing IL1B gene expression to ACTB or GAPDH (data not shown), therefore only data after ACTB normalization are presented. All gene amplifications were performed in triplicates. Primer sequences for PCR amplification are listed in (Table 3.5). Inter-replicate coefficients of variability were <10%.    101  3.2.6 Gene expression microarray analysis    RNA was extracted (monocytes) using the RNeasy kit (Qiagen) and quantified using a nanodrop (ThermoFisher). RNA quality was quantified by RNA integrity number (RIN) Bioanalyzer (Agilent). Illumina total prep RNA amplification kit was used to amplify and biotinylate the cRNA. Subsequently, this cRNA was loaded onto the Human HT-12 v4 BeadChip (Illumina) and run on an Illumina iScan System (Illumina, San Diego, CA, USA). Data obtained was extracted using GenomeStudio (v.2011.1) with the Gene Expression module (v1.9.0) and normalized using Bioconductor’s ‘lumi’ package in R studio [213]. After normalization, probes with a detection p value <0.01 in at least one of the samples, were selected for further analysis. These sets of probes were used for a principal component analysis (PCA) using the ‘prcomp’ function in R studio. Next, the contribution of detectable probes to functional outcomes (IL1B/IL6 transcript levels, IL-1β/IL-6 protein levels, pro-IL1β protein levels and Caspase-1 activity) was assessed using linear modeling. The multivariable linear modeling outcomes were computed using Bioconductor’s ‘limma’ package [178].   3.2.7 Pathway analysis   Differentially expressed gene probes at a false detection rate (FDR) <0.3, or probes that were associated with measured outcomes (by linear modeling) were tested for enrichment in pathways using Gene Set Enrichment Analysis (GSEA). Briefly, all detectable probes were ranked by fold change (unstim vs LPS) or association with measured outcome (p-value), and 102  then tested for enrichment against several databases, including canonical pathways from BioCarta, KEGG, WikiPathways, NetPath, and REACTOME. Ranked probes were simultaneously tested against multiple databases to reduce biases inherent to individual sources [214], and results were ranked by q-value.   3.2.8 Statistical analyses   Correlations were calculated using Spearman coefficients. A two-sided Mann-Whitney U Test was used for simple group comparisons, and a p value of 0.05 was considered as statistically significant. To perform hierarchal clustering analysis, a correlation matrix was, again, assembled based on gene expression and measured outcomes (Spearman’s rho value > 0.9). Outcomes were then clustered based in a Euclidian manner based on coefficients of correlation.   103  3.3 Results   3.3.1 Stability of immunological traits among individuals   To examine intra-individual immune variability, we first determined the stability of immune responses within individual subjects over time. To this end, blood was collected for isolation of mononuclear cells from the 12 male subjects and 12 female subjects in the pilot phase of the study, over a one-month interval. Samples were stimulated (LPS ± ATP) for measures of IL1B, IL6 gene expression, pro-IL-1 protein production, caspase-1 activation and IL-1β and IL-6 cytokine secretion as detailed in the methods (Illustration 3.1).   All five outcomes correlated over time, indicating remarkable preservation of immune responsiveness within individuals (Spearman coefficients ranging from 0.70 to 0.86; p<0.0001; Figure 3.1). In other words, individuals who were more responsive immunologically remained so for at least one month, raising interesting questions about the factors driving this individual responsiveness. On the other hand, some degree of variability was observed in these outcomes over time (Figure 3.1). When looking at these outcomes as a function of age, or sex, both variables did not significantly correlate with any of the outcomes ( Figure 3.2). Female subjects showed higher variability compared to males, particularly for caspase-1 activity, although averages between males and females were not statistically significantly ( Figure 3.2 G-L, Figure 3.3).    104   Illustration 3.1 Determining Variability in Functional Immune Responses  Schematic demonstrating approach used to estimate biological variability based on technical variability (replicate = instrumental and sample preparation plus temporal = sample storage).  105    Figure 3.1 Intra-Individual Correlations Between LPS-Stimulated IL-1β and IL-6 Responses in Adults Over 1 month (A) IL1B mRNA (Spearman r=0.7669, p< 0.0001), (B) pro-IL-1β intracellular protein (Spearman r=0.7731, p< 0.0001), (C) secreted IL-1β (Spearman r= 0.6963, p< 0.0001), (D) caspase-1 activity (Spearman r= 0.8590, p< 0.0001), (E) IL6 mRNA (Spearman r= 0.8028, p< 0.0001) and (F) secreted IL-6 (Spearman r= 0.7603, p< 0.0001).106   Figure 3.2 Impact of Age and Sex On IL-1β and IL-6 Outcome Measures Correlation of (A) IL1B mRNA, (B) pro- IL-1β, (C) secreted IL-1β, (D) caspase-1 activity, (E) IL6 mRNA and (F) secreted IL-6, with age. None of the correlation were statistically significant (Spearman’s test p-value >0.05 in all cases). (G-L) Levels of (G) IL1B mRNA, (H) pro- IL-1β, (I) secreted IL-1β, (J) caspase-1 activity, (K) IL6 mRNA and (L) secreted IL-6 showed no significant difference between males and female (Mann-Whitney U-test p-value >0.05 in all cases). The data in panels A-F were based on the second series of 97 adult male subjects. The data in panels G-L were based on pilot experiments comparing 12 male and 12 female subjects.107   Figure 3.3 Variability of Caspase-1 Activity by Sex Over One Month Data based on 12 male and 12 female subjects in pilot study.    108  3.3.2 Biological contribution of the variance in individuals’ immunological traits    To determine the source of the variability in outcomes, we performed several control experiments, as detailed in methods. Replicate experiments indicated that the variability due to technical (experimental) factors was minimal (Figure 3.4 A-D). For IL1B mRNA expression, prolonged storage of samples >3 months was an important source of technical variability; although this variability was still lower than the overall (total) variability (Figure 3.4 A). For IL-1 cytokine measures, the technical variability was negligible (Figure 3.4 C). Importantly, the total variability in outcome measures, over the course of a month was significantly higher than the technical variability seen in the replicates (Figure 3.4 blue vs green). Altogether, these experiments suggest that most of the variability we detected in the IL-1 outcome measures were due to biological factors rather than technical factors.  Next, we examined the relationship between each outcome. Secretion of IL-1β was most strongly predicted by caspase-1 activity (Figure 3.5 A-D), followed by expression of the IL1B mRNA and of the pro-IL1β protein (Figure 3.5 B,F). For IL-6, secretion was well predicted by its mRNA levels (Figure 3.5 E). 109   Figure 3.4 Estimates of Technical versus Biological Variability in Outcome Measures.  Bland-Altman plots of the replicate technical variability (green), temporal technical variability (blue) and total variability (red) in (A) IL1B mRNA, (B) pro- IL-1β, (C) secreted IL-1β, (D) caspase-1 activity. (E) Co-efficient of estimated biological variability over a one-month period for each outcome measures. Data is based on pilot samples from 12 male and 12 female adult subjects.  110   Figure 3.5 Correlation Between IL-6 And IL-1β Outcome Measures Following LPS Stimulation. (A) Intracellular pro-IL-1β did not correlate with IL1B mRNA levels – Spearman r= 0.1460, p = 0.23) However, secreted IL-1β was significantly correlated to (B) IL1B mRNA levels (Spearman r= 0.5060, p < 0.0001) (C), pro- IL-1β (Spearman r=0.4711, p< 0.0001) (D) and caspase-1 activity (Spearman r=0.6269, p< 0.0001). (E) IL-6 secretion significantly correlated with levels of IL6 mRNA (Spearman r=0.4322, p= 0.0001).  (F) Unadjusted partial correlation with 95% confidence intervals showing the contribution of each upstream IL-1β outcomes to the secretion of the protein itself. Panels A, B, E and F were computed using data obtained from 73 healthy individuals. 97 healthy samples were used for panels C and D.111  3.3.3 Co-expressed transcriptomic networks after LPS stimulation   To understand the extent to which changes in the transcriptome predicted variability in measured outcomes, we analyzed LPS stimulated monocytes in a separate cohort of 73 adult males. For this analysis, we chose to focus on a mono-ethnic group of Caucasian males to limit the variance in outcome measures, and to preserve power within our achievable sample size. Transcriptome responses were analyzed using an Illumina Human HT12 array covering 46,000 gene probes at the genome-wide level. In transcriptome analyses, about 40% of all gene probes (57% of detectable probes) on the array were differentially expressed after 5 hours of LPS, using a False-Discovery Rate cut-off of 0.01 (Figure 3.6, Table 3.1). First, a clustering analysis was conducted to identify chemokines and cytokines genes that were expressed co-linearly. IL1B and IL6 mRNA clustered closely with a small group of chemokines, amongst which CCL20 is the closest (Figure 3.7).  A Gene Set Enrichment Analysis (GSEA) of differentially expressed genes between LPS stimulated and unstimulated samples was conducted. Significantly enriched top pathways are listed in Table 3.2. Unsurprisingly, pathways related to innate immune functions are highly enriched upon LPS stimulation.   112   Figure 3.6 Differentially Expressed Genes Associated with LPS stimulation  (A) Venn diagram showing that of 47217 total probes on our array, 70% (33140) of were detectable, with nearly 40% (18993) significantly differentially expressed after LPS stimulation (computed using linear models of microarray – limma, with a false detection rate cut-off of 1%). (B) Distribution of samples shown on principal components 1 and 2, including all unstimulated (blue) and stimulated samples (red). (C) Volcano plot showing changes in expression for all genes between the unstimulated and the LPS-stimulated samples. Significant probes (p-value <0.01) are shown in red (unaffected probes in black). Vertical lines at x =1 and -1 show 2x and -2x fold change, respectively. Probes showing the highest and lowest expression (top and bottom 0.05%) are listed by gene name. 73 LPS stimulated and 64 unstimulated individual monocytes were used for the data shown in panels A-C.  113   Table 3.1 Top 25 Statistically Significant Differentially Expressed Genes (LPS vs Unstim) Linear modeling analysis of all detectable probes between 73 LPS stimulated and 64 unstimulated monocyte samples from white Caucasian male cohort.    Gene Name Probe IDLog(2) Fold ChangeAverage Expression P.ValueAdjusted P.valueIL6 ILMN_1699651 6.32 10.24 5.68E-141 9.42E-137CCL4L2 ILMN_1716276 6.15 10.29 3.18E-140 3.52E-136CCL2 ILMN_1720048 6.05 10.14 5.29E-144 1.75E-139LOC728835 ILMN_3235832 5.92 10.45 6.91E-140 5.73E-136CCL20 ILMN_1657234 5.83 10.34 3.96E-136 1.64E-132IL1A ILMN_1658483 5.54 9.95 1.54E-137 8.50E-134CCL4L1 ILMN_2100209 5.40 10.46 2.35E-111 8.03E-109SERPINB2 ILMN_2150851 5.17 10.04 3.95E-125 6.54E-122C15ORF48 ILMN_1805410 5.15 10.28 2.24E-126 4.64E-123SERPINB2 ILMN_2150856 5.04 10.27 7.91E-126 1.54E-122TNFAIP6 ILMN_1785732 4.91 9.66 2.64E-128 6.74E-125LOC730249 ILMN_1838319 4.81 9.39 3.01E-107 6.48E-105CTSL1 ILMN_1812995 4.73 10.94 3.57E-125 6.23E-122CXCL2 ILMN_1682636 4.72 10.40 7.36E-123 8.71E-120FOS ILMN_1669523 -4.65 10.78 3.28E-117 2.09E-114CXCL1 ILMN_1787897 4.58 9.87 2.18E-120 2.09E-117CTSL1 ILMN_2374036 4.48 11.11 1.83E-119 1.60E-116IL1F9 ILMN_2158713 4.47 9.20 2.21E-120 2.09E-117SOD2 ILMN_2406501 4.36 10.63 2.03E-121 2.24E-118TNF ILMN_1728106 4.35 9.97 5.52E-116 2.95E-113IL1RN ILMN_1689734 4.24 9.70 7.31E-98 7.55E-96IRAK2 ILMN_1745964 4.23 9.27 1.42E-136 6.70E-133IL7R ILMN_1691341 4.23 9.19 5.44E-111 1.78E-108CXCL5 ILMN_2171384 4.23 9.07 1.53E-85 6.93E-84IL1RN ILMN_1774874 4.14 10.35 2.38E-117 1.58E-114114   Figure 3.7 Hierarchal Clustering of Chemokine/Cytokine Genes Based on Association with LPS Stimulated Transcriptome Dendrogram represents the clustering of chemokine and cytokine transcripts based on how well they correlate with each other in the LPS stimulated dataset. Data in this figure was obtained after correlation of 73 LPS stimulated and 64 unstimulated monocyte samples from healthy adults, with the stimulated chemokine and cytokine transcripts shown above.   