The Role of Probiotic Lactobacillus in Immune Regulation and Modulation of the Vaginal Microbiota During Pregnancy

Siwen Yang

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Physiology University of Toronto

Siwen Yang

Doctor of Philosophy

Department of Physiology University of Toronto

2015

Abstract

Preterm birth (PTB) occurs in 10% of all pregnancies globally. Premature babies have a mortality rate 40 times higher than term infants. Approximately 25-30% of PTB can be attributed to intrauterine infection/inflammation. A disturbance of the vaginal microbiota as observed in bacterial vaginosis (BV) is associated with an increased risk of PTB. Treatment of preterm labor with antibiotics is largely ineffective, and probiotic lactobacilli have been proposed as a potential preventive therapy for BV and PTB. The objectives of this thesis were to assess 1) the effect of Lactobacillus rhamnosus GR-1 (GR-1) and its supernatant

(GR-1 SN) on the prevention of lipopolysaccharide (LPS)-induced PTB and systemic and intra-uterine cytokine and chemokine profiles in pregnant CD-1 mice, 2) the effect of GR-1 on the mouse vaginal microbiota, and 3) the effect of GR-1 and L. reuteri RC-14 on the cervico-vaginal cytokine profile and vaginal microbiota in pregnant women with an abnormal Nugent score. Pregnant mice were pre-treated with intra-peritoneal injections of

GR-1 SN or oral GR-1 live bacteria prior to intrauterine injection of LPS in two separate studies. The expression of cytokines and chemokines in the maternal plasma, amniotic fluid and intrauterine tissues were then measured. The vaginal microbiota was also determined in animals treated with oral GR-1 live bacteria. Pre-treatment with GR-1 SN, but not with GR-1 live bacteria, reduced LPS-induced PTB and inflammation in pregnant mice. The vaginal microbiota of pregnant mice was altered with oral GR-1 live bacteria. A randomized, double blind placebo-controlled trial was conducted, in which pregnant women with an abnormal

Nugent score in their first trimester of pregnancy received orally either placebo or GR-1 and

RC-14 for 12 weeks. Their cervico-vaginal cytokine profile and vaginal microbiota was then determined. Oral GR-1 and RC-14, at the dose and duration used, did not change the cytokine profile and vaginal microbiota of pregnant women with an abnormal Nugent score.

We conclude that L. rhamnosus GR-1 supernatant, but not the live bacteria, may have the potential to serve as a prophylactic therapy for inflammation-associated conditions during pregnancy, including PTB.

iii

Dedication

To my beloved parents, for their continuous support and unconditional love.

Acknowledgements

With sincere respect, I would like to express my gratitude to my supervisor, Dr. Alan

Bocking, for his continuous guidance and unfaltering belief in me for the past years. Your professional work ethics served as a role model to me. I appreciate your patience, understanding and support through the tough times. I am forever thankful for your encouragement and valuable ideas that make my PhD experience meaningful and productive.

My deepest thanks to my co-supervisor, Dr. John Challis, for sharing his wealth of knowledge in physiology and providing his continuous support throughout the years. I am grateful for your constructive recommendations while challenging me to think beyond my intellectual comfort zone.

Many thanks to members of my advisory committee for keeping me on the right track to the completion of my projects. I would like to thank Dr. Stephen Lye, for offering his advice on the physiological aspect of my project. I would also like to thank Dr. Sung Kim, for sharing his knowledgeable insights in immunology. I am thankful to Dr. Gregory Gloor for his advice on interpreting sequencing data. His excellent teaching skills made learning R less nerve wrecking.

I would like to extend my gratitude to Dr. Gregor Reid for sharing his knowledge on probiotics and his willingness to devote time to engage in my work. I would like to acknowledge members of his laboratory, Shannon, Jordan, Grace, Leslie, Amy, Camilla and

Yige for making me feel at home during my stay in London. Special thanks to Ms Shannon

Seney, Mr Rod McPhee and Ms Amy McMillian for providing Nugent scores for my project.

I would like to acknowledge members of Dr. Gloor’s laboratory, Jean and Julia, for helping me learn R. Special thanks to David Carter at the London Research Institute for his help with

Illumina Sequencing.

Sincere thanks to examiners of my qualifying exam, Drs Michelle Letarte and Theodore

Brown, and examiners of my CIHR grant proposal course, Drs Lee Adamson, Denise

Belsham and Clifford Librach, for your critical evaluations of my project and for your valuable suggestions at the examinations.

I would like to thank members of the VOGUE team who have generously contributed their ideas and time discussing my project. I would also like to thank Dr. Laurent Briollais, for offering his help with the statistical analysis of my project. I would like to thank the research nurses, Ms Mary-Jean Martin and Ms Tara Maria Rocco, of Mount Sinai Hospital for the recruitment of participants and collection of vaginal swabs, as well as the volunteer participants for their generous contribution of samples for the project.

I am lucky to have the great companion from members of the Bocking lab and the Lye lab.

Thank you all for providing such a supportive and enjoyable working environment. My special thanks to Dr. Wei Li for mentoring me during times of technical difficulties, and my deepest gratitude to Dr. Oksana Shynolva, for both your scientific insights and for offering me emotional support. Many thanks to executive assistants, Ms Elaine Dwek and Ms Beverly

Bessey, for being miracle creators. No matter how busy their bosses’ schedules were, you can always accommodate my continuous requests to schedule meetings.

I would like to thank the Department of Physiology for being a haven of intellectual freedom, and the wonderful staffs of the department, especially Ms Colleen Shea and Ms Rosalie Pang for your continuous support with administrative issues. I would also like to acknowledge the support of the funding agencies, the Genesis Research Foundation, the University of Toronto and Mount Sinai Hospital, for supporting my education and recognizing the importance of my project.

I am grateful for the company and steadfast support of my best friends, Sally Shi and Lydia

Zhou. Thank you girls for the wonderful moments we had together and for always believing in me. Special thanks to Han Li, for taking care of me like a big sister.

Finally, I would like to dedicate my work to my beloved parents, who are my source of strength. Thank you for your unconditional love and for your support in realizing my dreams.

Thank you for your guidance in life and for teaching me the most important aspects of life are to have Peace, Happiness and Health.

vii

Table of Contents

3. Table of Contents Abstract ...... ii Dedication ...... iv Acknowledgements ...... v Table of Contents ...... viii List of Abbreviations ...... xii List of Figures ...... xiv List of Tables ...... xvii

1 General Introduction ...... 2 1.1 Human Pregnancy and Parturition ...... 2 1.1.1 Anatomy of the Intra-uterine Environment ...... 3 1.2 Mechanisms of Human Parturition ...... 8 1.2.1 Prostaglandins (PGs) ...... 12 1.2.2 Matrix Metalloproteinase (MMPs) ...... 12 1.2.3 Cytokines and Chemokines ...... 15 1.3 Preterm Birth ...... 24 1.3.1 Epidemiology ...... 24 1.3.2 Etiology ...... 25 1.3.3 Infection Routes ...... 25 1.3.4 Infection and/or Inflammation- induced PTB ...... 26 1.3.5 Current Treatment Approaches ...... 28 1.3.6 Animal Models of Preterm Birth ...... 29 1.4 Vaginal Microbiota and Preterm Birth ...... 30 1.4.1 The human vaginal microbiota ...... 30 1.4.2 Bacterial Vaginosis ...... 32 1.5 Probiotics ...... 33 1.5.1 Safety and Compliance ...... 34 1.5.2 Lactobacilli ...... 35 1.6 Summary ...... 37

viii

2. Rationale and Hypotheses ...... 40 2.1 Rationale ...... 40 2.2 Hypotheses ...... 41

3. Probiotic Lactobacillus rhamnosus GR-1 supernatant (GR-1 SN) prevents Lipopolysaccharide (LPS)-induced preterm birth and reduces inflammation in pregnant CD-1 mice...... 43 3.1 Introduction ...... 43 3.2 Material and Methods ...... 44 3.2.1 Animals ...... 44 3.2.2 L. rhamnosus GR-1 supernatant preparation ...... 45 3.2.3 Intra-uterine injection of LPS by mini-laparotomy ...... 45 3.2.4 Dose effect of LPS on PTB rate (Set 1) ...... 45 3.2.5 Effect of GR-1 supernatant on the timing of LPS-induced PTB (Set 2) ...... 46 3.2.6 Effect of GR-1 supernatant on cytokines and chemokines (Set 3) ...... 46 3.2.7 Fetal Sex ratios (Set 4) ...... 47 3.2.8 Cytokine assay ...... 47 3.2.9 Maternal progesterone measurement ...... 47 3.2.10 Sex determination by PCR ...... 48 3.2.11 Statistical Analyses ...... 48 3.3 Results ...... 48 3.3.1 GR-1 SN reduced LPS-induced PTB (Set 2) ...... 48 3.3.2 GR-1 SN attenuated LPS induced cytokines and chemokines (Set 3) ...... 49 3.3.3 Plasma progesterone (Set 3) ...... 50 3.3.4 Fetal sex ratio (Set 4) ...... 50 3.4 Comment ...... 50

4. Oral Probiotic Lactobacillus rhamnosus GR-1 stimulates systemic and intrauterine production of cytokines and chemokines and modulates the vaginal microbiota in pregnant CD-1 mice...... 69 4.1 Introduction ...... 69 4.2 Material and Methods ...... 71 4.2.1 Animals ...... 71

4.2.2 Lactobacillus rhamnosus GR-1 preparation ...... 71 4.2.3 Intra-uterine injection of LPS by mini-laparotomy ...... 72 4.2.4 Oral administration of GR-1 by oral gavage ...... 72 4.2.5 Effect of oral GR-1 on the timing of LPS-induced PTB (Set 1) ...... 72 4.2.6 Effect of oral GR-1 on the gestational length (Set 2) ...... 73 4.2.7 Effect of oral GR-1 on cytokines and chemokines (Set 3) ...... 73 4.2.8 Effect of oral GR-1 on the vaginal and cecal microbiota (Set 4) ...... 73 4.2.9 Cytokine Assay ...... 74 4.2.10 Maternal progesterone measurement ...... 74 4.2.11 DNA isolation and V6 ribosomal DNA PCR amplification ...... 75 4.2.12 Sequencing ...... 75 4.2.13 Statistical Analysis ...... 75 4.3 Results ...... 76 4.3.1 Effect of oral GR-1 on the incidence of LPS-induced PTB and gestational length (Set 1 and Set 2) ...... 76 4.3.2 Effect of oral GR-1 on the cytokines and chemokines (Set 3) ...... 77 4.3.3 Maternal plasma progesterone (Set 3) ...... 78 4.3.4 Vaginal and Cecal Microbiota (Set 4) ...... 78 4.3.5 Effect of oral GR-1 on the vaginal microbiota (Set 4) ...... 79 4.3.6 Effect of oral GR-1 on the cecal microbiota (Set 4) ...... 79 4.4 Comment ...... 79

5. Effect of oral probiotics Lactobacillus rhamnosus GR-1® and Lactobacillus reuteri RC-14® on the vaginal microbiota and cervico-vaginal cytokines and chemokines in low risk pregnant women with an intermediate or high Nugent score...... 112 5.1 Introduction ...... 112 5.2 Materials and Methods ...... 114 5.2.1 Study Participants ...... 114 5.2.2 Study groups and randomization ...... 114 5.2.4 Probiotic Strains ...... 115 5.2.5 DNA Isolation and PCR amplification of V6 region of 16S rDNA ...... 116 5.2.6 Sequencing ...... 116

5.2.7 Protein Extraction and Cytokine/Chemokine Multiplex Assay ...... 117 5.2.8 Statistical Analyses ...... 117 5.3 Results ...... 118 5.3.1 Pre-randomization characteristics ...... 118 5.3.2 Pregnancy Outcomes ...... 119 5.3.3 Compliance to the treatment protocol ...... 119 5.3.4 Effect of oral probiotic GR-1 and RC-14 on the Nugent score ...... 120 5.3.5 Effect of oral probiotic GR-1 and RC-14 on the vaginal microbiota ...... 120 5.3.6 Effect of GR-1 and RC-14 on the concentrations of cervico-vaginal cytokines/chemokine ...... 121 5.4 Comment ...... 122

6. General Discussion ...... 141

List of References ...... 152

List of Appendices ...... 178

List of Abbreviations

11β-HSD-1 11β-Hydroxysteroid Dehydrogenase-1 ACTH Adrenocorticotropic Hormone ANOVA Analysis of Variance BV Bacterial Vaginosis CAP Contraction-Associated Protein COX-2 Cyclooxygenase-2 CRH Corticotropin-Releasing Hormone CSF Colony Stimulating Factors DC Dendritic Cells dNK Decidual Natural Killer ECM Extracellular Matrix HPA Hypothalamic-Pituitary-Adrenal IFN Interferon IL Interleukin JAK/STAT Janus Kinases and Signal Transducers and Activators of Transcription KC KC Keratinocyte Chemo-attractant Km Factor for converting mg/kg dose to mg/m2 dose L.rhamnosus Lactobacillus rhamnosus LPS LPS Lipopolysaccharide MLCK MLCK Myosin Light-Chain Kinase MMP MMP Matrix Metalloproteinase MRS MRS de Man, Rogosa and Sharpe NF-κB Nuclear Factor-Kappa B NK Natural Killer OT Oxytocin OTR Oxytocin Receptor PCR Polymerase Chain Reaction PG Prostaglandin PGDH Prostaglandin 15-Hydroxy Dehydrogenase

xii

PGE2 Prostaglandin E2 PGF2α Prostaglandin F2α PGHS Prostaglandin H Synthase PPROM Preterm Premature Rupture of the Membranes PTB Preterm Birth PTB Preterm Delivery PTL Preterm Labor PTGS2 Prostaglandin-Endoperoxide Synthase 2 SD Standard Deviation SDI Shannon Diversity Index SEM Standard Error of the Mean SMC Smooth Muscle Cell Th T-helper TL Term Labor TLR Toll-Like Receptor TNF-α Tumor Necrosis Factor-Alpha

xiii

List of Figures

Figure 1-1 Anatomy of the intra-uterine environment...... 7

Figure 1-2 Proposed mechanisms that underlie relaxation and contraction of the myometrium during pregnancy or labor...... 10

Figure 1-3 The proposed pathway of human parturition...... 14

Figure 3-1 Experimental design to investigate the effect of GR-1 supernatant (GR-1 SN) on the timing of LPS-induced PTB (Set 2)...... 54

Figure 3-3 Cumulative frequency plot showing the percentage of pregnant CD-1 mice that delivered at various gestational days following four different treatments (Set 2)...... 56

Figure 3-7 Histogram showing maternal plasma progesterone concentrations for different treatment groups (Set 3)...... 60

Figure 4-1 Probiotic Lactobacillus dose translation from a human dose to a mouse equivalent dose based on the body surface area (Km) and weight...... 85

xiv

Figure 4-2 Experimental design to investigate the effect of oral GR-1 on the timing of LPS- induced PTB (Set 1)...... 86

Figure 4-3 Experimental design to investigate the effect of oral GR-1 on the gestational length (Set 2)...... 87

Figure 4-4 Experimental design to investigate the effect of oral GR-1 on cytokines and chemokines (Set 3)...... 88

Figure 4-5 Experimental design to investigate the effect of oral GR-1 on the vaginal and cecal microbiota (Set 4)...... 89

Figure 4-7 Histogram showing the concentration of pro-inflammatory cytokines IL-1α, IL- 1β, IL-6, IL-17, IL-12p40, IL-12p70, TNFα and IFN-γ in the fetal membranes, placenta, decidua and myometrium of pregnant mice that received saline and GR-1 at 109 cfu via oral gavage (Set 3)...... 91

Figure 4-8 Histogram showing the concentration of anti-inflammatory cytokines IL-2, IL-4, IL-10 and IL-13 in the maternal plasma (MP) and amniotic fluid (AF) of pregnant mice that received varying doses of GR-1 (Set 3)...... 92

Figure 4-11 Histogram showing the concentration of chemokines CCL2, CCL3, CCL4, CCL5, CCL11, CXCL1 in the fetal membranes, placenta, decidua and myometrium of pregnant mice that received saline and oral GR-1 at 109 cfu (Set 3)...... 95

Figure 4-14 Stacked barplots showing the vaginal and cecal bacterial compositions of pregnant CD-1 mice that received either oral saline or GR-1...... 98

Figure 4-15 Scatterplot showing the Shannon diversity index (SDI) of the vaginal and cecal microbiota of pregnant CD-1 mice...... 99

Figure 5-1 Consort flow chart of pregnant women enrolled in the study...... 126

Figure 5-2 Stacked bar plot showing the vaginal microbiota clustered by bacteria similarity in pregnant women prior to treatment, at 13 weeks gestation (n=66)...... 127

Figure 5-4 Stacked bar plots showing the vaginal microbiota across pregnancy clustered by bacteria similarity in pregnant women who received either placebo (n=34) or probiotic (n=32) treatment...... 129

Figure 5-5 Scatterplot showing the Shannon Diversity Index (SDI) across gestations in pregnant women who received either placebo or probiotic treatment...... 130

Figure 5-6 Scatterplots showing the concentrations of cervico-vaginal cytokines IL-4, IL-10 and CSF3 across gestation in pregnant women who received either placebo or probiotic treatment...... 131

Figure 6-1 Changes in sytemic and intrauterine cytokines after treatment with Lactobacillus rhamnosus GR-1 supernatant or live bacteria...... 149

Figure 6-2 LPS-induced sytemic and intrauterine cytokines that were dampened with GR-1 supernatant pretreatment...... 150

xvi

List of Tables

Table 3-1 Delivery outcome of pregnant CD-1 mice that delivered preterm following different doses of LPS intrauterine injection (Set 1)...... 61

Table 3-2 Litter size and fetal weight of neonates born to pregnant CD-1 mice that received different treatments (Set 2)...... 62

Table 3-3 Baseline cytokine and chemokine concentrations in the maternal plasma, myometrium, amniotic fluid and placenta of pregnant CD-1 mice (Set 3)...... 63

Table 3-4 Cytokine and chemokine concentrations in the maternal plasma of pregnant CD-1 mice following different treatments (Set 3)...... 64

Table 3-5 Cytokine and chemokine concentrations in the myometrium of pregnant CD-1 mice following different treatments (Set 3)...... 65

Table 3-6 Cytokine and chemokine concentrations in the amniotic fluid of pregnant CD-1 mice following different treatments (Set 3)...... 66

Table 3-7 Cytokine and chemokine concentrations in the placenta of pregnant CD-1 mice following different treatments (Set 3)...... 67

Table 4-1 Delivery outcome of pregnant CD-1 following different treatments in Set 1...... 100

Table 4-2 Litter size and fetal weight of live term neonates born to pregnant CD-1 mice at term that received different treatments in Set 1...... 101

Table 4-3 Hours to delivery, litter size and fetal weight of neonates born to pregnant CD-1 mice that received saline or oral GR-1 (Set 2)...... 102

Table 4-4 Summary table of cytokines and chemokines in the maternal plasma, amniotic fluid and intrauterine tissues following varying doses of oral GR-1 treatment...... 103

Table 4-5 Maternal plasma progesterone concentrations in pregnant CD-1 mice with varying dose of GR-1 (Set 3) ...... 104

Table 4-6 Bacteria genera unique to the cecal and vaginal tissues of saline-treated pregnant CD- 1 mice...... 105

xvii

Table 4-7 Bacteria genera present in both the cecal and vaginal tissues of saline-treated pregnant CD-1 mice...... 106

Table 4-8 Bacteria at different taxonomic levels that have statistically significant higher abundance in the vaginal tissues than in the cecal tissues of saline-treated pregnant CD-1 mice...... 107

Table 4-9 Bacteria at different taxonomic levels that have statistically significant higher abundance in the cecal tissues than in the vaginal tissues of saline-treated pregnant CD-1 mice...... 108

Table 4-10 Bacteria at different taxonomic levels that decreased significantly with oral GR-1 treatment in the vaginal tissues of pregnant CD-1 mice...... 109

Table 4-11 Bacteria at different taxonomic levels that increased significantly with oral GR-1 treatment in the vaginal tissues of pregnant CD-1 mice...... 110

Table 5-1 Characteristics of pregnant women randomized at 13 weeks gestation...... 132

Table 5-2 Pregnancy outcomes...... 133

Table 5-3 Compliance of women in the probiotic and placebo groups...... 134

Table 5-4 Nugent scores of pregnant women across pregnancy in the probiotic and placebo groups...... 135

Table 5-5 The relative to mean abundance of vaginal bacterial species in pregnant women with a BV (7-10) or an intermediate (4-6) Nugent score at 13 weeks gestation...... 136

Table 5-6 The relative to mean abundance of vaginal bacteria species that decreased across gestation in pregnant women treated with placebo or probiotics...... 137

Table 5-7 The relative to mean abundance of vaginal bacterial species that increased across gestation in pregnant women treated with placebo or probiotics...... 138

Table 5-8 Summary table of cervico-vaginal cytokines and chemokines across gestation in pregnant women who received either placebo or probiotic treatment...... 139

xviii

Chapter One

General Introduction

Part of the contents of this chapter (Section 1.3 to 1.5) was published in Front Immunol. 2015 Feb;6:62 and appears here with the permission of the journal (authorization attached). My role involves manuscript preparation.

1 2

Chapter 1

1 General Introduction

1.1 Human Pregnancy and Parturition

There are four phases in human pregnancy: uterine quiescence (phase 0), contraction- associated protein (CAP)-activated myometrium (Phase 1), uterotonins stimulated myometrium (Phase 2) and uterine involution (Phase 3) (Challis et al., 2000).

During pregnancy, uterine quiescence (Phase 0) is maintained by high levels of signaling molecules, including progesterone, relaxin and prostacyclin (Challis et al., 2000). Progesterone, a steroid hormone produced by the placenta, dampens inflammation produced by inflammatory cytokines and prostaglandins (PGs), which would otherwise induce parturition prior to term (Parizek et al., 2014). In addition, progesterone suppresses the production of estrogen and PGs, thereby reducing smooth muscle cell contractility (Parizek et al., 2014). During pregnancy, progesterone receptor type B (PR-B) dominates (Parizek et al., 2014). The binding of progesterone to PR-B promotes an anti-inflammatory environment and maintains uterine quiescence (Tan et al., 2012). Prior to parturition, functional progesterone withdrawal is observed when the expression of PR-A increases with a concomitant decrease in the expression of PR-B (Tan et al., 2012). The inhibitory effect of progesterone on estrogen, PGs and myometrial contraction is then removed (Mesiano et al., 2002). Furthermore, mechanical stretch caused by the growing fetus results in an up- regulation in the expression of CAPs including oxytocin receptors (OTR), connexin-43 (Cx- 43), PGF2a and its receptors (FP) (Gibb and Challis, 2002). The CAPs activate the myometrium (Phase 1), making it receptive to stimulation by uterotonins such as OT and PGs (Phase 2) (Gibb and Challis, 2002). This results in the production of forceful myometrial contractions, essential for delivery of the fetus and the placenta (Gibb and Challis, 2002).

The fetus also secretes signaling molecules that determine the timing of parturition. Activation of the fetal hypothalamic-pituitary-adrenal (HPA) axis results in an increased

3 production of fetal adrenal cortisol, which suppresses progesterone production and promotes estrogen production (Marciniak et al., 2011). These mediators then promote uterine contractions and initiate the inflammatory cascade leading to parturition (Marciniak et al., 2011). Uterine involution (Phase 3), which occurs after the delivery of the fetus, is mediated by the effect of OT (Challis et al., 2000).

1.1.1 Anatomy of the Intra-uterine Environment

A. Myometrium

The human myometrium, comprised primarily of uterine myocytes, lies between the endometrium (innermost) and the perimetrium of the uterine wall (Coad, 2011). Myometrium produces several uterotonins and inflammatory cytokines, which stimulate the circular smooth muscle layer of the myometrium to produce intense and synchronous uterine contractions during labor (Shynlova et al., 2009). Animal studies revealed uterine myocytes are highly plastic smooth muscle cells (SMCs), which undergo phenotypic changes from a contractile state to a synthetic state, and proliferate during pregnancy (Shynlova et al., 2009). Under the influence of high circulating levels of progesterone and increased mechanical stretch from the fetus, myometrial SMCs proliferate by hypertrophy and remodel to accommodate the growing fetus (Shynlova et al., 2009). When progesterone responsiveness wanes near term, the myometrial SMCs switch from a synthetic state to a contractile state and are sensitive to the stimulation of uterotonins (Shynlova et al., 2009). Postpartum (after delivery), myometrium returns to a phenotype similar to its non-pregnant state (Shynlova et al., 2009).

B. Decidua

The decidua forms the maternal side of the fetal-maternal interface. The decidua parietalis and decidua basalis contact the non-invasive chorion trophoblast cells and the invasive extravillous trophoblast cells respectively (Coad, 2011) (Figure 1-1, page 7). Decidualization

4 is initiated by a rising level of progesterone, even in the absence of a conceptus, and is the process whereby endometrial stromal cells near the spiral arteries undergo morphological, biochemical, and functional changes into decidual stromal cells (DSCs) (Oreshkova et al., 2012). The elongated endometrium stromal fibroblast cells differentiate into enlarged round- shaped secretory DSCs, which can synthesize extracellular matrix components (laminin and fibronectin), hormones, cytokines and matrix metalloproteinase (MMPs) (Oreshkova et al., 2012). In early pregnancy, DSCs participate in the exchange of nutrients, gas and waste with the developing embryo, until the placenta becomes fully functional (Coad, 2011). The DSCs also ensure a controlled trophoblast invasion (Oreshkova et al., 2012). When decidualization is absent, placenta accreta results (Jauniaux et al., 2012). DSCs contain high proportions of resident leukocytes, and nearly 40% of the first trimester decidua is made up of leukocytes (Houser et al., 2012). Decidual leukocytes are important in normal placental development and the regulation of immune responses at the maternal-fetal interface (Houser et al., 2012). Among these decidual leukocytes, nearly 60% are decidual Natural Killer (dNK) cells, 25% are macrophages, 10-20% are T cells and the rest are dendritic cells (DCs) (Houser et al., 2012). The primary role of dNK cells is to initiate vascular remodelling necessary to ensure adequate placental blood flow (Wallace et al., 2012). Decidual macrophages and T cells express inflammatory cytokines and chemokines, which activate and amplify the inflammatory pathways leading to parturition (Houser et al., 2012). Women in term labor (TL) have an accumulation of decidual macrophages in comparison to women at term not in labor (Hamilton et al., 2012).

C. Placenta

The human placenta is composed of extensive branching and densely packed chorionic villi containing fetal blood vessels (Blackburn, 2012). The terminal villi, which make up the majority of the placenta, are the sites for maternal-fetal exchange (Blackburn, 2012). The stem or anchoring villi stabilize the villous tree and the intermediate villi are located between the stem villi and the terminal villi (Blackburn, 2012). Specialized cells of the placenta are called trophoblast cells, which comprise the outer layer of the blastocyst (Blackburn, 2012). The human placenta contains 15-30 cotyledons, which are separations of the decidua basalis

5 divided by placental septa (Blackburn, 2012). The cotyledons contain many chorionic villi, which are finger-like structures formed when the trophoblast cells undergo hyperplasia during implantation (Blackburn, 2012). The placenta is in contact with both maternal and fetal tissues. The outer layer of placental trophoblast cells is continuous with the decidua basalis (Blackburn, 2012). On the fetal side, the placenta is covered by a thin membranous structure called the chorionic plate that is continuous with the fetal membranes (Blackburn, 2012).

The intervillous space is filled with maternal blood, which is separated from the fetal circulation by several layers of tissues (Blackburn, 2012). They are (1) the microvillous membrane of the syncytiotrophoblast, (2) the syncytiotrophoblast cells, (3) the basal membrane of the syncytiotrophoblast, (4) the connective tissue mesenchyme of the villus, and (5) the epithelium of the fetal blood vessel (Blackburn, 2012). The inner mesenchymal core of the chorionic villi contains the umbilical cords and is formed from extraembryonic primitive mesoderm (Blackburn, 2012). Two umbilical arteries that spiral around the umbilical vein deliver deoxygenated blood from the fetus to the placenta (Blackburn, 2012). The arteries branch radially onto the chorionic plate and the chorionic vessels branch into many villous lobular arteries, which branch further into smaller vessels (Blackburn, 2012). This extensive branching makes the placenta an extensively vascularized organ.

The placenta serves both metabolic and endocrine functions. Gases, nutrients, and waste products are exchanged across the endothelial cells between the fetus and the mother (Blackburn, 2012). The placenta also synthesizes estrogen, progesterone, human chorionic gonadotropin (hCG) and cytokines that contribute to either pregnancy quiescence or the onset of parturition (Blackburn, 2012).

D. Fetal Membranes

The fetal membranes (amnion, chorion, trophoblast and decidua) surround and protect the developing fetus during pregnancy (Myatt and Sun, 2010) (Figure 1-1). The amnion is made up of amniotic epithelium and amniotic mesoderm, which later divides into the basal

6 membrane (Coad, 2011). Adjacent to the amnion is the chorion, which is composed of vascularized chorionic mesoderm and a basement membrane (Coad, 2011). The chorion is separated from the decidua by extravillous trophoblast cells (Coad, 2011). The fetal membranes contribute to amniotic fluid turnover, form a barrier between maternal and fetal compartments, and produce signaling molecules that contribute to labor initiation (Myatt and Sun, 2010). Locally produced mediators in the fetal membranes include PGs, glucocorticoids, pro-inflammatory cytokines and surfactant proteins (Myatt and Sun, 2010). The majority of PGs are produced in the fetal membranes and PG synthesis is segregated from its metabolism in different compartments of the fetal membranes (Myatt and Sun, 2010). During pregnancy, PGs produced in the amnion and chorion by PG synthases (PGHS) are metabolized by 15-hydroxy PG dehydrogenase (PGDH) in the chorion trophoblast (Keelan et al., 2003). A limited amount of PGs reach the myometrium and uterine quiescence is maintained. The fetal membranes also synthesize and metabolize glucocorticoids that increase surfactant synthesis to promote fetal lung maturation, and in turn trigger labor initiation (Myatt and Sun, 2010)..

E. Amniotic Fluid

The amniotic fluid cushions the fetus from potential external trauma and maintains a constant temperature in the uterus. The amniotic fluid also accommodates fetal movements, which are important to musculoskeletal structure development (Brace and Wolf, 1989). In addition, the amniotic fluid serves as a medium for the exchange of secreted cytokines, PGs, fetal adrenal cortisol and surfactant proteins between the umbilical vessels and the fetus (Brace and Wolf, 1989). The amniotic fluid volume increases in pregnant women from an initial 1.5 ml at 7 weeks of gestation to 770 ml at 28 weeks of gestation (Brace and Wolf, 1989). The change in the volume is minimal between 29 and 37 weeks of gestation, and after 34 weeks of gestation, the volume decreases (Brace and Wolf, 1989).

Amniotic cavity Decidua basalis Decidua parietalis Core of mesoderm Fetal membranes

Intervillous space (maternal blood) Multinucleated syntiocytotrophoblast Villous Mononucleated cytotrophoblast cytotrophoblast

Endometrial vessels Fetal circulation Extravillous trophoblast

Anchoring villus Chorionic mesoderm mesoderm Chorionic Basement membrane membrane Basement Fibroblast Layer Layer Fibroblast Intermediate Spongy Layer Layer Spongy Intermediate Compact Stromal Layer Layer Stromal Compact Basement membrane membrane Basement Amniotic epithelium epithelium Amniotic Amniotic fluid fluid Amniotic Placenta septum

Amniochorionic membrane

Myometrium

Cervix Decidua Chorion Amnion

Trophoblast Vagina

Figure 1-1 Anatomy of the intra-uterine environment.

The image is modified with permission from The New England Journal of Medicine: Goldenberg RL, Hauth JC, Andrews WW. (2000) Intrauterine infection and preterm delivery. 342 (20):1500-7. Copyright Massachusetts Medical Society.

1.2 Mechanisms of Human Parturition

Human parturition (labor), term and preterm, is driven by positive feed-forward cascades of inflammation produced by increasing levels of PGs and inflammatory cytokines (Challis et al., 2009). Labor is initiated by factors including uterine mechanical stretch, fetal endocrine signals and intrauterine infection (Challis et al., 2009). The balance of pro and anti- inflammatory cytokines, produced by CD4+ T helper (Th) cells, is important in predicting pregnancy outcomes (Challis et al., 2009). In early pregnancy, a modest Th1 pro- inflammatory environment promotes successful implantation and placentation (Wilczynski, 2005). As pregnancy progresses, there is a predominance of Th2 anti-inflammatory cytokines including IL-4 and IL-10, which maintain uterine quiescence (Wilczynski, 2005). A disruption of the Th1/Th2 balance favoring the predominance of Th1 pro-inflammatory cytokines such as IL-1, IL-6, and TNFα may be responsible for some cases of PTL (Challis et al., 2009).

A. Uterine Stretch

Mechanical stretch imposed by the fetus increases the expression of CAPs and causes uterine activation (Gibb and Challis, 2002). The CAPs promote increased myocyte contractility, excitability and intercellular communication (Gibb and Challis, 2002). The myometrial cells with an increase in the expression of CAPs are sensitized to uterotonin stimulation, which leads to the production of coordinated and forceful contractions (Gibb and Challis, 2002) (Figure 1-2, page 10).

During pregnancy, the myocytes are maintained in a relaxed state by the following factors: (1) a high intracellular electrochemical potential, (2) an elevated level of intracellular cyclic AMP (cAMP), and (3) actins in the globular form (Smith, 2007). During pregnancy, the myometrial cell surfaces are abundant with β2 and β3-sympathomimetic receptors, which promote the opening of potassium (K+) channels (Smith, 2007). The efflux of K+ leads to an increase in the intracellular electrochemical potential, which decreases the likelihood of a depolarization and reduces myocyte excitability (Smith, 2007). A high level of cAMP

9 activates protein kinase A (PKA) that enhances phosphodiesterase activity. Phosphodiesterase dephosphorylates and inactivates the myosin light chain kinase (MLCK) and causes calcium (Ca2+) re-uptake by the sarcoplasmic reticulum (SR) (Smith, 2007). The intracellular Ca2+ can no longer bind calmodulin to form a complex that activates the MLCK and causes myosin binding to actin, and the subsequent generation of uterine contractions (Smith, 2007).

In response to the mechanical stretch at the time of labor, myocytes establish physical and endocrine connections that promote coordinated and forceful uterine contractions. Tension development is achieved when actins convert into filamentous forms and attach to the underlying matrix via focal points in the cell membranes (Smith, 2007). An increase in gap junctions such as connexin (Cx)-43 permits the rapid transmission of action potentials and synchronous contractions over the entire uterus (Smith, 2007). With an increase in the expression of receptors for PGE and PGF, the myometrium is more responsive to PGs. The binding of PGs promotes the opening of ligand-gated calcium channels, which allow Ca2+ influx from extracellular space (Smith, 2007). Furthermore, the binding of OT to the OTR activates phospholipase C and inositol triphosphate (IP3) (Smith, 2007). IP3 subsequently promotes Ca2+ release from SR. An increase in intracellular Ca2+ concentration and a decrease in K+ efflux due to reduced expressions of β2 and β3-sympathomimetic receptors at labor, reduce the intracellular electronegativity and lead to depolarization (Smith, 2007). The intracellular Ca2+ forms a complex with calmodulin and activates the MLCK (Smith, 2007). Subsequently, the MLCK phosphorylates the myosin light chain and promotes ATPase activity, which enables myosin binding to actin and the development of uterine contractions (Smith, 2007).

Increased mechanical stretch also causes an increase in PGs and inflammatory cytokines, which in turn lead to the enhanced expression of PR-A (Jiang et al., 2012). PR-A, which lacks an N-terminal-activating domain, represses the activation and activity of some PR-B dependent genes (Madsen et al., 2004). The functional progesterone withdrawal removes the inhibition on estrogen and PGs, and increases their bioavailability to induce uterine contraction (Mesiano et al., 2002).

Figure 1-2 Proposed mechanisms that underlie relaxation and contraction of the myometrium during pregnancy or labor. Reproduced with permission from The New England Journal of Medicine: Smith R (2007) Parturition. 356: 271-283. Copyright Massachusetts Medical Society.

B. Hypothalamic-pituitary-adrenal (HPA) axis

High levels of circulating CRH produced by the periventricular nucleus in the fetal hypothalamus and the placenta are associated with the timing of labor (Voltolini and Petraglia, 2014). Furthermore, the level of circulating binding protein for CRH (CRHBP) levels falls, increasing the bioavailability of CRH (Voltolini and Petraglia, 2014). Binding of CRH to transmembrane G protein-coupled CRH type 1 receptor activates the fetal HPA axis and stimulates the fetal anterior pituitary to produce adrenocorticotrophic hormone (ACTH) (Voltolini and Petraglia, 2014) (Figure 1-3, page 14). ACTH causes the fetal adrenal gland to release the glucocorticoid cortisol (Voltolini and Petraglia, 2014). The placenta regulates the bioavailability of cortisol. During pregnancy, cortisol is converted into inactive cortisone by placental 11β-Hydroxysteroid dehydrogenase 2 (11β-HSD2), and at labor, inactive cortisone is converted into cortisol by placental 11β-HSD1 (Challis et al., 2000). In response to CRH, the fetal adrenal gland also produces dehydroepiandrosterone sulphate (DHEAS), which is an important substrate for placental estrogen synthesis (Voltolini and Petraglia, 2014). Elevated fetal cortisol increases the production of surfactant protein A (SP-A) and phospholipids, which stimulate fetal lung maturation. Furthermore, SP-A released into the amniotic fluid activates macrophages and stimulates the production of inflammatory mediators in the adjacent fetal membranes, which eventually lead to parturition (Smith, 2007). Fetal cortisol and CRH potentiate myometrial contractions by increasing the expression of PG receptors (Smith, 2007). Furthermore, fetal cortisol and CRH increase the synthesis of PGs by prostaglandin endoperoxide H synthases (PTGS) or cyclo-oxygenase (COX)-2 expressed in the amnion and chorion, and decrease the metabolism of PGs by prostaglandin dehydrogenase (PGDH) expressed in the chorionic trophoblast cells (Smith, 2007). In turn, PGs increase cortisol by upregulating placental 11β-HSD1 and downregulating 11β-HSD2 (Challis et al., 2000). Furthermore, CRH also stimulates the secretion of placental matrix metalloproteinase (MMP)-9, which contributes to fetal membrane rupture and cervical dilatation (Li and Challis, 2005).

1.2.1 Prostaglandins (PGs)

The production of prostaglandins (PG), comprised of 20-carbon chain unsaturated fatty acids, starts with phospholipase A2 (PLA2) cleaving the membrane phospholipids to release unesterified arachodonic acids (AA) (Keelan et al., 2003). Through the action of PTGS, AAs are converted into endoperoxide products, which are ultimately converted into primary PGs

(PGE2, PGF2α, PGD2 and prostacyclin/PGI2) through a series of isomerase reactions (Keelan et al., 2003).

The synthesis of PGs is regulated by the activity of constitutively expressed COX 1, inducible COX2 and PG synthases, while the metabolism of PGs is regulated by PGDH (Keelan et al., 2003). During pregnancy, high levels of 15-hydroxyprostaglandin dehydrogenase metabolize PGs in the chorion, decidua, placenta, myometrium and cervix, and maintain pregnancy quiescence (Olson and Ammann, 2007; Giannoulias et al., 2002). The expression of 15-hydroxyprostaglandin dehydrogenase diminishes in the chorionic trophoblast cells with the onset of parturition (Olson and Ammann, 2007), exposing the decidua, cervix and myometrium to PGE2. Concomitantly, an increase in the expression of COX2 in response to inflammatory stimuli in the amnion, choriodecidua and myometrium (Slater et al., 1999a; Slater et al., 1999b), and an increase in the expression of microsomal PGE synthases in the myometrium (Astle et al., 2007), lead to increased production of PGs. Subsequently, elevated PGs either directly promote myometrial contractility or through the stimulation of MMPs, cause fetal membrane rupture, cervical ripening and placental detachment (Olson and Ammann, 2007).

1.2.2 Matrix Metalloproteinase (MMPs)

MMPs are zinc-dependent enzymes that catalyze the degradation of collagen constituted- extracellular matrix of the cervix, fetal membranes, placenta and the uterus (Olgun and Reznik, 2010). MMPs are involved in normal parturition as well as in infection-triggered rupture of fetal membranes and preterm birth (PTB) (Maymon et al., 2001; Olgun et al., 2010). An increase in the amniotic fluid level of MMP-3 is associated with term and preterm

13 parturition, and with microbial invasion of the amniotic cavity (Park et al., 2003). Mid- trimester elevation of amniotic fluid MMP-8 is a risk factor for early spontaneous preterm delivery (PTB) less than 32 weeks, and MMP-8 at a level higher than >23 ng/mL predicts imminent PTB (Yoon et al., 2001). The non-specific MMP inhibitor, GM6001 reduces endotoxin induced PTB in the mouse, suggesting that one or more MMPs are critical in the pathogenesis of infection-associated PTB (Koscica et al., 2007).

Not in labor Labor

Amniochorionic) )membrane) PG

DHEA-S Cortisol E2

Adrenal

Lung CRH ACTH

Pituitary

PG PG

Fetal Myometrial Membrane contractions rupture MMP

PGDH Positive Feedback Cervical Pro-inflammatory Dilatation PGHS-2 Cytokines/Chemokines Anti-inflammatory Connexin-43 Cytokines

Immune cells

Figure 1-3 The proposed pathway of human parturition. During pregnancy (not in labor), high levels of prostaglandin dehydrogenase (PDGH) metabolize prostaglandins (PGs) and maintain pregnancy quiescence (Smith, 2007). At the time of labor, the expression of PGDH diminishes while the expression of PGHS-2 increases in response to elevated pro-inflammatory cytokines, exposing the intrauterine tissues to increasing levels of PGs. Placental corticotropin-releasing hormone (CRH) activates the fetal HPA axis and stimulates the fetal anterior pituitary to produce adrenocorticotrophic hormone (ACTH). ACTH causes the fetal adrenal gland to release the glucocorticoid cortisol. Furthermore, in response to CRH, the fetal adrenal gland produces dehydroepiandrosterone sulphate (DHEA-S), an important substrate for estrogen (E2) synthesis. Fetal cortisol and CRH increase the expression of PGs. Elevated PG promotes myometrial contractility, increases the expression of gap junctions (connexin-43), stimulates the expressions of matrix metalloproteinase (MMPs) and pro-inflammatory cytokines. Many positive feed-forward cascades underlie the process of parturition. The image is modified with permission from The New England Journal of Medicine: Goldenberg RL, Hauth JC, Andrews WW. (2000) Intrauterine infection and preterm delivery. 342 (20):1500-7. Copyright Massachusetts Medical Society.

1.2.3 Cytokines and Chemokines

Cytokines are small soluble proteins that function as signaling molecules in a paracrine or autocrine fashion, and are produced mainly by activated immune cells in the presence of antigens, microbial or viral products (Christiaens et al., 2008). The binding affinity between cytokines and their receptors is usually high (Km = 1010 - 1012 M-1); therefore, very low concentrations of cytokines (usually in picomolar) are sufficient to elicit a physiological change (Mak, 2006). Cytokines are hydrophilic and bind to cell surface receptors to initiate downstream intracellular signaling, which leads to altered cell functions (Mak, 2006).

Cytokines regulate the innate response, the adaptive response, and the growth and differentiation of hematopoietic cells (Mak, 2006). One cytokine can cross-regulate other cytokine(s) and/or their receptors in either an agonistic or an antagonistic fashion (Mak, 2006). This agonistic relationship creates a cascade of myometrial receptivity and a coordinated action responsible for increased myometrial contractility during labor (Christiaens et al., 2008). Anti-inflammatory cytokines have been observed to repress the over-expression of pro-inflammatory cytokines (Christiaens et al., 2008).

Cytokines may be classified based on either their structure motifs or their physiological functions. Structurally, interleukin (IL)-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12, IL-13, IL-15 and Interferon (IFN)-γ belong to the 4α helix family. CCL2, CCL3, CCL4, CCL5 and CCL11 share the CC chemokine motif, while IL-8 and CXCL10 possess the CXC chemokine motif (Mak, 2006). Alternatively, cytokines can be classified based on their functions as pro-inflammatory cytokines, anti-inflammatory cytokines, chemokines or growth factors (Mak, 2006). In this dissertation, cytokines are discussed based on their general functions and their roles in pregnancy are outlined below.

The implantation of a blastocyst and placentation in early pregnancy are predominantly pro-inflammatory processes (Dekel et al., 2014). As pregnancy progresses, the expression of pro-inflammatory cytokines is inhibited by anti-inflammatory cytokines, and the intrauterine environment switches to predominantly anti-inflammatory at the feto-

16 maternal interface (Challis et al., 2009). At the time of labor, a pro-inflammatory milieu is predominant which promotes uterine contractions through the interaction with the PG- signaling pathway (Challis et al., 2009). At each stage of pregnancy, inflammation is tightly controlled. Excessive inflammatory responses can lead to adverse pregnancy outcomes, including PTB, spontaneous abortion, fetal growth restriction, and hypertensive disorders (Challis et al., 2009).

A. Pro-Inflammatory Cytokines

Interleukin-1 (IL-1) and IL-1 receptor antagonist IL-1 is produced by macrophages, neutrophils, epithelial cells and endothelial cells (Mak, 2006). Of the two isoforms of IL-1 (IL-1α and IL-1β), IL-1β is present in greater abundance (Mak, 2006). IL-1β has been associated with the process of implantation, decidualization and labor (Geisert et al., 2012) and is detected in the culture-conditioned media of preimplantation human embryo (Baranao et al., 1997). Women who experience habitual abortion have decreased expressions of IL-1β and IL-6 in the endometrium (von Wolff et al., 2000). In the presence of steroid hormone, IL-1β induces the expression of Insulin like growth factor binding protein 1, a marker of decidualization, in the baboon stromal fibroblasts (Strakova et al., 2000). The output of cervico-vaginal IL-1β has been reported to increase with approaching term labor (Imai et al., 2001), although a recent study did not observe such an increase (Heng et al., 2014a).

IL-1 receptor antagonist (IL-1ra) binds to both IL-1 receptor type I and II, but does not lead to a signal transduction; therefore, IL-1ra serves as a competitive inhibitor that limits IL-1- induced inflammation (Arend et al., 1998). The bioavailability of IL-1ra is several thousand times higher than IL-1, and the cervico-vaginal output of IL-1ra decreases significantly with impending term labor (Heng et al., 2014a).

Interleukin-2 (IL-2) IL-2, produced primarily by T-helper lymphocytes type 1 (Th1), promotes the growth and differentiation of lymphocytes, macrophages and oligodendrocytes (Mak, 2006). The role of

IL-2 in pregnancy and parturition is still unclear. IL-2 has been reported to inhibit IL-1β- induced PGE2 production in human amnion cells, and in cultured chorion and decidua cells (Coulan et al., 1993a, 1993b). The role of IL-2 in labor in unknown.

Interleukin-6 (IL-6) IL-6, which belongs to the family of gp130 cytokines, is produced by macrophages, T-cells, mononuclear phagocytes, vascular endothelial cells and intra-uterine tissues (Mak, 2006). IL- 6 plays a role in acute phase reactions, hematopoiesis, differentiation and maturation of immune cells (B cells, T cells and macrophages) (Mak, 2006). In pregnancy, IL-6 has been suggested to be important in implantation, placentation and labor (Markert et al., 2011). Mice deficient in IL-6 have reduced fertility and a reduced number of viable implantation sites (Robertson et al., 2000). High levels of IL-6 have been detected in the invasive cytotropblast cells (Das et al., 2002), and IL-6 has been shown to activate MMPs in the trophoblast (Meisser et al., 1999), suggesting it might play a role in the process of trophoblast invasion. High concentrations of IL-6 have been observed in the maternal plasma and amniotic fluid of women in labor (Unal et al., 2011). Elevated maternal plasma IL-6, secreted in a pulsatile fashion, is also associated with increased uterine contractility during the active phase of labor (Papatheodorou et al., 2013).

Interleukin-12 (IL-12) IL-12 (or IL-12p70), a 70 kD heterodimer composed of p35 and p40 peptides, is produced primarily by monocytes and macrophages (Mak, 2006). IL-12 is crucial for the differentiation of Th0 cells into Th1 cells, which are capable of generating IFN-γ (Mak, 2006). IL-12 has been shown to limit trophoblast invasion by down regulating the expression of MMPs and up-regulating their inhibitors, tissue inhibitor of metalloproteinase-1 in JEG-3 cells, possibly through the production of IFN-γ (Karmakar et al., 2004). In addition, IL-12 enhances the cytotoxicity of CD8+ T cells and NK cells, which play important roles in cell- mediated immune response against potential pathogens (Freeman et al., 2012). Pregnant women with an elevated plasma concentration of IL-12 and low plasma IL-18 have an increased risk of PTL (Ekelund et al., 2008).

Interleukin-17 (IL-17) IL-17, a glycosylated homodimeric polypeptide, is produced by memory CD4+ T cells in the peripheral blood, decidua and placenta (Nakashima et al., 2010; Pongcharoen et al., 2007). IL-17 stimulates the production of IL-6, IL-8 and colony stimulating factor (CSF) 2 in fibroblasts and endothelial cells (Mak, 2006). IL-17 also promotes the processes of trophoblast invasion and angiogenesis, both of which are important in the establishment of placental vasculature (Pongcharoen et al., 2006). The plasma levels of IL-17 increase in third trimester healthy pregnant women (Martinez-Garcia et al., 2011).

Interferon-gamma (IFN-γ) IFN-γ, a type II interferon, is produced by mitogen-activated lymphocytes (Micallef et al., 2014). It possesses anti-pathogenic and anti-proliferative properties (Mak, 2006). It promotes pathogen elimination, possibly through the activation of macrophages and the subsequent production of TNFα and IL-12 (Mak, 2006). IFN-γ, derived from dNK cells, inhibits trophoblast cell growth and invasion in the mouse (Ain et al., 2003). In addition, IFN-γ is important in the process of extravillous trophoblast invasion into human first trimester decidua (Lockwood et al., 2014). However, high levels of IFN-γ have been associated with miscarriage and inhibition of angiogenesis (Micallef et al., 2014). IFN-γ has been shown to reduce the expression of COX-2 and the production of PGE2 in term and preterm placenta, in keeping with functional withdrawal of IFN-γ being involved in labor (Hanna et al., 2004).

Tumor Necrosis Factor alpha (TNFα) TNFα, initially produced as a transmembrane prohormone, is activated after the N-terminal 76 amino acids are proteolytically cleaved by TNF convertase (Mak, 2006). TNFα is produced by macrophages, lymphocytes, fibroblasts, neutrophils, endothelial cells and intrauterine issues (Mak, 2006). TNFα binds to TNF receptor-1 and 2, and is involved in the activation of cell death, cell proliferation and inflammation (Mak, 2006). TNFα is a major pro-inflammatory cytokine that underlies the inflammatory process leading to the initiation of labor (Christiaens et al., 2008). The concentration of TNFα in the amniotic fluid remains low throughout human pregnancy and sharply increases at term labor (Hayashi et al., 2008). In addition, TNFα induces the production of PGE2 in cultured human chorion, amnion and

19 decidual cells, as well as the expression of MMPs in cultured human chorion, myometrium and cervical smooth muscle cells (Christiaens et al., 2008).

Mice lacking the genes for pro-inflammatory cytokine IL-6 have delayed parturition, whereas mice with receptors for IL-1 and TNFα knockout are less susceptible to bacterially induced PTL (Robertson et al., 2010; Hirsch et al, 2006). These findings suggest IL-1, IL-6 and TNFα are important in the pathogenesis of inflammation-associated labor.

B. Anti-inflammatory Cytokines

Interleukin-4 (IL-4) IL-4, also known as B-cell activating factor-1, is produced by T helper lymphocytes Type 2 (Th2), mast cells, basophils, eosinophils and intrauterine tissues (Mak, 2006). IL-4 is important in the differentiation and activation of B cells, and the differentiation of naive Th0 cells into Th2 cells (Chatterjee et al., 2014). The production of IL-4 increases in the peripheral blood mononuclear cells throughout pregnancy and low levels of IL-4 have been suggested to contribute to higher incidences of infertility, spontaneous abortion, PTB, and preeclampsia (Chatterjee et al., 2014). IL-4 antagonizes the production of IL-1β, TNFα and PGE2 by human peritoneal macrophages (Hart et al., 1991).

Interleukin-5 (IL-5) IL-5, also known as eosinophil differentiation factor, is a homodimeric cytokine produced by eosinophils, Th2 cells, NK cells and mast cells (Mak, 2006). IL-5 promotes the survival, differentiation and activation of eosinophils (Mak, 2006). The role of IL-5 in pregnancy and parturition is unknown.

Interleukin-9 (IL-9) IL-9, produced by Th2 cells, plays a role in T lymphocyte proliferation and hematopoiesis (Mak, 2006). The role of IL-9 in pregnancy and parturition is unknown.

Interleukin-10 (IL-10) IL-10, produced by lymphocytes, macrophages, dendritic cells, placental and decidual mononuclear cells, is an important anti-inflammatory cytokine that contributes to uterine quiescence during pregnancy (Cheng and Sharma, 2014). IL-10 inhibits the production of many pro-inflammatory cytokines by inhibiting the Nuclear Factor-Kappa B (NF-κB) signalling pathway and activating the Janus Kinases and Signal Transducers and Activators of Transcription (JAK-STAT) and Phosphatidylinositol-3kinase (PI3K-Akt) signalling pathways (Cheng and Sharma, 2014). IL-10 also blocks the expression of major histocompatibility complex (MHC) class II and confers immune tolerance (Cheng and Sharma, 2014). The placental expression of IL-10 is significantly reduced around the time of labor (Cheng and Sharma, 2014).

Interleukin-13 (IL-13) IL-13, produced by Th2 cells, shares 30% sequence homology with IL-4 and therefore shares similar anti-inflammatory properties (Mak, 2006). The concentration of IL-13 has been detected in first trimester human trophoblast cells (Naruse et al., 2010). In human amnion- derived WISH (Wistar Institute, Susan Hayflick) cells, IL-13 inhibits the production of IL-8 and PGE2 (Keelan and Mitchell, 1998).

Interleukin-15 (IL-15) IL-15 is produced by activated monocytes of human intrauterine tissues (Mak, 2006). IL-15 stimulates the growth of NK cells and activates peripheral blood T lymphocytes (Mak, 2006), and IL-15 also stimulates the production of angiogenic factors such as IFN-γ in decidual NK cells (Murphy et al., 2009). Increased expression of IL-15 in human decidua has been associated with recurrent miscarriage in women (Toth et al., 2010). The amniotic fluid concentration of IL-15 is higher in pregnant women in the third trimester compared to the second trimester (Klimkiewicz et al., 2012), and elevated IL-15 produced by human fetal membranes has been found in in women who delivered preterm (Fortunato et al., 1998).

Among the anti-inflammatory cytokines, IL-10 is thought to be a key anti-inflammatory modulator of labor. Exogenous administration of IL-10 to mice deficient in the IL-10 gene reduces the incidence of PTB (Robertson et al., 2006).

C. Chemokines

Chemokines are a subclass of cytokines that stimulate the migration and activation of immune cells. Chemokines are classified into different groups based on the conserved cysteine residues near the N terminus: (1) CC chemokines (CCL2, CCL3, CCL4, CCL5 and CCL11), and (2) CXC chemokines (CXCL8 and CXCL10) (Mak, 2006).

CXCL8 CXCL8, also known as IL-8, is produced by monocytes, lymphocytes, fibroblasts, epithelial cells and endothelial cells in response to stimulation by pro-inflammatory cytokines. CXCL8 recruits and activates primarily neutrophils, basophils and T cells (Mak, 2006). CXCL8 activates neutrophils to generate reactive oxygen radicals, which can lead to tissue damage. In mice, CXCL8 is not present; instead, keratinocyte chemo-attractant (KC) recruits neutrophils (Mak, 2006). The expression of CXCL8 increases in human myometrium, choriodecidua and amnion, in association with labor (Elliott et al., 2000; Elliott et al., 2001)

CXCL10 CXCL10, also known as IFN-γ inducible protein 10 (IP-10), is produced upon stimulation with IFN-γ (Mak, 2006). CXCL10 is a chemo-attractant for activated T cells (Mak, 2006). The release of CXCL10 by dNK cells has been associated with the process of tissue building and remodelling of the blood vessels during early pregnancy (Vacca et al., 2013). Elevated expressions of CXCL10 mRNA and protein have been observed in choriodecidua from women in term labor (Hamilton et al., 2013).

CCL2 CCL2, also known as monocyte chemotactic protein-1 (MCP-1), is produced by monocytes, endothelial cells and intrauterine tissues (Mak, 2006). CCL2 is responsible for the recruitment of monocytes and their differentiation into macrophages (Mak, 2006). CCL2 stimulates the production of several pro-inflammatory cytokines, and the expression of CCL2 increases in the myometrium in pregnant women at term labor (Esplin et al., 2005).

CCL3 and CCL4 CCL3 and CCL4, also known as macrophage inflammatory protein (MIP)-1α and -1β respectively, are produced by macrophages, dendritic cells and lymphocytes (Mak, 2006). CCL3 and CCL4 recruit granulocytes and activate neutrophils, eosinophils and basophils (Mak, 2006). Both CCL3 and CCL4 can induce the synthesis and release of pro- inflammatory cytokines IL-1, IL-6 and TNFα from activated macrophages (Mak, 2006) CCL4 also promotes the migration of trophoblast cells (Hannan et al., 2006). The concentration of CCL3 in the amniotic fluid (Dudley et al., 1996) and the concentrations of CCL3 and CCL4 in the decidual leukocytes increase in pregnant women at term labor (Hamilton et al., 2013).

CCL5 CCL5, also known as Regulated upon activation, normal T-cell expressed and secreted (RANTES), is a chemotactic factor for T cells, eosinophils, basophils and lymphocytes (Mak, 2006) The myometrial concentration of CCL5 is down-regulated in women with prolonged pregnancy compared to women who delivered at term (Pabona et al., 2014). CCL5 is a pro-implantation factor as it increases regulatory T lymphocytes, favors the survival of trophoblast cells, confers maternal tolerance of fetal allograft and induces apoptosis of maternally activated T cells (Perez and Ramhorst, 2013). The role of CCL11 in labor is unknown.

CCL11 CCL11, also known as Eotaxin, recruits eosinophils and stimulates the migration of the extra-villous trophoblast (EVT) cells (Mak, 2006). The invasion by EVT cells into the maternal uterine decidual vessels is important to establish adequate placental blood flow to the fetus (Chau et al., 2013). The role of CCL11 in labor is unknown.

Chemokines are known to stimulate recruited immune cells to produce pro-inflammatory cytokines, which further amplify inflammatory responses in labor (Christiaens et al., 2008). IL-8, CXCL10, CCL2, CCL3 and CCL4 are elevated in the intrauterine tissues and/or amniotic fluid of women in labor, whereas the role of CCL5 and CCL11 in labor is unclear.

D. Growth Factors

Interleukin-3 (IL-3) IL-3, produced by T lymphocytes, mast cells, eosinophils, neurons and astrocytes, promotes growth and maturation of the hematopoietic progenitor cells into all cell types (Mak, 2006). IL-3 also recruits mature basophils in allergic reaction and promotes the differentiation and invasiveness of human trophoblast cells (Di Simone et al., 2000). The role of IL-3 in the labor process is unknown.

Colony Stimulated Factor 2 (CSF2) and CSF3 CSF is secreted by activated T cells, macrophages, mast cells, NK cells, stromal cells, endothelial cells and placental cells (Mak, 2006). CSF induces the proliferation and differentiation of hematopoietic stem cells into monocytes and granulocytes, including neutrophils, basophils and eosinophils (Mak, 2006). CSF2 acts on the bone marrow to increase the generation of hematopoietic precursor cells, and stimulates them to differentiate into granulocytes and monocytes (Mak, 2006). Both CSF2 and CSF3 play an important role in early pregnancy by promoting normal embryonic development, successful implantation and normal placentation (Robertson, 2007b; Furmento et al., 2014). CSF2 does not appear to contribute to the labor process since the level of CSF2 in the amniotic fluid is not different between women in term labor and those not in labor (Hayashi et al., 2006). In contrast, an increase in the concentration of CSF3 has been detected in the cervix of women during labor, suggesting a role of CSF3 in cervical remodeling (Sennstrom et al., 2000).

FGF basic FGF, expressed in the human placenta, is a potent inducer of angiogenesis (Mak, 2006). Angiogenesis is important for normal implantation and placentation. FGF also plays a role in the proliferation, differentiation, migration and invasion of human placental trophoblast cells (Anteby et al., 2005). The role of FGF-b in labor remains unknown.

Platelet Derived Growth Factor-bb (PDGF-bb) PDGF-bb, a pro-angiogenic factor produced primarily by platelets, regulates cell growth and division (Mak, 2006). PDGF-bb is important in the growth of uterine smooth muscle cells, as

24 well as vascular remodeling during pregnancy (Keyes et al., 1996). PDGF-bb has been suggested to be important in the migration of endometrial stromal cells, which is important to the implantation process (Schwenke et al., 2013). The role of PDGF-bb in labor remains unknown.

Vascular Endothelial Growth Factor (VEGF) VEGF, a pro-angiogenic factor, is ubiquitously expressed in vascularized organs like the placenta (Mak, 2006). VEGF stimulates the differentiation, proliferation and migration of endothelial cells (Mak, 2006). In addition, VEGF is important to decidual growth and the extravasation of white blood cells into the decidua, which subsequently contributes to the inflammatory process leading to term labor (Elfayomy and Almasry, 2014).

The role of growth factors in the labor process remains to be elucidated.

1.3 Preterm Birth

1.3.1 Epidemiology

Human preterm birth (PTB), defined as delivery prior to 37 weeks of gestation, is observed in approximately one in every ten pregnancies globally (Blencowe et al., 2012). Infants born preterm have a mortality rate 40 times higher than term infants; moreover, premature babies are at a greater risk of suffering from long-term health problems including cerebral palsy and respiratory disorders (Oskoui et al., 2013; Brostrom et al., 2013). Raising a functionally impaired premature infant places both emotional stress on parents and financial burden on society. The cost of neonatal intensive care has been estimated to be at least $26.2 billion in 2005 in the United States (Behrman, 2007). Hospital inpatient admissions cost for children born very premature (<28 weeks of gestation) during the first 10 years of their life is 20 times greater than for those born at term (Petrou, 2005). Obstetric interventions and the use of assisted reproduction techniques account for the rise in PTB. Despite advances in the healthcare system, PTB incidence has not decreased.

1.3.2 Etiology

Risk factors that contribute to PTB include physiological aspects such as a previous history of PTB, short cervical length, carrying a male fetus, the overall poor health status of the mother, advanced maternal age, lower body weight, being a smoker and socioeconomic factors such as educational background, social class, and race (Goldenberg et al., 2008a). The etiology of PTB is multifactorial and largely unknown: 50% of the cases are idiopathic while 20-40% are iatrogenic, where the presence of conditions such as pre-clampsia or intrauterine growth restriction (IUGR) requires delivery. The remaining 25-30% can be attributed to intrauterine infection and/or inflammation (Goldenberg et al., 2008a).

1.3.3 Infection Routes

The uterine environment during pregnancy is not sterile, and microorganisms can invade the uterus through the fallopian tube in a retrograde fashion from the abdominal cavity, haematogeneously via the placenta and most commonly, ascending through the cervix and vagina (Goldenberg et al., 2000). It has been proposed that once microorganisms reach the maternal intrauterine tissues, they can secrete phospholipase A2 to act on membrane phospholipids and through a series of catalytic reactions, primary PGs are formed (discussed in Section 1.2.1). Bacterial endotoxin such as lipopolysaccharides (LPS) found on the outer membrane of Gram-negative bacteria can stimulate PG production (Timmons et al., 2014). Binding of LPS to Toll-like receptor 4 (TLR4), a specific pattern recognition receptor, activates the NFкB pathway to induce an increase in pro-inflammatory cytokine and chemokine gene expression in intrauterine tissues (amnion, chorion and decidua), macrophages and endothelial cells. These inflammatory mediators in turn increase uterine contractility by either directly upregulating PG production, or indirectly via altering levels of enzymes involved in PG biosynthetic pathways such as increasing PTGS-2 in amnion and decidual stroma cells, and decreasing PGDH in chorion trophoblast cells (Smith, 2007). Pro- inflammatory cytokines stimulate each other as well as PG in a feed-forward cascade, such that they stimulate and accelerate the production of each other, hence amplifying the

26 inflammatory response. Furthermore, pro-inflammatory cytokines enhance the expression of MMPs, leading to fetal membrane rupture and cervical dilatation (Smith, 2007).

Fetal responses to infection and/or inflammation also play a role in PTL initiation. Microorganisms can cross an intact chorioamniotic membrane and create intra-amniotic inflammation, a condition termed the Fetal Inflammatory Response Syndrome (FIRS) (Gotsch et al., 2007). Elevated IL-6 has been observed in the umbilical cord blood in preterm neonates who had FIRS (Buhimschi et al., 2009). A recent study in asymptomatic women with PPROM has found the umbilical cord blood level of lipopolysaccharide (LPS)-binding protein (LBP), which can bind to plasma LPS, was significantly higher in preterm neonates who had FIRS (Pavcnik-Arnol et al., 2014). Pathogenic microorganisms such as Ureaplasma urealyticum and Mycoplasma hominis have been isolated from the umbilical cord blood of very preterm newborns (Goldenberg et al., 2008b). Intrauterine infection has also been associated with activation of the fetal hypothalamic-pituitary-adrenal (HPA) axis, increased cortisol biosynthesis and decreased cortisol metabolism to inactive cortisone by 11β-HSD2 in the placenta (Gravett et al., 2000). Together, sustained stimulation of fetal cortisol on placental CRH increases PG production, which in turn promotes uterine contractility and PTL (Voltolini and Petraglia, 2014).