115   Table 3.2 Top 20 Pathways Associated with Differentially Expressed Genes (LPS versus Unstim; Gene Set Enrichment Analysis) Pathways were ranked by statistical significance (p value). Including only genes with >2-fold change in differential expression (unstim vs LPS; FDR <1%).   Pathway NameSet sizeCandidates contained% Candidates in Set p-value q-valuePathway SourceImmune System 1174 142 12.10% 1.87E-11 4.19E-08 ReactomeTNF alpha Signaling Pathway 93 26 27.96% 6.63E-10 7.43E-07 WikipathwaysRheumatoid arthritis - Homo sapiens (human) 89 25 28.09% 1.27E-09 8.69E-07 KEGGTNF signaling pathway - Homo sapiens (human) 110 28 25.45% 1.55E-09 8.69E-07 KEGGTCR 244 44 18.03% 5.35E-09 2.28E-06 NetPathOsteoclast differentiation - Homo sapiens (human) 131 30 22.90% 6.10E-09 2.28E-06 KEGGLegionellosis - Homo sapiens (human) 55 18 32.73% 1.90E-08 6.05E-06 KEGGHemostasis 493 69 14.00% 2.31E-08 6.05E-06 ReactomeSpinal Cord Injury 116 27 23.28% 2.43E-08 6.05E-06 WikipathwaysChemokine signaling pathway - Homo sapiens (human) 189 36 19.05% 3.03E-08 6.78E-06 KEGGNuclear Receptors Meta-Pathway 314 50 15.92% 4.03E-08 8.22E-06 WikipathwaysInfluenza A - Homo sapiens (human) 175 34 19.43% 5.05E-08 9.44E-06 KEGGCytokine Signaling in Immune system 376 56 14.89% 6.08E-08 1.05E-05 ReactomeChemokine receptors bind chemokines 60 18 30.00% 8.69E-08 1.39E-05 ReactomePertussis - Homo sapiens (human) 75 20 26.67% 1.48E-07 2.21E-05 KEGGMAPK Signaling Pathway 168 32 19.05% 1.74E-07 2.44E-05 WikipathwaysNOD-like receptor signaling pathway - Homo sapiens (human) 57 17 29.82% 2.18E-07 2.66E-05 KEGGAP-1 transcription factor network 70 19 27.14% 2.23E-07 2.66E-05 PIDNF-kappa B signaling pathway - Homo sapiens (human) 91 22 24.18% 2.36E-07 2.66E-05 KEGGSelenium Micronutrient Network 83 20 24.10% 2.37E-07 2.66E-05 WikipathwaysToll-like receptor signaling pathway - Homo sapiens (human) 106 24 22.64% 2.53E-07 2.70E-05 KEGG116  3.3.4 Cluster of co-expressed genes predictive of IL-1 and IL-6 outcomes   To identify clusters of genes potentially regulating the production of the IL-1 and IL-6 cytokines, we then performed a pair-wise comparison of the expression levels of significantly expressed genes, following by a clustering of correlative expression patterns (Spearman’s rho value > 0.9) (Figure 3.8). This analysis identified four major clusters of gene expression variability within the unstimulated and LPS stimulated monocyte transcriptome (Cluster 1, and related 1a, 1b, and cluster 2). Based on a Gene Set Enrichment Pathway Analysis, Cluster 1 included primarily genes involved in basic transcription machinery. Cluster 1a included genes involved in Type II IFN signaling (Table 3.3). Cluster 1b involved genes not related to any known pathway (according to the databases listed in methods). Cluster 2 included genes involved in ribosomal activity, protein synthesis and the electron transport chain.  Next, we determined whether these co-varying four main gene clusters predicted IL-1β and IL-6 outcomes (Figure 3.9). Cluster 1a, previously identified as composed of type II IFN signaling genes predicted mainly IL1B and IL6 mRNA levels in LPS stimulated samples. Cluster 1b, which was linked to no known pathways, predicted secretion of IL-1β and IL-6 proteins. Cluster 1 and 2 were not associated with any of specific outcomes (Figure 3.9).    117   Figure 3.8 Co-Expression Matrix-Based Network of Unstimulated and LPS Stimulated Genes Two co-expression matrices of all detectable probes in the unstimulated and LPS stimulated samples were created. Nodes that had a Spearman’s rho >0.9 were networked using edges. Networked nodes unique to the unstimulated transcriptome (green), LPS stimulated transcriptome (red) or common to both conditions (blue) are shown above. Clusters 1, 1a, 1b and 2 were defined using a visual cut-off of the above data. Several small independent nodes that did not cluster were also detected but not shown for simplicity. 73 LPS stimulated and 64 unstimulated monocyte samples were used for this analysis.   118   Table 3.3 Top 2 Pathways Associated with Genes in Each Main Cluster in Figure 3.8 Pathways were ranked by statistical significance (p value). Overall, total 8, 39, 0 and 82 pathways were obtained for clusters 1, 1a 1b and 2, respectively, by Gene Set Enrichment Analysis. No known pathway corresponded to genes in cluster 1b.   Pathway Name Set SizeCandidates contained p-value q-value Pathway Source Cluster #Generic Transcription Pathway 541 14 (2.6%) 1.32E-05 0.00176 Reactome Cluster 1Gene Expression 1214 19 (1.6%) 0.000348 0.0232 Reactome Cluster 1Influenza A - Homo sapiens (human) 175 16 (9.1%) 1.81E-18 1.86E-16 KEGG Cluster 1aType II interferon signaling (IFNG) 37 10 (27.0%) 9.44E-17 4.86E-15 Wikipathways Cluster 1aRibosome - Homo sapiens (human) 137 35 (25.9%) 1.14E-40 2.37E-38 KEGG Cluster 2Electron Transport Chain 103 29 (28.2%) 9.20E-35 7.01E-33 Wikipathways Cluster 2119   Figure 3.9 Significance of Associations Between Expression Levels of Genes Composing Each of the Main Gene Clusters (Figure 3.8) and IL-1 or IL-6 Outcomes. Gene probes (LPS stimulated or unstimulated expression levels) that compose each cluster were tested against their association with each IL-1 or IL-6 outcomes (Figure 3.8). Clusters 1a (A) and 1b (B) included many statistically significant genes whose expression was associated with LPS stimulated mRNA and protein levels, respectively. Genes in cluster 1 (C) and 2 (D) did not show major overlap with any outcomes. Data in this figure was obtained after linear modelling of 73 LPS stimulated and 64 unstimulated monocyte samples from healthy adults with their respective IL-1β and IL-6 outcomes. P values graphed on a log10 scale.       120  3.3.5 Transcriptome events predictive of IL-1β and IL-6 outcomes   Next, we performed linear modeling to identify genes or pathways associated with IL-1β and IL-6 outcomes. 3275 and 4718 genes were associated with the most downstream outcomes: IL-1β and IL-6 secretion, respectively (Figure 3.10 A, B). A Gene Set Enrichment Analysis of these genes revealed that they mainly overlap with pathways related to cytokine/chemokine signaling as well as response to infections (Table 3.4).  Hierarchal clustering of the overall correlations between each IL-1 and IL-6 outcomes, and the whole transcriptome indicated that LPS stimulated outcomes cluster away from unstimulated outcomes, with the exception of caspase-1 activity (Figure 3.11). The mRNA outcomes (i.e. ILB and IL6 gene expression) also clustered away from protein outcomes (i.e. secreted IL-1β and IL-6) (Figure 3.11), implying distinct transcriptional factors regulating proximal versus more downstream cytokine production events.   A significant number of genes (LPS stimulated) predicted IL1B, IL6 mRNA and their corresponding protein outcomes (>10-15%). Surprisingly, however, fewer genes (4%) predicted caspase-1 activity (Figure 3.10 A). The relatively poor association between caspase-1 activity and transcriptome events was surprising to us given that it is the main determinant of an individual’s overall IL-1 response, and responsiveness over time (Figure 3.11, and Figure 3.1 D). Altogether, these results suggest a major dissociation between transcriptome events and caspase-1 activation.  121  Consistent with our findings that caspase-1 activity clusters away from all other outcomes (Figure 3.11), none of the genes that predict IL-1β secretion were associated with caspase-1 activity (Figure 3.12). These data suggest that the high level of variability in caspase-1 activation arises from events that are largely independent from transcriptome changes. Overall, it can be speculated that unidentified transcriptome-independent events infer a significant variability to this outcome.    122   Figure 3.10 Number of Genes Associated with IL-1β and IL-6 Outcomes (LPS) (A) Venn diagram of genes (LPS) that were associated with IL-1β outcomes. Approximately 27% of all genes were associated with IL-1β secretion, IL1B mRNA, caspase-1 activity and/or pro- IL-1β levels. (B) Venn diagram of genes that were associated with IL-6 outcomes. Data based on 73 LPS stimulated monocyte samples from healthy adults, with their respective IL-1β and IL-6 outcome.   123   Table 3.4 Top 10 Pathways Associated with Genes Predictive of IL-1 or IL-6 outcomes Pathways were ranked by statistical significance (p value). Based on overlapping genes predictive of all IL-1 or IL-6 outcomes in Venn diagram in Figure 3.9 (linear modeling).   Pathway Name Set size Candidates contained % Candidates in Set p-value q-value Pathway Source OutcomeChemokine signaling pathway - Homo sapiens (human) 189 10 5.29% 0.000327 0.0354 KEGG IL-1βtpo signaling pathway 25 4 16.00% 0.000383 0.0354 BioCarta IL-1βepo signaling pathway 11 3 27.27% 0.000419 0.0354 BioCarta IL-1βInositol phosphate metabolism 46 5 10.87% 0.000451 0.0354 Reactome IL-1βil-2 receptor beta chain in t cell activation 48 5 10.42% 0.000551 0.0354 BioCarta IL-1βil 3 signaling pathway 12 3 25.00% 0.000552 0.0354 BioCarta IL-1βCytokines and Inflammatory Response 29 4 13.79% 0.0006 0.0354 Wikipathways IL-1βmiR-targeted genes in epithelium - TarBase 327 13 3.98% 0.00073 0.0368 Wikipathways IL-1βIL2 signaling events mediated by STAT5 30 4 13.33% 0.000786 0.0368 PID IL-1βIL-4 Signaling Pathway 54 5 9.26% 0.000951 0.0368 Wikipathways IL-1βTCR 244 52 21.31% 2.34E-08 5.52E-05 NetPath IL-6Legionellosis - Homo sapiens (human) 55 17 30.91% 9.29E-06 0.011 KEGG IL-6Salmonella infection - Homo sapiens (human) 86 22 25.58% 1.47E-05 0.0116 KEGG IL-6Signaling by Rho GTPases 404 64 15.84% 3.31E-05 0.0156 Reactome IL-6miR-targeted genes in squamous cell - TarBase 154 29 18.83% 0.000318 0.075 Wikipathways IL-6Shigellosis - Homo sapiens (human) 65 16 24.62% 0.000341 0.075 KEGG IL-6Spliceosome - Homo sapiens (human) 133 26 19.55% 0.000349 0.075 KEGG IL-6miR-targeted genes in epithelium - TarBase 327 51 15.60% 0.000362 0.075 Wikipathways IL-6EGFR1 453 66 14.57% 0.000368 0.075 NetPath IL-6RHO GTPases Activate WASPs and WAVEs 36 11 30.56% 0.000394 0.075 Reactome IL-6124   Figure 3.11 Hierarchal Clustering of LPS Stimulated IL-1β and IL-6 Outcomes with Unstimulated (blue) or Stimulated (red) Transcriptome  Aside from distinct clustering between unstimulated and LPS stimulated transcriptome, protein and mRNA level outcomes distinctly cluster. Data in this figure is based on 73 LPS stimulated and 64 unstimulated monocyte samples, with their respective IL-1β and IL-6 outcomes.   125    Figure 3.12 Significance of Associations Between Expression Levels of Genes Associated with IL-1 Secretion, and Other IL-1/IL-6 Outcomes P-values of association for LPS stimulated (red) or unstimulated (green) gene expression levels (each dot represent a single gene probe). Data in this figure was obtained from linear modelling of 73 LPS stimulated and 64 unstimulated monocyte samples from healthy adults with their respective IL-1 and IL-6 outcomes. Values graphed on a log10 scale.   126  Primer name Sequence IL1B-forward 5’- GGC CCT AAA CAG ATG AAG TGC TCC-3’ IL1B-reverse 5’-GGT GCT CAG GTC ATT CTC CTG G-3’ IL6-forward 5'-GCA GAT GAG TAC AAA AGT CCT GA-3' IL6-reverse 5'-TTC TGT GCC TGC AGC TTC-3' BACT-forward 5’-TCC TAT GTG GGC GAC GAG G-3’ BACT-reverse 5’-GGT GTT GAA GGT CTC AAA CAT G-3’ GAPDH-forward 5’-CGT CAA GGC TGA GAA CGG GA-3’ GAPDH-reverse 5’-ATC AGC AGA GGG GGC AGA GA-3’ Table 3.5 Primer Sequences Used for PCR Amplification   127  3.4 Discussion   To understand how gene expression drives changes in immune responses, we studied LPS responses in monocytes, focusing on two main inflammatory cytokine pathways: IL-1β and IL-6. We investigated to what extent transcriptomic signatures can explain the variability in these responses in LPS stimulated monocytes isolated from a mono-ethnic cohort of healthy adult males. Monocytes are a primary source of these two cytokines in human blood [215]. A major strength of this study is the detailed evaluation of potential sources of variability in our experiments, using highly standardized experimental conditions, over a relatively short window of time.  Our data show that the capacity of an individual’s immune cells to produce IL-6 or IL-1 in response to LPS remains stable over a period of 1 month. Surprisingly, we find that caspase-1 activity clusters well away from other measured outcomes (Figure 3.11). Finally, we list a group of co-variable genes that are associated with the variability in the pathway components examined. Surprisingly, Gene Set Enrichment Analysis of co-variant transcript clusters reveals several non-immune pathways; these clusters were comprised of genes related to the electron transport chain, ribosome function, as well as a cluster of unknown function. Enrichment of these genes highlights the role of protein synthesis and metabolic function in regulating immune function, both in unstimulated, as well as LPS stimulated monocytes. The enriched pathways support the notion that post-transcriptional mechanisms, such as protein synthesis, control immune reactivity. 128  Understanding the source of variability in human immune responses has been of interest to immunologists for decades. In 1989, Endres et al. showed that LPS stimulated PBMCs from healthy individuals led to “low” and “high” producers of IL-1β [216]. Over the years, several other studies have demonstrated that inter-individual variability in cytokine production exists independent of technical variability and longitudinal intra-individual variation [217]–[219]. In our study, we confirm that the technical experimental variability in responses in our human cohort accounts for a marginal portion of the overall variability with the greatest proportion due to biology Figure 3.1). Unlike previous studies, we were not able to associate sex and age to LPS dependent responses ( Figure 3.2) [220]. However, some of our functional outcomes showed increased variability in female subjects over the short-term (Figure 3.3). These data, in conjunction with previously documented studies demonstrating variability associated with sex, led us to restrict our cohort to a group of Caucasian males [221]. In addition to being variable, IL-1β and IL-6 pathway components are interdependent. In line with previous analysis, we showed that IL-6 mRNA levels correlate well with secretion of their corresponding proteins (Figure 3.5 E) [222]. However, the relationship between secreted IL-1β and its pathway components is more complex. As published, we observed a dissociation between IL1B mRNA and pro-IL1β protein (Figure 3.5 A) [51], [223]. On the other hand, secretion of IL-1β correlated well with levels of pro-IL1β (Figure 3.5 C) and active caspase-1 (Figure 3.5 D/F) [51], [212]. The co-variability of these components led us to hypothesize that common areas of the transcriptomic landscape would lead to the variability seen in the secretion of IL-1β, the production of pro-IL1β and levels of active caspase-1.  129  The transcriptomic changes upon stimulation with LPS are consistent with previous studies (Figure 3.6) [224]–[226]. In 2005, Wurfel et al. stimulated whole blood with LPS to study the inter-individual variability between “high” and “low” responders. The candidate genes identified here, and in other studies, included several genes that were involved in interferon type II signaling [227], [228]. In addition to identifying a similar co-variable cluster of genes (Figure 3.8), we have also identified previously undescribed clusters in unstimulated and stimulated monocytes, which could explain the variability in pathway components after stimulation (Figure 3.8, Figure 3.9). Furthermore, components of TLR signaling pathways have often been viewed as the underlying cause of variability in the immune response [229]–[231]. Our data shows that a section of the transcriptome is associated with the variability in all pathway components and is composed of genes associated with innate immune responses (Table 3.4). However, upon hierarchal clustering of the association scores between our outcomes and the transcriptome, we noticed that variability in the IL1B mRNA levels clusters closer to IL6 mRNA response than to IL1β pathway components (Figure 3.11). Pathway analysis of genes common to all protein-level outcomes reveals variability in the processes that control translation.  A surprising finding was that genes that are associated with caspase-1 are not associated with transcriptomic events that predict IL-1β secretion, despite the fact that caspase-1 activity is a strong predictor of IL-1β secretion (Figure 3.5, Figure 3.11, Figure 3.12). Moreover, our observation that caspase-1 activity was highly conserved within individuals over time confirms that the variability in this measure is unlikely to be of experimental nature (Figure 3.1, Figure 3.4), and is likely determined by factors intrinsic to the individual. While 130  these findings may appear paradoxical, they suggest an important role for post-transcriptional events in the regulation of caspase-1 activity and IL-1β cytokine production. These findings also fit with what is known regarding caspase-1 activity, as it is ultimately regulated via potassium efflux.  Our data suggest that the variability in functional responses (namely IL-1β secretion) are heavily dependent on post-transcriptional mechanisms related to protein translation/degradation, and metabolism. This hypothesis is strengthened by the finding that genes related to ribosome function are enriched in co-expression cluster analysis, both in unstimulated and LPS stimulated conditions. Indeed, there is increasing evidence that innate immune responses are regulated post-transcriptionally [232], [233] (reviewed in [160]).  Overall, my findings suggest that immune variability is regulated at the level of protein synthesis/ribosome function; this is highlighted by the fact that the process of translation can be dampened in order to reduce inflammation [233], additionally, the regulation of ribosome proteins appears to be another level at which inflammation can be controlled [232]. In addition, the lack of association of active caspase-1 levels with the transcriptome suggests that post-transcriptomic events may modulate the secreted IL-1β levels.  This study demonstrates that while simple immune outcomes (such as IL-6 secretion) have strong associations with gene expression data, the more complex outcomes of caspase-1 activity fails to be predicted by transcriptomic events, even though this outcome is the main rate-limiting step for IL-1β secretion. It is important to note that while caspase-1 associated genes did not associate with predictors of IL-1β production, secretion of IL-1β can be predicted by transcriptomic changes, to some degree. This somewhat paradoxical finding highlights the 131  importance of selecting an appropriate functional readout when comparing transcriptomic data to phenotype.   132  Chapter 4: Interpretation, significance and future directions    4.1 Interpretation   I have shown in my thesis that innate immune functions vary considerably among healthy populations as well as across the developmental spectrum, likely to adapt to distinct age-adapted environmental challenges. I also propose a mechanism for the global suppression of PRR functions, or alarmin signals, as the immune system develops in utero. In term newborns, innate immune functions need to be rapidly turned on, as environmental and microbial exposures accumulate. The developmental signals that trigger a shift from a state of suppressed innate immune reactivity in early fetuses, to a progressive activation during the third trimester of gestation have yet to be described. Likewise, future studies are required to understand what produces the differences in innate immune reactivity between healthy adults (even though these differences are nearly as profound as what can be observed in fetuses). My thesis work provides important insights into the regulation of immune responses in humans. This knowledge may also help understand susceptibility to adult diseases involving dysregulation of inflammation, such as asthma, auto-immune diseases and malignancies. In Chapter 2 of my thesis, I confirmed that inflammatory cytokine secretion is broadly impaired in the mononuclear cells of preterm neonates, particularly those cytokines necessary for anti-fungal immune responses. Prior to this project, anti-fungal research in preterm neonates focused predominately on clinical risk factors and epidemiology, with few studies examining the molecular mechanisms that regulate immune responses in preterm infants [19]. 133  Infants born prematurely are particularly vulnerable to invasive fungal infections; factors associated with vulnerability include low birthweight, exposure to broad-spectrum antibiotics, and colonization status. Overall, these studies suggest that disruptions to physical barriers, perturbations of normal flora, and invasive medical interventions, as well as nosocomial infections make preterm infants more susceptible to infections by Candida spp. Chapter 2 of my thesis reinforces the finding that PRR responses are also attenuated in this population, and that this attenuation extends to anti-fungal immune responses. Additionally, I propose a fundamental mechanism by which innate immune responses are globally suppressed in preterm monocytes. This finding is important as up to this point, there was no unifying mechanism that adequately explained the broad immune dysregulation observed in preterm infants. In this thesis, I analyzed the response of preterm monocytes to Candida ex-vivo. My results suggest that innate immune responses are down-regulated through a defect in mTOR activity impacting the ability of preterm immune cells to mount an efficient glycolytic and translation response necessary during immune activation. This novel mechanism provides an interesting explanation why innate immune response are globally suppressed at this age as it would likely impact multiple receptor and signaling pathways converging towards the production of cytokines. This remains to be determined in future studies. Although therapeutics to restore immune function in these neonates do not currently exist, these insights open interesting therapeutic possibilities to restore immune function in these infants through a modulation of metabolic/mTOR pathways. The knowledge gained from understanding the ontogeny of immune regulation may also provide insight into the regulation of inflammation in 134  adults, as the same metabolic and translational mechanisms that are suppressive in utero likely influence the degree of reactivity in healthy human subjects. Genome-wide transcriptome analysis revealed large scale changes in transcripts related to metabolism and protein synthesis in preterm monocytes, relative to infants born at term, and healthy adults. Surprisingly, cytokine gene expression levels in all three age groups were not significantly different. These results suggest a global regulation at the post-transcriptional level. While my findings provide a mechanism by which potentially harmful inflammation is curtailed in the fetus, several aspects of this mechanism remain unclear. Pulse labelling experiments indicate a decreased level of protein synthesis in preterm monocytes. However, to survive, a certain amount of protein synthesis must occur. Thus, the question is raised: Is the overall rate of translation lowered in preterm monocytes, or is there a process of selection, that favours the translation of certain transcripts over others? Preterm immune cells appear “primed” transcriptionally, but not translationally. To address this question, it would be interesting to study the phenotype of immune cells even earlier in gestation, in the second or even first trimester. This poses a major practical challenge in getting enough cells to study; this technical barrier could be potentially overcome with the use of single cell RNA-sequencing that requires a very low input cell number and that also does not require any a priori knowledge of these cells for analysis contrary to flow cytometry which requires labeling for known cell markers. My findings suggest an active suppression of mTOR as a mechanism to ablate glycolysis and thus, protein synthesis. Transcriptomic findings suggest that the DNA damage inducible transcription factor 4-like (DDIT4L) molecule may represent a novel developmental 135  regulation of immune reactivity is exciting and certainly worth exploring in an in vivo animal model. In Chapter 3 of my thesis, I further examined how the transcriptome can predict immune functions; this time in healthy adults. The aim of this study was to better understand the source of inter-individual innate immune variability. In other words, I examined the variability of inflammatory cytokine secretion between individuals, and across time. I discovered that innate immune cytokine secretion is relatively stable over the course of 1 month. Additionally, I conclude that the variability observed in cytokine secretion was not due to technical variability. Transcriptomic analysis of our Caucasian male cohort suggests distinct gene expression clusters that predict simple cytokine outcomes, namely IL-1β and IL-6 secretion. In contrast, the rate-limiting step of caspase-1 activity was not associated with transcriptomic clusters that predicted IL-1β secretion. Caspase-1 cleavage and activation is dependent on potassium efflux [234] and is therefore controlled post-transcriptionally. I found that while some outcomes showed strong association with transcript levels (e.g. IL-6 secretion), the more complex functional outcome of caspase-1 activity was associated with relatively few genes. However, associating outcomes with gene expression clusters allowed for the identification of transcriptomic signatures that predict cytokine secretion.  