1.3.4 Infection and/or Inflammation- induced PTB

The etiology of PTB is multifactorial, with inflammation during pregnancy being one of its causes. The predominance of pro-inflammatory cytokines has been proposed to be responsible for the early onset of labor or PTL (Challis et al., 2009). The inflammatory cascade is further amplified by an increase in the expression of chemokines, which attract decidua leukocytes to produce additional pro-inflammatory cytokines (Hamilton et al., 2013).

The production of various cytokines has been studied in the amniotic fluid, cervico-vaginal secretions and maternal plasma. Levels of IL-1β, IL-6, IL-8 and TNFα are elevated in amniotic fluid and cervical fluid of women at risk of PTL, especially those with intra- amniotic infection (El-Bastawissi et al., 2000; Hitti et al., 2001; Jun et al., 2000; von

Mincwitz et al., 2000; Holst et al., 2011). CCL3, CCL4, CCL5 in the amniotic fluid and CCL2 in the cervical fluid are also significantly higher in PTL women with microbial invasion of the amniotic cavity (Holst et al., 2011). Elevated levels of IL-6 are found in the amniotic fluid and umbilical vein of infants born to mothers with chorioamnionitis (Holst et al., 2011; Chaiworapongsa et al., 2002).

High levels of circulating plasma IL-1β, IL-6 and IL-8 have been observed in women with PPROM in the presence of chorioamnionitis at 22-36 weeks gestation (von Minckwitz et al., 2000). Increased levels of plasma TNF-α, IL-12 and IL-18 are also detected in women at risk of recurrent spontaneous PTB (Vogel et al., 2007). However, recent studies have found IL-6 in the amniotic fluid and cervico-vaginal fluid, but not in plasma, are associated with spontaneous PTB (Wei et al., 2010). It appears the presence of cytokines and chemokines at the maternal–fetal interface, including intrauterine tissues, amniotic fluid or cervico-vaginal fluid, are more representative of the pathology of PTB than are levels in maternal plasma.

Anti-inflammatory cytokines maintain pregnancy quiescence by inhibiting the production of pro-inflammatory cytokines and PGs (Challis et al., 2009). IL-10 expression in the placenta is lower in women who give birth preterm with chorioamnionitis compared to samples obtained from women who underwent elective terminations in their second trimester of pregnancy (Hanna et al., 2006). The same finding has been observed in women in term labor with chorioamnionitis compared to women at term not in labor (Hanna et al., 2006). Mid- trimester amniotic fluid concentrations of IL-10 are not different between preterm and term delivery (Puchner et al., 2011), while cervico-vaginal levels of IL-4 and IL-10 are often below the level of detection using current assays (Vogel et al., 2007). There is an association between elevated plasma IL-10 with an increased risk of preeclampsia or intrauterine growth restriction (Ferguson et al., 2014). Overall, the positive and negative predictive values of any single specific cytokine or chemokine for PTB is limited (Menon et al., 2014) although the examination of interactions with a multifactor dimensionality reduction analysis between multiple cytokines within maternal–fetal compartments, rather than a single cytokine, may better predict the risk of PTB (Bhat et al., 2014). Other factors that need to be taken into account when analyzing cytokine profiles include ethnicity of the study population, maternal

28 body mass index (BMI), a previous history of PTB, whether anti-inflammatory medications were taken and psychological status (Velez et al., 2008; Cator et al., 2014). For instance, compared to women who deliver at term, amniotic fluid levels of IL-1β and TNF-α were higher in African American women, but not in Caucasian women, who delivered preterm (Velez et al., 2008).

1.3.5 Current Treatment Approaches

A non-invasive diagnostic test with a high positive predictive value and a high negative predicative value is needed to differentiate between true and false PTL. Recently, Heng et al discovered that a set of nine genes, together with maternal clinical data, could accurately predict whether 70% of participants would or would not have a spontaneous PTB within 48 hours of hospital admission. This method for the diagnosis of PTL outperformed the traditional fetal fibronectin test (Heng et al., 2014b).

The efficacy and safety of interventions to prevent PTB are largely unsatisfactory. The efficacy of drugs that act as antagonists or inhibitors of oxytocin receptors, PGHS-2, prostaglandin PTGFR receptors, or phosphodiesterase (PDE4) are yet to be determined (Papatsonis et al., 2013; Lopez et al., 2007). Though effective, the safety of treatment with progesterone and progestational agents remains unclear (Jayasooriya and Lamont, 2009; O’Brien and Lewis., 2009). Antibiotics have been proposed to prevent infection-mediated PTB; however, antibiotic treatment has limited success at preventing PTB, yielding mixed results (Subramaniam et al., 2012; Oliver and Lamont, 2013). In practice, the use of antibiotics to reduce PTB is limited to women with abnormal genital tract biota and administration has to be early in pregnancy (< 22 completed weeks of gestation) before substantial inflammatory damage occurs (Oliver and Lamont, 2013). The use of metronidazole might even increase the incidence of PTB (Shennan et al., 2006). Current interventions of infection-mediated PTB are aimed at treatment rather than prevention for the management of PTB. Since intrauterine infection may remain asymptomatic until PTL or premature rupture of membranes (Goldenberg et al., 2000), safe and effective prophylactic intervention may be more appropriate.

1.3.6 Animal Models of Preterm Birth

Animal models are essential research tools for investigating pathways that promote preterm parturition and for testing potential therapeutic interventions. Mammalian animal models of PTB include mouse, rat and rabbit. In these species, involution of the corpus luteum and a subsequent decline in maternal plasma progesterone precedes the onset of labor, which is not observed in humans (Elovitz and Mrinalini, 2004). Similarly, in the sheep model of PTB, a decrease in progesterone production and an increase in estradiol production as a result of fetal cortisol-induced synthesis of placental enzymes eventually lead to parturition (Elovitz and Mrinalini, 2004). However, in non-human primates and in the human, a functional progesterone withdrawal precedes the process of parturition (Elovitz and Mrinalini, 2004).

The mouse has been extensively used to study human PTB because the mechanisms of murine parturition share many similarities with human parturition, including the pro-labor roles played by pro-inflammatory cytokines, chemokines, PGs and MMPs. In addition to being inexpensive, small in size, having a short gestational period (19-20 days) and the ability to tolerate surgery, the mouse confers advantages such as the possibility of genetic manipulation to help discern the pathways involved in parturition (Elovitz and Mrinalini, 2004). Genetically manipulated mice that are deficient in key genes promote parturition defects. By studying these mice, some of the pathways involved in parturition have been found to be redundant for term labor (Elovitz and Mrinalini, 2004). Elovitz et al (2003) developed an intrauterine approach for the investigation of LPS-induced PTL in CD-1 mice. A localized model of intrauterine inflammation or infection is clinically useful since most women with PTL do not display symptoms of systemic illness such as significant increases in white blood cell counts, C-reactive protein or temperature (Goldenberg et al., 2008a). A localized intrauterine infection/inflammation using intrauterine injection of LPS mimics more accurately what is most commonly observed in the human. In mice, this method results in high rates of preterm delivery with little or no maternal mortality (Elovitz et al., 2003).

1.4 Vaginal Microbiota and Preterm Birth

1.4.1 The human vaginal microbiota

The vaginal microbiota composition is dynamic throughout a woman’s life. Before puberty, it is dominated by anaerobic bacteria (Farage and Maibach, 2006). Rising estrogen levels at puberty lead to an increase in mucosal glycogen production whose metabolized substrates support vaginal colonization with lactobacilli (Spear et al., 2014). This is one reason the vagina is highly colonized by lactobacilli during the reproductive years and pregnancy (Romero et al., 2014a; Ravel et al., 2011). At menopause, lactobacilli abundance decreases coinciding with a reduction in circulating estrogen concentration (Gupta et al., 2006; Hummelen et al., 2011).

Lactobacilli are Gram-positive, facultative anaerobic bacteria, whose adherence to the vaginal mucosal epithelia appears to form an important line of defense against potential pathogens (Othman et al., 2007). In the vast majority of pregnant healthy women, lactobacilli dominate (Romero et al., 2014a; Aagaard et al., 2012). Several important aspects of the vaginal microbiota have been uncovered recently, particularly by sequencing PCR-amplified universal 16S ribosomal DNA (rDNA): (1) The healthy vaginal microbiota is dominated by a few Lactobacillus species (Lamont et al., 2011); (2) The detection of L. iners, Atopobium vaginae and BV-associated bacteria 1, 2 and 3 (BVAB), is apparent in women with BV (Lamont et al., 2011; Verstraelen et al., 2004; Fredricks et al., 2005).

The 16S ribosomal RNA gene is highly conserved in prokaryotic bacteria and is most widely targeted in vaginal microbiome studies (Romero et al., 2014a, Romero et al., 2014b, Gloor et al., 2010), although the cpn60 gene (Chaban et al., 2014) and the rpoB gene have also been studied (Vos et al., 2012). Various methods are available to identify bacteria using the 16S rRNA gene. These include denaturing gradient gel electrophoresis (DGGE), fluorescence in situ hybridization (FISH), terminal-restriction fragment length polymorphism (T-RFLP), quantitative polymerase chain reaction (qPCR) and microarray. However, these detection methods often target specific bacteria and do not provide sufficient resolution to characterize

31 microbial communities (Ling et al,. 2010). High-throughput sequencing technologies such as 454 pyrosequencing and Illumina sequencing provide greater sequencing depth for the identification of bacterial taxa and their relative abundance (Ling et al., 2010). There are nine hyper-variable regions (V1 to V9), separated by the conserved regions, in the 16S rRNA gene. Sequencing these short variable region(s) provides sufficient taxonomic information and allows identification to the species level. It has been shown that full-length sequencing missed 58% of the genera identified by V6 (Huse et al, 2008).

Several variable regions have been used in human vaginal microbiome studies, including V1- V2 (Romero et al., 2014a), V3-V5 (Walther-Antonio et al., 2014), and V6 (Gloor et al., 2010). There are both pros and cons to using each of the variable regions for the study of human vaginal microbiome. For microbiome studies in this thesis, I chose the V6 region as it provides high distinguishing power for Lactobacillus spp. in the vagina (Gloor et al., 2010), since one of my goals is to determine the relative abundance of vaginal Lactobacillus spp. after exogenous lactobacilli administration. However, it is important to recognize a limitation to using the V6 region such as the inability to detect Mycoplasma hominis, Ureaplasma parvum, and Ureaplasma urealyticum (Gloor et al., 2010).

Although relatively few 16S ribosomal DNA (rDNA) studies have been used with samples from pregnant women, indications are that the microbiota does fluctuate during this time. Some researchers have suggested that there are up to five different community state types (CSTs) of bacteria, clusters generated based on similarity in vaginal bacterial composition, in asymptomatic pregnant and non-pregnant women (Romero et al., 2014a; Ravel et al., 2011). Three of the CSTs (I, II, III) are dominated by Lactobacillus spp., namely L. iners, L. crispatus, or L. jensenii and/or L. gasseri. Two others, CST IV-A and CST IV-B have a low relative abundance of Lactobacillus spp. and are composed of species within the genera Peptoniphilus, Anaerococcus, Corynebacterium, Finegoldia and Prevotella (CST IV-A), and Atopobium, Sneathia, Gardnerella, Ruminococcaceae, Parvimonas and Mobiluncus (CST IV-B) (Romero et al., 2014a). Such studies have suggested that the vaginal microbiota composition of pregnant women has a higher abundance of L. vaginalis, L. crispatus, L.

32 gasseri and L. jensenii, but lower CST IV-B bacteria, and is more stable than non-pregnant women (Romero et al., 2014a; Walther-Antonio et al., 2014; Aagaard et al., 2012), with L. crispatus in particular, promoting stability (Verstraelen et al., 2009). This remains to be verified, but it may be due to hormonal changes. With advancing gestational age, the relative abundance of Lactobacillus spp. increases while that of anaerobic or strict-anaerobic microbial species decreases (Romero et al., 2014b).

1.4.2 Bacterial Vaginosis

BV is a polymicrobial dysbiosis, characterized by an alteration in the endogenous vaginal microbiota with an absent or decreased proportion of lactobacilli and dominance of G. vaginalis, Prevotella bivia, Mobiluncus spp., Mycoplasma hominis and A. vaginae (Schwebke et al., 2014; Ugwumadu, 2002). In many clinical units, the diagnosis of BV involves using a Gram stain Nugent scoring system with or without the Amsel criteria (a vaginal pH > 4.5, an amine fishy odour when vaginal fluid is mixed with potassium chloride, the presence of clue cells) (Nugent et al., 1991). A Nugent score of 7-10 is seen microscopically as a near absence of rod shaped lactobacilli and a high abundance of pathogenic morphotypes is considered BV (Nugent et al., 1991). Sequencing of the vaginal microbiota of women with BV reveals a diverse array of bacteria, including the presence of L. iners (Fredricks et al., 2005; Jackobsson and Forsum, 2007). Improvement in diagnostic accuracy for BV can be accomplished by using a DNA level of ≥109 copies/mL for G. vaginalis and ≥108 copies/mL for A. vaginae (Menard et al., 2008).

The prevalence of BV can vary between populations, but it remains common during pregnancy, where it is associated with a 40% increase in the risk of PTB (Ugwumuda, 2002). Women with an abnormal vaginal biota in their first trimester of pregnancy have a higher risk of delivering preterm (Donders et al., 2009). Although an earlier Cochrane Review (McDonald et al., 2007) suggested that antibiotic treatment of abnormal vaginal biota (intermediate biota or BV) before 20 weeks of gestation may reduce the risk of PTB, a recent Cochrane Review concluded that antibiotic treatment of BV does not reduce the risk of PTB,

33 regardless of when (before 20 weeks or after 20 weeks of gestation) the treatment is given (Brocklehurst et al., 2013). Some of these organisms possess sialidase activity, which has been associated with an increased risk of PTB (Smayevsky et al., 2001). Sialidases are hydrolytic enzymes that play a role in down-regulating the innate response by degrading immunoglobin-A (IgA), and it has been used in some diagnostic kits for this reason. Higher LPS concentrations, mostly from P. bivia (Aroutcheva et al., 2008), and the concentrations of pro-inflammatory cytokines IL-1β, IL-6 and IL-8 are elevated in the cervico-vaginal fluid of pregnant women with BV (Mitchell and Marrazzo, 2014).

In African American and Hispanic women, a higher abundance of Mycoplasma spp. and a lower abundance of BVAB3 is associated with an increased risk of PTB in the second trimester (Wen et al., 2014). This is unlikely due to race per se, but rather cultural and social influences. Other pathogens, such as Leptotrichia, Sneathia, BVAB1 and Mobiluncus spp appear in higher abundance prior to 16 weeks gestation in women with a previous history of PTB and who deliver preterm (Nelson et al., 2014). Yet, such findings are not universal, and other studies, albeit small, report no difference in the vaginal microbial composition between women who have a spontaneous PTB and those who deliver at term (Romero et al., 2014b; Hyman et al., 2014).

1.5 Probiotics

Probiotics are defined as "live microorganisms which when administered in adequate amounts, confer a health benefit on the host" (FAO/WHO, 2001). A number of meta- analyses of clinical trials with probiotics have confirmed that probiotics are both safe and effective for the treatment and/or prevention of numerous infectious and/or inflammatory diseases (Goldenberg et al., 2013; Yang et al., 2014a; Grin et al., 2013). Lactobacillus and Bifidobacterium are the most commonly studied probiotics. Bifodobacteria are present in intestinal biota, but they can also be detected in the vagina. Lactobacilli play a potential beneficial role in human reproduction and maintenance of healthy urinary and reproductive tracts (Reid et al., 2015).

Probiotics are used to treat many gastrointestinal diseases including necrotizing enterocolitis and ulcerative colitis. A systematic review of randomized, controlled trials reported a decrease in the incidence of necrotizing enterocolitis with probiotic lactobacilli and/or bifidobacteria supplementation in preterm and very low birth weight neonates (Deshpande et al., 2010). Probiotic Bifidobacterium breve and galacto-oligosaccharide improves the clinical condition in patients with ulcerative colitis (Ishikawa et al., 2011). Probiotic yogurt containing a combination of Lactobacillus rhamnosus GG, Bifidobacterium lactis and L. acidophilus reduces the incidence of antibiotic-associated diarrhea in children (Fox et al., 2015). Prenatal supplementation of probiotic bifodobacteria to the mothers and postnatally to the infants decreases the risk of developing atopic dermatitis in infants (Enomoto et al., 2014).

The use of antibiotics to treat BV in non-pregnant and pregnant women remains the method of choice, unchanged for many decades, and still too often ineffective. Metronidazole and clindamycin, by far the most commonly used agents, do not restore vaginal lactobacilli abundance, which may account for relapses in some women; and prolonged use promotes the development of drug resistance. The need for new treatments for BV that restore microbiota homeostasis and acidity without undesirable side effects has led investigators to study probiotics. Human studies have provided evidence that probiotic lactobacilli can reduce BV recurrence and increase lactobacilli abundance in the vagina of non-pregnant women (Reid et al., 2003a; Homayouni et al., 2014). The use of lactobacilli as an adjuvant therapy to antibiotics also shows promise in lowering BV recurrence rates (Bodean et al., 2013). Indeed, the adjunctive use of L. rhamnosus GR-1 and L. reuteri RC-14 with metronidazole improves the cure of BV (Maritinez et al., 2009; Anukam et al., 2006).

1.5.1 Safety and Compliance

Probiotic intervention in pregnancy is generally acceptable with good compliance among pregnant women (Lindsay et al., 2014). A recent meta-analysis of randomized clinical trials found that the use of probiotics Lactobacillus and Bifidobacterium during pregnancy had no

35 effect on the incidence of Caesarean section, birth weight, or gestational age and there were no adverse effects (Dugoua et al., 2009).

1.5.2 Lactobacilli

Lactobacilli are gram-positive facultative anaerobic bacteria that dominate the vaginal microbiota of women of reproductive age (Ravel et al., 2011). Probiotic lactobacilli are used most commonly to maintain healthy vaginal and urogenital tracts.

1.5.2.1 Route of Administration

Oral administration of 109 - 1011 colony-forming units (cfu) of lactobacilli is the standard dose believed to be required for passage through the intestine and subsequent improvement of gut and vaginal health (Othman et al., 2007; Homayouni et al., 2014; Morelli et al., 2004; Reid, 2001a). There are many variables that influence vaginal colonization by lactobacilli including glycogen levels, substances used in vaginal washing, the use of antibiotics and the ability of lactobacilli to produce substances such as hydrogen peroxide (Vallor et al., 2001; Mirmonsef et al., 2014). The oral administration of L. acidophilus and L. bifidus has been reported to be more effective than the vaginal route in reducing BV occurrence in antibiotic- treated non-pregnant women (Bodean et al., 2013). However, the probiotic composition of the oral capsule was different from the vaginal capsule (L. rhamnosus, L. acidophilus, S thermophilus and L. bulgaricus) in that study, and the potential mechanism seems unclear. Furthermore, the treatment duration was longer for patients who received the oral capsule than those who received vaginal capsules (Bodean et al., 2013). An advantage of the oral route is that it may reduce pathogen ascendance from the rectum to perineum and vagina, while some women may perceive the intra-vaginal approach to be the more invasive instillation of microbes.

1.5.2.2 Potential Mechanisms of Action

L. rhamnosus GR-1 and L. reuteri RC-14 (GR-1 and RC-14) persist up to 19 days in the human vagina following intra-vaginal administration (Gardiner et al., 2002). Exogenous lactobacilli colonization appears to be transient and lactobacilli exert their anti-pathogenic properties indirectly via a number of mechanisms. These include the production of antimicrobial substances, competitive exclusion with pathogenic bacteria and fungi, acidification of the vagina, and modulation of the immune system (Reid and Bocking, 2003b). Endogenous lactobacilli maintain the vaginal pH <4.5 by metabolizing glycogen secreted by vaginal mucosal epithelia and produce lactic acid, which is a potent microbicide against potential reproductive tract infections (O’Hanlon et al., 2013). The acidic environment of a healthy vagina creates a hostile environment for BV-associated pathogens while favoring lactobacilli growth (O’Hanlon et al., 2013; Borges and Teixeira, 2014). It may also help to prevent viruses, such as HIV, from infecting the host (Petrova et al., 2013).

The anti-inflammatory property of lactobacilli is important in control of mucosal and systemic inflammation (Kemgang et al., 2014). L. rhamnosus GR-1 supernatant (GR-1 SN) enhances IL-10 and colony stimulating factor 3 (CSF3) production in mouse macrophages (Kim et al., 2006). In primary human placental trophoblast cells, GR-1 SN increases IL-10 and CSF3 production via JAK/STAT and MAPK pathways; down-regulates LPS-induced TNFα output through c-Jun-N-terminal kinases (JNKs) inhibition and increases the expression of the PG metabolizing enzyme PGDH in a sex-dependent fashion (Yeganegi et al., 2009; Yeganegi et al., 2010; Yeganegi et al., 2011).

The effect of lactobacilli on the immune system and their vaginal colonization ability are species and strain specific. In the mouse gastrointestinal (GI) tract, L. plantarum and L. rhamnosus GG exacerbate inflammation and the development of DSS-induced colitis while L. paracasei is protective (Mileti et al., 2009). In the human vagina, L. rhamnosus GR-1 and L. reuteri RC-14 (GR-1 and RC-14) but not the intestinal probiotic L. rhamnosus GG persists up to 19 days following intra-vaginal administration of either GR-1 and RC-14 or GG (Gardiner et al., 2002). Intra-vaginal instillation of L. rhamnosus GR-1 up-regulates some

37 antimicrobial activity in premenopausal women (Krijavainen et al., 2008). A combination of B. bifidum, B. infantis, L. acidophilus, L. casei, L. salivarius and Lactococcus lactis has been reported to provide a wider antimicrobial spectrum and greater stimulation of IL-10 production along with suppression of pro-inflammatory cytokines in cultured human peripheral blood mononuclear cells compared to the individual strains (Timmerman et al., 2007). A combination of Bacteriocin like inhibitory substances (BLIS) from the L. rhamnosus L60 and L. fermentum L23 can reduce the growth of group B streptococcal isolates obtained from pregnant women more effectively than each Lactobacillus strain alone (Ruiz et al., 2012).

Lipoteichoic Acid (LTA) on the cell surface of lactobacilli can also stimulate macrophages to secrete immune-mediators. Improved anti-inflammatory activity in a murine model of colitis in vivo has been observed when LTA is either removed or modified (D-alanylation) (Grangette et al., 2005; Claes et al., 2010; Mohamadzadeh et al., 2011). The supernatant of lactobacilli also has anti-inflammatory properties in cultured human placental trophoblast cells, decidual cells, monocytes and macrophages (Yeganegi et al., 2009; Yeganegi et al., 2010; Yeganegi et al., 2011; Li et al., 2014; Lin et al., 2008). These studies imply that the administration of supernatant from lactobacilli may promote desirable effects and represent an alternative for the prevention and treatment of inflammatory disorders, such as some cases of PTB. The identification of these bioactive metabolite(s) remains to be achieved.

1.6 Summary

Bacterial vaginosis, which is characterized by a depletion of lactobacilli in the vaginal microbiota of pregnant women, contributes to an increased risk of PTB (Donders et al., 2009). Bacterial endotoxin induced over-expression of pro-inflammatory cytokines and chemokines stimulate the onset of PTL (Challis et al., 2009). Probiotic Lactobacillus rhamnosus GR-1 has been shown to possess anti-inflammatory properties in cultured human intra-uterine tissues (Yeganegi et al., 2009; Yeganegi et al., 2010; Yeganegi et al., 2011; Li et al., 2014) and lactobacilli also have the ability to reverse BV in non-pregnant women (Reid et al., 2003a). The role of lactobacilli in immune regulation and modulation of the

38 vaginal microbiota in pregnant women with BV remains unknown. Furthermore, the potential of lactobacilli as a prophylactic therapy for PTB has not been directly examined. This thesis evaluated the potential of both Lactobacillus rhamnosus GR-1 and L. reuteri RC- 14 live bacteria and its supernatant in the prevention of PTB in a mouse model. In addition, the effects of live Lactobacillus rhamnosus GR-1 bacteria on the reversal of BV, the concentration of cytokines and chemokines and the vaginal microbiota in pregnant women diagnosed with an intermediate or high Nugent score were also investigated.

Chapter Two

Rationale and Hypotheses

39 40

Chapter 2

2. Rationale and Hypotheses

2.1 Rationale

Preterm birth (PTB) occurs in 11% of all pregnancies worldwide with premature infants at a higher risk of developing adverse long-term health outcomes (Blencowe et al., 2012; Oskoui et al., 2013; Brostrom et al., 2013). Inflammation is one of the major contributing factors to both infection-mediated PTB and spontaneous PTB (Challis et al., 2009). Antibiotic administration to prevent PTB has been unsuccessful.

Probiotics, defined as “live microorganisms which, when administered in adequate amounts, confer a health benefit on the host”, have been used to treat inflammatory conditions in the gastro-intestinal and genito-urinary tracts (FAO/WHO, 2001; Reid et al., 2015). Lactobacillus spp. are commensal to the human vagina and intestinal tracts. Lactobacilli can modulate immune responses, reduce pathogenic adherence and/or produce bacteriocins to discourage pathogen growth (Reid and Bocking., 2003b). Pro-inflammatory cytokines and chemokines contribute to the onset of PTB in both humans and animals. Previous studies demonstrate that L. rhamnosus GR-1 supernatant has anti-inflammatory properties in cultured human intra-uterine tissues (Yeganegi et al., 2009; Yeganegi et al., 2010; Yeganegi et al., 2011; Li et al., 2014).

Since the most common route of intrauterine infection is thought to be ascending through the vagina, I induced localized inflammation and PTB with intra-uterine LPS injection. I investigated the effect of L. rhamnosus GR-1 supernatant on lipopolysaccharide (LPS)- induced PTB and concentrations of cytokines and chemokines. Since intra-uterine tissues and circulating leukocytes are potential sources of cytokines, outputs of cytokines and chemokines were measured in the maternal plasma, amniotic fluid as well as various intrauterine tissues.

Bacterial Vaginosis, characterized by the absence of lactobacilli, has been associated with an increase in the risk of PTB (Donders et al., 2009). Oral Lactobacillus rhamnosus GR-1 and L. reuteri RC-14 reduce the recurrence of bacterial vaginosis (BV) by restoring the indigenous lactobacilli in the vagina of non-pregnant women (Reid et al., 2003a). It remains to be determined whether Lactobacillus rhamnosus GR-1 and L. reuteri RC-14 have a potential beneficial effect on pregnant women with BV.

I evaluated the potential of using live Lactobacillus rhamnosus GR-1 bacteria and its supernatant as prophylactic treatments for the prevention of PTB in a mouse model. I also investigated the effect of live Lactobacillus rhamnosus GR-1 bacteria on the reversal of BV, the concentrations of cervico-vaginal cytokines and chemokines and the vaginal microbiota in pregnant women diagnosed with an intermediate or high Nugent score.

2.2 Hypotheses

I hypothesize that (a) An intra-peritoneal administration of L. rhamnosus GR-1 supernatant (GR-1 SN) reduces the incidence of LPS-induced PTB, as well as systemic and intrauterine cytokine and chemokine concentrations in pregnant CD-1 mice (CHAPTER 3).

(b) Oral administration of live L. rhamnosus GR-1 bacteria reduces LPS-induced PTB, lowers systemic and intrauterine cytokine and chemokine concentrations and alters the vaginal and cecal microbiota of pregnant CD-1 mice (CHAPTER 4).

(c) In pregnant women with an intermediate or BV Nugent score on vaginal gram stain smears, oral administration of L. rhamnosus GR-1 and L. reuteri RC-14 will

I. return an abnormal Nugent score to a normal Nugent score, II. reduce the cervico-vaginal concentrations of pro-inflammatory cytokines and chemokines and, III. alter the vaginal microbial profiles (CHAPTER 5)

Chapter Three

Probiotic Lactobacillus rhamnosus GR-1 supernatant (GR-1 SN) prevents Lipopolysaccharide (LPS) -induced preterm birth and reduces inflammation in pregnant CD-1 mice.

The contents were published in Am J Obstet Gynecol. 2014 Jul;211(1): 44.e1-12 and appear here with the permission of the journal (authorization attached). My role involved experiment design, conduct and result analyses as well as manuscript preparation.

42 43

Chapter 3

3. Probiotic Lactobacillus rhamnosus GR-1 supernatant (GR-1 SN) prevents Lipopolysaccharide (LPS)-induced preterm birth and reduces inflammation in pregnant CD-1 mice.

3.1 Introduction

Preterm birth (PTB) occurs in 11% of all pregnancies worldwide with premature infants at a higher risk of developing adverse long-term health outcomes (Blencowe et al., 2012; Brostrom et al., 2013; Oskoui et al., 2013). Inflammation is a contributing factor to both infection-mediated PTB as well as spontaneous PTB, and the most common route of infection is thought to be ascending through the vagina (Goldenberg et al., 2000; Goldenberg et al., 2008a). In this study, I administered lipopolysaccharide (LPS) to pregnant CD-1 mice as a model for both infection and inflammation-associated PTB since LPS activates Toll-like receptor 4 mediated inflammatory pathways (Elovitz et al., 2003.; Koga and Mor, 2010). Antibiotic administration to prevent PTB has been unsuccessful (Subramaniam et al., 2012), possibly since they do not replenish vaginal lactobacilli. In addition, prolonged antibiotic use promotes resistant bacterial strains (Beigi et al., 2004).

Probiotic, defined as “live microorganisms which, when administered in adequate amounts, confer a health benefit on the host”, have been used to treat inflammatory conditions in the gastro-intestinal and genito-urinary tracts (FAO/WHO, 2001; Reid et al., 2015; Othman et al., 2007). Probiotic lactobacilli, a genus commensal to human vagina and intestinal tracts, reduce the recurrence of bacterial vaginosis (BV) in non-pregnant women (Reid et al., 2003a), and are associated with a 40% increase in the risk of PTB (Donders et al., 2009). Lactobacilli can modulate immune responses, reduce pathogenic adherence and/or produce bacteriocins to discourage pathogen growth (Reid and Bocking, 2003b). Our previous studies have also demonstrated that L. rhamnosus GR-1 supernatant have anti-inflammatory

44 properties (Yeganegi et al., 2009; Yeganegi et al., 2010; Yeganegi et al., 2011; Li et al., 2014; Kim et al., 2006).

Cytokines play a pivotal role in PTB in both humans and animals and the predominance of anti- over pro-inflammatory cytokines is important to pregnancy maintenance (Challis et al., 2009). Cytokines can act as regulators of the innate and adaptive immune systems as well as hematopoiesis (Elgert, 2009). Intra-uterine tissues and circulating leukocytes are potential sources of cytokines (Young et al., 2002). Chemokines can attract decidual leukocytes, which in turn recruit additional pro-inflammatory cytokines to amplify the inflammatory cascade (Esplin et al., 2005; Hamilton et al., 2013).

The effect of lactobacilli on PTB and inflammatory responses in pregnant CD-1 mice in vivo is unknown. In this study, I test the hypothesis that GR-1 supernatant will attenuate LPS induced PTB and also profile systemic and intra-uterine immune markers in LPS-treated mice with and without GR-1 SN treatment. Lastly, I examine whether the effect of GR-1 SN on LPS-induced PTB is dependent on changes in maternal plasma progesterone or sex of the fetus.

3.2 Material and Methods

3.2.1 Animals

Female HSD:ICR (CD-1) outbred mice (8-12 weeks old; Harlan Laboratories) were bred; the morning of vaginal plug detection was designated Gestational Day (GD) 1. The normal gestational length of pregnant CD-1 mice is 19-20 days. Animals were handled in accordance with guidelines of the Canadian Council for Animal Care and all procedures were approved by the Animal Care Committee of Toronto Center for Phenogenomics (Animal Use Protocol #0164). Animals were housed in a pathogen-free, humidity controlled 12 h light:12 h dark cycle animal facility with free access to food and water. I performed 4 sets of independent experiments and used a total of 155 animals.

3.2.2 L. rhamnosus GR-1 supernatant preparation

GR-1 was grown for 8-10 hours anaerobically at 37oC in de Man, Rogosa, and Sharpe

(MRS) broth (BD, Ontario) to an optical density of ~0.9 at 600nm (representing ~108 -109 cfu/mL of bacteria), and centrifuged at 3000 rpm for 10 min at 25oC. The supernatant (GR-1 SN) was filtered twice with 0.22 µm filters to remove residual bacteria. I used 200µL of GR- 1 SN, representing ~2x107 -108 cfu/mL of bacteria for intraperitoneal (i.p) injection, since in previous studies, i.p injection of ~107 cfu of GR-1 increased anti-inflammatory cytokine G- CSF production in mice (Martins et al., 2011).