The lack of an association between transcriptome and caspase-1 activation highlights the fact that many biological functions are controlled post-transcriptionally and reinforces the thought that while transcriptomic studies are powerful tools, they do not fully capture the complex interactions that determine functional outcomes. Interestingly, as we previously 136  observed in preterm infants, metabolic function and protein synthesis pathways were among the pathways that were associated with cytokine secretion.  4.2  Significance   A major unanswered question is how innate immune responses are regulated at different stages in life; most studies to date have focused on one stage of life (preterm, term, or adult), and usually in the context of pathology. In this thesis, I provide important insights into the regulation of innate immune function across development. Chapter 2 of my thesis is the first study that links preterm immune responses with metabolic regulation; although the interplay between metabolism and immune function is well understood [125], [126], [129], [144], developmental suppression of innate immunity by limiting metabolic function is a novel finding. Most studies to date regarding preterm immune function has focused on characterizing specific immune dysfunctions; while many immune deficits have been identified, no unifying mechanism of regulation has ever been proposed. In this thesis I propose a mechanism by which immune function is globally suppressed. I also propose that the immune suppression observed is a preserved mechanism that serves to prevent preterm birth as well as organ damage [62]–[66].  My research opens novel therapeutic possibilities for the use of metabolism–regulating drugs to reprogram the preterm immune system to better fight neonatal infections. Another potential therapeutic application for my findings is the field of vaccine adjuvants. The efficacy of vaccines can vary widely in the first months of life [235]. By manipulating cellular 137  metabolism, I speculate that the efficacy of early-life vaccines can be enhanced, allowing for earlier immunizations, and enhanced protection. The strategy of targeting metabolic skewing for therapeutic benefit is supported by many studies that have utilized inhibition of glycolysis to target malignancies. Additionally, inhibition of oxidative phosphorylation, and enhancement of anaerobic glycolysis has shown promise in reducing ischemia-reperfusion injuries in the heart and brain [236]–[238]. In a recent screening study, several FDA approved drugs have been shown to enhance glycolysis/reduce oxidative phosphorylation [239]. A major challenge in investigating therapeutic agents in preterm infants is the high degree of heterogeneity in this population, as well as the lack of appropriate controls. I hypothesize that enhancement of glycolytic function in preterm monocytes would restore inflammatory cytokine secretion by rescuing protein synthesis pathways. To test this hypothesis, I would treat preterm cord blood monocytes with previously identified enhancers of glycolysis, and measure cytokine secretion, metabolic activity, and rates of protein synthesis.   Chapter 3 of my thesis links variability in cytokine secretion to discrete transcriptomic clusters. These findings bridge the gap between transcriptomic studies and functional outcomes; I show that there is a large degree of discrepancy between functional outcomes such as cytokine secretion, caspase-1 activity, and transcriptional changes. I also demonstrate a link between metabolism and protein synthesis pathways in a non-disease model of immune regulation. Many genome-wide association studies to date have linked polymorphisms in specific genes to phenotypic alterations, such as diabetes. These associations have ultimately led to the discovery of several therapeutics being developed (reviewed in [240]). The major caveat with GWAS is that many of the genetic loci associated with a specific phenotype do not 138  contain genes, or alternatively, contain multiple genes. eQTL studies aim to overcome the weaknesses of GWAS by linking genetic polymorphisms to alterations in gene expression. While this approach overcomes a major limitation of GWAS, it introduces a new problem: alterations in gene expression do not always necessitate a change in phenotype. While eQTL studies have provided insights into the mechanisms of several inflammatory diseases [241] such as Crohn’s disease [241], asthma [241], and coronary artery disease [241], my findings caution against relying solely on using changes in gene expression as a functional readout. It has long been known that there is a poor correlation between transcriptome and proteome. Post-transcriptional mechanisms play a large role in regulating immune responses; this finding is reinforced by my finding that cytokine secretion is severely dampened in preterm monocytes, despite high levels of transcripts. The high degree of reliance on transcriptional changes as a stand-in for phenotypic alterations has largely been due to the difficulty and expense of doing large scale proteomics research, as well as the relative ease and affordability of transcriptomic studies.  My findings highlight the need for the use of complimentary techniques, to confirm transcriptomic findings at the phenotypic level, either through proteomic methods, or using functional readouts such as cell composition and enzymatic activity. Recent studies have sought to address this weakness using proteomics, this new approach has been coined protein Quantitative Trait Loci (pQTL) studies. These studies have utilized high throughput mass-spectrometry, as well as antibody based array tools to link genomic variations to protein abundance [159], [242], [243]. pQTL studies address the major drawbacks of eQTL studies; 139  however, protein abundance still fails to capture more complex outcomes, such as caspase-1 activity levels and cellular composition. Although existing eQTL studies have been incredibly helpful in understanding the regulation of gene expression, my findings show that key regulators of innate immunity, such as caspase-1 activity levels, do not associate well with changes in gene expression; as such, although caspase-1 is a key regulator of inflammation, the mechanisms regulating its activity cannot be examined with transcriptomic methods.  4.3  Contribution to current literature    Studies to date have consistently described suppression of immune function in preterm infants [54], [55], [57], [118]–[120]; however, a unifying mechanism for this suppression has not been proposed. Additionally, there have been no studies that explain with any detail the mechanisms that bring innate immune functions online as development progresses. My thesis links metabolism and protein synthesis to immune function in developing fetuses via regulation by mTOR, a master regulator of metabolism and protein synthesis. This proposed mechanism provides a molecular explanation for the rapid upregulation of innate immune function as a fetus approaches full term. The why and what of preterm immunity has been previously investigated; in chapter 2 of my thesis, the major question was how innate immunity is globally suppressed. Studies examining inter-individual immune variability in humans to date have mostly focused on pathogenic events. However, few studies have investigated inter-individual 140  responses in healthy humans. Of the papers that exist, most focus on linking variability to the genetic polymorphisms [244], [245]; while these findings are important and shed light on one level of regulation, findings showing environmental influences, such as age, seasonality, and microbiome, greatly influence inflammatory cytokine responses [246], [247] suggesting a more dynamic level of regulation also exists [121]. Recent evidence has emerged showing that the degree of influence of genetics on phenotype varies widely based on cell type, with monocyte traits being more influenced by the environment [248].   A recent paper by Brodin et al. examined immune function in twins, and concluded that cumulative environmental exposures drive most immune differences in genetically similar individuals [169]; in particular, the authors showed that IL-1β secretion is largely influenced by non-genetic factors. In Chapter 3 of my thesis, I show that some functional outcomes cannot be predicted by transcriptome events, and that metabolism and protein synthesis are post-transcriptional mechanisms that may modulate these responses. My findings in Chapter 3 also highlight the disconnect between transcript abundance and functional outcomes [159], [249], [250].  4.4  Limitations and future directions   While preterm infants present a unique opportunity to study the developmental regulation of innate immunity, they also present many technical challenges. The very limited blood volume that is available from preterm placentas was a major limitation to my studies. This made it quite challenging to repeat experiments that required large numbers of cells (polysome profiling, Western blots, and Seahorse metabolic flux experiments). Additionally, 141  with such a heterogenous population, there was a large degree of variability in blood composition. Due to the limited number of samples available, I was not powered to investigate the relationship between cause of preterm birth (i.e. ascending infections, hypertension, etc.), and immune function; although previous studies suggest that antenatal corticosteroids, and chorioamnionitis do not play a major role in immune reactivity [57].  