3.2.3 Intra-uterine injection of LPS by mini-laparotomy

Intrauterine injection of LPS was given via mini-laparotomy on GD 15 as previously described (Elovitz et al., 2003). Mice were anesthetized with isoflurane inhalation and given analgesic buprenorphine (0.1mg/kg). An incision (~1cm) was made to expose the lower segments of the uterine horns. Saline (100µL) or LPS (Escherichia coli 055:B5, Sigma- Aldrich, St. Louis) dissolved in 100µL saline was injected between the two lowest gestational sacs of either the left or right uterine horn. Fascia and skin were closed with 4.0 vicryl sutures and staples respectively. Mice were housed in individual cages.

3.2.4 Dose effect of LPS on PTB rate (Set 1)

A dose response for LPS was established using saline, 25µg, 65µg, 125µg or 250µg of LPS (n=10 per group) to determine the lowest dose that produced 100% PTB. PTB was defined as the delivery of at least one pup within 48 hours of LPS injection. LPS 125µg was the lowest dose that resulted in almost 100% PTB and was therefore chosen for subsequent experiments.

3.2.5 Effect of GR-1 supernatant on the timing of LPS-induced PTB (Set 2)

Mice were randomly assigned to receive Saline, GR-1 SN, LPS 125µg or LPS 125µg+GR-1 SN (n=9-17 per group, Figure 3-1). Animals were given two doses of 200 µL GR-1 SN or saline intra-peritoneally, at 24 hours (GD14) and 15-30 minutes (GD15) prior to intrauterine LPS or saline injection (GD15). In our preliminary experiments, I did not observe any effect of orally administered GR-1 SN on PTB, and I chose not to administer GR-1 vaginally because of concerns of possible vaginal leakage. Given our previous experiments whereby i.p injection of GR-1 SN caused immune responses in non-pregnant mice as well as in vitro (Martins et al., 2011; Yeganegi et al., 2009), I chose to give GR-1 SN i.p in these studies. Saline was given to mice in the control group since I did not observe an effect with MRS pretreatment on LPS-induced PTB in preliminary experiments. Animals were monitored hourly until term for the delivery of pups, and the time of delivery was recorded.

3.2.6 Effect of GR-1 supernatant on cytokines and chemokines (Set 3)

Mice were randomly assigned to receive Saline, GR-1 SN, LPS 125µg or LPS 125µg+GR-1 SN (n=10 per group, Figure 3-2). The majority of animals in Group 2 delivered between 10- 15 hours after LPS administration and therefore animals in Group 3 were euthanized with carbon dioxide 8 hours post LPS or saline injection for the collection of amniotic fluid, placental and myometrial tissue. Prior to euthanization, maternal blood was collected from anesthetized mice by cardiac puncture and plasma was obtained by centrifugation at 5,000 xg for 15 min at 4oC. Amniotic fluid was pooled from all gestational sacs and centrifuged to remove any cellular debris. Placental tissue was separated from decidua and fetal membranes in ice-cold PBS and pooled from all fetuses. Myometrium was separated from decidua and endometrium by scraping (Shynolva et al., 2013). All samples were flash-frozen in liquid nitrogen and stored at -80 oC. In a subgroup of animals (n=5), I measured progesterone concentrations in maternal plasma.

3.2.7 Fetal Sex ratios (Set 4)

Mice received LPS 125µg +GR-1 SN (n= 16) and were monitored for PTB. After delivery of at least one pup, animals were euthanized and individual fetal tails were collected and genotyped to determine fetal sex. For animals that delivered at term, tails from the neonates were collected. Fetuses and neonates were euthanized by cold anesthesia on ice followed by decapitation.

3.2.8 Cytokine assay

Cytokine and chemokine concentrations were determined using a mouse 23-multiplex cytokine assay (Appendix I, Biorad, Ontario) on a Luminex 200 cytometer and Bioplex HTF (Bio-Rad). The assay measured concentrations of interleukin (IL)-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-17, IFN-γ, CXCL 1, CCL2, CCL3, CCL4, CCL5, CCL11, TNFα, CSF2, and CSF3. Data analysis was performed using Bio-Plex Manager (version 5.0, Bio-Rad) and results are presented as concentrations (pg/mL). I randomly chose 7 animals to measure the concentration of cytokines and chemokines in the intra-uterine tissues. Tissues were crushed and homogenized in EDTA-free protease inhibitor containing RIPA lysis buffer (1mL per 0.5g of tissue). Homogenized samples were left on ice for 45 minutes before being centrifuged at 12,000 xg for 15 minutes at 4°C to collect the supernatant. Protein concentration was measured by Bradford assay kit (Bio-Rad, Ontario) with bovine serum albumin as standard. 250µg of total protein were used for the measurement of cytokines/chemokines in myometrium and placenta tissues.

3.2.9 Maternal progesterone measurement

Plasma progesterone concentration was measured with an Enzyme Immunoassay kit (Appendix II, Cayman Chemical Co, Michigan). Samples were diluted (400 v/v) with EIA buffer and assayed in duplicate. The intra and inter-assay coefficients of variation were 6.9 % and 12.1 % respectively.

3.2.10 Sex determination by PCR

DNA extracted from individual fetal tails was amplified using Sigma REDExtract-N-Amp Tissue PCR kit (XNAT, Sigma, St Louis). DNA isolated was amplified using primers Jarid1d FWD: GCACAGGACCTCAGGGACCCAG, Jarid1d REV: CAGAGGCATTCATCGATGAGG, Jarid 1c REV2: TGAGTTGGTACGACGAAGCTGCAG (Clapcote and Roder, 2005). PCR amplified products were resolved using a 2% agarose gel. Double bands (331 and 302bp) were seen for males and a single band (331bp) for females. Sex ratio was calculated by expressing the number of male fetuses over total number of fetuses.

3.2.11 Statistical Analyses

Statistical analysis was carried out using SigmaStat (version 3.5). Comparison of PTB rate was made with Fisher exact analyses (two tailed). Unpaired Student’s t test, two tailed, was used to compare sex ratios. Comparison of cytokine, chemokine and progesterone concentrations in multiple groups were carried out with One-way ANOVA or ANOVA on Ranks followed by Student Newman Keuls test as post hoc test. Data were tested for normality and equal variance and data are expressed as mean values ± SEM. Data were adjusted for false discovery rate using Benjamini Hochberg procedure and an adjusted p- value of p<0.05 was considered statistically significant.

3.3 Results

3.3.1 GR-1 SN reduced LPS-induced PTB (Set 2)

Intrauterine injection of LPS on GD15 resulted in dose-dependent PTB within 48 hours (Table 3-1). GR-1 SN significantly reduced the rate of PTB from 94% (16/17) in the LPS 125µg group to 57% (8/14) in the LPS 125µg+GR-1 SN group (p=0.028, Figure 3-3). One mouse in the LPS 125µg-treated group had all fetuses resorbed at term. Four out of six mice in the LPS125µg+GR-1 SN group delivered live pups at term, and in the remaining two mice,

49 all fetuses had resorbed at term. All animals in the saline and GR-1 SN control groups delivered live pups at term. The mean litter size was 13.0 ± 0.89 and the mean weight per pup was 1.72 ± 0.02 grams in the saline group, and these were not different in mice in other treatment groups (p>0.05, Table 3-2).

3.3.2 GR-1 SN attenuated LPS induced cytokines and chemokines (Set 3)

Baseline pro-inflammatory cytokine concentrations (IL-1α, IL-1β and IL-12p70) and LPS- induced increases (IL-1α, IL-1β, IL-6, IL-17) were highest in the myometrium (Table 3-3). Compared to other compartments, baseline chemokine concentrations (CXCL1, CCL2, CCL3, CCL4, CCL5, CCL11) were low in the maternal plasma but their production increased markedly (CCL2, CCL4, CCL5) with LPS stimulation (57-186 fold) (Table 3-3). LPS-increased both IL-4 and IL-10 concentrations in all compartments except amniotic fluid (Table 3-3). Among all cytokines measured, IL-6 and CSF3 had the greatest increases following LPS treatment (Table 3-3).

LPS significantly increased IL-1α, IL-6, IL-12p70, TNFα; CCL2, CCL3, CCL4, CCL5, CSF2 and CSF3 in the maternal plasma (Table 3-4), myometrium (Table 3-5), amniotic fluid (Table 3-6) and placenta (Table 3-7). LPS also significantly increased IL-1β, IL-10, IL- 12p40, IL-17, CCL11, IL-13, IFN-γ and IL-3 in the maternal plasma and myometrium (Table 3-4 and 3-5) but not in the amniotic fluid (Table 3-6). These cytokines/chemokines were also significantly elevated in the placenta following LPS except for IFN-γ and CCL11 that were below detection limits (Table 3-7). LPS increased IL-2, IL-4, IL-5, IL-9 and CXCL1 to various degrees in tissues and fluids (Table 3-4, Table 3-5, Table 3-6 and Table 3-7). IL-5 in the amniotic fluid and IL-9 in the plasma and amniotic fluid were below the limits of assay detection (Table 3-4 and Table 3-6).

Pretreatment with GR-1 SN significantly attenuated the LPS-induced elevation in pro- inflammatory cytokines IL-1β, IL-6, IL-12p40, TNFα as well as chemokines CCL4 and CCL5 in the plasma and IL-6, IL-12p70, IL-13, IL-17, TNFα in the myometrium (p<0.05, Figure 3-4 and 3-5). LPS-induced increases in all other cytokines, including IL-10, remained

50 elevated with GR-1 SN treatment (Figure 3-6). GR-1 SN treatment significantly decreased the LPS-induced elevation in IL-6, TNFα, CCL3 and CCL4 in the amniotic fluid; and IL-6 and IL-12p70 in the placenta (p<0.05, Figure 3-4 and 3-5). GR-1 SN alone increased placental IL-4 and IL-10 (p<0.05; Figure 3-6). There was no difference in the production of cytokines and chemokines between saline and GR-1 SN treated mice in the plasma, myometrium, placenta or amniotic fluid (Figure 3-4, Figure 3-5 and Figure 3-6).

3.3.3 Plasma progesterone (Set 3)

Maternal plasma progesterone concentration (saline: 68 ± 4.6 ng/ml) was significantly reduced by LPS 125µg (42 ± 7.4 ng/ml) as well as LPS 125µg + GR-1 SN (38 ± 4.5 ng/ml) treatment (p< 0.05; Figure 3-7). Mice that received GR-1 SN alone maintained high concentrations of plasma progesterone (59.1 ± 1.7 ng/ml) comparable to that of the saline group (p>0.05) (Figure 3-7). In order to confirm that these concentrations are identical, however, a sample size of 6 mice in each group would be required (power analysis test). With 5 animals, as shown here, I did not demonstrate a statistically significant difference.

3.3.4 Fetal sex ratio (Set 4)

Among mice that received both LPS 125µg and GR-1 SN, there was no difference in the percentage of male fetuses in mice that delivered preterm (55 ± 4.9%) with a litter size 11.0 ±0.84 (n=5) compared to those that delivered at term (49 ± 4.3%) with a litter size of 10.7 ±0.54 (n=11).

3.4 Comment

In this study, I have shown that Lactobacillus rhamnosus GR-1 supernatant (GR-1 SN) reduces LPS-induced PTB and dampens both systemic and intrauterine inflammation in pregnant CD-1 mice. I profiled multiple cytokines and chemokines in the maternal plasma, myometrium, placenta and amniotic fluid following LPS alone, and in combination with GR- 1 SN. The reduction of LPS-induced PTB by GR-1 SN was independent of changes in

51 circulating progesterone and fetal sex ratios.

I have focused our studies on the beneficial effects of the L. rhamnosus GR-1 strain, which decreases the recurrence risk of BV in non-pregnant women and modulates immune responses (Reid et al., 2003a; Yeganegi et al., 2009; Martins et al., 2011). L. rhamnosus also has anti-microbial activity against pathogens including E. coli (Reid and Bocking, 2003b). Analysis of the genomes and phenotypes of 100 L. rhamnosus strains has demonstrated the presence of two major geno-phenotypes, A and B with different carbohydrate metabolism and adherence properties and differing beneficial effects (Douillard et al., 2013). Geno- phenotype B L. rhamnosus strains, but not strains in geno-phenotype A, display traits that allow them to survive in the GI tract, including expression of mucus-binding pili and bile resistance. Although it is unclear which geno-phenotype GR-1 belongs to, Reid et al have demonstrated that GR-1 can adhere to urogenital and vaginal cells in vitro (Reid and Bruce, 2001b). I have administered GR-1 SN prior to LPS injection since it is likely that its protective benefit in humans would be in prevention of PTB as opposed to treatment once labor has started. A previous study has demonstrated p40, a 40kDa soluble protein purified from the supernatant of L. rhamnosus GG, reduces intestinal inflammation in dextran sulfate sodium (DSS)-induced colitis in mice (Yan and Polk, 2012). Preliminary experiments in our laboratory have identified a heat-sensitive protein-like molecule (>30kDa), which suppresses LPS-induced TNFα production to a comparable degree as unfractionated GR-1 SN in immortalized mouse macrophages (unpublished observations). GR-1 SN may also contain small molecules such as lactic acid that could account for its anti-pathogenic property. I propose that the active moiety(ies) in GR-1 SN, when given intra-peritoneally, activate signaling molecules which lead to the immune-modulatory effects observed.

In this study, LPS stimulated multiple cytokines and chemokines, resulting in systemic and intrauterine inflammation, consistent with previous studies using the same PTB mouse model (Yang et al., 2009; Shynlova et al., 2013). Although I found that LPS increased plasma IL- 12p70 and IL-17 in contrast to a previous study (Yang et al., 2009), this may be due to the higher LPS dose I used. In general, LPS induced the greatest changes in pro-inflammatory cytokine concentrations in the myometrium, suggesting myometrial immune alterations play

52 a key role in the onset of LPS-induced preterm labor. Despite low baseline maternal plasma CCL2 and CCL4 concentrations, they increased markedly with LPS stimulation, in keeping with their role in recruitment of peripheral immune cells. Although I cannot extrapolate our findings directly to the clinical setting, many of the cytokines and chemokines I report on have been implicated in the pathogenesis of human PTB (Sweet et al., 2007; Ekelund et al.. 2008; Ito et al., 2010; Weissenbacher et al.; 2013).

My findings provide evidence that GR-1 SN promotes an anti-inflammatory environment, consistent with our previous studies (Yeganegi et al., 2009; Yeganegi et al., 2011). GR-1 SN decreased LPS-induced TNFα concentrations in myometrium, maternal plasma and amniotic fluid but not the placenta. IL-6 concentrations were also markedly decreased by pretreatment with GR-1 SN in all tissues and fluids studied. IL-6 -/- mice have delayed PTB compared to wild-type mice, and mice with double knockouts for TNFα and IL-1 receptors are refractory to bacterially induced PTB (Hirsch et al., 2006; Robertson et al., 2010). Together with our findings, this suggests that both IL-6 and TNFα play an essential role in LPS-induced PTB in mice. Pretreatment with GR-1 SN did not alter LPS-induced IL-1α whereas the increase in IL-1β concentration was partially reduced by GR-1 SN in maternal plasma. Although IL-1 is involved in human PTB, IL-1 signaling may not play a critical role in murine PTB (Yoshimura and Hirsch 2005). GR-1 SN alone increased placental IL-10 and IL-4 concentrations, which have been shown to counteract inflammatory responses in mice (Robertson et al., 2007a). IL-10 prolongs gestation in LPS treated IL-10-/- C57BL/6 mice (Keelan et al., 1999). GR-1 SN also attenuated LPS-induced amniotic fluid CCL3 and CCL4, as well as plasma CCL4 and CCL5, suggesting GR-1 SN may play a role in reducing leukocyte recruitment to sites of inflammation. GR-1 SN did not decrease LPS-induced CCL2 and CXCL1 in plasma or myometrium, which would promote and maintain effective pathogen clearance. Our previous studies have indicated that Mitogen-Activated Protein Kinase and Janus Kinase/Signal Transducer and Activator Transcription pathways may be involved in the mechanism of action of GR-1 SN on LPS-induced cytokine productions (Yeganegi et al., 2010) and I propose that similar mechanisms may be important in vivo.

The role of hematopoietic cytokines CSF2 and CSF3 in human PTB is unclear although elevated cervico-vaginal fluid CSF2 concentrations have been shown to be associated with cervical shortening (Chandiramani et al., 2012). In our study, GR-1 SN attenuated LPS- induced myometrial CSF2 production but not in the placenta, amniotic fluid or maternal plasma. LPS increased CSF3 concentrations markedly in all tissues and fluids and this was maintained with prior GR-1 SN treatment. Since CSF3 possesses important anti- inflammatory properties (Martins et al., 2011; Yeganegi et al., 2011), this is an additional mechanism whereby GR-1 SN favors an anti-inflammatory environment.

Plasma progesterone decreased in animals treated with both LPS and LPS+GR-1 SN that is in keeping with a previous study (Fidel et al., 1998), and GR-1 SN alone had no effect on plasma progesterone concentrations. Unlike in humans, term parturition in mice occurs after luteolysis and is associated with a decline in plasma progesterone concentrations (Challis et al., 2000; Mesiano et al., 2002). However, it is not known whether a decline in progesterone is essential for mice to undergo PTB (Hirsch and Muhle, 2002; Elovitz and Wang, 2004; Romero and Stanczyk, 2013); I propose inflammation may be a more important factor than progesterone withdrawal in this PTB model.

Pregnant women carrying male fetuses are more susceptible to PTB than those with female fetuses (Challis et al., 2013). I have previously shown that LPS increases TNFα output and prostaglandin-endoperoxide synthase 2 expressions to a greater extent in trophoblast cells from pregnancies with a male fetus (Yeganegi et al., 2009; Yeganegi et al., 2011). In the current study, I did not observe any differences in fetal sex ratios between mice that delivered preterm and those that delivered at term when treated with LPS+GR-1 SN.

In summary, probiotic Lactobacillus rhamnosus GR-1 supernatant attenuates LPS-induced inflammation as well as the rate of PTB in pregnant mice. This study provides further evidence regarding the potential mechanisms whereby probiotic lactobacilli may reduce the risk of PTB, and hence supports the need for clinical trials to assess their efficacy.

Gestational Day (GD) Pregnant CD-1 mice 1 Intra-peritoneal (200µL)

……….#

14 Saline GR-1 Supernatant

15 15-30 min# Laparotomy LPS Saline LPS Intra-uterine Saline 125µg 125µg 16 infusion (100µL) n=9 n=17 n=9 n=14

17 Preterm Monitor for time of delivery

Term 18

19/20

Figure 3-1 Experimental design to investigate the effect of GR-1 supernatant (GR-1 SN) on the timing of LPS-induced PTB (Set 2).

Mice received an intra-peritoneal injection of either saline or GR-1 SN on GD 14. A second dose of saline or GR-1 SN was given on GD 15, approximately 15-30 minutes prior to an intrauterine injection of saline or LPS 125µg. Animals were monitored for the time of delivery in individual cages till term (GD19/20). Preterm delivery was defined as delivery of at least one pup 48 hours after intrauterine injection of LPS (GD 17).

Gestational Day (GD) Pregnant CD-1 mice 1 Intra-peritoneal (200µL)

……….#

GR-1 14 Saline Supernatant

Laparotomy 15 15-30 min# LPS Intra-uterine Saline LPS Saline 125µg infusion (100µL) 125µg

8 hours# n=10 n=10 n=10 n=10 16

Sample

Collection

17 Preterm • Maternal blood • Amniotic Fluid • Intra-uterine tissues Term 18

19/20

Figure 3-2 Experimental design to investigate the effect of GR-1 supernatant (GR-1 SN) on the concentration of cytokines and chemokines in the maternal plasma, amniotic fluid and intra-uterine tissues (Set 3). Mice received an intra-peritoneal injection of either saline or GR-1 SN on GD 14. A second dose of saline or GR-1 SN was given on GD 15, approximately 15-30 minutes prior to an intrauterine injection of saline or LPS 125µg. The maternal blood, amniotic fluid and intrauterine tissues (placenta and myometrium) were collected 8 hours after intrauterine injection of saline or LPS for the measurement of cytokines and chemokines.

100 LPS

80 a,b,c 60 LPS + GR-1 SN

% delivered 40 % delivered %

20 Saline 0 GR-1 SN 15 16 17 18 19 Gestationalgestational Day day

Figure 3-3 Cumulative frequency plot showing the percentage of pregnant CD-1 mice that delivered at various gestational days following four different treatments (Set 2). Preterm birth is defined as delivery of at least one pup within 48 hours of LPS injection. Four treatment groups are shown: LPS 125µg (solid triangle, n=17), LPS 125µg+GR-1 SN (open triangle, n=14), saline (open circle, n=9) and GR-1 SN (solid circle, n=9). The LPS group was compared with each of the three remaining groups using Fisher’s exact test. Statistical significance was denoted with different letters. a: p<0.0001 versus saline; b: p<0.0001 versus GR-1 SN; c: p=0.0281 versus LPS 125µg +GR-1 SN.

IL-1β Saline IL-6 ** b ** 8000 GR-1 SN 2500 b b LPS 125ug b 6000 LPS 125ug 2000 ** 4000 b + GR-1 SN b 2000 a b ** 1500 a b a 200 1000 NS *** c c c 100 a 500 c a a b Concentration (pg/mL) Concentration Concentration (pg/mL) Concentration a a a a a a c a a 0 0 Maternal Myometrium Placenta Amniotic Maternal Myometrium Placenta Amniotic Plasm a Fluid Plasma Fluid

** IL-12p40 IL-12p70 * 2000 500 b b 1500 400 1000 c b 500 c b 300 300 b 200 ** ab 200 ab a a b a,c a NS 100 b a,b a,b,c 100 b a a a a a a b a a Concentration (pg/mL) Concentration (pg/mL) Concentration a 0 0 Maternal Myometrium Placenta Amniotic Maternal Myometrium Placenta Amniotic Plasm a Fluid Plasma Fluid

TNFα ** IL-17 b 200 b 80 * * b 150 60 * c 100 b a 40 a c a a b a c 50 b 20 b a a b NS a a a b b

Concentration (pg/mL) Concentration a a a (pg/mL) Concentration a a 0 0 Maternal Myometrium Placenta Amniotic Maternal Myometrium Placenta Amnotic Plasma Fluid Plasm a Fluid

Figure 3-4 Histogram showing concentrations of pro-inflammatory cytokines IL-1β, IL- 6, IL-12p40, IL-12p70, TNFα and IL-17 in the maternal plasma (n=10 animals per treatment group), myometrium (n=7), placenta (n=7) and amniotic fluid (n=10) of pregnant CD-1 mice (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are 4 treatment groups: saline (white); GR-1 SN (light grey); LPS 125µg (dark grey) and LPS 125µg+GR-1 SN (black bars). Comparison within groups was assessed with 2 tailed, One Way ANOVA for IL-12p70 and TNFα in the myometrium and ANOVA on ranks followed by Newman Keuls test for all other cytokines. Statistical significance was denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

CCL3 Saline CCL4 * 3000 b GR-1 SN 600 b LPS 125ug b

2000 LPS 125ug b b b + GR-1 SN 400 b 1000 *** 300 b *** a c 200 a 200 b b b b b a a c a 100 c a a a Concentration (pg/mL) Concentration

Concentration (pg/mL) Concentration a a a a a a a a 0 0 Maternal Myometrium Placenta Amniotic Maternal Myometrium Placenta Amniotic Plasm a Fluid Plasma Fluid

CCL5 CSF2 ** * 2000 b 400 b 1500 c 300 1000 b 500 b b 200 c b b 100 b b b a b b 100 a a 75 a a a 50 a b b a a a a Concentration (pg/mL) Concentration 25 a a a a (pg/mL) Concentration a 0 0 Maternal Myometrium Placenta Amniotic Maternal Myometrium Placenta Amniotic Plasm a Fluid Plasm a Fluid

Figure 3-5 Histogram showing concentrations of chemokines CCL3, CCL4, CCL5 and hematopoietic factor CSF2 in the maternal plasma (n=10 animals per treatment group), myometrium (n=7), placenta (n=7) and amniotic fluid (n=10) of pregnant CD-1 mice (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are 4 treatment groups: saline (white); GR-1 SN (light grey); LPS 125µg (dark grey) and LPS 125µg+GR-1 SN (black bars). Comparison within groups was assessed with 2 tailed, One Way ANOVA on ranks followed by Newman Keuls post-hoc test. Statistical significance was denoted with different letters and as asterisks (*p<0.05; **p<0.01; ***p<0.001).

IL-4 IL-10 8 300 b a,c 250 b 6 a,b,c b b 200 a,b,c b a,b 150 4 b 50 b b a,c b c c 40 a,b b b b a 30 b 2 a b a a a 20 a a a a a Concentration (pg/mL) Concentration Concentration (pg/mL) Concentration 10 0 0 Maternal Myometrium Placenta Amnotic Maternal Myometrium Placenta Amnotic Plasm a Fluid Plasm a Fluid

Figure 3-6 Histogram showing concentrations of anti-inflammatory cytokines IL-4 and IL-10 in the maternal plasma (n=10 animals per treatment group), myometrium (n=7), placenta (n=7) and amniotic fluid (n=10) of pregnant CD-1 mice (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are 4 treatment groups: saline (white); GR-1 SN (light grey); LPS 125µg (dark grey) and LPS 125µg+GR-1 SN (black bars). Comparison within groups was assessed with 2 tailed, One Way ANOVA on ranks followed by Newman Keuls post-hoc test. Statistical significance was denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

80 a a 60 b b 40

0 Saline GR-1 SN LPS125µg LPS125µg +GR-1 SN Maternal Progesterone concentration (ng/mL)

Figure 3-7 Histogram showing maternal plasma progesterone concentrations for different treatment groups (Set 3). Results are mean values ± SEM and expressed in ng/mL (5 animals per group). Comparison within groups was assessed with 2 tailed, One Way ANOVA followed by Newman Keuls post-hoc test. Statistical significance was denoted as different letters (p<0.05).

Table 3-1 Delivery outcome of pregnant CD-1 mice that delivered preterm following different doses of LPS intrauterine injection (Set 1). Mice received 0 (saline), 25µg, 50µg, 125µg, and 250µg LPS (10 animals in each group). One mouse from each of the LPS 65µg group and LPS125µg group had all fetuses absorbed at term. Preterm delivery is defined as delivery of at least one pup within 48 hours of LPS injection. The saline group was compared with each of the four remaining groups using Fisher’s exact test. Statistical significance was denoted with different letters.

LPS dose No. of animals No. of animals (µg) delivered preterm delivered term

Saline 0 a 10 a

LPS 25 µg 4 a 6 a

LPS 65 µg 8 b 1 b

LPS 125 µg 9 b 0 b

LPS 250 µg 10 b 0 b

Table 3-2 Litter size and fetal weight of neonates born to pregnant CD-1 mice that received different treatments (Set 2). Results are mean values ± SEM. One-Way ANOVA followed by Newman Keuls post-hoc test was used to compare the groups with one another. Mice in the LPS 125µg group delivered preterm. Number of animals is indicated in brackets.

Group Litter size Weight per pup P-value (gram)

Saline 13.0 ± 0.89 (n=9) 1.72 ± 0.02 (n=9) > 0.05

GR-1 SN 12.4 ± 0.51 (n=9) 1.73 ± 0.02 (n=9) > 0.05

LPS 125 µg - - -

LPS 125 µg + 10.0 ± 1.04 (n=4) 1.69 ± 0.04 (n=4) > 0.05 GR-1 SN

Table 3-3 Baseline cytokine and chemokine concentrations in the maternal plasma, myometrium, amniotic fluid and placenta of pregnant CD-1 mice (Set 3). Results are mean values ± SEM and expressed in pg/mL (maternal plasma (n=10), myometrium (n=7), placenta (n=7) and amniotic fluid (n=10)). Increases in concentrations following LPS 125µg treatment are indicated in brackets (fold increase).

Cytokine/ Maternal Myometrium Amniotic Placenta Chemokine Plasma Fluid

IL-1α 9.8±2.4 (7) 679.6±170.6 (174) 1.8±0.6 (26) 357.2±222.8 (3) IL-1β 43.6±4.2 (4) 174.9±19.7 (37) 71.9±8.7 (1) 39.5±9.3 (22) IL-2 4.9±1.3 (4) 0.2±0.1 (33) 5.3±0.9 (0.5) < OOR (OOR) IL-3 0.1±0.01 (16) 7.4±2.9 (2) 7.9±1.5 (1) 1.1±0.2 (6) IL-4 0.7±0.1 (4) 1.1±0.2 (4) 3.5±0.6 (1) 0.9±0.1 (3) IL-5 5.9±2.5 (3) 2.4±0.4 (6) < OOR 2.6±0.7 (2) IL-6 11.8±1.8 (110) 6.2±1.2 (275) 9.3±1.1 (190) 5.7±0.7 (28) IL-9 < OOR 331.5±93.8 (1) < OOR 164.6±11.2 (3) IL-10 7.6±3.6 (26) 9.5±1.3 (4) 25.8±3.7 (1) 8.2±1.4 (3) IL-12p40 37.0±6.0 (37) 34.1±15.1 (11) 58.5±6.3 (1) 92.0±29.6 (2) IL-12p70 24.0±7.1 (3) 86.1±21.1 (5) 69.9±11.3 (1) 18.8±4.0 (5) IL-13 21.6±4.1 (4) 18.1±3.3 (11) 75.1±15.3 (1) 16.6±3.0 (6) IL-17 4.4±1.3 (3) 4.2±1.6 (15) 3.5±0.9 (2) 0.9±0.3 (7) CSF2 19.2±1.0 (4) 57.6±9.3 (5) 52.2±3.9 (2) 11.4±3.1 (16) CSF3 2409.5±289.8 (63) 437.2±142.5 (449) 214.8±72.4 (288) 478.7±96.0 (390) IFN-γ 0.6±0.1 (19) 5.1±0.6 (3) 3.0±0.3 (1) < OOR CXCL1 48.1±8.7 (77) 116.3±51.7 56.1±4.7 (220) 1855.3±142.2 CCL2 50.7±14.1 (149) 132.2±40.0 (138) 1508.2±157.1 (2) 77.8±20.0 (8) CCL3 1.4±0.3 (48) 47.2±14.8 (49) 21.6±4.9 (10) 67.2±13.6 (15) CCL4 2.7±0.7 (57) 31.8±5.5 (9) 10.4±1.7 (37) 15.3±1.4 (4) CCL5 8.7±2.4 (186) 6.6±1.3 (49) 10.1±0.8 (7) 1.9±0.6 (8) CCL11 183.8±33.5 (5) 227.8±30.0 (4) 618.8±26.5 (1) < OOR TNFα 19.3±2.3 (4) 38.4±11.3 (3) 51.0±6.0 (3) 7.5±1.4 (5)

< OOR: out of range (below lowest detectable concentration) > OOR: out of range (above highest detectable concentration)

Table 3-4 Cytokine and chemokine concentrations in the maternal plasma of pregnant CD-1 mice following different treatments (Set 3). Results are mean values ± SEM and expressed in pg/mL for each treatment group (10 animals per group). Comparison within groups was assessed with One Way ANOVA for IL- 13 and ANOVA on ranks followed by Newman Keuls post-hoc test for all other cytokines. Statistical significance is denoted by different letters (p<0.05).

Maternal Plasma Saline GR-1 SN LPS125µg LPS 125µg Cytokine +GR-1 SN

a a b b IL-1α 9.8±2.4 16.7±4.0 69.5±6.0 72.5±9.4 IL-1β 43.6±4.2a 56.0±7.9a 181.2±21.9b 98.8±8.2c IL-2 4.9±1.3a 4.9±1.4a 17.6±4.1b 16.7±0.9b IL-3 0.1±0.01a 0.1±0.01a 1.6±0.3b 0.9±0.1b IL-4 0.7±0.1a 0.7±0.1a 2.9±0.7b 2.2±0.3b IL-5 5.9±2.5a 3.4±0.3a 16.5±4.3b 8.9±1.2b IL-6 11.8±1.8a 12.6±1.8a 1300.0±324.4b 362.8±74.9c IL-9 < OOR < OOR < OOR < OOR IL-10 7.6±3.6a 10.8±1.8a 196.5±27.2b 224.5±33.6b IL-12p40 37.0±6.0a 43.2±4.7a 1384.8±280.8b 278.1±74.9c IL-12p70 24.0±7.1a 19.2±8.3a 64.4±13.9b 37.5±7.6b IL-13 21.6±4.1a 30.9±3.5a 76.9±6.1b 65.8±4.1b IL-17 4.4±1.3a 2.3±0.4a 13.9±1.1b 14.4±2.7b CSF2 19.2±1.0a 27.3±3.9a 85.8±3.7b 93.2±3.6b CSF3 2409.5±289.8a 3338.3±515.1a 150981.9±32647.3b 233038.8±63249.0b IFN-γ 0.6±0.1a 0.8±0.2a 11.5±2.9b 18.3±6.9b CXCL1 48.1±8.7a 37.5±8.9a 3702.7±748.8b 3366.0±1096.5b CCL2 50.7±14.1a 132.9±34.7a 7570.9±1186.8b 5916.8±833.6b CCL3 1.4±0.3a 3.6±1.2a 66.9±4.3b 57.8±7.4b CCL4 2.7±0.7a 3.6±0.5a 136.3±18.9b 45.7±8.0c CCL5 8.7±2.4a 15.4±3.8a 1622.4±182.3b 922.1±45.0c CCL11 183.8±33.5a 121.7±32.3a 831.8±91.5b 754.2±175.3b TNFα 19.3±2.3a 25.3±3.4a 80.3±11.1b 50.6±2.4c

< OOR: out of range (below lowest detectable concentration)

Table 3-5 Cytokine and chemokine concentrations in the myometrium of pregnant CD- 1 mice following different treatments (Set 3). Results are mean values ± SEM and expressed in pg/mL for each treatment group (7 animals per group). Comparison within groups was assessed with One Way ANOVA for IL-4, IL- 12p70, IFN-γ and TNFα, and ANOVA on ranks followed by Newman Keuls post-hoc test for all other cytokines. Statistical significance is denoted by different letters (p<0.05).

Myometrium Saline GR-1 SN LPS125µg LPS 125µg Cytokine +GR-1 SN

IL- 1α 679.6±170.6 a 974.8±70.2 a 118194.4±112940.0 b 34851.9±20441.5 b IL-1β 174.9±19.7a 373.8±117.8a 6422.0±837.8b 4991.7±1111.0b IL-2 0.2±0.1a 0.19±0.1a 6.6±3.9b 4.1±1.4b IL-3 7.4±2.9a,b 6.2±2.5a 15.6±2.3b 10.2±1.3a,b IL-4 1.1±0.2a 1.6±0.1a 4.4±0.4b 3.8±0.2b IL-5 2.4±0.4a 4.3±1.6a 13.5±7.2b 9.8±3.5b IL-6 6.2±1.2a 12.1±2.3a 1704.4±494.6b 291.1±67.5c IL-9 331.5±93.8a 248.5±113.1a 329.6±72.5a 196.5±85.5a IL-10 9.5±1.3a 8.9±0.5a 36.9±4.3b 37.8±4.8b IL-12p40 34.1±15.1a 35.5±6.3a 362.9±81.0b 400.3±121.9b IL-12p70 86.1±21.1a 99.0±19.2a 400.4±39.3b 295.4±36.7c IL-13 18.1±3.3a 24.7±5.9a 206.8±21.5b 133.1±19.3c IL-17 4.2±1.6a 3.8±1.4a 63.0±16.7b 16.2±3.3c CSF2 57.6±9.3a 58.2±7.9a 313.4±50.2b 168.8±15.0c CSF3 437.2±142.5a 1434.5±834.1a 196003.8±66522.5b 184154.1±118930.5b IFN-γ 5.1±0.6a 4.6±0.9a 17.8±1.3b 13.9±1.3c CXCL1 116.3±51.7 232.8±118.9 > OOR 39265.4±6534.7 CCL2 132.2±40.0a 195.2±82.5a 18245.8±5702.7b 45343.6±23004.8b CCL3 47.2±14.8a 148.4±49.7a 2299.8±471.2b 1981.6±349.7b CCL4 31.8±5.5a 48.1±12.1a 300.3±53.9b 319.1±86.4b CCL5 6.6±1.3a 27.5±10.7a 325.7±46.8b 313.8±72.9b CCL11 227.8±30.0a 233.5±23.5a 861.7±203.5b 670.7±154.4b TNFα 38.4±11.3a 32.6±10.5a 128.7±19.2b 78.8±10.6a

> OOR: out of range (above highest detectable concentration)

Table 3-6 Cytokine and chemokine concentrations in the amniotic fluid of pregnant CD-1 mice following different treatments (Set 3). Results are mean values ± SEM and expressed in pg/mL for each treatment group (10 animals per group). Comparison within groups was assessed with One Way ANOVA for IL- 1β, IL-2, IL-12p70, CSF2 and IFN-γ, and ANOVA on ranks followed by Newman Keuls post-hoc test for all other cytokines. Statistical significance is denoted by different letters (p<0.05).

Amniotic Fluid Saline GR-1 SN LPS125µg LPS 125µg Cytokine +GR-1 SN

IL- 1α 1.8±0.6 a 2.6±1.1 a 46.4±9.1 b 35.8±8.7 b IL-1β 71.9±8.7a 75.5±11.2a 75.9±13.3a 65.4±9.0a IL-2 5.3±0.9a 5.5±0.4a 2.9±0.5b 3.7±0.7a,b IL-3 7.9±1.5a 7.6±1.3a 10.2±1.6a 7.9±1.6a IL-4 3.5±0.6a,b 1.9±0.3b 5.1±0.9a,c 4.0±0.9a,b,c IL-5 < OOR < OOR < OOR < OOR IL-6 9.3±1.1a 9.2±0.9a 1767.4±584.6b 365.6±35.9c IL-9 < OOR < OOR < OOR < OOR IL-10 25.8±3.7a,b 22.7±3.0b 40.6±5.6a,c 33.0±4.4a,b,c IL-12p40 58.5±6.3a 51.3±5.0a 73.0±6.7a 65.2±5.8a IL-12p70 69.9±11.3a,b 46.7±7.8b 93.1±7.4a,c 76.3±8.5a,b,c IL-13 75.1±15.3a 90.8±13.1a 112.4±15.9a 77.3±8.5a IL-17 3.5±0.9a 2.0±0.7a 6.2±1.6a 4.1±1.0a CSF2 52.2±3.9a 49.7±4.9a 97.6±15.0b 88.3±8.6b CSF3 214.8±72.4a 297.9±118.6a 61939.9±28767.9b 33742.1±11824.3b IFN-γ 3.0±0.3a 4.2±0.5a 4.7±0.8a 4.0±0.4a CXCL1 56.1±4.7a 65.9±12.8a 12298.9±3378.9b 9714.0±3461.3b CCL2 1508.2±157.1a 1251.1±113.3a 3538.2±819.7b 2678.4±538.3b CCL3 21.6±4.9a 23.2±3.5a 222.8±33.8b 75.5±16.4c CCL4 10.4±1.7a 13.2±2.0a 389.5±138.1b 138.1±47.9c CCL5 10.1±0.8a 8.4±0.8a 66.6±11.5b 47.3±8.9b CCL11 618.8±26.5a 572.8±82.4a 578.6±116.2a 481.7±94.9a TNFα 51.0±6.0a 40.3±3.3a 162.8±17.5b 89.5±16.4c

< OOR: out of range (below lowest detectable concentration)

Table 3-7 Cytokine and chemokine concentrations in the placenta of pregnant CD-1 mice following different treatments (Set 3). Results are mean values ± SEM and expressed in pg/mL for each treatments group (7 animals per group). Comparison within groups was assessed with 2 tailed, One Way ANOVA for IL-12p40, CCL5 and TNFα, and ANOVA on ranks followed by Newman Keuls post-hoc test for all other cytokines. Statistical significance is denoted by different letters (p<0.05).

Placenta Saline GR-1 SN LPS125µg LPS 125µg Cytokine +GR-1 SN

a a b b IL-1α 357.2±222.8 398.5±196.0 905.8±130.6 661.2±89.6 IL-1β 39.5±9.3a 139.8±83.8a 875.2±275.9b 662.1±23.7b IL-2 < OOR < OOR < OOR < OOR IL-3 1.1±0.2a 2.8±1.6a 7.1±2.9b 5.3±1.8b IL-4 0.9±0.1a 3.6±1.2b 2.3±0.1c 2.2±0.1c IL-5 2.6±0.7a 3.1±0.6a 6.2±1.5b 5.8±1.0b IL-6 5.7±0.7a 9.25±4.3a 169.2±22.0b 42.7±4.2c IL-9 164.6±11.2a 147.7±29.8a 551.2±217.7b 511.0±138.9b IL-10 8.2±1.4a 14.2±1.6b 25.9±1.4b 22.8±2.0b IL-12p40 92.0±29.6a 120.8±33.7a,b 227.1±33.8b 162.5±25.9a,b IL-12p70 18.8±4.0a 37.9±13.6a 93.7±12.3b 34.7±9.8a IL-13 16.6±3.0a 41.5±21.5a 96.9±10.2b 94.6±7.7b IL-17 0.9±0.3a 1.74±0.4a 6.3±1.3b 4.8±1.5b CSF2 11.4±3.1a 57.9±39.8a 187.3±35.2b 110.2±9.4b CSF3 478.7±96.0a 2400.1±1706.7a 186810.4±54388.5b 192855.0±126303.9b IFN-γ < OOR < OOR < OOR < OOR CXCL1 1855.3±142.2 3846.7±2108.0 > OOR > OOR CCL2 77.8±20.0a 89.7±6.9a 649.2±214.6b 877.2±229.6b CCL3 67.2±13.6a 121.3±31.6a 992.8±200.7b 910.2±150.6b CCL4 15.3±1.4a 17.1±2.3a 58.6±11.9b 51.1±4.9b CCL5 1.9±0.6a 2.6±0.8a 15.3±2.1b 12.4±1.7b CCL11 < OOR < OOR < OOR < OOR TNFα 7.5±1.4a 10.5±2.6a 39.2±6.0b 30.9±3.9b

< OOR: out of range (below lowest detectable concentration) > OOR: out of range (above highest detectable concentration)

Chapter Four

Oral Probiotic Lactobacillus rhamnosus GR-1 stimulates systemic and intrauterine production of cytokines and chemokines and modulates the vaginal microbiota in pregnant CD-1 mice.

I would like to thank Dr. Gregory Gloor for his advice on the analyses of sequencing data and for his help on filtering and organizing the data into operational taxonomic unit tables. I would like to express my gratitude to Dr. David Carter at the Robarts Research Institute (London, Ontario, Canada) for performing the Ion torrent sequencing.

68 69

Chapter 4

4. Oral Probiotic Lactobacillus rhamnosus GR-1 stimulates systemic and intrauterine production of cytokines and chemokines and modulates the vaginal microbiota in pregnant CD-1 mice.

4.1 Introduction

Cytokines and chemokines play a pivotal role in infection/inflammation-induced preterm labor (PTL) (Challis et al., 2009). The intra-uterine tissues (amnion, chorion, placenta, decidual and myometrium), as well as the leukocytes infiltrating these tissues, are potential sources of cytokines and chemokines (Young et al., 2002). Pro- and anti-inflammatory cytokines balance the production of one another throughout pregnancy and during labor (Keelan et al., 2003). A shift to a pro-inflammatory bias ends uterine quiescence and leads to the onset of parturition (Challis et al., 2009). Chemokines recruit immune cells, phagocytize pathogens and induce pathogenic cell lysis (Hamilton et al., 2013). Recruited immune cells can also produce more pro-inflammatory cytokines, which amplify the inflammatory cascade leading to PTL (Hamilton et al., 2013). Chemokines such as IL-8 in the cord blood, cervical and amniotic fluid are increased in association with preterm birth (PTB) and cervical ripening (Sennstrom et al., 2000; Jacobsson et al., 2005; Matoba et al., 2009).

An abnormal vaginal microbiota, such as that found in bacterial vaginosis (BV) that is characterized by a depletion of endogenous vaginal lactobacilli, has been associated with an increased risk of PTB (Donders et al., 2009). Probiotics are defined as “live microorganisms which, when administered in adequate amounts, confer a health benefit to the host” (FAO/WHO, 2001). Lactobacillus spp. are normal commensals of the human vaginal microbiota, and have been used to treat urogenital infections and reduce bacterial vaginosis (BV) occurrence (Reid, 2012, Reid, 2001a). Previous studies have demonstrated that probiotic L. rhamnosus GR-1 supernatant (GR-1 SN) increases anti-inflammatory cytokine IL-10 production while reducing lipopolysaccharides (LPS)-induced pro-inflammatory

70 cytokine TNF-α production in cultured human placental trophoblast cells (Yeganegi et al., 2009; Yeganegi et al., 2011) and decidual cells (Li et al., 2014).

L. rhamnosus GR-1 and L. reuteri RC-14 live bacteria taken orally at a daily dose of 109 to 1010 colony-forming units (cfu) decrease BV relapse and re-establish the vaginal ecosystem in non-pregnant women (Reid et al., 2003a; Reid, 2012). I wished to investigate whether a higher dose of lactobacilli leads to an improved efficacy, using varying doses of lactobacilli in pregnant CD-1 mice in this study. The mouse gut microbiota resembles that of the adult human from the family to the phylum taxonomic level, with Bacteroidetes and Firmicutes being the most abundant phyla in both human and mouse gut microbiota (Kostic et al., 2013). The vaginal microbiota of non-pregnant BALB/cJ mice (Meysick and Garber, 1992) and the vaginal microbiota of women diagnosed with BV share the common characteristic of low lactobacilli abundance, and this gives the opportunity to detect potential lactobacilli colonization after an exogenous administration of oral lactobacilli in mice.

In Chapter 3 of this thesis, I demonstrated that GR-1 SN, harvested from approximately 108 – 109 cfu of GR-1 live bacteria per mL, reduces LPS-induced PTB in mice. Therefore, in this study, I treated mice with 109 cfu of GR-1 orally. Furthermore, oral administration of 109 - 1011 cfu of lactobacilli is the dose range used in previous studies to improve human vaginal health (Mastromarino et al., 2013; Homayouni et al.. 2014). To calculate the mouse equivalent dose, I used the following factors (Km of mouse = 3; Km of human = 37) to account for the difference in the body surface area, and I also took into consideration the difference in body weight (mouse ≈ 20g; human ≈ 60kg) (Reagan-Shaw et al., 2008). The detailed calculations are shown in Figure 4-1. The normal gestational length of pregnant CD- 1 mice is 19-20 days, and gestational day (GD) 9 to GD 15 is equivalent to the second trimester in human pregnancy. I chose to treat the pregnant mice for 7 consecutive days (GD 9-15) since this corresponds to the 12 weeks treatment protocol used in pregnant women in the next chapter (Chapter 5).

Mice are widely used to study the mechanisms underlying human PTB and an established model of intrauterine infection using 250µg of LPS to induce 100% preterm delivery in

71 pregnant CD-1 mice with no maternal mortality has been developed (Elovitz et al., 2003). Building on our previous studies that demonstrated GR-1 SN reduces LPS-induced PTB as well as systemic and intrauterine inflammation (Chapter 3), in this study, I evaluated whether oral GR-1 live bacteria has similar anti-inflammatory properties in the mouse. I hypothesized that oral GR-1 can reduce LPS-induced PTB, and GR-1 alone can increase anti-inflammatory cytokines in the plasma, amniotic fluid and intrauterine tissues in pregnant CD-1 mice. Furthermore, I hypothesized that oral GR-1 would modulate both the mouse vaginal and cecal (gut) microbiota.

4.2 Material and Methods

4.2.1 Animals

Female HSD:ICR (CD-1) outbred mice (8-12 weeks old; Harlan Laboratories) were bred and the morning of vaginal plug detection was designated as GD 1. Animals were handled in accordance with guidelines of the Canadian Council for Animal Care and all procedures were approved by the Animal Care Committee of Toronto Center for Phenogenomics (Animal Use Protocol #0164). Animals were housed in a pathogen-free, humidity controlled 12 h light:12 h dark cycle animal facility with free access to food and water. I performed 4 sets of independent experiments with a total of 180 animals.

4.2.2 Lactobacillus rhamnosus GR-1 preparation

Lactobacillus rhamnosus GR-1 (GR-1) was grown for 8-10 hours anaerobically at 37 oC in de Man, Rogosa, and Sharpe (MRS) broth (Becton Dickinson, Ontario) to an optical density of ~0.9 at 600 nm (representing ~108 -109 cfu per mL of bacteria), and then centrifuged at 3000 rpm for 10 min at 25 oC. The GR-1 pellet was then washed twice with sterile saline, centrifuged, and re-suspended in saline to obtain a final concentration of 108, 109 or 1010 cfu.

4.2.3 Intra-uterine injection of LPS by mini-laparotomy

Intrauterine injection of LPS was given via mini-laparotomy on GD 15 as previously described (Elovtiz et al., 2003). Mice were anesthetized with isoflurane inhalation and given analgesic buprenorphine (0.1mg/kg). An incision (~1cm) was made to expose the lower segments of the uterine horns. Sterile saline (100µL) or LPS (Escherichia coli 055:B5, Sigma-Aldrich, St. Louis) dissolved in 100µL sterile saline was injected between the two lowest gestational sacs of either the left or right uterine horn. Fascia and skin were closed with 4.0 vicryl sutures and staples, respectively. Mice were housed in individual cages.

4.2.4 Oral administration of GR-1 by oral gavage

Mice received 100-300µL of either 109 cfu of GR-1 or saline by oral gavage using an autoclaved animal feeding needle (Richtree, NY, USE) once daily from GD 9 to GD 15. In this study, I chose the oral route over the vaginal route so I could more accurately measure the dose administered since GR-1 inoculum may leak after vaginal instillation. GR-1 was given via oral gavage instead of in drinking water because GR-1 live bacteria sediment to the bottom of the water bottle with time.

4.2.5 Effect of oral GR-1 on the timing of LPS-induced PTB (Set 1)

Mice were randomly assigned to receive either saline or GR-1 via oral gavage once daily from GD 9 to GD15. On GD15, approximately 30 minutes after the last dose of GR-1 or saline, the animals were divided to receive saline, LPS 25µg or LPS 50µg via mini- laparotomy (Elovitz et al., 2003). A separate group of animals (sham group) received neither oral gavage nor mini-laparotomy. There were seven groups in Set 1, with 11 animals in each group (Figure 4-2). Animals were then housed in individual cages and monitored hourly until term (GD 19-20) for the delivery of pups. Time (hours) to delivery, fetal weight and litter size were recorded. PTB was defined as delivery of at least one pup within 48 hours (GD 17) of LPS injection.

4.2.6 Effect of oral GR-1 on the gestational length (Set 2)

Mice received either 100-300µL of GR-1 (108, 109 or 1010 cfu) or saline by oral gavage once daily from GD 9 to 15, and were housed in individual cages, monitored hourly until term (GD 19-20) for the delivery of pups. A separate group of animals (sham group) did not receive oral gavage. There were five groups in Set 2, with 11 animals in each group (Figure 4-3). Time (hours) to delivery, fetal weight and litter size were recorded.

4.2.7 Effect of oral GR-1 on cytokines and chemokines (Set 3)

Mice were randomly assigned into four groups (Figure 4-4). The animals received 100- 300µL of 1) saline (n=13), 2) GR-1 108 (n=7), 3) GR-1 109 (n=8) or 4) GR-1 1010 cfu (n=6) by oral gavage once daily from GD 9 to 15. After the last dose of GR-1 or saline on GD 15, mice were anesthetized by isoflurane inhalation and maternal blood was collected by cardiac puncture. Blood was centrifuged at 5,000 xg for 15 min at 4oC and plasma was transferred into a clean tube and stored at -80 oC. Mice were then euthanized in a carbon dioxide chamber, and both uterine horns were dissected and kept in ice-cold phosphate buffer solution (PBS). The amniotic fluid was collected using a 1mL syringe with 27-gauge needle, pooled from all gestational sacs and centrifuged to remove any cellular debris before stored at -80 oC. Placental tissue was separated from decidua and fetal membranes in ice-cold PBS. Myometrium was obtained by scraping off the endometrium on a petri dish cover, which was kept on top of ice. Each intra-uterine tissue (fetal membrane, placental, myometrial and decidual tissues) was dissected with sterile instruments and pooled from all fetuses in a given mouse. All tissue samples were flash-frozen in liquid nitrogen and stored at -80 oC.

4.2.8 Effect of oral GR-1 on the vaginal and cecal microbiota (Set 4)

Mice were randomly assigned to two groups (Figure 4-5). They received 100-300µL of either saline or 109 cfu of GR-1 by oral gavage (n=7 per group) once daily from GD 9 to GD 15. After the last dose of GR-1 or saline on GD 15, mice were euthanized with carbon dioxide. The vagina was everted using sterile tweezers and ~1/3 cm of the vaginal tissue was removed.

The cecum pouch (~3/4 cm) long was identified at the beginning of the large intestine and was dissected free. Both vaginal and cecal tissues were stored at -20 oC. In one vaginal sample and one cecal sample (saline group) and one cecal sample (GR-1 group), insufficient sample was available for analysis.

4.2.9 Cytokine Assay

Cytokine and chemokine concentrations were determined in duplicate using a mouse 23- multiplex cytokine assay (Appendix I, Biorad, Ontario) on a Luminex 200 cytometer and Bioplex HTF (Bio-Rad). The assay measured concentrations of interleukin (IL)-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-17, IFN-γ, CXCL 1, CCL2, CCL3, CCL4, CCL5, CCL11, TNFα, CSF2, and CSF3. Data analysis was performed using Bio-Plex Manager (version 5.0, Bio-Rad) and results are presented as concentrations (pg/mL). There were two plasma samples and two amniotic fluid samples that had insufficient protein for assay. Samples from 7 animals in the saline group and 7 samples in the GR-1 109 cfu group were used for intrauterine tissue analyses. Tissues were crushed and homogenized in EDTA-free protease inhibitor containing RIPA lysis buffer (1mL per 0.5g of tissue). Homogenized samples were left on ice for 45 minutes before being centrifuged at 12,000 xg for 15 minutes at 4°C to collect the supernatant. Protein concentration was measured by Bradford assay kit (Bio-Rad, Ontario) with bovine serum albumin as standard. 250µg of total protein was used for the measurement of cytokines and chemokines in the tissues.

4.2.10 Maternal progesterone measurement

Plasma progesterone concentration was measured with an Enzyme Immunoassay kit (Appendix II, Cayman Chemical Co, Michigan). Samples were diluted 400X with EIA buffer and assayed in duplicate. The intra and inter-assay coefficients of variation were 5.4 % and 11.2 % respectively.

4.2.11 DNA isolation and V6 ribosomal DNA PCR amplification

DNA was isolated from the vaginal and cecal tissues using PowerSoil DNA Isolation Kit (Appendix III, VWR, Ontario, Canada). The bacterial DNA was PCR amplified using barcoded primers targeting the V6 region of 16S ribosomal DNA (rDNA) with colorless GO- Taq hot start master mix (Promega, Ontario, Canada) for 25 repeating cycles of 95°C, 55°C and 72°C for 1 minute each step. The amplified products were then quantified using the QuBit broad-range double-stranded DNA fluorometric quantitation reagent kit (Life technologies, Ontario, Canada). Samples were pooled at equal molar concentrations and purified using a Wizard PCR Clean-Up Kit (Promega, Ontario, Canada) prior to sequencing.

4.2.12 Sequencing

Barcoded DNA was sequenced in pairs on the Ion Torrent platform (316 DNA chips, 12 samples per chip) at the Robarts Research Institute (Western University, Canada). The sequence results were provided in a fastq format. All sequences were filtered and a table of counts was generated for each sample containing sequences grouped at 97% operational taxonomic unit (OTU) and 100% identical sequence unit identity. Sequences were then classified to distinct taxonomic species using the online Ribosomal Database Project (http://rdp.cme.msu.edu/seqmatch/seqmatch_intro.jsp). Sequences not identical across all best matches were marked as unclassified.

4.2.13 Statistical Analysis

Statistical analyses of the cytokine, chemokine and progesterone data were carried out using SigmaStat (version 3.5). Comparison of PTB rate was made with Fisher exact analyses (two tailed). One-Way ANOVA was used to detect a difference in gestational length, litter size and fetal weight following different treatments. Comparison between multiple groups in the maternal plasma and amniotic fluid were carried out with One Way ANOVA followed by Tukey test. Kruskal-Wallis ANOVA on Ranks followed by Dunn’s method was used for data that were not normally distributed. Comparison between two groups in the intrauterine

76 tissues was performed with unpaired Student’s t-test or Mann-Whitney Rank Sum Test. The sequencing data was centered ratio logarithm transformed (Aitchison, 1986) before performing statistical analyses with R (version 3.0.1). Briefly, the geometric mean of the proportions of all species detected in a sample was computed. A ratio x was determined from the proportion of species i over the geometric mean. Then, the relative abundance of species i was calculated by taking natural logarithm of x. Both protein and sequencing data were tested for normality and equal variance and were expressed as mean values ± SEM or mean values ± SD. Data were adjusted for false discovery rate using Benjamini Hochberg procedure and an adjusted p-value of p<0.05 was considered statistically significant. The Shannon diversity index was calculated by first taking the proportion of a bacterial species relative to the total number of species detected in a given sample, and multiplied the value by the natural logarithm of this proportion. The product was then summed across all bacterial species, and multiplied by -1 (Magurran, 2003).

4.3 Results

4.3.1 Effect of oral GR-1 on the incidence of LPS-induced PTB and gestational length (Set 1 and Set 2)

Intrauterine injection of LPS 25µg on GD 15 resulted in 36% PTB (4 out of 11 animals) and pretreatment with oral GR-1 led to 64% PTB (7 out of 11 animals) (p>0.05, Table 4-1). Mice that received LPS 50µg had 100% PTB (11 out of 11 animals), and the incidence of PTB did not change with oral GR-1 pretreatment (p>0.05, Table 4-1). Animals in the sham, saline and GR-1 treated groups delivered live pups at term (Table 4-1). The mean litter size was 12.5 ± 0.37 and the mean weight per pup was 1.70 ± 0.09 grams in the sham group. These were not different between different treatment groups (p>0.05, Table 4-2).

The mean hours to delivery were 106 ± 3.3 hours, the mean litter size was 12.4 ± 0.36, and the mean weight per pup was 1.68 ± 0.08 grams in the sham group. These were not different in mice that received saline or different doses of GR-1 (p>0.05, Table 4-3).

4.3.2 Effect of oral GR-1 on the cytokines and chemokines (Set 3)

The concentrations of the pro-inflammatory cytokine TNFα in the maternal plasma increased with 108 and 109 cfu of oral GR-1 (p<0.05, Figure 4-6). The concentrations of plasma IL- 12p40, as well as IL-6 and IFN-γ in the amniotic fluid increased with GR-1 (1010 cfu) treatment (p<0.05, Figure 4-6). Concentrations of TNFα and IL-17 increased in the placenta, as did IL-12p70 in the fetal membranes and IL-1α in the myometrium with GR-1 (109 cfu) treatment (p<0.05, Figure 4-7). There was no significant change in any of the pro- inflammatory cytokines measured in the decidua with GR-1 treatment (p>0.05, Figure 4-7). There was no change in the concentration of IL-1β in the maternal plasma, amniotic fluid or any of the intrauterine tissues between the different treatment groups (p>0.05, Figure 4-6 and Figure 4-7).

The concentrations of the anti-inflammatory cytokines IL-2, IL-4 and IL-10 did not change in the maternal plasma, amniotic fluid, or in the placenta and decidua between the different treatment groups (p>0.05, Figure 4-8 and Figure 4-9). Concentrations of IL-10 increased in the fetal membranes and IL-4 in the myometrium with GR-1 (109 cfu) treatment (p<0.05, Figure 4-9). GR-1 decreased the concentration of IL-4 in the fetal membranes, IL-10 in the myometrium and IL-2 in the placenta (p<0.05, Figure 4-9). The concentration of IL-13 did not change in any compartment between the different treatment groups (p>0.05, Figure 4-8 and Figure 4-9).

GR-1, at a dose of 1010 cfu, significantly increased the concentrations of chemokine CCL2, CCL3, CCL4, CCL5 and CCL11 in the amniotic fluid (p<0.05, Figure 4-10). There was no change in chemokine concentrations in the maternal plasma, placenta or decidua, following GR-1 treatment (p>0.05, Figure 4-10 and Figure 4-11). The concentration of CCL5 increased with GR-1 (109 cfu) treatment in the fetal membranes, and decreased in the myometrium (p<0.05, Figure 4-11). GR-1 did not alter the concentration of CXCL1 in the maternal plasma, amniotic fluid or in any of the tissues (p>0.05, Figure 4-10 and Figure 4-11).

The concentration of hematopoietic factors CSF2 and CSF3 did not change in the maternal plasma, amniotic fluid, or tissues with GR-1 treatment alone at any dose (p>0.05, Figure 4- 12 and Figure 4-13). The concentration of IL-3 also did not change in the maternal plasma and amniotic fluid (p>0.05, Figure 4-12). The concentration of IL-3 increased with GR-1 (109 cfu) treatment in the fetal membranes and placenta, and decreased in the myometrium (p<0.05, Figure 4-13).

The concentrations of IL-5 and IL-9 were below the detection limits of the assay in the maternal plasma, amniotic fluid and tissues. With GR-1 treatment (1010 cfu), the concentrations of CCL4 and CCL11 in the maternal plasma (Figure 4-10) and CSF3 in the amniotic fluid (Figure 4-12) were below the limits of assay detection. The changes in cytokines and chemokines with GR-1 treatment are summarized in Table 4-4.

4.3.3 Maternal plasma progesterone (Set 3)

There was no difference in plasma progesterone concentrations between mice that received saline (39.5 ± 4.4 ng/ml) and varying doses of GR-1 (108 cfu: 50.9 ± 4.4 ng/ml; 109 cfu: 48.6 ± 4.9 ng/ml and 1010 cfu: 47.1 ± 7.0 ng/ml) (p>0.05, Table 4-5).

4.3.4 Vaginal and Cecal Microbiota (Set 4)

Sixty-two bacterial genera were detected in the vaginal tissues and 44 genera were identified in the cecal tissues (Table 4-6 and 4-7). There were 24 bacterial genera unique to the vaginal microbiota and 6 genera unique to the cecal microbiota (Table 4-6). The major bacterial orders in the cecum of saline-treated mice were Bacteroidales and Clostridiales, while Bacillales, Deinococcales and Pasteurellales dominated the vaginal microbiota in these mice (Figure 4-14). There was no difference in the Shannon diversity index (SDI) between the vaginal tissues and cecal tissues of saline-treated mice (p>0.05, Figure 4-15).

The relative mean abundance of 8 bacterial orders: Lactobacillales, Pseudomonadales, Actinomycetales, Enterobacteriales, Hydrogenophilales, Neisseriales, Xanthomonadales, and

Chromatiales were higher in the vaginal tissues than in the cecal tissues (Table 4-8). The relative mean abundance of 5 bacteria order Bacteroidales, Clostridiales, Deinococcales, Achleplasmatales and Opitutales were higher in the cecal tissues than in the vaginal tissues (Table 4-9). The significance differences between the vaginal and cecal tissues of saline treated pregnant CD-1 mice at lower taxonomic levels are summarized in Table 4-8 and Table 4-9.

4.3.5 Effect of oral GR-1 on the vaginal microbiota (Set 4)

The relative abundance of bacteria order Bacillales, Pseudomonadales, Burkholderiales, Hydrogenophilales decreased with oral GR-1 treatment (p<0.05, Table 4-10). Oral GR-1 significantly increased the relative abundance of bacteria order Bacteroidales and Clostridiales (p<0.05, Table 4-11). The significance differences at lower taxonomic levels in the vaginal microbiota between saline and GR-1 treated pregnant CD-1 mice are summarized in Table 4-10 and Table 4-11. There was no difference in the SDI ratio in the vaginal tissues between the saline and GR-1-treated mice (p>0.05, Figure 4-15).

4.3.6 Effect of oral GR-1 on the cecal microbiota (Set 4)

The SDI ratio was higher in the cecal tissues of GR-1 treated mice compared to saline treated mice (p<0.05, Figure 4-15). The oral administration of GR-1 to pregnant CD-1 mice had no effect on their cecal microbial profiles. There was no change in the abundance of Lactobacillus in the mouse vaginal and cecal tissues. Although the vaginal microbiota appears to resemble the cecal microbiota in pregnant mice treated with GR-1, there was no statistical significant difference between the two groups (p>0.05, Figure 4-14).

4.4 Comment

In this study, I have shown that the oral administration of Lactobacillus rhamnosus GR-1 live bacteria at a dose of 109 cfu does not reduce LPS-induced PTB nor does it have an effect

80 on the gestational length, fetal weight, litter size and the maternal circulating progesterone concentration in pregnant CD-1 mice. Oral GR-1 live bacteria given alone can modulate the systemic and intrauterine cytokines and chemokines.