An additional limitation I faced was the lack of an accurate model of preterm innate immunity. An overall limitation to the use of microarray data is that the data we obtained only provides a snapshot of what is happening in the cell at a specific time point; as the transcriptome is dynamic, with transcripts having a short half-life, data from multiple time-points would have provided a more nuanced picture of the events regulating innate immunity. However, technical limitations regarding cell numbers, as well as cost, prevented us from pursuing this. Another overall limitation of microarrays is the reliance on probes for detection of transcripts. As probes do not exist for every transcriptional product and splice form, we are unable to capture novel transcripts; this shortcoming can be overcome with RNA sequencing approaches. Additionally, recent advances in single cell RNA sequencing technologies may allow us to overcome the technical as well as ethical issues that prevent detailed examination of immune maturation in the post-natal period of preterm infants. In the future, we plan to further describe neonatal immune function, investigate ribosome function, and further dissect the unique metabolic phenotype of preterm infants. Work is currently underway to quantify ribosomes using flow cytometry, as well as fluorescent microscopy. Additionally, we plan to measure mitochondrial mass and potential in preterm monocytes. The purpose of these experiments is to determine the physiological nature of the 142  metabolic/translational deficits. Knowledge gained from these studies will be used to determine whether these deficits are reversible via pharmaceutical intervention. Ultimately, we would like to restore innate immune function in preterm monocytes by modulating metabolic function. Restoration of innate immune responses in these infants would allow them to better resist infection, as well as mount an effective vaccine response. Another aspect of future studies would be to utilize ribosome profiling techniques, combined with RNA sequencing to examine transcripts more closely; it will be particularly interesting to examine mRNA sequences that are translated vs those that are not; I predict that in the preterm infant, there are specific transcriptional sequences that allow some mRNAs to be translated rather than others.  A better understanding of the drivers of innate immune variability may someday allow us to identify individuals that are less responsive to vaccines, as well as more vulnerable to infections.      143  Bibliography [1] J. E. Lawn et al., “Every Newborn: progress, priorities, and potential beyond survival.,” Lancet (London, England), vol. 384, no. 9938, pp. 189–205, Jul. 2014. [2] B. J. Stoll et al., “Trends in Care Practices, Morbidity, and Mortality of Extremely Preterm Neonates, 1993-2012.,” JAMA, vol. 314, no. 10, pp. 1039–51, Sep. 2015. [3] J. Baardman, I. Licht, M. P. de Winther, and J. Van den Bossche, “Metabolic–epigenetic crosstalk in macrophage activation,” Epigenomics, vol. 7, no. 7, pp. 1155–1164, 2015. [4] A. T. Phan, A. W. Goldrath, and C. K. Glass, “Metabolic and Epigenetic Coordination of T Cell and Macrophage Immunity.,” Immunity, vol. 46, no. 5, pp. 714–729, May 2017. [5] B. Novakovic et al., “β-Glucan Reverses the Epigenetic State of LPS-Induced Immunological Tolerance.,” Cell, vol. 167, no. 5, p. 1354–1368.e14, 2016. [6] S. Saeed et al., “Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity.,” Science, vol. 345, no. 6204, p. 1251086, Sep. 2014. [7] B. Adkins, “Neonatal immunology: Responses to pathogenic microorganisms and epigenetics reveal an ‘immunodiverse’ developmental state,” Immunol. Res., vol. 57, no. 1–3, pp. 246–257, 2013. [8] O. M. de Goede et al., “Nucleated red blood cells impact DNA methylation and expression analyses of cord blood hematopoietic cells.,” Clin. Epigenetics, vol. 7, no. 1, p. 95, 2015. [9] O. M. de Goede, P. M. Lavoie, and W. P. Robinson, “Cord blood hematopoietic cells from preterm infants display altered DNA methylation patterns.,” Clin. Epigenetics, vol. 9, p. 39, 2017. [10] S. W. Brubaker, K. S. Bonham, I. Zanoni, and J. C. Kagan, “Innate Immune Pattern Recognition: A Cell Biological Perspective,” Annu. Rev. Immunol., vol. 33, no. 1, pp. 257–290, Mar. 2015. [11] M. D. Turner, B. Nedjai, T. Hurst, and D. J. Pennington, “Cytokines and chemokines: At the crossroads of cell signalling and inflammatory disease.,” Biochim. Biophys. Acta, vol. 1843, no. 11, pp. 2563–2582, Nov. 2014. [12] C. A. Hunter and S. A. Jones, “IL-6 as a keystone cytokine in health and disease.,” Nat. Immunol., vol. 16, no. 5, pp. 448–57, May 2015. [13] G. Lopez-Castejon and D. Brough, “Understanding the mechanism of IL-1β secretion.,” Cytokine Growth Factor Rev., vol. 22, no. 4, pp. 189–95, Aug. 2011. [14] A. Liston, E. J. Carr, and M. A. Linterman, “Shaping Variation in the Human Immune System.,” Trends Immunol., vol. 168, no. 4, pp. 1984–91, Aug. 2016. [15] S. Ganesan et al., “Caspase-8 Modulates Dectin-1 and Complement Receptor 3-Driven IL-1β Production in Response to β-Glucans and the Fungal Pathogen, Candida albicans.,” J. Immunol., vol. 193, no. 5, pp. 2519–30, Sep. 2014. [16] A. Tóth, E. Zajta, K. Csonka, C. Vágvölgyi, M. G. Netea, and A. Gácser, “Specific pathways mediating inflammasome activation by Candida parapsilosis.,” Sci. Rep., vol. 7, p. 43129, Feb. 2017. [17] M. S. Kelly, D. K. Benjamin, and P. B. Smith, “The epidemiology and diagnosis of invasive candidiasis among premature infants.,” Clin. Perinatol., vol. 42, no. 1, pp. 105–17, viii–ix, Mar. 2015. 144  [18] N. A. R. Gow, F. L. van de Veerdonk, A. J. P. Brown, and M. G. Netea, “Candida albicans morphogenesis and host defence: discriminating invasion from colonization.,” Nat. Rev. Microbiol., vol. 10, no. 2, pp. 112–22, Feb. 2012. [19] C. Michalski, B. Kan, and P. M. Lavoie, “Antifungal immunological defenses in newborns,” Front. Immunol., vol. 8, no. MAR, pp. 1–8, 2017. [20] J. M. Bliss, “Candida parapsilosis: an emerging pathogen developing its own identity.,” Virulence, vol. 6, no. 2, pp. 109–11, 2015. [21] M. Pammi, L. Holland, G. Butler, A. Gacser, and J. M. Bliss, “Candida parapsilosis is a significant neonatal pathogen: a systematic review and meta-analysis.,” Pediatr. Infect. Dis. J., vol. 32, no. 5, pp. e206-16, May 2013. [22] D. Trofa, A. Gácser, and J. D. Nosanchuk, “Candida parapsilosis, an emerging fungal pathogen.,” Clin. Microbiol. Rev., vol. 21, no. 4, pp. 606–25, Oct. 2008. [23] S. Aliaga et al., “Changes in the incidence of candidiasis in neonatal intensive care units.,” Pediatrics, vol. 133, no. 2, pp. 236–42, Feb. 2014. [24] A. Tóth et al., “Candida albicans and Candida parapsilosis induce different T-cell responses in human peripheral blood mononuclear cells.,” J. Infect. Dis., vol. 208, no. 4, pp. 690–8, Aug. 2013. [25] A. Prosser et al., “Phagocytosis of neonatal pathogens by peripheral blood neutrophils and monocytes from newborn preterm and term infants.,” Pediatr. Res., vol. 74, no. 5, pp. 503–10, Nov. 2013. [26] K. G. Destin, J. R. Linden, S. S. Laforce-Nesbitt, and J. M. Bliss, “Oxidative burst and phagocytosis of neonatal neutrophils confronting Candida albicans and Candida parapsilosis.,” Early Hum. Dev., vol. 85, no. 8, pp. 531–5, Aug. 2009. [27] J. R. Linden, M. A. Maccani, S. S. Laforce-Nesbitt, and J. M. Bliss, “High efficiency opsonin-independent phagocytosis of Candida parapsilosis by human neutrophils.,” Med. Mycol., vol. 48, no. 2, pp. 355–64, Mar. 2010. [28] S. Gonia et al., “Candida parapsilosis Protects Premature Intestinal Epithelial Cells from Invasion and Damage by Candida albicans.,” Front. Pediatr., vol. 5, no. March, p. 54, 2017. [29] M. Rosas et al., “The induction of inflammation by dectin-1 in vivo is dependent on myeloid cell programming and the progression of phagocytosis.,” J. Immunol., vol. 181, no. 5, pp. 3549–57, Sep. 2008. [30] M. G. Netea, L. a B. Joosten, J. W. M. van der Meer, B.-J. Kullberg, and F. L. van de Veerdonk, “Immune defence against Candida fungal infections.,” Nat. Rev. Immunol., vol. 15, no. 10, pp. 630–642, 2015. [31] B. N. Gantner, R. M. Simmons, and D. M. Underhill, “Dectin-1 mediates macrophage recognition of Candida albicans yeast but not filaments.,” EMBO J., vol. 24, no. 6, pp. 1277–86, Mar. 2005. [32] S. W. Kashem et al., “Candida albicans Morphology and Dendritic Cell Subsets Determine T Helper Cell Differentiation,” Immunity, vol. 42, no. 2, pp. 356–366, 2015. [33] M. G. Netea, N. A. R. Gow, L. a B. Joosten, I. Verschueren, J. W. M. van der Meer, and B. J. Kullberg, “Variable recognition of Candida albicans strains by TLR4 and lectin recognition receptors.,” Med. Mycol., vol. 48, no. 7, pp. 897–903, Nov. 2010. [34] M. G. Netea et al., “Immune sensing of Candida albicans requires cooperative recognition 145  of mannans and glucans by lectin and Toll-like receptors.,” J. Clin. Invest., vol. 116, no. 6, pp. 1642–50, Jun. 2006. [35] M. L. Gil and D. Gozalbo, “About the role of TLR2 and TLR4 in cytokine secretion by murine macrophages in response to Candida albicans.,” FEMS Immunol. Med. Microbiol., vol. 46, no. 1, pp. 1–2, Feb. 2006. [36] G. D. Brown, J. Herre, D. L. Williams, J. A. Willment, A. S. J. Marshall, and S. Gordon, “Dectin-1 mediates the biological effects of beta-glucans.,” J. Exp. Med., vol. 197, no. 9, pp. 1119–24, May 2003. [37] P. R. Taylor et al., “Dectin-1 is required for beta-glucan recognition and control of fungal infection.,” Nat. Immunol., vol. 8, no. 1, pp. 31–8, Jan. 2007. [38] S. Altmeier, A. Toska, F. Sparber, A. Teijeira, C. Halin, and S. LeibundGut-Landmann, “IL-1 Coordinates the Neutrophil Response to C. albicans in the Oral Mucosa.,” PLoS Pathog., vol. 12, no. 9, p. e1005882, Sep. 2016. [39] H. R. Conti and S. L. Gaffen, “Host responses to Candida albicans: Th17 cells and mucosal candidiasis.,” Microbes Infect., vol. 12, no. 7, pp. 518–27, Jul. 2010. [40] C. E. Zielinski et al., “Pathogen-induced human TH17 cells produce IFN-γ or IL-10 and are regulated by IL-1β.,” Nature, vol. 484, no. 7395, pp. 514–8, Apr. 2012. [41] M. G. Netea et al., “Human dendritic cells are less potent at killing Candida albicans than both monocytes and macrophages,” Microbes Infect., vol. 6, no. 11, pp. 985–989, 2004. [42] O. Gross et al., “Syk kinase signalling couples to the Nlrp3 inflammasome for anti-fungal host defence.,” Nature, vol. 459, no. 7245, pp. 433–6, May 2009. [43] S. I. Gringhuis et al., “Dectin-1 is an extracellular pathogen sensor for the induction and processing of IL-1β via a noncanonical caspase-8 inflammasome.,” Nat. Immunol., vol. 13, no. 3, pp. 246–54, Mar. 2012. [44] O. Gross et al., “Card9 controls a non-TLR signalling pathway for innate anti-fungal immunity.,” Nature, vol. 442, no. 7103, pp. 651–6, 2006. [45] E.-O. Glocker et al., “A homozygous CARD9 mutation in a family with susceptibility to fungal infections.,” N. Engl. J. Med., vol. 361, no. 18, pp. 1727–35, Oct. 2009. [46] P. M. Lavoie and O. Levy, “Chapter 125: The Mononuclear Phagocyte System [In Press],” in Fetal and Neonatal Physiology, 5th Editio., R. Polin, S. Abman, D. Rowitch, and W. Benitz, Eds. Philadelphia: Elsevier, 2015. [47] N. P. Davies, A. G. Buggins, R. J. Snijders, P. N. Noble, D. M. Layton, and K. H. Nicolaides, “Fetal leucocyte count in rhesus disease.,” Arch. Dis. Child., vol. 67, no. 4 Spec No, pp. 404–6, Apr. 1992. [48] C. C. Yost et al., “Impaired neutrophil extracellular trap (NET) formation: a novel innate immune deficiency of human neonates.,” Blood, vol. 113, no. 25, pp. 6419–27, Jun. 2009. [49] A. S. Byrd et al., “NETosis in neonates: Evidence of a reactive oxygen species-independent pathway in response to fungal challenge,” J. Infect. Dis., vol. 213, no. 4, pp. 634–639, Feb. 2016. [50] L. Ziegler-Heitbrock, “Monocyte subsets in man and other species,” Cell. Immunol., vol. 289, no. 1–2, pp. 135–139, May 2014. [51] A. A. Sharma et al., “Impaired NLRP3 inflammasome activity during fetal development regulates IL-1β production in human monocytes.,” Eur. J. Immunol., vol. 45, no. 1, pp. 238–49, Jan. 2015. 