Pretreatment with oral GR-1 live bacteria does not alter the incidence of LPS-induced PTB, which is in contrast to our previous findings that GR-1 SN reduces LPS-induced PTB (Yang et al., 2014b). Oral GR-1 alone stimulates the production of both systemic and intra-uterine pro-inflammatory cytokines and chemokines. These findings are consistent with a previous report that demonstrated in human decidual cells, L. rhamnosus CNCM I-4036 stimulate the production of various pro-inflammatory cytokine and chemokines (Bermudez-Brito et al., 2014). One plausible explanation is that the lipoteichoic acid (LTA) on the cell surface of lactobacilli live bacteria may stimulate immune cells such as macrophages to secrete inflammatory cytokines. When LTA is removed or modified (D-alanylation), improved anti- inflammatory activity has been observed in a murine model of colitis (Grangette et al., 2005; Claes et al., 2010; Mohamadzadeh et al., 2011).

The concentration of various pro-inflammatory cytokines increased following GR-1 treatment in the maternal plasma, amniotic fluid as well as in the intra-uterine tissues. In the maternal plasma, pro-inflammatory cytokine was observed to increase significantly starting at the lowest dose of GR-1 (108 cfu) (Figure 4-6), whereas a higher dose (1010 cfu) was needed to elicit an increase in the concentration of various pro-inflammatory cytokines and chemokines in the amniotic fluid (Figure 4-6 and Figure 4-10). Inflammatory cytokines including TNFα, IL-1, and IL-6 have been implicated in the pathogenesis of human PTB (Challis et al., 2009). Mice with the TNFα and IL-1 receptors knocked out (Hirsch et al., 2006) or mice deficient in the IL-6 gene have delayed PTB compared to wild-type mice (Robertson et al., 2010). In this study, a number of chemokines increased in the amniotic fluid with GR-1 treatment, including CCL2 and CCL4. Elevated levels of CCL2 have been observed in the mid-trimester amniotic fluid of women who delivered preterm (La Sala et al., 2012). The concentration of amniotic fluid CCL4 has also been noted to be higher in women with clinical signs of intrauterine infection and/or inflammation compared to women who were asymptomatic (Weissenbacher et al., 2013). In this study, GR-1 increased the

81 production of IFN-γ, which has been shown to have antimicrobial properties and promotes pathogen elimination (Mak, 2006). GR-1 also increased the concentration of CCL2, which is important in pathogen phagocytosis (Mak, 2006). The concentration of IL-12 increased with GR-1 treatment. IL-12 is important for the differentiation of Th0 cells into Th1 cells, which play an important role in cell-mediated immune responses (Mak, 2006). Taken together, GR- 1 given orally promotes some degree of systemic and intra-uterine inflammation in pregnant CD-1 mice.

Despite an increase in the concentration of TNFα and IL-6 with GR-1 treatment, labor was not initiated, even at the highest doses of GR-1 (1010 cfu) used in this study. Furthermore, the elevation in plasma TNFα was not sustained at a higher dose of GR-1 (1010 cfu), suggesting there may be a degree of tolerance to excess GR-1 stimulation. Although the anti- inflammatory cytokines IL-4 and IL-10 decreased with GR-1 treatment in fetal membranes and myometrium respectively, the concentrations of IL-10 and IL-4 in the fetal membranes and myometrium increased respectively, suggesting the anti-inflammatory cytokines may interact with each other. This would be consistent with a previous study that has shown that IL-4 dampens the production of IL-10 in dendritic cells (Yao et al., 2005). Oral GR-1 also increased IL-2 in the placenta. IL-2 has been demonstrated to inhibit IL-1β-induced PGE2 production in human amnion cells, and PGE2 production in culture chorion and decidua cells (Coulam et al., 1993a; Coulam et al., 1993b).

Oral GR-1 increased placental IL-17 and IL-3 concentrations. IL-17 promotes the process of trophoblast invasion and angiogenesis, which are important in the establishment of placental vasculature (Pongcharoen et al., 2006; Pongcharoen et al., 2007) and IL-3 is involved in the differentiation and invasiveness of human trophoblast cells (Di Simone et al., 2000). Taken together, these findings suggest that GR-1 may also affect the processes of angiogenesis and placenta development in pregnant mice.

In this study, 44% (30 out of 68) of the identified bacterial genera were unique to either the vaginal microbiota or the cecal microbiota of pregnant CD-1 mice although the common bacteria genera shared by both the cecal and vaginal microbiota were present in different

82 relative abundances. Bacteroides and Barnesiella were found in both the vaginal and cecal microbiota (Table 4-7) but the relative abundance of the two genera were significantly higher in the cecal tissues than in the vaginal tissues (Table 4-9). These observations are in agreement with a previous study, which reported the cecal and vaginal microbiota in non- pregnant BALB/cJ mice each have its own distinct bacterial genera although there is some overlap with each other (Barfod et al., 2013). These investigators have reported that the vaginal and cecal microbiota only have bacterial genera Ruminococcus in common and that three bacterial genera, Robinsoniella, Parasutterella and Ramlibacter, are unique to the cecal microbiota (Barfod et al., 2013). In contrast, I have identified 38 bacterial genera shared by both the vaginal and cecal microbiota (Table 4-7). It is known that different strains of mice, C3H, Balb/c, Nude FoxN1nu and C57BL/6J mice display their own unique gut microbiomes (Gutierrez-Orozco et al., 2015). In the human, the vaginal microbiome between pregnant and non-pregnant women has also been found to be different, with a higher abundance of various Lactobacillus spp. observed in pregnant women (Romero et al., 2014a). There is little information known about the vaginal microbiome of pregnant mice versus non-pregnant mice in different strains. The microbiome analysis is also influenced by the choice of PCR primer that targets the 16S rDNA (Kuczynski et al., 2012). Primers that target the V6 region have been reported to overestimate species richness (Youssef et a., 2009). Therefore, the differences between the previous study (Barfod et al., 2013) and this study might be the result of the use of a different strain of mice (CD-1 versus BALB/cj), the pregnancy status (pregnant versus non-pregnant mice), a difference in sequencing protocol (primers target the V6 region versus the V3-4 region of 16S rDNA) or a difference in housing conditions.

In this study, Clostridiales and Bacteroidales, which belong to the phyla Firmicutes and Bacteroidetes respectively, were found to dominate the cecal microbiota of pregnant CD-1 mice. This is in agreement with a previous study, which found that Firmicutes and Bacteroidetes were the most abundant phyla in the cecal microbiota of non-pregnant BALB/cJ mice (Barfod et al., 2013). I have found Pasteurellales, Bacillales and Deinococcales, which belong to the phyla Proteobacteria, Firmicutes, and Deinococcus respectively, to be present in greater abundances in the vaginal tissues of pregnant CD-1 mice, which is consistent with the previous study that found Proteobacteria, Firmicutes,

Bacteroidetes, Actinobacteria and Cyanobacteria to be the major phyla in the vaginal microbiota of non-pregnant BALB/cJ mice (Barfod et al., 2013). I found the species diversity was higher in the cecal tissue with GR-1 treatment, suggesting GR-1 may promote the growth of other bacteria in the mouse gut. In addition, there was high variability in the species diversity in the vaginal microbiota of saline-treated mice (Figure 5-15). Some of the animals received a higher volume of saline than others (range: 100-300µL), and it is possible that excessive urination may have affected the vaginal species richness through dilution. Since GR-1 bacteria, which were resuspended in saline, have high viscosity, this dilutional effect would not have been observed in the GR-1-treated mice (Figure 5-15).

Oral GR-1 altered the mouse vaginal microbiota in this study. The relative abundance of bacterial order Bacillales decreased with oral GR-1 treatment, with representative genera in Bacillales including Staphylococcus. Certain strains of Staphylococcus such as S. aureus rectovaginal colonization have been associated with an increased risk of infections in pregnant women (Top et al., 2012). Oral GR-1 also decreased the relative abundances of bacteria orders Pseudomonadales, Burkholderiales, Hydrogenophilales, which collectively belong to the phylum Proteobacteria. Many disease-causing bacteria can be found within this phylum, including Escherichia (urinary tract infection), Salmonella (enteritis and typhoid fever), Vibrio (cholera), and Helicobacter (gastritis). Inflammatory conditions such as inflammatory bowel diseases have been associated with an increased abundance of Proteobacteria in the human gut (Mukhopadhya et al., 2012). Furthermore, an increase in the proportion of Proteobacteria in the stool of third trimester pregnant women has been associated with an increase in various pro-inflammatory cytokines (Koren et al., 2012).

In contrast, oral GR-1 increased the relative abundance of bacteria order Bacteroidales, which belongs to the phylum Bacteroidetes. It has been observed that Bacteroidetes is present in higher abundance in lean people than in obese people, and the abundance of Bacteroidetes increases in adults on low-calorie weight loss diets (Ley et al., 2006). Oral GR-1 also significantly increased the relative abundance of bacterial order Clostridiales and the genus Clostridium. The genus Clostridium contains species such as C. botulinum toxin A, which interestingly has been shown to inhibit oxytocin-induced uterine contractions in

84 cultured human myometrial cells (Burd et al., 2009). Many species belonging to the same genus of bacteria may have diverging effects on the host. There may also be potential interactions between the bacteria, as well as between multiple microbiomes at different body sites. Future metabolomic studies will help provide functional interpretations to the changes in the vaginal microbiota observed in this study.

It has been previously shown that oral GR-1 and RC-14 colonize the vagina of non-pregnant women (Anukam et al., 2006). It is not known however whether the lactobacilli given orally persist in the gut and are later transmitted to the vagina due to the proximity of rectum, or the Lactobacillus strains transiently colonize the gut and induce the gut epithelium to produce signaling molecules, which in turn alter the vaginal environment to favor the growth of the Lactobacillus spp. In this study, oral GR-1 altered the mouse vaginal microbiota, but not the cecal microbiota, and there was no difference in the relative abundance of Lactobacillus rhamnosus after GR-1 treatment. These findings suggest that GR-1 taken orally did not persist in the gut and instead, GR-1 may induce signaling mediators to modulate the vaginal environment, inducing changes in cytokines and altering the growth of bacteria other than lactobacilli.

In summary, this study has demonstrated that oral GR-1 live bacteria have immune- stimulatory properties in pregnant CD-1 mice, which is different from the anti-inflammatory effect observed with GR-1 SN. A high dose of GR-1 live bacteria may have adverse effects due to its inflammatory stimulation evident particularly in the amniotic fluid. Furthermore, oral GR-1 modulates the vaginal but not the cecal microbiota, suggesting the potential mechanism of GR-1 whereby probiotic lactobacilli exert its effect is primarily through the secretion of signalling molecules. Findings in this study suggest that the supernatant of lactobacilli, rather than its live bacterial counterpart, may be more appropriate for the prevention of PTB.

Mouse (cfu/kg) = Human (cfu/kg) X Human Km Animal Km

Human: ~60kg Animal Km: 3 Human Km: 37

Mouse (cfu/kg) = 109 to 1011 cfu per 60kg X 37/3

= ~ 2x109 to 1011 cfu

Each mouse weighs approximately 20g,

Per Mouse = 2 X 109 to 1011 cfu 1000/20

= 2 X 108 to 109 cfu

Figure 4-1 Probiotic Lactobacillus dose translation from a human dose to a mouse equivalent dose based on the body surface area (Km) and weight. Km: factor for converting mg/kg dose to mg/m2 dose. The equation is modified from Reagan-Shaw S, Nihal M, Ahmad N (2008) Dose translation from animal to human studies revisited. FASEB J 22: 659-661.

Gestational Day (GD) Pregnant CD-1 mice

1 0ral Gavage (100-300µL) ……….#

Saline GR-1 (109 cfu) 9

11 7 consecutive days (GD 9 -15) Once daily 12

14 Laparotomy/ Intra-uterine infusion (100µL) GD 15 15 Saline LPS 25µg LPS 50µg Saline LPS 25µg LPS 50µg 16 n=11 n=11 n=11 n=11 n=11 n=11 Preterm 17 Monitor for time of delivery Term 18 Sham No oral gavage or 19/20 laparotomy

n=11

Figure 4-2 Experimental design to investigate the effect of oral GR-1 on the timing of LPS-induced PTB (Set 1). Pregnant mice were given saline or GR-1 (109 cfu) via oral gavage once daily from GD 9 to GD 15. Animals were then divided to receive intra-uterine injection of saline, LPS (25µg) or LPS (50µg). Mice in the sham group did not receive any experimental procedures (oral gavage or mini-laparotomy). Mice were monitored for the time of delivery in individual cages until term (GD 19/20). Preterm delivery was defined as delivery of at least one pup 48 hours after intrauterine injection of LPS (GD 17).

Gestational Day (GD) Pregnant CD-1 mice

1 0ral Gavage (100-300µL) ……….#

Saline GR-1 (108 cfu) GR-1 (109 cfu) GR-1 (1010 cfu) 9 n=11 n=11 n=11 n=11 10

11 7 consecutive days (GD 9-15) Once daily 12

Preterm 17 Monitor for time of delivery Term 18 Sham No oral gavage 19/20 n=11

Figure 4-3 Experimental design to investigate the effect of oral GR-1 on the gestational length (Set 2). Pregnant mice were given saline or three increasing doses of GR-1 (108 cfu, 109 cfu, 1010cfu) via oral gavage once daily from GD 9 to GD 15. Mice in the sham group did not receive any experimental procedures. Mice were monitored for the time of delivery in individual cages until term (GD 19/20).

Gestational Day (GD) Pregnant CD-1 mice

1 0ral Gavage (100-300µL) ……….#

Saline GR-1 (108 cfu) GR-1 (109 cfu) GR-1 (1010 cfu) 9 n=13 n=7 n=8 n=6

11 7 consecutive days (GD 9-15) Once daily 12

14 Cytokine and chemokine protein measurement 15

16 Maternal Plasma ✔ ✔ ✔ ✔ ……….# Amniotic Fluid ✔ ✔ ✔ ✔ 17 Tissues ✔ - ✔ - 18

19/20

Figure 4-4 Experimental design to investigate the effect of oral GR-1 on cytokines and chemokines (Set 3). Pregnant mice were given saline or three increasing doses of GR-1 (108 cfu, 109 cfu, 1010cfu) via oral gavage once daily from GD 9 to GD 15. Maternal plasma and amniotic fluid were collected from mice in all four groups on GD 15. Intra-uterine tissues (fetal membranes, placenta, decidua and myometrium) were harvested from mice in the saline and GR-1 109 cfu groups.

Gestational Day (GD) Pregnant CD-1 mice

1 0ral Gavage (100-300µL) ……….#

GR-1 (109 cfu) 9 Saline n=7 n=7 10

11 7 consecutive days (GD 9-15) Once daily 12

Sample • Vaginal Tissue 15 • Cecal Tissue Collection 16

19/20

Figure 4-5 Experimental design to investigate the effect of oral GR-1 on the vaginal and cecal microbiota (Set 4). Pregnant mice were given either saline of GR-1 (109 cfu) via oral gavage once daily from GD 9 to GD 15. Vaginal and cecal tissues were collected on GD 15 for sequencing analysis.

IL-1α IL-1β Saline 15 800 GR-1 108 cfu GR-1 109 cfu 600 10 10 GR-1 10 cfu

400

5 200 Concentration (pg/mL) Concentration (pg/mL) 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

IL-6 ** IL-17 800 a a a b 100 700 80 600 60 500 40 400 40 30 30 20 20 10

Concentration (pg/mL) 10 Concentration (pg/mL) 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

IL-12p40 IL-12p70 1000 ** 250 b 800 200

600 a,b 150

400 a,b 100 a 200 50 Concentration (pg/mL) Concentration (pg/mL) 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

TNFα IFN-γ ** 1500 10 a a a b 1250 ** 8 1000 a b b a 6 750 4 500 2 250 Concentration (pg/mL) Concentration (pg/mL) 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

Figure 4-6 Histogram showing the concentration of pro-inflammatory cytokine IL-1α, IL-1β, IL-6, IL-17, IL-12p40, IL-12p70, TNFα and IFN-γ in the maternal plasma (MP) and amniotic fluid (AF) of pregnant mice that received varying doses of GR-1 (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are four treatment groups: saline (white, MP n=13, AF n=12); GR-1 108 cfu (light grey, MP n=7, AF n=6); GR-1 109 cfu (dark grey, MP n=6, AF n=8) and GR-1 1010 cfu (black bars, MP n=6, AF n=6). Comparison within groups was assessed with Kruskal-Wallis test followed by Dunns post- hoc test. Statistical significance was denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

IL-1α ** IL-1β 1350 a b 400 1100 Saline 850 GR-1 109 cfu 600 300 350 100 200 20 15 10 100 Concentration (pg/mL)

Concentration (pg/mL) 5 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

IL-6 IL-17 30 20 25 ** 15 20 a b 15 10

10 5 5 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

IL-12p40 ** IL-12p70 70 125 a b 60 100 50 75 40 30 50 20 25

10 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

TNFα IFN-γ 1000 ** 6 a b 800 4 600

400 2 200 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes Figure 4-7 Histogram showing the concentration of pro-inflammatory cytokines IL-1α, IL-1β, IL-6, IL-17, IL-12p40, IL-12p70, TNFα and IFN-γ in the fetal membranes, placenta, decidua and myometrium of pregnant mice that received saline and GR-1 at 109 cfu via oral gavage (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are two treatment groups, saline (white, n=7) and GR-1 at 109 cfu (dark grey, n=5) in the fetal membranes. There are 6 samples in the saline group and 6 samples in the GR-1 group for the placenta, decidua and myometrium. Comparison between the two groups was assessed with Mann-Whitney test. Statistical significance is denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

IL-2 IL-4 Saline 65 50 8 GR-1 10 cfu 55 9 45 GR-1 10 cfu 40 10 35 GR-1 10 cfu 25 15 30 5 5 20 4 3 10 2 Concentration (pg/mL) Concentration (pg/mL) 1 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

IL-13 IL-10 250 200

200 150

150 100 100 50 50 Concentration (pg/mL) Concentration (pg/mL) 0 0 Maternal Amniotic Maternal Amniotic Fluid Plasma Fluid Plasma

Results are mean values ± SEM and expressed in pg/mL. There are four treatment groups: saline (white, MP n=13, AF n=12); GR-1 108 cfu (light grey, MP n=7, AF n=6); GR-1 109 cfu (dark grey, MP n=6, AF n=8) and GR-1 1010 cfu (black bars, MP n=6, AF n=6). Comparison within groups was assessed with Kruskal-Wallis test followed by Dunns post- hoc test.

** IL-2 IL-4 35 a b 15 Saline 9 25 GR-1 10 cfu 12 15 ** 9 * 5 a b 5 a b 4 6 3 2 3 Concentration (pg/mL) 1 Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

IL-13 IL-10 ** 100 125 a b 80 100 * 60 75 a b 40 50

20 25 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

Figure 4-9 Histogram showing the concentration of anti-inflammatory cytokines IL-2, IL-4, IL-10 and IL-13 in the fetal membranes, placenta, decidua and myometrium of pregnant mice that received saline and oral GR-1 at 109 cfu (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are two treatment groups, saline (white, n=7) and GR-1 at 109 cfu (dark grey, n=5) in the fetal membranes. There are 6 samples in the saline group and 6 samples in the GR-1 group for the placenta, decidua and myometrium. Comparison between the two groups was assessed with Mann-Whitney test. Statistical significance was denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

CCL2 ** CCL3 *** 10000 a a a b 200 Saline b 8000 175 GR-1 108 cfu 6000 GR-1 109 cfu 150 4000 GR-1 1010 cfu 2000 125 a,b 100 1000 a,b 800 75 600 50 a 400 25 Concentration (pg/mL) Concentration (pg/mL) 200 0 0 Maternal Amniotic Maternal Amniotic Plasma Fluid Plasma Fluid

CCL4 ** CCL5 500 a a a b 50 ** 450 a a a b 400 40 350 300 30 250 150 125 20 100 75 50 10 Concentration (pg/mL) 25

CCL11 CXCL1 400 *** 120 b 300 90 a,b a,b 200 60 a 100 30 Concentration (pg/mL) Concentration (pg/mL)

Figure 4-10 Histogram showing the concentration of chemokines CCL2, CCL3, CCL4, CCL5, CCL11, CXCL1 in the maternal plasma (MP) and amniotic fluid (AF) of pregnant mice that received varying doses of GR-1 (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are four treatment groups: saline (white, MP n=13, AF n=12); GR-1 108 cfu (light grey, MP n=7, AF n=6); GR-1 109 cfu (dark grey, MP n=6, AF n=8) and GR-1 1010 cfu (black bars, MP n=6, AF n=6). Comparison within groups was assessed with Kruskal-Wallis test followed by Dunns post- hoc test. Statistical significance is denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

CCL2 CCL3

1000 175 Saline GR-1 109 cfu 150 800 125 600 100

400 75 50 200 25 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

CCL4 CCL5 150 30 25 120 20 90 * 15 ** a b 60 10 a b

30 5 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

CCL11 CXCL1 200 2000

150 1500

100 1000

50 500 Concentration (pg/mL) Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

Results are mean values ± SEM and expressed in pg/mL. There are two treatment groups, saline (white, n=7) and GR-1 at 109 cfu (dark grey, n=5) in the fetal membranes. There are 6 samples in the saline group and 6 samples in the GR-1 group for the placenta, decidua and myometrium. Comparison between the two groups was assessed with Mann-Whitney test. Statistical significance is denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

CSF2 CSF3 3500 300 Saline 8 3000 GR-1 10 cfu 250 9 GR-1 10 cfu 2500 10 200 GR-1 10 cfu 2000 150 1500 100 1000 50 500 Concentration (pg/mL) Concentration (pg/mL)

IL-3 35 25 15 5 5 4 3 2

Concentration (pg/mL) 1 0 Maternal Amniotic Plasma Fluid

Figure 4-12 Histogram showing the concentration of hematopoietic factors CSF2, CSF3 and IL-3 in the maternal plasma (MP) and amniotic fluid (AF) of pregnant mice that received varying doses of GR-1 (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are four treatment groups: saline (white, MP n=13, AF n=12); GR-1 108 cfu (light grey, MP n=7, AF n=6); GR-1 109 cfu (dark grey, MP n=6, AF n=8) and GR-1 1010 cfu (black bars, MP n=6, AF n=6). Comparison within groups was assessed with Kruskal-Wallis test followed by Dunns post- hoc test. Statistical significance is denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

CSF2 CSF3 100 15000 Saline 9 10000 GR-1 10 cfu

75 5000

50 500 400 300 25 200

Concentration (pg/mL) 100 Concentration (pg/mL) 0 0 Fetal Placenta Decidua Myometrium Fetal Placenta Decidua Myometrium Membranes Membranes

IL-3 15 * a b 12 ** a b 9

6 * 3 a b Concentration (pg/mL) 0 Fetal Placenta Decidua Myometrium Membranes

Figure 4-13 Histogram showing the concentrations of hematopoietic factors CSF2, CSF3 and IL-3 in the fetal membranes, placenta, decidua and myometrium of pregnant CD-1 mice that received saline and oral GR-1 at 109 cfu (Set 3). Results are mean values ± SEM and expressed in pg/mL. There are two treatment groups in the intrauterine tissues: saline (white) and GR-1 109 cfu (dark grey). There are two treatment groups, saline (white, n=7) and GR-1 at 109 cfu (dark grey, n=5) in the fetal membranes. There are 6 samples in the saline group and 6 samples in the GR-1 group for the placenta, decidua and myometrium. Comparison between the two groups was assessed with Mann- Whitney test. Statistical significance is denoted with different letters and with asterisks (*p<0.05; **p<0.01; ***p<0.001).

Vaginal Vaginal Cecal Cecal Saline GR-1 109 cfu Saline GR-1 109 cfu n=6 n=7 n=6 n=6

Figure 4-14 Stacked barplots showing the vaginal and cecal bacterial compositions of pregnant CD-1 mice that received either oral saline or GR-1. Each bar represents the vaginal or cecal microbiota of a single mouse and corresponds to the identification number labeled below each bar. Bacterial order found in >1% abundance are represented by a unique color, and orders that have <1% abundance are pooled into a single fraction at the top of the bar in dark blue.

3 *

1 Shannon Index Diversity 0 Saline GR-1 Saline GR-1 (Vaginal) (Vaginal) (Cecal) (Cecal) n=6 n=7 n=6 n=6

Figure 4-15 Scatterplot showing the Shannon diversity index (SDI) of the vaginal and cecal microbiota of pregnant CD-1 mice. Results are mean values ± SD and expressed in SDI ratios. Comparisons between the saline and GR-1 groups in the vaginal and in the cecal tissues, as well as between the two saline groups were assessed with Mann Whitney’s test. Statistical significance is denoted with an asterisk (*p<0.05).

100

Table 4-1 Delivery outcome of pregnant CD-1 following different treatments in Set 1. Preterm delivery is defined as delivery of at least one pup within 48 hours of intrauterine injection of LPS. For the delivery outcome results, the LPS 25µg group was compared with each of the following four groups using Fisher’s exact test (sham, saline, 109 cfu, and LPS 25µg + GR-1 109 cfu group with 11 animals in each group). The LPS 50µg group was compared with each of the following four groups using Fisher’s exact test (sham, saline, 109 cfu, and LPS 50µg + GR-1 109 cfu group with 11 animals in each group). Statistical significance is denoted with different letters.

Group No. of animals delivered No. of animals delivered preterm term

Sham 0 a 11 a

Saline 0 a 11 a

GR-1 109 cfu 0 a 11 a

LPS 25 µg 4 b 7 b

LPS 25 µg + GR-1 109 cfu 7 b 4 b

LPS 50 µg 11 c 0 c

LPS 50 µg + GR-1 109 cfu 11 c 0 c

101

Table 4-2 Litter size and fetal weight of live term neonates born to pregnant CD-1 mice at term that received different treatments in Set 1. Litter size and fetal weight data are expressed in mean values ± SEM. One-Way ANOVA followed by Tukey test was used to compare the groups with one another (p>0.05). Pregnant mice in the LPS 25µg group (4 out of 11), LPS 25µg+ GR-1 109 cfu group (7 out of 11), LPS 50µg group (11 out of 11) and the LPS 50µg + GR-1 109 cfu group (11 out of 11) delivered preterm and there were no surviving pups.

Group Litter size Weight per pup P-value (gram)

Sham 12.5±0.37 (n=11) 1.70 ±0.09 (n=11) > 0.05

Saline 12.7 ±0.47 (n=11) 1.84 ±0.13 (n=11) > 0.05

GR-1 109 cfu 12.5 ±0.43 (n=11) 1.73 ±0.11 (n=11) > 0.05

LPS 25 µg 11.4 ±0.61 (n=7) 1.51 ±0.14 (n=7) > 0.05

LPS 25 µg + GR-1 109 cfu 12.3 ±0.48 (n=4) 1.80 ±0.04 (n=4) > 0.05

LPS 50 µg - - > 0.05

LPS 50 µg + GR-1 109 cfu - - > 0.05

102

Table 4-3 Hours to delivery, litter size and fetal weight of neonates born to pregnant CD-1 mice that received saline or oral GR-1 (Set 2). Results are mean values ± SEM and expressed in hours (n=11 per group). One-Way ANOVA followed by Tukey test was used to compare the groups with one another.

Group Hours to Litter size Weight per P value delivery pup (gram)

Sham 106 ± 3.3 12.4 ± 0.36 1.68 ± 0.08 > 0.05

Saline 106 ± 3.3 12.1 ± 0.53 1.84 ± 0.13 > 0.05

GR-1 109 cfu 100 ± 3.0 12.2 ± 0.44 1.63 ± 0.11 > 0.05

GR-1 109 cfu 99 ± 2.4 12.5 ± 0.53 1.45 ± 0.10 > 0.05

GR-1 1010 cfu 101 ± 3.3 12.5 ± 0.45 1.69 ± 0.11 > 0.05

103

Table 4-4 Summary table of cytokines and chemokines in the maternal plasma, amniotic fluid and intrauterine tissues following varying doses of oral GR-1 treatment. An upward arrow indicates a significant increase and a downward arrow indicates a significant decrease following GR-1 treatment, when compared to mice that received saline. A dash (-) denotes no significant difference is observed. The numerical value in brackets indicates the dose of GR-1 at which significance is achieved. (8) 108 cfu; (9) 109 cfu; (10) 1010 cfu.

Pro- Maternal Amniotic Fetal Placenta Decidua Myometrium inflammatory Plasma Fluid Membranes cytokines IL-1α - - - - - ! (9)

IL-1β ------

IL-6 - ! (10) - - - -

IL-17 - - - ! (9) - -

IL-12p40 ! (10) - - - - -

IL-12p70 - - " (9) - - -

TNFα ! (8,9) - - ! (9) - -

IFNγ - ! (10) - - - -

Anti- Maternal Amniotic Fetal Placenta Decidua Myometrium inflammatory Plasma Fluid Membranes cytokines IL-2 - - ! (9) - - -

IL-4 - - " (9) - - ! (9)

IL-10 - - ! (9) - - " (9)

IL-13 ------

Chemokines Maternal Amniotic Fetal Placenta Decidua Myometrium Plasma Fluid Membranes

CCL2 - ! (10) - - - -

CCL3 - ! (10) - - - -

CCL4 - ! (10) - - - -

CCL5 - ! (10) ! (9) - - " (9)

CCL11 - ! (10) - - - -

CXCL1 ------

Hematopoietic Maternal Amniotic Fetal Placenta Decidua Myometrium Factors Plasma Fluid Membranes

CSF2 ------

CSF3 ------

IL-3 - - ! (9) ! (9) - " (9)

104

Table 4-5 Maternal plasma progesterone concentrations in pregnant CD-1 mice with varying dose of GR-1 (Set 3) Results are mean values ± SEM and expressed in ng/mL (n=6 per group). Comparison within groups was assessed with Kruskal Wallis test followed by Dunns post-hoc test (p > 0.05).