146  [52] E. R. Krow-Lucal, C. C. Kim, T. D. Burt, and J. M. McCune, “Distinct functional programming of human fetal and adult monocytes.,” Blood, vol. 123, no. 12, pp. 1897–904, Mar. 2014. [53]  a. S. Grumach, M. E. Ceccon, R. Rutz,  a. Fertig, and M. Kirschfink, “Complement profile in neonates of different gestational ages,” Scand. J. Immunol., vol. 79, no. 4, pp. 276–281, 2014. [54] E. Förster-Waldl et al., “Monocyte toll-like receptor 4 expression and LPS-induced cytokine production increase during gestational aging,” Pediatr. Res., vol. 58, no. 1, pp. 121–124, Jul. 2005. [55] K. Sadeghi et al., “Immaturity of infection control in preterm and term newborns is associated with impaired toll-like receptor signaling.,” J. Infect. Dis., vol. 195, no. 2, pp. 296–302, Jan. 2007. [56] T. Strunk et al., “Responsiveness of human monocytes to the commensal bacterium Staphylococcus epidermidis develops late in gestation.,” Pediatr. Res., vol. 72, no. 1, pp. 10–8, Jul. 2012. [57] A. A. Sharma et al., “Hierarchical maturation of innate immune defences in very preterm neonates,” Neonatology, vol. 106, no. 1, pp. 1–9, Jan. 2014. [58] P. M. M. Lavoie et al., “Profound Lack of Interleukin (IL)–12/IL‐23p40 in Neonates Born Early in Gestation Is Associated with an Increased Risk of Sepsis,” J. Infect. Dis., vol. 202, no. 11, pp. 1754–1763, Dec. 2010. [59] V. J. Philbin et al., “Imidazoquinoline Toll-like receptor 8 agonists activate human newborn monocytes and dendritic cells through adenosine-refractory and caspase-1-dependent pathways.,” J. Allergy Clin. Immunol., vol. 130, no. 1, p. 195–204.e9, Jul. 2012. [60] E. A. Marchant, G. K. Boyce, M. Sadarangani, and P. M. Lavoie, “Neonatal Sepsis due to Coagulase-Negative Staphylococci,” Clin. Dev. Immunol., vol. 2013, p. 586076, 2013. [61] A. G. Cuenca, J. L. Wynn, L. L. Moldawer, and O. Levy, “Role of Innate Immunity in Neonatal Infection,” Am. J. Perinatol., vol. 30, no. 2, pp. 105–112, Jan. 2013. [62] M. Cappelletti, S. Della Bella, E. Ferrazzi, D. Mavilio, and S. Divanovic, “Inflammation and preterm birth.,” J. Leukoc. Biol., vol. 99, no. 1, pp. 67–78, 2016. [63] S. Malaeb and O. Dammann, “Fetal inflammatory response and brain injury in the preterm newborn.,” J. Child Neurol., vol. 24, no. 9, pp. 1119–26, Sep. 2009. [64] M. W. Kemp, “Preterm birth, intrauterine infection, and fetal inflammation,” Front. Immunol., vol. 5, no. DEC, p. 574, 2014. [65] K. M. Adams Waldorf and R. M. McAdams, “Influence of infection during pregnancy on fetal development.,” Reproduction, vol. 146, no. 5, pp. R151-62, 2013. [66] I. Adams-Chapman et al., “Neurodevelopmental outcome of extremely low birth weight infants with Candida infection,” J. Pediatr., vol. 163, no. 4, 2013. [67] E. A. Marchant et al., “Attenuated innate immune defenses in very premature neonates during the neonatal period,” Pediatr. Res., vol. 78, no. 5, pp. 492–497, Nov. 2015. [68] N. Jain and W. A. Walker, “Diet and host-microbial crosstalk in postnatal intestinal immune homeostasis,” Nat. Rev. Gastroenterol. Hepatol., vol. 12, no. 1, pp. 14–25, 2015. [69] M. Fulde and M. W. Hornef, “Maturation of the enteric mucosal innate immune system during the postnatal period,” Immunol Rev, vol. 260, no. 1, pp. 21–34, 2014. [70] A. J. Battersby and D. L. Gibbons, “The gut mucosal immune system in the neonatal 147  period,” Pediatr Allergy Immunol, vol. 24, no. 5, pp. 414–421, 2013. [71] C. L. Maynard, C. O. Elson, R. D. Hatton, and C. T. Weaver, “Reciprocal interactions of the intestinal microbiota and immune system,” Nature, vol. 489, no. 7415, pp. 231–241, 2012. [72] A. A. Sharma, L. Chew, M. Ladd, R. Jen, and P. M. Lavoie, “Ex vivo purification and characterization of human invariant Natural Killer T cells.,” J. Immunol. Methods, vol. 373, no. 1–2, pp. 1–7, Oct. 2011. [73] L. Loh, M. a Ivarsson, J. Michaëlsson, J. K. Sandberg, and D. F. Nixon, “Invariant natural killer T cells developing in the human fetus accumulate and mature in the small intestine.,” Mucosal Immunol., vol. 7, no. October 2013, pp. 1–11, 2014. [74] M. Ladd et al., “Natural killer T cells constitutively expressing the interleukin-2 receptor α chain early in life are primed to respond to lower antigenic stimulation,” Immunology, vol. 131, no. 2, pp. 289–299, 2010. [75] E. Esplugues et al., “Control of TH17 cells occurs in the small intestine,” Nature, vol. 475, no. 7357, pp. 514–518, Jul. 2011. [76] C. B. Wilson, E. Rowell, and M. Sekimata, “Epigenetic control of T-helper-cell differentiation.,” Nat. Rev. Immunol., vol. 9, no. 2, pp. 91–105, Feb. 2009. [77] S. de Roock et al., “Defective TH17 development in human neonatal T cells involves reduced RORC2 mRNA content.,” J. Allergy Clin. Immunol., vol. 132, no. 3, p. 754–756.e3, Sep. 2013. [78] K. K. Dijkstra, S. B. E. A. Hoeks, B. J. Prakken, and S. de Roock, “TH17 differentiation capacity develops within the first 3 months of life.,” J. Allergy Clin. Immunol., vol. 133, no. 3, p. 891–4.e5, Mar. 2014. [79] T. Korn, E. Bettelli, M. Oukka, and V. K. Kuchroo, “IL-17 and Th17 Cells,” Annu. Rev. Immunol., vol. 27, no. 1, pp. 485–517, 2009. [80] B. S. S. Sheridan et al., “γδ T Cells Exhibit Multifunctional and Protective Memory in Intestinal Tissues,” Immunity, vol. 39, no. 1, pp. 184–195, 2013. [81] D. R. McDonald, “TH17 deficiency in human disease,” J. Allergy Clin. Immunol., vol. 129, no. 6, pp. 1429–1435, 2012. [82] A. Schuetz et al., “Initiation of ART during early acute HIV infection preserves mucosal Th17 function and reverses HIV-related immune activation.,” PLoS Pathog., vol. 10, no. 12, p. e1004543, Dec. 2014. [83] S. Cypowyj, C. Picard, L. Maródi, J.-L. Casanova, and A. Puel, “Immunity to infection in IL-17-deficient mice and humans,” Eur. J. Immunol., vol. 42, no. 9, pp. 2246–2254, Sep. 2012. [84] S. Vanden Eijnden, S. Goriely, D. De Wit, M. Goldman, and F. Willems, “Preferential production of the IL-12(p40)/IL-23(p19) heterodimer by dendritic cells from human newborns.,” Eur. J. Immunol., vol. 36, no. 1, pp. 21–6, Jan. 2006. [85] J. Mirpuri et al., “Commensal Escherichia coli reduces epithelial apoptosis through IFN-alphaA-mediated induction of guanylate binding protein-1 in human and murine models of developing intestine.,” J. Immunol., vol. 184, no. 12, pp. 7186–7195, 2010. [86] S. Fukuda et al., “Bifidobacteria can protect from enteropathogenic infection through production of acetate,” Nature, vol. 469, no. 7331, pp. 543–547, 2011. [87] M.-C. Arrieta et al., “Early infancy microbial and metabolic alterations affect risk of 148  childhood asthma.,” Sci. Transl. Med., vol. 7, no. 307, p. 307ra152, Sep. 2015. [88] B. Cao, M. J. Stout, I. Lee, and I. U. Mysorekar, “Placental Microbiome and Its Role in Preterm Birth,” Neoreviews, vol. 15, no. 12, pp. 537–545, 2014. [89] V. S. Kuppala, J. Meinzen-Derr, A. L. Morrow, and K. R. Schibler, “Prolonged Initial Empirical Antibiotic Treatment is Associated with Adverse Outcomes in Premature Infants,” J. Pediatr., vol. 159, no. 5, pp. 720–725, 2011. [90] H. S. Deshmukh et al., “The microbiota regulates neutrophil homeostasis and host resistance to Escherichia coli K1 sepsis in neonatal mice.,” Nat. Med., vol. 20, no. 5, pp. 524–530, 2014. [91] S. Arboleya et al., “Intestinal microbiota development in preterm neonates and effect of perinatal antibiotics.,” J. Pediatr., vol. 166, no. 3, pp. 538–44, Mar. 2015. [92] J. E. Berrington, C. J. Stewart, S. P. Cummings, and N. D. Embleton, “The neonatal bowel microbiome in health and infection.,” Curr. Opin. Infect. Dis., vol. 27, no. 3, pp. 236–43, Jun. 2014. [93] T. Drell et al., “The development of gut microbiota in critically ill extremely low birth weight infants assessed with 16S rRNA gene based sequencing,” Gut Microbes, vol. 5, no. 3, pp. 304–312, 2014. [94] V. Mai et al., “Fecal Microbiota in Premature Infants Prior to Necrotizing Enterocolitis,” PLoS One, vol. 6, no. 6, p. e20647, Jun. 2011. [95] A. L. Morrow et al., “Early microbial and metabolomic signatures predict later onset of necrotizing enterocolitis in preterm infants,” Microbiome, vol. 1, no. 1, p. 13, 2013. [96] E. C. Claud, L. Lu, P. M. Anton, T. Savidge, W. A. Walker, and B. J. Cherayil, “Developmentally regulated IkappaB expression in intestinal epithelium and susceptibility to flagellin-induced inflammation.,” Proc. Natl. Acad. Sci. U. S. A., vol. 101, no. 19, pp. 7404–8, May 2004. [97] K. Alfaleh, J. Anabrees, D. Bassler, and T. Al-Kharfi, “Probiotics for prevention of necrotizing enterocolitis in preterm infants,” in Cochrane Database of Systematic Reviews, no. 3, Wiley-Blackwell, 2011, p. CD005496. [98] S. Trend et al., “Leukocyte Populations in Human Preterm and Term Breast Milk Identified by Multicolour Flow Cytometry.,” PLoS One, vol. 10, no. 8, p. e0135580, 2015. [99] S. Trend et al., “Antimicrobial protein and Peptide concentrations and activity in human breast milk consumed by preterm infants at risk of late-onset neonatal sepsis.,” PLoS One, vol. 10, no. 2, p. e0117038, 2015. [100] B. D. W. Chow, J. L. Reardon, E. O. Perry, S. S. Laforce-Nesbitt, R. Tucker, and J. M. Bliss, “Host Defense Proteins in Breast Milk and Neonatal Yeast Colonization.,” J. Hum. Lact., vol. 32, no. 1, pp. 168–73, Feb. 2016. [101] T. Strunk, T. Kollmann, and S. Patole, “Probiotics to prevent early-life infection,” Lancet Infect. Dis., vol. 15, no. 4, pp. 378–9, Apr. 2015. [102] A. G. Cuenca et al., “TRIF-Dependent Innate Immune Activation Is Critical for Survival to Neonatal Gram-Negative Sepsis.,” J. Immunol., vol. 194, no. 3, pp. 1169–77, 2015. [103] J. L. Wynn et al., “Defective innate immunity predisposes murine neonates to poor sepsis outcome but is reversed by TLR agonists,” Blood, vol. 112, no. 5, pp. 1750–1758, 2008. [104] T. Strunk, T. Inder, X. Wang, D. Burgner, C. Mallard, and O. Levy, “Infection-induced inflammation and cerebral injury in preterm infants.,” Lancet. Infect. Dis., vol. 14, no. 8, 149  pp. 751–62, Aug. 2014. [105] K. D. Kronforst et al., “A neonatal model of intravenous Staphylococcus epidermidis infection in mice <24 h old enables characterization of early innate immune responses.,” PLoS One, vol. 7, no. 9, p. e43897, 2012. [106] W. Storm, “Transient bacteremia following endotracheal suctioning in ventilated newborns.,” Pediatrics, vol. 65, no. 3, pp. 487–90, Mar. 1980. [107] J. L. LeFrock, A. S. Klainer, W. H. Wu, and H. Turndorf, “Transient Bacteremia Associated With Nasotracheal Suctioning,” JAMA, vol. 236, no. 14, pp. 1610–1, Oct. 1976. [108] B. a. Chang et al., “Early inflammation in the absence of overt infection in preterm neonates exposed to intensive care,” Cytokine, vol. 56, no. 3, pp. 621–626, 2011. [109] V. Chau, D. E. McFadden, K. J. Poskitt, and S. P. Miller, “Chorioamnionitis in the pathogenesis of brain injury in preterm infants.,” Clin. Perinatol., vol. 41, no. 1, pp. 83–103, Mar. 2014. [110] L. F. Gentile et al., “Protective Immunity and Defects in the Neonatal and Elderly Immune Response to Sepsis,” J. Immunol., vol. 192, no. 7, pp. 3156–3165, 2014. [111] L. F. Gentile et al., “Host Responses to Sepsis Vary in Different Low-Lethality Murine Models,” PLoS One, vol. 9, no. 5, p. e94404, 2014. [112] P. a. Efron, A. M. Mohr, F. a. Moore, and L. L. Moldawer, “The future of murine sepsis and trauma research models,” J. Leukoc. Biol., vol. 98, no. 6, pp. 945–52, Dec. 2015. [113] C. L. Smith et al., “Identification of a human neonatal immune-metabolic network associated with bacterial infection,” Nat. Commun., vol. 5, pp. 1–15, 2014. [114] M. Cernada et al., “Genome-Wide Expression Profiles in Very Low Birth Weight Infants With Neonatal Sepsis.,” Pediatrics, vol. 133, no. 5, pp. e1203–e1211, 2014. [115] S. G. Kallapur et al., “Chronic fetal exposure to Ureaplasma parvum suppresses innate immune responses in sheep.,” J. Immunol., vol. 187, no. 5, pp. 2688–95, Sep. 2011. [116] B. W. Kramer, S. G. Kallapur, T. J. Moss, I. Nitsos, J. P. Newnham, and A. H. Jobe, “Intra-amniotic LPS modulation of TLR signaling in lung and blood monocytes of fetal sheep.,” Innate Immun., vol. 15, no. 2, pp. 101–7, Apr. 2009. [117] F. Gotsch et al., “The fetal inflammatory response syndrome.,” Clin. Obstet. Gynecol., vol. 50, no. 3, pp. 652–83, Sep. 2007. [118] A. A. Sharma, R. Jen, A. Butler, and P. M. Lavoie, “The developing human preterm neonatal immune system: a case for more research in this area.,” Clin. Immunol., vol. 145, no. 1, pp. 61–8, Oct. 2012. [119] B. Kan, H. R. Razzaghian, and P. M. Lavoie, “An Immunological Perspective on Neonatal Sepsis,” Trends Mol. Med., vol. 22, no. 4, pp. 290–302, 2016. [120] N. Marr et al., “Attenuation of respiratory syncytial virus-induced and RIG-I-dependent type I IFN responses in human neonates and very young children.,” J. Immunol., vol. 192, no. 3, pp. 948–57, Feb. 2014. [121] S. F. Martin, “Adaptation in the innate immune system and heterologous innate immunity,” Cell. Mol. Life Sci., vol. 71, no. 21, pp. 4115–4130, 2014. [122] H. Bisgaard, K. Bønnelykke, and J. Stokholm, “Immune-mediated diseases and microbial exposure in early life.,” Clin. Exp. Allergy, vol. 44, no. 4, pp. 475–81, Apr. 2014. [123] S. Elahi et al., “Immunosuppressive CD71+ erythroid cells compromise neonatal host defence against infection.,” Nature, vol. 504, no. 7478, pp. 158–62, Dec. 2013. 150  [124] S.-C. Cheng et al., “mTOR- and HIF-1 -mediated aerobic glycolysis as metabolic basis for trained immunity,” Science (80-. )., vol. 345, no. 6204, pp. 1250684–1250684, Sep. 2014. [125] B. Everts et al., “TLR-driven early glycolytic reprogramming via the kinases TBK1-IKKɛ supports the anabolic demands of dendritic cell activation.,” Nat. Immunol., vol. 15, no. 4, pp. 323–32, 2014. [126] A. F. McGettrick and L. A. J. O’Neill, “How metabolism generates signals during innate immunity and inflammation,” J. Biol. Chem., vol. 288, no. 32, pp. 22893–22898, 2013. [127] J. Rius et al., “NF-kappaB links innate immunity to the hypoxic response through transcriptional regulation of HIF-1alpha.,” Nature, vol. 453, no. 7196, pp. 807–11, Jun. 2008. [128] S. K. Biswas and A. Mantovani, “Orchestration of metabolism by macrophages,” Cell Metabolism, vol. 15, no. 4. Elsevier Inc., pp. 432–437, 2012. [129] Y. Fujieda et al., “Inflammation and Resolution Are Associated with Upregulation of Fatty Acid β-Oxidation in Zymosan-Induced Peritonitis,” PLoS One, vol. 8, no. 6, p. e66270, 2013. [130] M. G. Netea, J. Quintin, and J. W. M. Van Der Meer, “Trained immunity: A memory for innate host defense,” Cell Host Microbe, vol. 9, no. 5, pp. 355–361, 2011. [131] T. Horng, “mTOR trains heightened macrophage responses,” Trends Immunol., vol. 36, no. 1, pp. 1–2, Jan. 2015. [132] E. Pennisi, “Immunology. Metabolic shift may train immune cells.,” Science, vol. 345, pp. 1550–1551, 2014. [133] J. W. M. van der Meer, L. A. B. Joosten, N. Riksen, and M. G. Netea, “Trained immunity: A smart way to enhance innate immune defence,” Mol. Immunol., vol. 68, no. 1, pp. 40–44, 2015. [134] M. G. Netea et al., “Trained immunity: A program of innate immune memory in health and disease,” Science (80-. )., vol. 352, no. 6284, p. aaf1098-aaf1098, 2016. [135] S.-C. Cheng et al., “mTOR- and HIF-1 -mediated aerobic glycolysis as metabolic basis for trained immunity,” Science (80-. )., vol. 345, no. 6204, p. 1250684, 2014. [136] D. C. Ifrim et al., “Trained immunity or tolerance: opposing functional programs induced in human monocytes after engagement of various pattern recognition receptors.,” Clin. Vaccine Immunol., vol. 21, no. 4, pp. 534–45, Apr. 2014. [137] T. F. Liu, V. T. Vachharajani, B. K. Yoza, and C. E. McCall, “NAD+ -dependent sirtuin 1 and 6 proteins coordinate a switch from glucose to fatty acid oxidation during the acute inflammatory response,” J. Biol. Chem., vol. 287, no. 31, pp. 25758–25769, Jul. 2012. [138] A. Kauppinen, T. Suuronen, J. Ojala, K. Kaarniranta, and A. Salminen, “Antagonistic crosstalk between NF-κB and SIRT1 in the regulation of inflammation and metabolic disorders.,” Cell. Signal., vol. 25, no. 10, pp. 1939–48, Oct. 2013. [139] P. Millet, V. Vachharajani, L. McPhail, B. Yoza, and C. E. McCall, “GAPDH Binding to TNF- mRNA Contributes to Posttranscriptional Repression in Monocytes: A Novel Mechanism of Communication between Inflammation and Metabolism,” J. Immunol., vol. 196, no. 6, pp. 2541–2551, 2016. [140] S.-C. Cheng et al., “Broad defects in the energy metabolism of leukocytes underlie immunoparalysis in sepsis.,” Nat. Immunol., vol. 17, no. 4, pp. 406–13, Apr. 2016. [141] M. D. Buck et al., “T cell metabolism drives immunity,” J. Exp. Med., vol. 212, no. 9, pp. 151  1345–1360, 2015. [142] S. Galván-Peña and L. A. J. O’Neill, “Metabolic reprograming in macrophage polarization.,” Front. Immunol., vol. 5, no. September, p. 420, 2014. [143] B. Kelly and L. A. J. O’Neill, “Metabolic reprogramming in macrophages and dendritic cells in innate immunity,” Cell Res., vol. 25, no. 7, pp. 771–784, Jul. 2015. [144] S. C.-C. Huang et al., “Cell-intrinsic lysosomal lipolysis is essential for alternative activation of macrophages,” Nat. Immunol., vol. 15, no. August, pp. 1–12, 2014. [145] H. a Mintz et al., “MORPHOLOGICAL AND BIOCHEMICAL STUDIES OF ISOLATED MITOCHONDRIA FROM FETAL, NEONATAL, AND ADULT LIVER AND FROM NEOPLASTIC TISSUES,” J. Cell Biol., vol. 34, no. 2, pp. 513–523, Aug. 1967. [146] G. Mészáros et al., “Altered mitochondrial response to activation of T-cells in neonate,” Acta Physiol. Hung., vol. 102, no. 2, pp. 216–227, 2015. [147] T. Honzik et al., “Activities of respiratory chain complexes and pyruvate dehydrogenase in isolated muscle mitochondria in premature neonates,” Early Hum. Dev., vol. 84, no. 4, pp. 269–276, 2008. [148] C. Valcarce, J. M. Izquierdo, M. Chamorro, and J. M. Cuezva, “Mammalian adaptation to extrauterine environment: mitochondrial functional impairment caused by prematurity.,” Biochem. J., vol. 303 ( Pt 3, no. 1 994, pp. 855–62, 1994. [149] M. Tsubahara et al., “Glucose metabolism soon after birth in very premature infants with small- and appropriate-for-gestational-age birth weights,” Early Hum. Dev., vol. 88, no. 9, pp. 735–738, 2012. [150] S. Illsinger et al., “Energy metabolism in umbilical endothelial cells from preterm and term neonates,” J. Perinat. Med., vol. 39, no. 5, pp. 587–593, 2011. [151] P. Brauner et al., “Induction of Uncoupling Protein 3 Gene Expression in Skeletal Muscle of Preterm Newborns,” Pediatr. Res., vol. 53, no. 4, pp. 691–697, 2003. [152] L. Wenchich, J. Zeman, H. Hansíková, R. Plavka, W. Sperl, and J. Houstek, “Mitochondrial energy metabolism in very premature neonates.,” Biol. Neonate, vol. 81, no. 4, pp. 229–35, Nov. 2002. [153] N. a. Oey, J. P. N. Ruiter, T. Attié-Bitach, L. IJlst, R. J. a. Wanders, and F. a. Wijburg, “Fatty acid oxidation in the human fetus: Implications for fetal and adult disease,” J. Inherit. Metab. Dis., vol. 29, no. 1, pp. 71–75, 2006. [154] M.-C. C. Alexandre-Gouabau et al., “Maternal and cord blood LC-HRMS metabolomics reveal alterations in energy and polyamine metabolism, and oxidative stress in very-low birth weight infants,” J. Proteome Res., vol. 12, no. 6, pp. 2764–2778, 2013. [155] G. M. Tannahill et al., “Succinate is an inflammatory signal that induces IL-1β through HIF-1α.,” Nature, vol. 496, no. 7444, pp. 238–42, 2013. [156] H. Wen and J. P.-Y. P. Y. Ting, “Agitation by Suffocation: How Hypoxia Activates Innate Immunity via the Warburg Effect,” Cell Metab., vol. 17, no. 6, pp. 814–815, 2013. [157] N. Nandagopal and P. P. Roux, “Regulation of global and specific mRNA translation by the mTOR signaling pathway,” Translation, vol. 3, no. 1, p. e983402, 2015. [158] C. G. Proud, “mTORC1 regulates the efficiency and cellular capacity for protein synthesis.,” Biochem. Soc. Trans., vol. 41, no. 4, pp. 923–6, 2013. [159] A. Ghazalpour et al., “Comparative analysis of proteome and transcriptome variation in mouse.,” PLoS Genet., vol. 7, no. 6, p. e1001393, Jun. 2011. 152  [160] R. Rauscher and Z. Ignatova, “Tuning innate immunity by translation.,” Biochem. Soc. Trans., vol. 43, no. 6, pp. 1247–52, 2015. [161] C. a Piccirillo, E. Bjur, I. Topisirovic, N. Sonenberg, and O. Larsson, “Translational control of immune responses: from transcripts to translatomes.,” Nat. Immunol., vol. 15, no. 6, pp. 503–11, 2014. [162] P. Pierre, “Immunity and the regulation of protein synthesis: surprising connections.,” Curr. Opin. Immunol., vol. 21, no. 1, pp. 70–7, Feb. 2009. [163] R. J. Argüello, C. Rodriguez Rodrigues, E. Gatti, and P. Pierre, “Protein synthesis regulation, a pillar of strength for innate immunity?,” Curr. Opin. Immunol., vol. 32, pp. 28–35, Feb. 2015. [164] B. Mazumder, X. Li, and S. Barik, “Translation control: a multifaceted regulator of inflammatory response.,” J. Immunol., vol. 184, no. 7, pp. 3311–9, Apr. 2010. [165] A. G. Hinnebusch, “The scanning mechanism of eukaryotic translation initiation.,” Annu. Rev. Biochem., vol. 83, pp. 779–812, 2014. [166] J. R. Buchan and R. Parker, “Eukaryotic stress granules: the ins and outs of translation.,” Mol. Cell, vol. 36, no. 6, pp. 932–41, Dec. 2009. [167] R. Yamashita et al., “Comprehensive detection of human terminal oligo-pyrimidine (TOP) genes and analysis of their characteristics,” Nucleic Acids Res., vol. 36, no. 11, pp. 3707–3715, 2008. [168] V. Gandin et al., “nanoCAGE reveals 5’ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs.,” Genome Res., vol. 26, no. 5, pp. 636–48, 2016. [169] P. Brodin et al., “Variation in the human immune system is largely driven by non-heritable influences.,” Cell, vol. 160, no. 1–2, pp. 37–47, Jan. 2015. [170] C. P. Martinez-Jimenez et al., “Aging increases cell-to-cell transcriptional variability upon immune stimulation.,” Science, vol. 355, no. 6332, pp. 1433–1436, 2017. [171] P. M. Lavoie and O. Levy, “Mononuclear Phagocyte System,” in Fetal and Neonatal Physiology, Elsevier, 2017, p. 1208–1216.e3. [172] D. J. Dowling and O. Levy, “Ontogeny of early life immunity.,” Trends Immunol., vol. 35, no. 7, pp. 299–310, Jul. 2014. [173] T. Strunk, A. Currie, P. Richmond, K. Simmer, and D. Burgner, “Innate immunity in human newborn infants: prematurity means more than immaturity.,” J. Matern. Fetal. Neonatal Med., vol. 24, no. 1, pp. 25–31, Jan. 2011. [174] B. Ferwerda et al., “Human dectin-1 deficiency and mucocutaneous fungal infections.,” N. Engl. J. Med., vol. 361, no. 18, pp. 1760–7, Oct. 2009. [175] P. M. Lavoie et al., “Influence of common non-synonymous Toll-like receptor 4 polymorphisms on bronchopulmonary dysplasia and prematurity in human infants.,” PLoS One, vol. 7, no. 2, p. e31351, 2012. [176] 1000 Genomes Project Consortium et al., “A global reference for human genetic variation.,” Nature, vol. 526, no. 7571, pp. 68–74, Oct. 2015. [177] A. Panda, J. Martindale, and M. Gorospe, “Polysome Fractionation to Analyze mRNA Distribution Profiles,” BIO-PROTOCOL, vol. 7, no. 3, pp. 1–11, 2017. [178] M. E. Ritchie et al., “limma powers differential expression analyses for RNA-sequencing and microarray studies.,” Nucleic Acids Res., vol. 43, no. 7, p. e47, Apr. 2015. 153  [179] G. Bindea et al., “ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks.,” Bioinformatics, vol. 25, no. 8, pp. 1091–3, Apr. 2009. [180] P. Shannon et al., “Cytoscape: a software environment for integrated models of biomolecular interaction networks.,” Genome Res., vol. 13, no. 11, pp. 2498–504, Nov. 2003. [181] J. W. Yu et al., “MALT1 Protease Activity Is Required for Innate and Adaptive Immune Responses,” PLoS One, vol. 10, no. 5, p. e0127083, 2015. [182] S. E. Turvey et al., “The CARD11-BCL10-MALT1 (CBM) signalosome complex: Stepping into the limelight of human primary immunodeficiency,” J. Allergy Clin. Immunol., vol. 134, pp. 276–284, 2014. [183] E. J. E. L. Pearce and E. J. E. L. Pearce, “Metabolic pathways in immune cell activation and quiescence.,” Immunity, vol. 38, no. 4, pp. 633–43, Apr. 2013. [184] T. Weichhart, M. Hengstschläger, and M. Linke, “Regulation of innate immune cell function by mTOR,” Nat. Rev. Immunol., vol. 15, no. 10, pp. 599–614, 2015. [185] W. K. E. Ip, N. Hoshi, D. S. Shouval, S. Snapper, and R. Medzhitov, “Anti-inflammatory effect of IL-10 mediated by metabolic reprogramming of macrophages.,” Science, vol. 356, no. 6337, pp. 513–519, 2017. [186] M. N. Corradetti, K. Inoki, and K.