Treatment Saline GR-1 108 cfu GR-1 109 cfu GR-1 1010 cfu

Progesterone 39.5 ± 4.4 50.9 ± 4.4 48.6 ± 4.9 47.1 ± 7.0 concentration

105

Table 4-6 Bacterial groups unique to the cecal and vaginal tissues of saline-treated pregnant CD-1 mice. Presence of the bacteria is denoted with + while absence of the bacteria is denoted with –.

cecal vaginal Bacterial Group (saline) (saline) Actinobacteria;Actinobacteria;Actinomycetales;Corynebacteriaceae;Corynebacterium - + Actinobacteria;Actinobacteria;Actinomycetales;Micrococcaceae;Micrococcus - + Firmicutes;Bacilli;Bacillales;Bacillaceae1;Anoxybacillus - + Firmicutes;Bacilli;Bacillales;Bacillaceae1;Geobacillus - + Firmicutes;Bacilli;Bacillales;Staphylococcaceae;Staphylococcus - + Proteobacteria;Betaproteobacteria;Hydrogenophilales;Hydrogenophilaceae;Hydrogenophilus - + Proteobacteria;Betaproteobacteria;Hydrogenophilales;Hydrogenophilaceae;Petrobacter - + Proteobacteria;Betaproteobacteria;Neisseriales;Neisseriaceae;Neisseria - + Proteobacteria;Gammaproteobacteria;Pseudomonadales;Moraxellaceae;Acinetobacter - + Proteobacteria;Gammaproteobacteria;Pseudomonadales;Pseudomonadaceae;Pseudomonas - + Actinobacteria;Actinobacteria;Actinomycetales;Actinomycetaceae;Actinomyces - + Actinobacteria;Actinobacteria;Actinomycetales;Microbacteriaceae;Microbacterium - + Actinobacteria;Actinobacteria;Actinomycetales;Nocardioidaceae;Marmoricola - + Actinobacteria;Actinobacteria;Actinomycetales;Micrococcaceae;Rothia - + Actinobacteria;Actinobacteria;Coriobacteriales;Coriobacteriaceae;Atopobium - + Bacteroidetes;Sphingobacteria;Sphingobacteriales;Cytophagaceae;Hymenobacter - + Deinococcus-Thermus;Deinococci;Thermales;Thermaceae;Thermus - + Firmicutes;Bacilli;Bacillales;Bacillales_IncertaeSedisXI;Gemella - + Firmicutes;Bacilli;Lactobacillales;Aerococcaceae;Aerococcus - + Firmicutes;Bacilli;Lactobacillales;Enterococcaceae;Enterococcus - + Proteobacteria;Alphaproteobacteria;Rhizobiales;Methylobacteriaceae;Methylobacterium - + Proteobacteria;Gammaproteobacteria;Enterobacteriales;Enterobacteriaceae;Shigella - + Proteobacteria;Gammaproteobacteria;Pasteurellales;Pasteurellaceae;Actinobacillus - + Proteobacteria;Gammaproteobacteria;Pseudomonadales;Moraxellaceae;Psychrobacter - + Firmicutes;Bacilli;Bacillales;Alicyclobacillaceae;Alicyclobacillus + - Actinobacteria;Actinobacteria;Coriobacteriales;Coriobacteriaceae;Slackia + - Firmicutes;Clostridia;Clostridiales;Clostridiales_IncertaeSedisXIII;Clostridiaceae + - Firmicutes;Clostridia;Clostridiales;Ruminococcaceae;Ruminococcus + - Firmicutes;Clostridia;Halanaerobiales;Halobacteroidaceae;Halobacteroidaceae + - Firmicutes;Clostridia;Thermoanaerobacterales;Thermoanaerobacteraceae;Moorella + -

106

Table 4-7 Bacterial groups present in both the cecal and vaginal tissues of saline-treated pregnant CD-1 mice.

cecal vaginal Bacterial Group (saline) (saline) Bacteroidetes;Bacteroidia;Bacteroidales;Bacteroidaceae;Bacteroides + + Bacteroidetes;Bacteroidia;Bacteroidales;Porphyromonadaceae;Barnesiella + + Bacteroidetes;Bacteroidia;Bacteroidales;Porphyromonadaceae;Candidatus + + Bacteroidetes;Bacteroidia;Bacteroidales;Porphyromonadaceae;Gram-negative + + Bacteroidetes;Bacteroidia;Bacteroidales;Porphyromonadaceae;Porphyromonadaceae + + Bacteroidetes;Bacteroidia;Bacteroidales;Porphyromonadaceae;Porphyromonas + + Bacteroidetes;Bacteroidia;Bacteroidales;Rikenellaceae;Alistipes + + Bacteroidetes;Sphingobacteria;Sphingobacteriales;Sphingobacteriaceae;Parapedobacter + + Deinococcus-Thermus;Deinococci;Deinococcales;Deinococcaceae;Deinococcus + + Firmicutes;Bacilli;Bacillales;Bacillaceae1;Bacillus + + Firmicutes;Bacilli;Bacillales;Paenibacillaceae1;Paenibacillus + + Firmicutes;Bacilli;Lactobacillales;Lactobacillaceae;Lactobacillus + + Firmicutes;Bacilli;Lactobacillales;Leuconostocaceae;Leuconostoc + + Firmicutes;Bacilli;Lactobacillales;Streptococcaceae;Lactococcus + + Firmicutes;Bacilli;Lactobacillales;Streptococcaceae;Streptococcus + + Firmicutes;Clostridia;Clostridiales;Clostridiaceae1;Clostridium + + Firmicutes;Clostridia;Clostridiales;Clostridiales_IncertaeSedisXII;Fusibacter + + Firmicutes;Clostridia;Clostridiales;Clostridiales_IncertaeSedisXII;Peptostreptococcaceae + + Firmicutes;Clostridia;Clostridiales;Eubacteriaceae;Acetobacterium + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Clostridiales + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Clostridium + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Eubacterium + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Lachnospiraceae + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Oribacterium + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Ruminococcus + + Firmicutes;Clostridia;Clostridiales;Lachnospiraceae;Shuttleworthia + + Firmicutes;Clostridia;Clostridiales;Ruminococcaceae;Clostridiales + + Firmicutes;Clostridia;Clostridiales;Ruminococcaceae;Clostridium + + Firmicutes;Clostridia;Clostridiales;Ruminococcaceae;Lactobacillales + + Firmicutes;Erysipelotrichia;Erysipelotrichales;Erysipelotrichaceae;Clostridium + + Firmicutes;Erysipelotrichia;Erysipelotrichales;Erysipelotrichaceae;Turicibacter + + Proteobacteria;Betaproteobacteria;Burkholderiales;Sutterellaceae;Parasutterella + + Proteobacteria;Deltaproteobacteria;Desulfovibrionales;Desulfovibrionaceae;Bilophila + + Proteobacteria;Gammaproteobacteria;Oceanospirillales;Halomonadaceae;Haererehalobacter + + Proteobacteria;Gammaproteobacteria;Oceanospirillales;Halomonadaceae;Halomonas + + Tenericutes;Mollicutes;Acholeplasmatales;Acholeplasmataceae;Flavescence + + Verrucomicrobia;Opitutae;Opitutales;Opitutaceae;Opitutus + + Verrucomicrobia;Verrucomicrobiae;Verrucomicrobiales;Verrucomicrobiaceae;Akkermansia + +

107

Table 4-8 Bacteria at different taxonomic levels that have statistically significant higher abundance in the vaginal tissues than in the cecal tissues of saline-treated pregnant CD- 1 mice. Results are mean values ± SD and expressed in centered logarithm transformed ratios. Comparison between the saline (vaginal) group (n=6) and saline (cecal) group (n=6) was assessed with Mann-Whitney test. Statistical significance is denoted with different letters (p<0.05).

Bacteria Taxonomy Saline Mean Relative abundance Fold (vaginal) (ratio) Diff Changes Mean % Saline Saline (2 -diff) Abundance (vaginal) (cecal) ORDER Lactobacillales 13.2±12.1 7.3±3.2 a 3.3±1.5 b -4.0 16 Pseudomonadales 4.0±3.9 5.5±0.9 a -5.1±3.8 b -10.6 1552 Actinomycetales 7.7±10.0 5.3±1.9 a -4.0±3.1 b -9.3 630 Enterobacteriales 11.3±24.1 3.5±6.4 a -3.8±2.6 b -7.3 158 Hydrogenophilales 1.6±2.4 2.4±2.7 a -5.1±2.3 b -7.5 181 Neisseriales 0.4±0.9 -0.6±3.0 a -5.1±3.5 b -4.5 23 Xanthomonadales 0.2±0.2 -1.1±4.1 a -5.1±2.7 b -4.0 16 Chromatiales 0.5±0.9 -1.3±3.9 a -5.1±2.1 b -3.8 14 FAMILY Bacillaceae1 12.2±15.7 7.9±2.8 a -1.2±1.4 b -9.1 549 Propionibacteriaceae 2.7±2.3 4.8±2.2 a -2.9±1.5 b -7.7 208 Staphylococcaceae 2.2±2.2 4.8±1.6 a -4.1±1.7 b -8.9 478 Moraxellaceae 3.0±3.7 4.3±3.2 a -4.1±4.7 b -8.3 315 Comamonadaceae 1.5±1.3 4.3±2.4 a -4.1±4.6 b -8.4 338 Corynebacteriaceae 3.9±3.3 a -4.1±4.0 b -8.0 256 Hydrogenophilaceae 1.6±2.4 3.1±2.7 a -4.1±2.2 b -7.2 147 Micrococcaceae 0.1±0.2 1.8±1.0 a -4.1±3.5 b -5.8 56 Pseudomonadaceae 0.3±0.3 1.7±2.4 a -4.1±3.0 b -5.8 56 Burkholderiales incertae_sedis 0.2±0.3 0.5±2.0 a -4.1±2.7 b -4.5 23 GENERA Anoxybacillus 2.2±1.2 6.0±1.3 a -4.1±2.4 b -10.1 1097 Staphylococcus 2.2±2.2 5.6±1.7 a -4.1±2.0 b -9.7 832 Acinetobacter 2.5±3.5 4.6±3.2 a -4.1±6.0 b -8.6 388 Hydrogenophilus 1.5±2.4 3.5±2.8 a -4.1±2.4 b -7.6 194 Corynebacterium 4.0±8.4 3.0±4.9 a -4.1±5.2 b -7.1 137 Pseudomonas 0.1±0.1 2.5±2.5 a -4.1±3.3 b -6.5 91 Micrococcus 0.1±0.1 1.5±1.0 a -4.1±3.8 b -5.6 49 Petrobacter 0.1±0.0 0.9±1.8 a -4.1±3.2 b -5.0 32 Neisseria 0.4±0.9 0.8±3.0 a -4.1±3.7 b -4.9 30 Geobacillus 0.2±0.5 0.5±2.7 a -4.1±2.7 b -4.6 24

108

Table 4-9 Bacteria at different taxonomic levels that have statistically significant higher abundance in the cecal tissues than in the vaginal tissues of saline-treated pregnant CD- 1 mice. Results are mean values ± SD and expressed in centered logarithm transformed ratios (n=6 per group). Comparison between the saline (vaginal) group (n=6) and saline (cecal) group (n=6) was assessed with Mann-Whitney test. Statistical significance is denoted with different letters (p<0.05).

Saline Mean Relative abundance Fold Bacteria Taxonomy (vaginal) Diff Changes Mean % Saline Saline (2diff) Abundance (vaginal) (cecal) ORDER Bacteroidales 7.1±9.0 5.2±2.2 a 10.2±1.3 b 5.0 32.2 Clostridiales 4.0±3.3 4.9±2.0 a 9.3±1.6 b 4.4 20.7 Deinococcales 0.5±0.6 1.2±2.1 a 6.0±1.4 b 4.8 27.6 Acholeplasmatales 0.1±0.3 -2.1±4.2 a 4.6±3.9 b 6.8 109.7 Opitutales 0.1±0.2 -4.2±3.3 a 2.2±2.6 b 6.4 84.7 FAMILY Lachnospiraceae 2.7±1.8 5.0±2.3 a 10.0±1.7 b 4.9 30.6 Porphyromonadaceae 5.6±8.1 4.9±3.5 a 10.5±1.4 b 5.5 46.5 Paenibacillaceae1 0.9±0.7 3.5±2.0 a 7.0±1.4 b 3.6 11.7 Bacteroidaceae 0.8±0.7 3.2±1.5 a 9.1±1.9 b 5.9 59.0 Ruminococcaceae 0.7±1.1 2.7±1.7 a 6.7±1.2 b 4.0 15.6 Deinococcaceae 0.5±0.6 1.8±2.1 a 7.0±1.5 b 5.2 35.8 Clostridiaceae1 0.3±0.3 1.5±1.7 a 6.0±3.0 b 4.6 23.9 Rikenellaceae 0.2±0.1 0.7±1.9 a 6.4±4.1 b 5.8 53.9 Prevotellaceae 0.5±0.6 0.6±3.6 a 7.0±1.6 b 6.4 83.4 Acholeplasmataceae 0.1±0.3 -1.5±4.4 a 5.7±3.9 b 7.2 142.3 Sutterellaceae 0.1±0.3 -2.6±3.5 a 3.7±3.7 b 6.4 83.0 Opitutaceae 0.1±0.2 -3.5±3.4 a 3.3±2.7 b 6.8 109.9 GENERA Barnesiella 3.4±5.6 4.8±3.2 a 9.4±1.1 b 4.6 25.0 Bacteroides 0.8±0.7 4.0±1.6 a 9.1±1.6 b 5.1 35.1 Clostridiales 0.6±0.9 2.8±2.3 a 7.0±1.2 b 4.2 18.0 Deinococcus 0.5±0.6 2.6±2.2 a 7.0±1.3 b 4.4 21.3 Porphyromonadaceae 0.2±0.3 2.0±1.9 a 5.1±3.8 b 3.0 8.2 Candidatus 0.4±0.8 0.5±3.6 a 6.5±1.7 b 6.0 63.4 Porphyromonas 0.1±0.1 -0.3±2.5 a 4.7±0.7 b 5.0 32.2 Alistipes 0.1±0.1 -1.2±3.4 a 5.4±3.3 b 6.6 96.6 Parasutterella 0.2±0.3 -1.9±3.4 a 3.7±3.4 b 5.6 49.4

109

Table 4-10 Bacteria at different taxonomic levels that decreased significantly with oral GR-1 treatment in the vaginal tissues of pregnant CD-1 mice. Results are mean values ± SD and expressed in centered logarithm transformed ratios. Comparison between the saline group (n=6) and GR-1 group (n=7) was assessed with Mann- Whitney test. Statistical significance is denoted with different letters (p<0.05).

Saline Mean Relative Fold Bacteria Taxonomy Mean % abundance Diff Changes abundance Saline GR-1 (2diff) ORDER Bacillales 15.4±14.4 7.8±2.6 a 5.1±0.4 b -2.7 0.16 Pseudomonadales 4.0±3.9 5.5±0.9 a 1.3±3.8 b -4.2 0.05 Burkholderiales 2.2±1.8 4.5±1.4 a 2.5±1.4 b -2.1 0.24 Hydrogenophilales 1.6±2.4 2.4±2.7 a -0.6±2.3 b -3.0 0.12 FAMILY Bacillaceae1 12.2±15.7 7.9±2.8 a 3.3±1.4 b -4.5 0.04 Propionibacteriaceae 2.7±2.3 4.8±2.2 a 2.1±1.5 b -2.7 0.15 Staphylococcaceae 2.2±2.2 4.8±1.6 a 2.5±1.7 b -2.4 0.19 Comamonadaceae 1.5±1.3 4.3±2.4 a -1.7±4.6 b -6.1 0.01 Micrococcaceae 0.1±0.2 1.8±1.0 a -3.1±3.5 b -4.9 0.03 Burkholderiales_incertae_sedis 0.2±0.3 0.5±2.0 a -4.8±2.7 b -5.2 0.03 GENERA Anoxybacillus 2.2±1.2 6.0±1.3 a 2.1±2.4 b -4.0 0.06 Staphylococcus 2.2±2.2 5.6±1.7 a 2.9±2.0 b -2.6 0.16 Micrococcus 0.1±0.1 1.5±1.0 a -2.7±3.8 b -4.2 0.05 Petrobacter 0.1±0.0 0.9±1.8 a -3.7±3.2 b -4.6 0.04 Geobacillus 0.2±0.5 0.5±2.7 a -4.0±2.7 b -4.5 0.05

110

Table 4-11 Bacteria at different taxonomic levels that increased significantly with oral GR-1 treatment in the vaginal tissues of pregnant CD-1 mice. Results are mean values ± SD and expressed in centered logarithm transformed ratios (saline group: n=6; GR-1 group: n=7). Comparison between the saline group (n=6) and GR-1 group (n=7) was assessed with Mann-Whitney test. Statistical significance is denoted with different letters (p<0.05).

Saline Mean Relative abundance Fold Bacteria Taxonomy Mean % Diff Changes abundance Saline GR-1 (2diff) ORDER Bacteroidales 7.1±9.0 5.2±2.2 a 7.6±1.3 b 2.4 5.13 Clostridiales 4.0±3.3 4.9±2.0 a 7.2±1.6 b 2.3 4.90 FAMILY Bacteroidaceae 0.8±0.7 3.2±1.5 a 6.2±1.9 b 2.9 7.72 Halomonadaceae 0.0±0.1 -2.8±2.5 a 1.3±2.1 b 4.0 16.54 GENERA Bacteroides 0.8±0.7 4.0±1.6 a 6.7±1.6 b 2.7 6.43 Clostridium 0.4±0.3 3.1±1.3 a 5.8±0.9 b 2.7 6.38 Porphyromonas 0.1±0.1 -0.3±2.5 a 2.8±0.7 b 3.1 8.35

Chapter Five

Effect of oral probiotic Lactobacillus rhamnosus GR-1® and Lactobacillus reuteri RC-14® on the vaginal microbiota and cervico-vaginal cytokines and chemokines in low risk pregnant women with an intermediate or high Nugent score.

I am grateful to the research nurses, Ms Mary-Jean Martin and Ms Tara Maria Rocco, of Mount Sinai Hospital for the recruitment of participants and the collection of vaginal swabs, and staff at the Centre for Mother, Infant, and Child Research (Sunnybrook health Sciences Centre, Toronto, Canada) for randomization of the participants and statistical analyses of pregnancy outcomes. I would like to thank Dr. Laurent Briollais for his advice on statistical analyses as well as members of the CIHR Vaginal Microbiome (VOGUE) team for the discussion of idea and finding. I would also like to thank Dr. Gregory Gloor for his advice on the analyses of sequencing data and for his help on filtering and organizing the data into operational taxonomic unit tables. I would like to express my gratitude to Dr. David Carter at the Robarts Research Institute (London, Ontario, Canada) for performing the Illumina sequencing. I would like to thank Ms Shannon Seney, Mr Rod McPhee and Ms Amy McMillan for providing the Nugent scores.

111 112

Chapter 5

5.1 Introduction

A healthy human vaginal microbiota, characterized by the dominance of Lactobacillus spp., plays an important role in reproductive health and disease. Several studies have shown that lactobacilli prevent the overgrowth of pathogens by secreting antibacterial hydrogen peroxide, lactic acid, and bacteriocins (Reid and Bocking, 2003b). Bacterial vaginosis (BV), an altered vaginal microbiota associated with preterm birth (PTB), is characterized by a depletion of lactobacilli and an overgrowth of facultative anaerobic bacteria such as Gardnerella vaginalis, Atopobium vaginae, Prevotella spp., Mobiluncus spp. and Mycoplasma hominis (Donders et al., 2009; Donati et al., 2010). Cytokines and chemokines play pivotal roles in PTB, and the predominance of pro-inflammatory cytokines over anti- inflammatory cytokines, observed during an ascending infection, is associated with the early onset of labor (Keelan et al., 2003; Challis et al., 2009). BV is associated with elevated vaginal concentrations of pro-inflammatory cytokine Interleukin (IL)-1β and chemokine IL-8, both of which are elevated in the amniotic fluid and cervical fluid of women with microbial invasion of the amniotic cavity and preterm delivery (Balkus et al., 2010; Holst et al., 2011). A Gram stain Nugent score of 7-10 and/or the presence of three of the Amsel criteria is indicative of BV: a vaginal pH > 4.5, an amine fishy odour when vaginal fluid is mixed with potassium chloride, the presence of clue cells, or milky homogenous discharge (Nugent et al., 1991; Reid and Bocking, 2003b).

The use of high throughput sequencing techniques to characterize the human vaginal microbiota overcomes several limitations of traditional culture-based techniques, including

113 the failure to detect uncultivable microorganisms and underestimation of the vaginal diversity (Gloor et al.. 2010; Hummelen et al., 2010; Srinivasan et al.. 2012). Several studies have employed sequencing methods to characterize the vaginal microbiota of healthy pregnant (Romero et al., 2014a; Romero et al., 2014b; Aagaard et al., 2012), and non- pregnant women (Gloor et al., 2010; Hummelen et al., 2010; Srinivasan et al., 2012). In this study, I used the Illumina MiSeq sequencing platform to identify phylogenetically diverse microorganisms to the species level in pregnant women with an intermediate or high Nugent score, using primers that target V6 region of the 16S ribosomal DNA (rDNA) (Gloor et al., 2010).

Probiotics are defined as “live microorganisms which, when administered in adequate amounts, confer a health benefit on the host” (FAO/WHO, 2001). Probiotic lactobacilli, administered through either the oral or vaginal route, ameliorate BV and replenish lactobacilli abundance in the vaginal biota of non-pregnant women (Homayouni et al., 2014). Oral administration of lactobacilli confers additional health benefits, such as reduction of urinary tract infection (Reid et al., 2015; Reid, 2001a; Walsh et al., 2014). The rationale for selecting probiotic Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 (GR-1 and RC-14) to improve the abnormal vaginal biota was derived from a previous study in non- pregnant women, in which treatment with GR-1 and RC-14 with a similar dosing range (109 cfu) reduced BV occurrence and recurrence (Reid et al., 2003a).

Our previous studies have demonstrated the supernatant of L. rhamnosus GR-1 (GR-1 SN) possesses anti-inflammatory properties in cultured human intrauterine tissues (Yeganegi et al., 2009; Yeganegi et al., 2011; Li et al., 2014), mouse macrophages (Kim et al., 2006) and can reduce inflammation-associated PTB in pregnant mice (Yang et al., 2014b) (Chapter 3). To date, the effect of oral probiotic supplementation in modulating the vaginal microbiota and cervico-vaginal cytokines and chemokines in pregnant women diagnosed with an abnormal Nugent score remains unknown. I hypothesize that pregnant women with an abnormal Nugent score will revert to a normal Nugent score with oral GR-1 and RC-14 treatment; oral probiotics will dampen the cervico-vaginal concentration of pro-inflammatory cytokines and chemokines, and will modulate the vaginal microbiota in these women.

114

5.2 Materials and Methods

5.2.1 Study Participants

Pregnant women were recruited from low risk antenatal clinics at Mount Sinai Hospital, Toronto, Canada. Women were over 18 years of age, were prior to 17 weeks of gestation, had singleton pregnancies, and were able to provide informed consent. Women who had multi-fetal pregnancies, fetal complications, maternal history of previous PTB, second trimester loss, significant maternal medical, surgical complications or HIV were excluded. A Dacron swab was placed in the posterior fornix or lateral vaginal wall (avoiding cervical mucous) for 10 seconds, and smears applied to the microscope slides were Gram-stained and scored according to the Nugent criteria (Nugent et al., 1991). A letter of No Objection was obtained from Health Canada for the use of probiotic lactobacilli and the study was approved by the Ethics Review Board of Mount Sinai Hospital (Research Ethics Board Approval Number: 08-005-A).

5.2.2 Study groups and randomization

A total of 328 women were consented and screened between May 2012 and October 2013 for the presence of an intermediate (4-6) or high (7-10) Nugent score at the time of their routine vaginal speculum examination between 12 to 16 weeks of pregnancy (on average 13.3 weeks gestation). Of the 328 women screened, 86 women had a Nugent score ≥ 4 (Figure 5-1). In order to detect a difference between a BV prevalence of 30% in the probiotic group and 60% in the placebo group at the end of treatment protocol, a sample size of 40 pregnant women in each group was needed. Z test was used to determine the sample size with alpha=0.05 and power =0.8. The sample size was increased to 43 in each group to compensate for 5% of women lost to follow-up. Following informed consent, they were randomized using a web- based randomization service to receive by mouth, two identical looking capsules per day containing either GR-1 and RC-14 (n=43) or placebo (n=43) for 12 weeks. The choice of oral administration over vaginal administration was based on a previous study in non- pregnant women, in which treatment with the same lactobacilli strains (GR-1 and RC-14) at

115 a similar dose (109 cfu), reduced BV occurrence and recurrence (Reid et al., 2003a). Vaginal swabs were collected at 13, 28 and 35 weeks gestation and analyzed for Nugent score, cytokine and chemokines, and vaginal microbiota. The characteristics of women at the time of randomization (13 weeks gestation) are summarized in Table 5-1. Fourteen women were lost to follow-up or withdrew from the study, 3 women had taken less than 25% of the 168 capsules over the 12-week treatment period, and there was insufficient sample for analysis in 3 women (Figure 5-1). After excluding these women, there were 32 women in the probiotic group and 34 women in the placebo group with samples available for the sequencing analysis (Figure 5-1). There were insufficient samples in 2 additional women for the cytokine protein measurements, so there were 31 women in the probiotic group and 33 women in the placebo group with samples available for the cytokine assay (Figure 5-1).

5.2.3 Nugent score

Vaginal swab smears were graded on a 10-point scale based on the presence or absence of various bacterial morphotypes, including Lactobacillus spp. (gram-positive rods), and pathogenic Gardnerella vaginalis (small gram-variable rods) and Bacteroides spp. (small gram-negative rods). A score of 0-3 was considered a normal vaginal microbiota, with high abundance of Lactobacillus spp.; a score of 4-6 represented an intermediate biota with higher proportions of non-Lactobacillus morphotypes, and a score of 7-10 was considered BV, with the near absence of Lactobacillus morphotypes and high abundance of the pathogenic morphotypes (Nugent et al., 1991). The smears were analyzed by three experienced observers in Dr. Gregor Reid’s laboratory (Lawson Research Institute, London, Canada).

5.2.4 Probiotic Strains

Lactobacillus rhamnosus, GR-1® and Lactobacillus reuteri, RC-14® (GR-1 and RC-14) and placebo capsules were provided by Chr Hansen, Denmark. The probiotic capsules contained at least 5x109 viable cells per capsule (or 2.5x109 cells of GR-1 and 2.5x109 cells of RC-14) freeze-dried in gelatin capsules each containing 180 mg of powder. Anhydrous dextrose and

116 potato starch were used as fillers to adjust for variations in the amount of microbial culture used, microcrystalline cellulose was used as binder and magnesium stearate was used as lubricant (manufacturer’s manual).

5.2.5 DNA Isolation and PCR amplification of V6 region of 16S rDNA

Vaginal swabs were equilibrated in 800µL phosphate buffer saline (PBS) on ice and vortexed for 1 min. The swab was removed and DNA extracted with a Qiagen Stool Extraction Kit (Appendix III, Qiagen, Toronto, Canada). Bacterial DNA was amplified with barcoded primers targeting the V6 region of the 16S rDNA (Robarts Research Institute, Western University, Canada). PCR amplification was performed with colorless GO-Taq hot start master mix (Promega, Canada) for 25 repeating cycles of 95°C, 55°C and 72°C for 1 minute each step. The amplified products were quantified using a QuBit broad-range double- stranded DNA fluorometric quantitation reagent kit (Life technologies, Canada). Samples were pooled at equal molar concentrations and purified using Wizard PCR Clean-Up Kit according to manufacturer’s instructions (Promega, Canada).

5.2.6 Sequencing

Barcoded DNA was sequenced in pairs on the MiSeq Illumina platform at the Robarts Research Institute (Western University, London, Canada). The V6L and V6R primers included a unique 12bp sequence tag to barcode each sample. The primers used were: V6L-5′- ACACTCTTTCCCTACACGACGCTCTTCCGATCTNNNNNNNNNNNNCWACGC GARGAACCTTACC-3′ and V6R-5′-CGGTCTCGGCATTCCTGCTGAACCGCTCTTCCG ATCTNNNNNNNNACRACACGAGCTGACGAC-3′, where the italicized sequences are the Illumina MiSeq sequencing primers and the bold font denotes the universal 16S rRNA gene primers. The sequence results were provided in the fastq format. All sequences were filtered and a table of counts was generated for each sample containing sequences grouped at 97% operational taxonomic unit (OTU) and 100% identical sequence unit identity. The sequences

117 were then classified to distinct taxonomic species using the online Ribosomal Database Project (http://rdp.cme.msu.edu/seqmatch/seqmatch_intro.jsp). Sequences not identical across all best matches were marked as unclassified.

5.2.7 Protein Extraction and Cytokine/Chemokine Multiplex Assay

Vaginal swabs were equilibrated in Tris-HCl buffer (pH 7.5) with 150 mmol/L NaCl, 1mmol/L phenylmethylsulfonyl fluoride (Sigma), 0.05% Tween-20 (Sigma) and a protease inhibitor cocktail tablet (Roche) for 30 min at 4 oC and vortexed every 10 min. The swab was removed and the buffer centrifuged at 16,000 × g for 15 min at 4 oC. The supernatant was then stored at -80oC in aliquots until further analysis. IL-1 receptor antagonist (IL-1rα), IL- 1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12p70, IL-13, IL-15, IL-17, basic Fibroblast Growth Factor (bFGF), Colony Stimulating Factor (CSF) 2, CSF3, Interferon (IFN)-γ, CXCL10, CCL2, CCL3, CCL4, CCL5, CCL11, Platelet-Derived Growth Factor (PDGF)-bb, Tumor Necrosis Factor (TNF)-α and Vascular Endothelial Growth Factor (VEGF) were measured with a 27 human multiplex cytokine/chemokine kit according to manufacturer’s instructions (Appendix I, Biorad, Canada).

5.2.8 Statistical Analyses

Unpaired Student’s t test (two tailed) or Chi-square test was carried out using SigmaStat (version 3.5) to compare 1) pre-randomization characteristics; 2) pregnancy outcomes; 3) compliance of women to the treatment protocol; 4) the percentage of women who reversed to a normal Nugent score between the placebo and probiotic groups; (5) the microbial profiles between women with an intermediate Nugent score and a high Nugent score, prior to treatment. Two-Way Repeated Measure ANOVA followed by Holm Sidak method was carried out using SigmaStat (version 3.5) to test for treatment and gestational effects on microbial profiles and the concentrations of cytokines and chemokines. For sequencing data, centered ratio logarithm transformation was performed as described previously (Aitchison, 1986). Briefly, the geometric mean of the proportions of all species detected in a sample was

118 computed. A ratio x was determined from the proportion of species i over the geometric mean. Then, the relative abundance of species i was calculated by taking natural logarithm of x. Statistical analysis of the sequencing data was carried out using R (version 3.0.1). Generalized Estimation Equation Model was used for data that did not follow the normal distribution. Data were adjusted for false discovery rate using Benjamini Hochberg procedure and an adjusted p-value of p<0.05 was considered statistically significant. Data were tested for normality and equal variance and were expressed as mean values ± standard deviation (SD). The Shannon diversity index was calculated by first taking the proportion of a bacteria species relative to the total number of species detected in a sample, and multiplying it by the natural logarithm of this proportion. The product was then summed across all bacteria species, and multiplied by -1 (Magurrant 2003).

5.3 Results

5.3.1 Pre-randomization characteristics The mean maternal age was 33.8 ± 4.2 years old and the mean pre-pregnancy body mass index was 22.5 ± 3.2 for the women in the probiotic group at 13 weeks gestation, and these characteristics were not different for the women in the placebo group (34.4 ± 3.3 years old and BMI: 22.4 ± 3.1) (Table 5-1). In both groups, vaginal swabs used for screening of an abnormal Nugent score were taken at 13 weeks gestation. Over 55% of the women were Caucasian in both groups. Other ethnicities included South and East Asian, Black and Hispanic.

Forty out of the 43 women (93%) in both the placebo and the probiotic groups had a natural conception (Table 5-1). Seventeen pre-existing conditions were reported in 14 women randomized to the probiotic group and 27 pre-existing conditions were reported in 21 women in the placebo group. A total of 16 previous surgeries were reported in 14 women in the probiotic group and 34 surgeries were reported in 22 women in the placebo group. Fourteen women randomized to the probiotic group and 22 women in the placebo group were on medications at the beginning of the treatment protocol. Thirty-five out of 43 women (81.4%) in the probiotic group and 41 out of 43 women (95.4%) in the placebo group reported

119 ingesting probiotic containing fermented food during pregnancy. These characteristics were not different between the placebo and probiotic groups (p>0.05, Table 5-1).

5.3.2 Pregnancy Outcomes

Antibiotics were taken by 13.9% of the women in the probiotic group and 11.6% of the women in the placebo group during pregnancy for various indications (p>0.05, Table 5-2). There was no difference in antibiotic administration during labor in both groups (46.3% in the probiotic group and 37.2% in the placebo group, p>0.05). 19.5% of the women had induction of labor and 80.5% of the women had a vaginal delivery in the probiotic group. These percentages were not different in the placebo group. The mean gestational age at delivery was 39.1 ± 1.4 weeks in the probiotic group, and this was not different in the placebo group (39.4 ± 0.9 weeks, p>0.05, Table 5-2). The mean birth weight was 3340 ± 433 grams in the probiotic group, and this was not different in the placebo group (3351 ± 463 grams, p>0.05). There was 1 infant with intrauterine growth restriction (IUGR) in the placebo group, and 2 infants in the probiotic group delivered at 34 weeks gestation in association with premature rupture of membranes (Table 5-2). In 1 of these infants, the Apgar score was less than 7 at 5 minutes. There was no difference in the fetal sex distribution or cord blood pH between the two groups. There was also no difference in the number of women who experienced symptoms such as vaginal itching, vaginal discharge and vaginal odour during the 12-week treatment period between the two groups. There were no adverse reactions to the probiotics or placebo reported.

5.3.3 Compliance to the treatment protocol

At the end of the 12 week treatment period, there were on average 13 pills (7.7%) left in the bottles returned by women in the probiotic group and 9 pills (5.4%) remaining in the bottles returned by women in the placebo group (Table 5-3). Twenty-six out of 32 women (81.2%) in the probiotic group and 29 out of 34 women in the placebo group (86.3%) had taken more than 75% of the total pills (168 pills) (p > 0.05, Table 5-3). The remaining women had taken

120 more than 50% of the total pills. Three women have taken less than 25% of the total pills and were excluded from subsequent analyses.

5.3.4 Effect of oral probiotic GR-1 and RC-14 on the Nugent score

The primary outcome of this trial was to evaluate changes in the Nugent score among women with an abnormal Nugent score after oral probiotic supplementation, in comparison to placebo-treated women. There were 11 out of 32 women (34.4%) in the probiotic group and 11 out of 34 women (32.3%) in the placebo group, who reversed to a normal Nugent score at 28 weeks gestation (p>0.05, Table 5-4). The percentages were similar in both groups at 35 weeks gestation (p>0.05).

5.3.5 Effect of oral probiotic GR-1 and RC-14 on the vaginal microbiota

A total of 93 distinct bacterial species were detected at 13 weeks gestation (Table 5-5). The most abundant species were Lactobacillus iners, Lactobacillus crispatus, Gardnerella vaginalis and Atopobium vaginae across pregnancy. Thirty of 66 women had a single bacterial species (A. vaginae, n=4; L. jensenii, n=1; iners, n=12, L crispatus, n=9 and G. vaginalis, n=4), which dominated more than 40% of their vaginal microbiota at 13 weeks gestation (Figure 5-2). In the remaining women, the vaginal microbiota was dominated by a mixture of different bacterial species. The vaginal microbiota of pregnant women with an intermediate Nugent score (n=42) and those with a BV Nugent score (n=24) at 13 weeks gestation are shown in Figure 5-3. The vaginal microbiota at the time of study entry (13 weeks gestation) were not different between these women (p > 0.05, Table 5-5) and therefore, these results were pooled for all in subsequent analyses.