-L. Guan, “The stress-inducted proteins RTP801 and RTP801L are negative regulators of the mammalian target of rapamycin pathway.,” J. Biol. Chem., vol. 280, no. 11, pp. 9769–72, Mar. 2005. [187] M. F. Princiotta et al., “Quantitating protein synthesis, degradation, and endogenous antigen processing,” Immunity, vol. 18, no. 3, pp. 343–354, 2003. [188] S. Arandjelovic and K. S. Ravichandran, “Phagocytosis of apoptotic cells in homeostasis.,” Nat. Immunol., vol. 16, no. 9, pp. 907–17, 2015. [189] J. Guinea, “Global trends in the distribution of Candida species causing candidemia.,” Clin. Microbiol. Infect., vol. 20 Suppl 6, no. 6, pp. 5–10, Jun. 2014. [190] J. Chen et al., “Epidemiology of and risk factors for neonatal candidemia at a tertiary care hospital in western China,” BMC Infect. Dis., vol. 16, no. 1, p. 700, 2016. [191] J. Chen et al., “Integrated measures for prevention of invasive Candida infections in preterm infants in a Chinese neonatal intensive care unit,” Am. J. Infect. Control, vol. 43, no. 12, pp. 1321–1325, Dec. 2015. [192] D. K. Benjamin et al., “Neonatal candidiasis among extremely low birth weight infants: risk factors, mortality rates, and neurodevelopmental outcomes at 18 to 22 months.,” Pediatrics, vol. 117, no. 1, pp. 84–92, Jan. 2006. [193] J. L. Robinson et al., “Characteristics and outcome of infants with candiduria in neonatal intensive care - a Paediatric Investigators Collaborative Network on Infections in Canada (PICNIC) study.,” BMC Infect. Dis., vol. 9, p. 183, Nov. 2009. [194] L. Saiman et al., “Risk factors for candidemia in Neonatal Intensive Care Unit patients,” Pediatr. Infect. Dis. J., vol. 19, no. 4, pp. 319–324, Apr. 2000. [195] M. J. Marakalala et al., “Differential adaptation of Candida albicans in vivo modulates immune recognition by dectin-1.,” PLoS Pathog., vol. 9, no. 4, p. e1003315, Apr. 2013. [196] S. Hontelez, A. Sanecka, M. G. Netea, A. B. van Spriel, and G. J. Adema, “Molecular view on PRR cross-talk in antifungal immunity.,” Cell. Microbiol., vol. 14, no. 4, pp. 467–74, 154  Apr. 2012. [197] J. R. Bermick et al., “Neonatal monocytes exhibit a unique histone modification landscape.,” Clin. Epigenetics, vol. 8, p. 99, 2016. [198] R. Elfakess, H. Sinvani, O. Haimov, Y. Svitkin, N. Sonenberg, and R. Dikstein, “Unique translation initiation of mRNAs-containing TISU element.,” Nucleic Acids Res., vol. 39, no. 17, pp. 7598–609, Sep. 2011. [199] C. C. Thoreen, L. Chantranupong, H. R. Keys, T. Wang, N. S. Gray, and D. M. Sabatini, “A unifying model for mTORC1-mediated regulation of mRNA translation,” Nature, vol. 485, no. 7396, pp. 109–113, 2012. [200] O. Larsson et al., “Distinct perturbation of the translatome by the antidiabetic drug metformin.,” Proc. Natl. Acad. Sci. U. S. A., vol. 109, no. 23, pp. 8977–82, Jun. 2012. [201] V. T. Vachharajani et al., “SIRT1 inhibition during the hypoinflammatory phenotype of sepsis enhances immunity and improves outcome.,” J. Leukoc. Biol., vol. 96, no. 5, pp. 785–96, Nov. 2014. [202] P. L. De Jager, N. Hacohen, D. Mathis, A. Regev, B. E. Stranger, and C. Benoist, “ImmVar project: Insights and design considerations for future studies of ‘healthy’ immune variation.,” Semin. Immunol., vol. 27, no. 1, pp. 51–7, Feb. 2015. [203] S. Kim et al., “Characterizing the genetic basis of innate immune response in TLR4-activated human monocytes.,” Nat. Commun., vol. 5, p. 5236, Oct. 2014. [204] B. P. Fairfax et al., “Innate immune activity conditions the effect of regulatory variants upon monocyte gene expression.,” Science, vol. 343, no. 6175, p. 1246949, Mar. 2014. [205] E. J. Carr et al., “The cellular composition of the human immune system is shaped by age and cohabitation.,” Nat. Immunol., vol. 17, no. 4, pp. 461–468, Apr. 2016. [206] J.-L. Casanova, “Human genetic basis of interindividual variability in the course of infection.,” Proc. Natl. Acad. Sci. U. S. A., vol. 112, no. 51, pp. E7118-27, Dec. 2015. [207] J.-L. Casanova and L. Abel, “The genetic theory of infectious diseases: a brief history and selected illustrations.,” Annu. Rev. Genomics Hum. Genet., vol. 14, no. 85, pp. 215–43, 2013. [208] J.-L. Casanova, L. Abel, and L. Quintana-Murci, “Immunology taught by human genetics.,” Cold Spring Harb. Symp. Quant. Biol., vol. 78, pp. 157–72, 2013. [209] L. V Hooper, D. R. Littman, and A. J. Macpherson, “Interactions between the microbiota and the immune system.,” Science, vol. 336, no. 6086, pp. 1268–73, Jun. 2012. [210] R. J. W. Arts, M. S. Gresnigt, L. A. B. Joosten, and M. G. Netea, “Cellular metabolism of myeloid cells in sepsis.,” J. Leukoc. Biol., vol. 100, no. October, Jun. 2016. [211] N. Fodil, D. Langlais, and P. Gros, “Primary Immunodeficiencies and Inflammatory Disease: A Growing Genetic Intersection.,” Trends Immunol., vol. 37, no. 2, pp. 126–40, Feb. 2016. [212] A. Tang et al., “Inflammasome-Mediated IL-1b Production in Humans with Cystic Fibrosis,” PLoS One, vol. 7, no. 5, pp. 1–12, 2012. [213] P. Du, W. A. Kibbe, and S. M. Lin, “lumi: a pipeline for processing Illumina microarray.,” Bioinformatics, vol. 24, no. 13, pp. 1547–8, Jul. 2008. [214] M. A. Mooney and B. Wilmot, “Gene set analysis: A step-by-step guide.,” Am. J. Med. Genet. B. Neuropsychiatr. Genet., vol. 168, no. 7, pp. 517–27, Oct. 2015. [215] A. J. Mitchell, B. Roediger, and W. Weninger, “Monocyte homeostasis and the plasticity 155  of inflammatory monocytes.,” Cell. Immunol., vol. 291, no. 1–2, pp. 22–31. [216] S. Endres et al., “In vitro production of IL 1 beta, IL 1 alpha, TNF and IL2 in healthy subjects: distribution, effect of cyclooxygenase inhibition and evidence of independent gene regulation.,” Eur. J. Immunol., vol. 19, no. 12, pp. 2327–33, Dec. 1989. [217] S. C. Mueller, R. März, M. Schmolz, and B. Drewelow, “Intraindividual long term stability and response corridors of cytokines in healthy volunteers detected by a standardized whole-blood culture system for bed-side application.,” BMC Med. Res. Methodol., vol. 12, p. 112, Aug. 2012. [218] M. W. van der Linden, T. W. Huizinga, D. J. Stoeken, A. Sturk, and R. G. Westendorp, “Determination of tumour necrosis factor-alpha and interleukin-10 production in a whole blood stimulation system: assessment of laboratory error and individual variation.,” J. Immunol. Methods, vol. 218, no. 1–2, pp. 63–71, Sep. 1998. [219] P. Yaqoob, E. A. Newsholme, and P. C. Calder, “Comparison of cytokine production in cultures of whole human blood and purified mononuclear cells.,” Cytokine, vol. 11, no. 8, pp. 600–5, Aug. 1999. [220] S. M. Coyle, S. E. Calvano, and S. F. Lowry, “Gender influences in vivo human responses to endotoxin.,” Shock, vol. 26, no. 6, pp. 538–43, Dec. 2006. [221] L. Konecna, M. S. Yan, L. E. Miller, J. Schölmerich, W. Falk, and R. H. Straub, “Modulation of IL-6 production during the menstrual cycle in vivo and in vitro.,” Brain. Behav. Immun., vol. 14, no. 1, pp. 49–61, Mar. 2000. [222] W. Schraut, P. Wendelgass, J. C. Calzada-Wack, M. Frankenberger, and H. W. Ziegler-Heitbrock, “TNF gene expression in monocytes of low and high responder individuals.,” Cytokine, vol. 9, no. 3, pp. 206–11, Mar. 1997. [223] R. Schindler, B. D. Clark, and C. A. Dinarello, “Dissociation between interleukin-1 beta mRNA and protein synthesis in human peripheral blood mononuclear cells.,” J. Biol. Chem., vol. 265, no. 18, pp. 10232–7, Jun. 1990. [224] J. C. Boldrick et al., “Stereotyped and specific gene expression programs in human innate immune responses to bacteria.,” Proc. Natl. Acad. Sci. U. S. A., vol. 99, no. 2, pp. 972–7, Jan. 2002. [225] G. J. Nau, J. F. L. Richmond, A. Schlesinger, E. G. Jennings, E. S. Lander, and R. A. Young, “Human macrophage activation programs induced by bacterial pathogens.,” Proc. Natl. Acad. Sci. U. S. A., vol. 99, no. 3, pp. 1503–8, Feb. 2002. [226] I. V Yang et al., “Novel regulators of the systemic response to lipopolysaccharide.,” Am. J. Respir. Cell Mol. Biol., vol. 45, no. 2, pp. 393–402, Aug. 2011. [227] S. Blankley et al., “Identification of the key differential transcriptional responses of human whole blood following TLR2 or TLR4 ligation in-vitro.,” PLoS One, vol. 9, no. 5, p. e97702, 2014. [228] M. M. Wurfel et al., “Identification of high and low responders to lipopolysaccharide in normal subjects: an unbiased approach to identify modulators of innate immunity.,” J. Immunol., vol. 175, no. 4, pp. 2570–8, Aug. 2005. [229] D. M. Agnese et al., “Human toll-like receptor 4 mutations but not CD14 polymorphisms are associated with an increased risk of gram-negative infections.,” J. Infect. Dis., vol. 186, no. 10, pp. 1522–5, Nov. 2002. [230] F. M. S. Leoratti et al., “Variants in the toll-like receptor signaling pathway and clinical 156  outcomes of malaria.,” J. Infect. Dis., vol. 198, no. 5, pp. 772–80, Sep. 2008. [231] A. E. Medvedev et al., “Distinct mutations in IRAK-4 confer hyporesponsiveness to lipopolysaccharide and interleukin-1 in a patient with recurrent bacterial infections.,” J. Exp. Med., vol. 198, no. 4, pp. 521–31, Aug. 2003. [232] M. Ceppi et al., “Ribosomal protein mRNAs are translationally-regulated during human dendritic cells activation by LPS,” Immunome Res., vol. 5, no. 1, p. 5, Jan. 2009. [233] J. Schott, S. Reitter, J. Philipp, K. Haneke, H. Schäfer, and G. Stoecklin, “Translational Regulation of Specific mRNAs Controls Feedback Inhibition and Survival during Macrophage Activation,” PLoS Genet., vol. 10, no. 6, 2014. [234] R. Muñoz-Planillo, P. Kuffa, G. Martínez-Colón, B. L. Smith, T. M. Rajendiran, and G. Núñez, “K+ efflux is the common trigger of NLRP3 inflammasome activation by bacterial toxins and particulate matter.,” Immunity, vol. 38, no. 6, pp. 1142–53, Jun. 2013. [235] N. Wood and C.-A. Siegrist, “Neonatal immunization: where do we stand?,” Curr. Opin. Infect. Dis., vol. 24, no. 3, pp. 190–5, Jun. 2011. [236] J. L. Vanoverschelde, M. F. Janier, J. E. Bakke, D. R. Marshall, and S. R. Bergmann, “Rate of glycolysis during ischemia determines extent of ischemic injury and functional recovery after reperfusion.,” Am. J. Physiol., vol. 267, no. 5 Pt 2, pp. H1785-94, Nov. 1994. [237] S. Kishi et al., “Meclizine Preconditioning Protects the Kidney Against Ischemia-Reperfusion Injury.,” EBioMedicine, vol. 2, no. 9, pp. 1090–101, Sep. 2015. [238] M. Li et al., “A TIGAR-regulated metabolic pathway is critical for protection of brain ischemia.,” J. Neurosci., vol. 34, no. 22, pp. 7458–71, May 2014. [239] V. M. Gohil et al., “Nutrient-sensitized screening for drugs that shift energy metabolism from mitochondrial respiration to glycolysis.,” Nat. Biotechnol., vol. 28, no. 3, pp. 249–55, Mar. 2010. [240] E. D. E. Papathanassoglou, M. D. Giannakopoulou, and E. Bozas, “Genomic variations and susceptibility to sepsis.,” AACN Adv. Crit. Care, vol. 17, no. 4, pp. 394–422, 2006. [241] J. E. Peters et al., “Insight into Genotype-Phenotype Associations through eQTL Mapping in Multiple Cell Types in Health and Immune-Mediated Disease.,” PLoS Genet., vol. 12, no. 3, p. e1005908, Mar. 2016. [242] R. J. Hause et al., “Identification and validation of genetic variants that influence transcription factor and cell signaling protein levels,” Am. J. Hum. Genet., vol. 95, no. 2, pp. 194–208, 2014. [243] J. M. Chick et al., “Defining the consequences of genetic variation on a proteome-wide scale,” Nature, vol. 534, no. 7608, pp. 500–505, 2016. [244] Y. Li et al., “Inter-individual variability and genetic influences on cytokine responses to bacteria and fungi,” Nat. Med., vol. 22, no. 8, pp. 952–960, 2016. [245] S. K. Hall et al., “Correlation of polymorphic variation in the promoter region of the interleukin-1 beta gene with secretion of interleukin-1 beta protein.,” Arthritis Rheum., vol. 50, no. 6, pp. 1976–83, Jun. 2004. [246] R. ter Horst et al., “Host and Environmental Factors Influencing Individual Human Cytokine Responses,” Cell, vol. 167, no. 4, p. 1111–1124.e13, 2016. [247] M. Schirmer et al., “Linking the Human Gut Microbiome to Inflammatory Cytokine Production Capacity,” Cell, vol. 167, no. 4, p. 1125–1136.e8, 2016. [248] M. Mangino, M. Roederer, M. H. Beddall, F. O. Nestle, and T. D. Spector, “Innate and 157  adaptive immune traits are differentially affected by genetic and environmental factors.,” Nat. Commun., vol. 8, no. May 2016, p. 13850, Jan. 2017. [249] C. Vogel and E. M. Marcotte, “Insights into the regulation of protein abundance from proteomic and transcriptomic analyses,” Nat. Rev. Genet., vol. 13, no. 4, pp. 227–232, 2012. [250] Y. Liu, A. Beyer, and R. Aebersold, “On the Dependency of Cellular Protein Levels on mRNA Abundance,” Cell, vol. 165, no. 3, pp. 535–550, 2016.     158  Appendix    Figure A.1 Representative Gating Strategy for Pro-IL-1β MFI   159    Figure A.2 Age-Related Changes in Expression of Genes Involved in Mitochondrial Electron-Transport  Depiction of differential expression in gene involved in the Electron Transport Chain. Data from same samples as in Figure 3 and 4, with preterm ranging from 24 - 28 weeks gestation. Image was created using the PathVisio software at: https://www.pathvisio.org/about/cite-us/. Log fold change of genes significantly affect by age (Limma, unstimulated and LPS samples, FDR 5%) is depicted comparing preterm (left of boxes) and term (right of boxes) samples to adults. Blue/red indicate decreased/increased expression in newborns compared to adults. Scale represents z-score. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0365287/manifest

Comment

Related Items