The vaginal microbiota of pregnant women who received placebo (n=34) and those who received probiotics (n=32) at 13, 28 and 35 weeks gestation are shown in Figure 5-4. There was no difference in the vaginal microbiota between pregnant women in the placebo and probiotic groups at the end of the 12-week treatment protocol (28 weeks gestation), or at 35

121 weeks gestation (Table 5-6 and Table 5-7). There was no difference in the vaginal microbiota between the placebo and probiotic groups when data were grouped by ethnicity, pre- pregnancy BMI or when women whose vaginal microbiota were dominated by Lactobacillus spp were excluded (data not shown).

Lactobacillus rhamnosus was detected in 98% of the women (65 out of 66 women) at 13 weeks gestation, and its abundance did not alter with probiotic treatment. There were two women in the probiotic group who delivered at 34 weeks gestation in association with premature rupture of membranes. In one of these women, her vaginal microbiota was dominated by L. jensenii, and following probiotic treatment, her vaginal biota became more heterogeneous, with increased abundance of species including L. gasseri, G. vaginalis and Prevotella bivia (Figure 5-4). The other woman had a heterogenous vaginal microbiota initially, and with probiotic treatment, L. cripatus dominated her vaginal microbiota (Figure 5-4).

The relative mean abundance of 12 species including L. iners, L. acidophilus, G. vaginalis and A. vaginae decreased at 28 weeks and/or 35 weeks of gestation in the placebo group and/or the probiotic group, compared to 13 weeks of gestation (Table 5-6). In contrast, the relative mean abundance of 9 species increased across pregnancy (Table 5-7). There was no difference in the Shannon diversity index between the probiotic and placebo groups at 13, 28 or 35 weeks gestation (Figure 5-5).

5.3.6 Effect of GR-1 and RC-14 on the concentrations of cervico-vaginal cytokines/chemokine

The cervico-vaginal concentrations of cytokines and chemokines at the time of study entry (13 weeks gestation) were not different between pregnant women diagnosed with an intermediate or BV Nugent score (p>0.05, data not shown). Therefore, these data were combined in subsequent analyses. The concentration of cytokines and chemokines were not different between placebo (n=34) and probiotic-treated (n=32) women at 13, 28 or 35 weeks

122 gestation (p>0.05, Table 5-8).

Levels of pro-inflammatory cytokines IL-1β, IL-6, IL-12p70, IL-17, IFN-γ, TNFα, anti- inflammatory cytokines IL-9, IL-13, chemokines IL-8, CXCL10, CCL11, CCL2, CCL3, CCL4, CCL5, and growth/hematopoietic factors VEGF, PDGFbb, bFGF, CSF2 and IL-7 did not change throughout pregnancy (p>0.05, Table 5-8). The concentrations of the anti- inflammatory cytokine IL-4 in the placebo group and IL-10 in both probiotic and placebo groups increased slightly at 28 weeks gestation, but were not different at 35 weeks gestation, when compared to 13 weeks gestation (p<0.05, Figure 5-6). Concentration of the hematopoietic factor CSF3 decreased at 28 weeks in the probiotic group and at 35 weeks gestation in the placebo group, when compared to 13 weeks gestation (p<0.05, Figure 5-6). Concentrations of IL-2, IL-5, IL-15 and IL-1ra were outside the detection limit.

5.4 Comment

In this prospective, randomized, double blinded, and placebo-controlled trial, there was no difference in the pre-randomization characteristics, pregnancy outcomes and compliance to the treatment protocol between pregnant women in the placebo and the probiotic groups. Pregnant women were initially classified by their Nugent scores as either BV or Intermediate. However, since the vaginal microbiota of pregnant women diagnosed with a BV Nugent score did not differ from women with an intermediate Nugent score at 13 weeks gestation, we grouped these women for subsequent analyses. Furthermore, lactobacilli dominated the vaginal microbiota in more than one third of the pregnant women with an abnormal Nugent score, at 13 weeks gestation. Retrospectively, it was observed that some slides were of poor quality and the presence of peripheral blood mononuclear cells made scoring the slides difficult. The Nugent scoring system may not be the ideal approach for the diagnosis of asymptomatic BV although at the time this study was started, this was the gold standard. It has been shown that a DNA level of ≥109 copies/mL for G. vaginalis and ≥108 copies/mL for A. vaginae has a 95% sensitivity and positive predictive value, and 99% specificity and negative predictive value for the diagnosis of BV, which are higher than has been reported using the Nugent score (Menard et al., 2008).

123

There was no difference between pregnant women randomized to the probiotic group and the placebo group with regards to their mean maternal age, pre-pregnancy BMI, ethnicity distribution, mode of conception, and intake of antibiotics and/or fermented food during pregnancy. There was no difference in the vaginal microbiota between probiotic-treated and placebo-treated pregnant women at the end of the 12-week treatment period, nor at 35 weeks gestation. This was evident as well when I excluded women with high lactobacilli abundance in their vaginal microbiota prior to treatment (18 women in the placebo group and 22 women in the probiotic group). I based my probiotic dosage on a previous study in non-pregnant women, which demonstrated oral supplementation of GR-1 and RC-14 at 109 cfu restores the indigenous lactobacilli in women with recurrent BV (Reid et al., 2003a). Since this study was started, it has been demonstrated that pregnant women have a higher abundance of several Lactobacillus spp including L. crispatus, L. gasseri and L. jensenii, and a more resistent microbiota than non-pregnant women (Romero et al., 2014a; Aagaard et al., 2012). It is plausible the current dose (5 x 109 cfu) may not be sufficient to alter the vaginal microbiota in pregnant women, and that a higher dose is required.

In this longitudinal study, I characterized the vaginal microbiota in pregnant women with an abnormal Nugent score throughout pregnancy and observed that the vaginal microbiota is not static across gestation, in agreement with a previous study in pregnant women with a healthy vaginal biota (Romero et al., 2014b). Specifically, I observed a decline in the relative abundances of A. vaginae, A. rimae and G. vaginalis consistent with previous observations in pregnant women with a healthy biota (Romero et al., 2014b). In contrast to studies that target the V1-V3 (Romero et al., 2014b) and V3-V4 (Ling et al., 2010; Srinivasan et al.. 2012) regions of the 16S rDNA, I did not observe a change in the relative abundance of Gemella and Sneathia sanguinegens, and I did not detect the presence of Eggerthella spp., Parvimonas micra, BV associated bacteria 1 (BVAB1), BVAB2 or Ureaplasma parvum. The use of primers that targeted the V6 region in this study may have under- estimated the presence of these bacteria (Gloor et al., 2010; Hummelen et al., 2010). Using sequencing primers that target the cpn60 gene, it is possible to measure the abundance of Mollicutes, including Mycoplasma hominis, Ureaplasma parvum, and Ureaplasma

124 urealyticum in non-pregnant women (Chaban et al., 2014) as well as in pregnant women with an abnormal vaginal biota at 13 weeks gestation (Hill et al, unpublished data).

The relative abundance of L. iners and L. acidophilus across gestations decreased in this study, in contrast to a previous report that found an increase in the relative abundance of several Lactobacillus spp (L. crispatus, L. jensenii, L. gasseri and L. vaginalis) with advancing gestational age in women with a healthy biota (Romero et al., 2014b). It is important to distinguish that women in this study had an abnormal biota, which did not resolve in 65% of the women by 35 weeks gestation.

Previous studies in term cultured human intra-uterine tissues and in pregnant mice have demonstrated that Lactobacillus rhamnosus GR-1 supernatant possesses anti-inflammatory properties (Yeganegi et al., 2009; Yeganegi et al., 2011; Yang et al., 2014b; Li et al., 2014). In this study, oral GR-1 and RC-14 did not alter the cervico-vaginal concentrations of cytokines or chemokines. It is known that in non-pregnant women, exogenous lactobacilli colonization is transient (Gardiner et al., 2012). Alternatively, a higher dose of live lactobacilli is needed to produce sufficient bioactive metabolites to achieve similar effects in humans. Thirdly, it is possible that an underlying state of inflammation is required before lactobacilli exert an anti-inflammatory effect.

There was a shift towards an anti-inflammatory environment across gestation as evident by an increase in IL-4 and IL-10 concentrations at 28 weeks gestation. The levels of cervico- vaginal IL-4 and IL-10 were in comparable range with previous studies (Nenadic and Pavlovic, 2008; Chandiramani et al., 2012), and these observations are consistent with the hypothesis that a dampening of inflammation is important to the maintenance of uterine quiescence (Challis et al., 2009).

CSF3, which is important in placentation, neutrophil progenitors proliferation, differentiation and survival, decreased with advancing gestational age. CSF3 has also been shown to possess anti-inflammatory properties in cultured human placental trophoblast cells (Yeganegi et al., 2011). However, elevated maternal CSF3 concentrations have been associated with

125 spontaneous PTB in humans (Whitcomb et al., 2009). Taken together with the observations in this study, CSF3 appears to be anti-inflammatory and a decline in the cervico-vaginal concentration of CSF3 at 35 weeks gestation may promote the inflammatory responses that eventually lead to the initiation of labor.

There was 1 infant with intrauterine growth restriction (IUGR) in the placebo group, and 2 infants in the probiotic group were delivered at 34 weeks gestation in association with PPROM. In 1 of those infants, the apgar score was less than 7 at 5 minutes. There was no difference in fetal sex distribution between placebo-treated and probiotic-treated pregnant women. There were neither adverse side effects nor alterations in pregnancy outcomes with probiotic treatment, in agreement with a recent meta-analysis of randomized clinical trials, which demonstrated the use of probiotics Lactobacillus is safe during pregnancy (Dugoua et al., 2009).

This study provides a longitudinal overview of vaginal microbiota and cervico-vaginal cytokine profiles throughout pregnancy, which may serve as a baseline for future clinical trials that assess the efficacy of probiotic administration to pregnant women. In contrast to my initial hypotheses, at the current dose (5 x 109 cfu) and duration (12-weeks), oral GR-1 and RC-14 does not alter the Nugent score, vaginal microbiota or cervico-vaginal cytokine profiles in pregnant women with an abnormal Nugent score. Future trials should consider using a higher lactobacilli dose or for a longer duration that includes women with high-risk pregnancies. Future metabolomic studies investigating the function of bacterial species might shed light to the clinical relevance of the changes in various bacterial species observed as pregnancy progress.

126

Enrolment (n = 328) 13 weeks gestation Excluded (n= 242) Normal Nugent Score ♦!! ♦ 2 women with high Nugent score declined to be randomized Randomized (n= 86) Nugent score ≥ 4

Probiotics (n=43) 12 weeks, Placebo (n=43) Twice a day orally GR-1 and RC-14 ! 5 X 109 viable cells

(n=38) 28 weeks (n=38) ♦!!!1 withdrawal gestation ♦!!!5 lost to follow up ♦!!!4 lost to follow up

(n=32) 35 weeks (n=34) ♦!!!4 lost to follow up gestation ♦!4 non-compliant or ♦ !!2 non-compliant or insufficient samples insufficient samples

Figure 5-1 Consort flow chart of pregnant women enrolled in the study.

127 V1_109 V1_083 V1_250 V1_285 V1_311 V1_151 V1_289 V1_326 V1_271 V1_294 V1_290 V1_222 V1_328 V1_214 V1_314 V1_274 V1_275 V1_277 V1_280 V1_319 V1_266 V1_293 V1_212 V1_299 V1_315 V1_037 V1_094 V1_329 V1_179 V1_030 V1_168 V1_164 V1_106 V1_085 V1_147 V1_199 V1_170 V1_247 V1_209 V1_125 V1_258 V1_303 V1_296 V1_016 V1_295 V1_127 V1_088 V1_091 V1_090 V1_178 V1_041 V1_327 V1_038 V1_224 V1_322 V1_302 V1_118 V1_173 V1_058 V1_248 V1_053 V1_114 V1_213 V1_193 V1_138 V1_152

1.0

0.8

0.6

0.4 Microbiota fraction

0.2

0.0

Figure 5-2 Stacked bar plot showing the vaginal microbiota clustered by bacteria similarity in pregnant women prior to treatment, at 13 weeks gestation (n=66). Each bar represents the vaginal microbiota of a single woman and corresponds to the participant identification (ID) number labeled in the dendogram, clustered using average linkage cluster analysis. Species found in >1% abundance are represented by a unique color. Species with <1% abundance in the sample are pooled into a single fraction at the top of the bar in grey color. Women who have a single bacterial species which dominated more than 40% of their vaginal microbiota are identified with a color dot below their identification number that corresponds to the dominant species (Dark green, Atopobium vaginae, n=4; Very light blue, Lactobacillus (L.) jensenii, n=1; blue, L. iners, n=12; light blue, L crispatus, n=9 and red, Gardnerella vaginalis, n=4). Black rectangles are used to denote women with a BV Nugent score and white rectangles are used to identify women with an intermediate Nugent score.

128

BV (Nugent score of 7-10)

Intermediate (Nugent score of 4-6)

Figure 5-3 Stacked bar plots showing the vaginal microbiota clustered by bacteria similarity in pregnant women with a BV (n=24) or an intermediate (n=42) Nugent score prior to treatment, at 13 weeks gestation. Each bar represents the vaginal microbiota of one woman and corresponds to the identification number labeled in the dendogram, clustered using average linkage cluster analysis. A unique color is used to represent species found in >1% abundance. Species with <1% abundance are pooled into a fraction at the top in grey color.

129

Placebo (n=34) Probiotics (n=32) PTB & & & 13 wks V1_099 V1_300 V1_058 V1_053 V1_114 V1_151 V1_214 V1_314 V1_212 V1_299 V1_315 V1_277 V1_280 V1_326 V1_294 V1_290 V1_222 V1_328 V1_083 V1_250 V1_285 V1_170 V1_247 V1_258 V1_037 V1_179 V1_168 V1_329 V1_295 V1_296 V1_090 V1_178 V1_091 V1_127 V1_085 V1_199 gestation V1_109 V1_311 V1_274 V1_275 V1_319 V1_266 V1_293 V1_271 V1_289 V1_016 V1_303 V1_030 V1_094 V1_106 V1_164 V1_088 V1_147 V1_125 V1_209 V1_302 V1_118 V1_173 V1_038 V1_224 V1_322 V1_041 V1_327 V1_248 V1_213 V1_193 V1_138 V1_152 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4 Microbiota fraction Microbiota fraction

0.2 0.2

0.0 0.0

28 wks gestation

35 wks gestation

Figure 5-4 Stacked bar plots showing the vaginal microbiota across pregnancy Black& clustered**&discussion&about&Black&and&PTB??& by bacteria similarity in pregnant women who received either placebo (n=34) or probiotic (n=32) treatment. Each bar represents the vaginal microbiota of a single woman and corresponds to the identification number labeled in the dendogram, clustered using average linkage cluster analysis. Species found in >1% abundance are represented by a unique color and species that has <1% abundance are pooled into a single fraction at the top of the bar in grey color. Women were aligned in the same vertical column at 13, 28 and 35 weeks of gestation. Women who have undergone preterm birth (PTB) (n=2) in the probiotic group are denoted with white squares.

130

1 Shannon Index Diversity 0 13w 28w 35w 13w 28w 35w Placebo Probiotics

Figure 5-5 Scatterplot showing the Shannon Diversity Index (SDI) across gestations in pregnant women who received either placebo or probiotic treatment. Results are mean values ± SD and expressed in ratios. Comparisons between the probiotic (n= 32) and placebo (n= 34) groups at 13, 28 and 35 weeks gestation were assessed with Two Way Repeated Measure ANOVA followed by Holm-Sidak post hoc test (p>0.05).

131

IL-4 IL-10 7 35 6 30 5 25 4 20 3 2 15 2.0 15 1.5 10 1.0 5 Concentration (pg/mL) Concentration

Concentration (pg/mL) Concentration 0.5 0.0 0 13w 28w 35w 13w 28w 35w 13w 28w 35w 13w 28w 35w Placebo Probiotics Placebo Probiotics

CSF3 1000

750

500

250

-250 Concentration (pg/mL) Concentration -500 13w 28w 35w 13w 28w 35w Placebo Probiotics

Figure 5-6 Scatterplots showing the concentrations of cervico-vaginal cytokines IL-4, IL-10 and CSF3 across gestation in pregnant women who received either placebo or probiotic treatment. Results are mean values ± SD and expressed in picogram per milliliter. Comparison between the placebo group (n=33) and the probiotic group (n=31) was assessed with the Generalized Estimation Equation model in R.

132

Table 5-1 Characteristics of pregnant women randomized at 13 weeks gestation. Comparison between the probiotic group (n=32) and the placebo group (n=34) was performed with Student’s t-test or Chi-square (p>0.05).

Ethnicity is based on 32 women in the placebo group and 34 women in the probiotic group.

133

Table 5-2 Pregnancy outcomes.

Comparison between the probiotic group (n=32) and the placebo group (n=34) was performed with Student’s t-test or Chi-square (p>0.05).

Probiotic Group Placebo Group n = 41 n = 43 Antibiotics during pregnancy 1 6 (13.9%) 5 (11.6%) Antibiotics during labour and delivery 19 (46.3%) 16 (37.2%) Induction of labor 8 (19.5%) 9 (20.9%) Mode of Delivery Vaginal 33 (80.5%) 34 (79.1%) Spontaneous 28/33 (84.9%) 28/34 (82.4%) Assisted 5/33 (15.2%) 6/34 (17.7%)

C-section 8 (19.5%) 9 (20.9%) Emergency 6/8 (75.0%) 2/9 (22.2%) Labour 6/6 1/2 Elective 2/8 (25.0%) 7/9 (77.8%) Repeat 2/2 6/7 Abnormal presentations 1/2 1/7 Gestational age at delivery (weeks) 39.1 ± 1.4 39.4 ± 0.9

Birth weight (g) 3340 ± 433.68 3351 ± 463.49

IUGR Severe (<3rd centile) 2 0 1 (2.3%) Preterm birth ( < 37 weeks 2 0 gestation) 3 Apgar score <7 at 5 minutes 1 (2.4%) 0

Fetal Sex Male 19 (46.3%) 24 (55.8%) Female 22 (53.7%) 19 (44.2%) Cord blood pH 7.26 ± 0.07 7.26 ± 0.08

1 Antibiotics included penicillin, teva-cloxacillin, erythromycin, amoxicillin, macrobid, clindamycin, biaxin, ciprofloxacin, cephalexin and topical metronidazole. 2 The woman also had oligohydramnios. 3 Two women (not included in n=41) delivered at 34 weeks gestation in association of premature rupture of membranes.

134

Table 5-3 Compliance of women in the probiotic and placebo groups.

Comparison between the probiotic group (n=32) and the placebo group (n=34) was performed with Student’s t-test (p>0.05).

Compliance Probiotic Group Placebo Group n = 32 n = 34

Number of pills remaining in bottle at the 13 (0 - 28) 9 (0 - 26) end of 12 weeks of treatment (range)

Number of women who took > 50% of 6 (18.8%) 5 (14.7%) total pills in bottle ( < 84 pills remaining)

Number of women who took > 75% of 26 (81.2%) 29 (86.3%) total pills in bottle ( < 42 pills remaining)

135

Table 5-4 Nugent scores of pregnant women across pregnancy in the probiotic and placebo groups. Comparison between the probiotic group (n=32) and the placebo group (n=34) was performed with Student’s t-test (p>0.05).

Nugent Score Probiotic Group Placebo Group P value n = 32 n = 34

13 weeks gestation

• BV 11 (34.4%) 13 (38.2%) > 0.05

• Intermediate flora 21 (65.6%) 21 (61.8%) > 0.05

• Normal 0 0

28 weeks gestation

• BV 11 (34.4%) 4 (11.8%) > 0.05

• Intermediate flora 10 (31.3%) 19 (55.9%) > 0.05

• Normal 11 (34.4%) 11 (32.3%) > 0.05

35 weeks gestation

• BV 8 (25.0%) 10 (29.4%) > 0.05

• Intermediate flora 12 (37.5%) 12 (35.3%) > 0.05

• Normal 12 (37.5%) 12 (35.3%) > 0.05

136

Table 5-5 The relative to mean abundance of vaginal bacterial species in pregnant women with a BV (7-10) or an intermediate (4-6) Nugent score at 13 weeks gestation. Results are mean values ± SD and expressed in centered logarithm transformed ratios. Comparison between the BV (n=24) and Intermediate (n=42) groups was performed with Student’s t-test (p >0.05).

Species BV Intermediate Species BV Intermediate Gardnerella vaginalis 10.9±2.1 10.1±2.5 Anaerococcus hydrogenalis -1.6±1.4 -1.0±1.4 Atopobium vaginae 9.9±2.1 8.8±2.2 Gemella asaccharolytica -1.6±2.3 -2.1±0.9 Lactobacillus iners 9.8±2.6 10.6±2.3 TM7 phylum -1.6±2.5 -1.5±1.5 Lactobacillus crispatus 8.6±2.3 10.3±2.6 Bifidobacterium bifidum -1.7±2.1 -2.1±0.9 Lactobacillus jensenii 7.1±2.6 8.0±2.8 Peptoniphilus lacrimalis -1.8±1.0 -2.0±1.0 Veillonellaceae bacterium 6.9±3.3 4.6±2.1 Varibaculum cambriense -1.8±1.2 -1.6±1.1 Lactobacillus acidophilus 6.3±3.0 6.9±2.3 Staphylococcus epidermidis -1.8±1.4 -0.9±1.2 Lactobacillus gasseri 5.6±1.8 6.9±2.5 Arthrobacter albus -1.9±1.0 -1.6±1.1 Bifidobacterium breve 4.9±2.9 5.0±2.9 Facklamia hominis -1.9±1.0 -1.6±1.0 Bifidobacterium longum 4.2±2.7 2.7±3.1 Corynebacterium appendicis -1.9±1.0 -1.9±0.9 Atopobium rimae 3.8±2.8 2.0±1.6 Veillonella parvula -1.9±1.0 -1.8±1.3 Dialister micraerophilus 3.8±2.4 2.4±1.5 Clostridiales coagulans -1.9±1.1 -1.8±0.9 Prevotella timonensis 3.5±2.9 2.3±1.8 Prevotella corporis -1.9±1.2 -2.0±0.8 Dialister propionicifaciens 2.9±2.8 1.4±1.5 Mobiluncus curtisii -1.9±1.2 -1.9±0.9 Bacillus cereus 2.5±3.2 2.7±2.4 Corynebacterium coyleae -1.9±1.2 -1.9±0.9 Lactobacillus rhamnosus 2.3±1.7 1.9±1.7 Corynebacterium amycolatum -2.0±1.0 -1.6±0.9 Lactobacillus vaginalis 2.2±1.9 3.5±2.7 Campylobacter ureolyticus -2.0±1.0 -1.9±0.9 Prevotella bivia 2.1±3.0 0.7±1.8 Corynebacterium mucifaciens -2.0±1.1 -1.9±0.8 Streptococcus agalactiae 2.1±2.1 2.6±2.4 Peptostreptococcus anaerobius -2.0±1.1 -2.0±0.8 Desulfotomaculum halophilum 1.6±3.3 0.6±1.9 Porphyromonas asaccharolytica -2.0±1.3 -1.9±0.8 Prevotella amnii 1.3±4.2 -0.1±2.0 Brevibacterium ravenspurgense -2.0±1.4 -2.0±1.0 Enterobacter cloacae 1.1±2.9 1.1±1.8 Bifidobacterium dentium -2.0±1.6 -1.9±1.5 Escherichia coli 1.1±2.5 1.3±1.9 Anaerococcus murdochii -2.1±0.9 -2.0±0.6 Lactobacillus delbrueckii 0.9±3.3 0.5±2.5 Anaerococcus prevotii -2.1±0.9 -2.0±0.9 Sneathia sanguinegens 0.9±3.2 0.1±1.9 Lactobacillus sp.TS2gene -2.1±1.0 -1.8±1.2 Leuconostoc mesenteroides 0.8±3.3 1.3±2.5 Anaerococcus obesiensis -2.1±1.1 -1.6±1.4 Leptotrichia amnionii 0.8±3.3 0.0±1.9 Campylobacter rectus -2.1±1.1 -2.2±0.8 Lactococcus lactis 0.8±2.9 0.8±2.3 Anaerococcus tetradius -2.1±1.2 -2.0±1.2 Streptococcus anginosus 0.6±1.8 0.8±1.6 Propionimicrobium lymphophilum -2.1±1.2 -2.1±0.9 Alloscardovia omnicolens 0.5±3.0 0.2±2.5 Erythrobacter flavus -2.1±1.9 -1.2±2.7 Lactobacillaceae bacterium 0.5±3.0 0.9±3.0 Anaerococcus lactolyticus -2.2±0.9 -2.2±0.6 Corynebacterium jeikeium 0.0±1.4 -0.1±1.5 Bifidobacterium adolescenti -2.2±1.7 -1.8±1.5 Peptoniphilus|s|sp. S9 -0.1±1.6 -0.4±1.3 Actinomyces europaeus -2.3±0.9 -2.0±0.8 Prevotella bacterium -0.2±2.7 -0.6±1.5 Sideroxydans lithotrophicus -2.3±0.9 -2.2±0.6 Finegoldia magna -0.3±1.6 0.1±1.1 Clostridiales bacterium -2.4±0.8 -1.9±0.8 Streptococcus sobrinus -0.3±2.1 -1.0±1.5 Porphyromonas bennonis -2.4±0.9 -2.1±0.8 Morganella morga -0.4±2.7 -0.2±1.9 Prevotella denticola -2.4±1.2 -2.3±0.7 Streptococcus thermophilus -0.5±2.6 0.3±2.1 FirGemella haemolysans -2.4±1.2 -2.1±1.3 Bifidobacterium adolescentis -0.8±2.1 -1.0±2.1 Vulcanibacillus modesticaldus -2.5±0.8 -2.2±0.9 Prevotella micans -0.8±2.3 -1.5±1.3 Prevotella disiens -2.5±1.1 -1.9±1.3 Corynebacterium sundsvallense -1.0±1.5 -0.8±1.5 Lactobacillus brevis -2.5±1.2 -2.0±1.4 Lactobacillus coleohominis -1.1±2.2 -0.6±2.1 Cryptobacterium curtum -2.5±1.5 -2.1±1.3 Streptococcus pneumoniae -1.2±1.4 -0.2±2.3 Tannerella forsythia -2.6±0.7 -2.3±0.5 Corynebacterium pseudogenitalium -1.2±1.6 -0.9±1.7 Actinobaculum massiliense -2.6±0.9 -2.2±0.9 Prevotella melaninogenica -1.2±1.9 -1.0±1.8 Globicatella sanguinis -2.7±0.9 -2.3±0.7 Actinomyces neuii -1.3±2.7 -1.5±1.1 Helcococcus sueciensis -2.7±0.9 -2.3±0.6 Fusobacterium nucleatum -1.6±1.2 -1.6±1.1

137

Table 5-6 The relative to mean abundance of vaginal bacteria species that decreased across gestation in pregnant women treated with placebo or probiotics.

Results are mean values ± SD and expressed in centered logarithm transformed ratios. Comparisons between the placebo (n=34) and probiotic (n=32) groups at 13, 28 and 35 weeks gestation were assessed Generalized Estimation Equation model in R. Statistical significance within the placebo group (a’, b’ and c’) and within the probiotic group (a, b, and c) was denoted with different letters (p < 0.05).

Placebo Group (n=32) Probiotic Group (n=34)

Species 13 wks 28 wks 35 wks 13 wks 28 wks 35 wks p-value

a’ a’ b’ a a a Lactobacillus iners 10.7±2.7 10.5±2.5 10.1±2.5 9.8±2.1 9.8±2.5 9.5±2.4 6.5E-03

a’ a’ b’ a b b Gardnerella vaginalis 10.3±2.4 10.0±2.2 9.3±2.4 10.5±2.4 9.9±2.4 9.6±2.5 2.1E-09

a’ a’ b’ a b a,b Atopobium vaginae 9.4±2.2 9.3±2.2 8.8±2.2 9.0±2.3 8.6±2.4 8.7±2.4 1.8E-03

a’ b’ c’ a a a Lactobacillus acidophilus 6.6±2.6 5.7±2.7 5.3±2.7 6.8±2.5 6.4±2.4 6.4±2.4 5.6E-07

a’ a’ b’ a b b Atopobium rimae 2.8±2.2 2.3±2.3 1.6±2.3 2.5±2.3 1.9±2.0 1.7±2.6 1.7E-05

a’ b’c’ c’ a b b Bacillus cereus 2.8±2.5 1.6±2.4 1.9±2.2 2.5±2.9 1.5±2.9 1.7±2.9 1.3E-08

a’ b’ c’ a b b Lactobacillaceae bacterium 1.6±3.2 -0.7±1.4 -1.3±1.5 -0.2±2.5 -1.2±1.6 -1.4±1.2 7.9E-11

a’ b’ b’ a b b Escherichia coli 1.2±1.8 -0.7±1.7 -0.2±2.1 1.3±2.5 -0.4±2.1 0.0±1.9 1.8E-10

a’ b’ c’ a b b Desulfotomaculum halophilum 1.4±2.6 0.4±2.6 -0.8±2.4 0.5±2.5 -0.9±2.1 -1.1±2.3 1.9E-11

a’ b’ b’ a a a Streptococcus thermophilus -0.3±2.3 -2.4±1.3 -2.4±1.6 0.2±2.3 -1.7±1.6 -2.3±1.6 5.6E-17

a’ b’ b’ a b b Erythrobacter flavus -1.7±2.4 -3.0±1.1 -3.1±1.4 -1.4±2.6 -2.6±1.7 -2.9±1.6 5.6E-14

a’ a’ b’ a a,b b Prevotella denticola -2.5±0.8 -2.8±1.8 -3.4±1.0 -2.2±1.0 -2.7±1.2 -3.1±1.1 2.5E-07

138

Table 5-7 The relative to mean abundance of vaginal bacterial species that increased across gestation in pregnant women treated with placebo or probiotics. Results are mean values ± SD and expressed in centered logarithm transformed ratios. Comparisons between the placebo (n=34) and probiotic (n=32) groups at 13, 28 and 35 weeks gestation were assessed Generalized Estimation Equation model in R. Statistical significance within the placebo group (a’, b’ and c’) and within the probiotic group (a, b, and c) was denoted with different letters (p < 0.05).

Placebo Group (n=32) Probiotic Group (n=34)

Species 13 wks 28 wks 35 wks 13 wks 28 wks 35 wks p-value

a’ b’ c’ a b b Corynebacterium pseudogenitalium -0.9±1.5 0.4±1.6 1.0±1.8 -1.1±1.8 1.0±2.4 1.0±1.7 1.7E-16

a’ a’ b’ a b b Facklamia hominis -1.6±1.0 -1.4±1.4 -0.8±1.8 -1.8±1.1 -0.9±1.6 -1.0±1.6 4.2E-05

a’ b’ c’ a b b Corynebacterium amycolatum -1.9±0.9 -1.1±1.5 -0.5±1.8 -1.6±1.0 -0.7±1.7 -0.3±1.1 8.1E-10

a’ b’ c’ a b b Clostridiales coagulans -1.8±0.9 -1.0±1.7 -0.3±2.0 -1.9±1.0 -1.3±1.8 -1.3±1.7 2.9E-07

a’ b’ b’ a b b Varibaculum cambriense -1.7±1.0 -0.4±1.8 0.0±1.7 -1.6±1.3 -0.7±1.5 -0.5±1.8 5.1E-12

a’ b’ b’ a a b Campylobacter ureolyticus -2.0±0.9 -1.8±1.7 -1.1±1.5 -1.6±1.6 -1.6±1.6 -1.4±2.1 4.3E-06

a’ b’ b’ a b b Corynebacterium coyleae -2.2±1.4 -1.1±2.4 -1.4±2.4 -2.1±1.1 -1.7±1.6 -1.7±1.5 1.9E-03

a’ b’ b’ a a a Prevotella disiens -0.9±1.5 0.4±1.6 1.0±1.8 -1.1±1.8 1.0±2.4 1.0±1.7 2.8E-04

a’ a’ b’ a a b Cryptobacterium curtum -1.6±1.0 -1.4±1.4 -0.8±1.8 -1.8±1.1 -0.9±1.6 -1.0±1.6 5.3E-09

139

Table 5-8 Summary table of cervico-vaginal cytokines and chemokines across gestation in pregnant women who received either placebo or probiotic treatment. Results are mean values ± SD and expressed in picogram per milliliter. Data have equal variance but were not normally distributed. Comparison between the placebo group (n=33) and the probiotic group (n=31) was assessed with the Generalized Estimation Equation model in R. Statistical significance within the placebo group (a’, b’ and c’) and within the probiotic group (a, b, and c) is denoted with different letters (p<0.05).

Placebo Group (n = 33) Probiotic Group (n = 31)

13 wks 28 wks 35 wks 13 wks 28 wks 35 wks IL-1β 121.3±186.6 a’ 80.7±171.6 a’ 72.2±166.8 a’ 199.7±404.2 a 66.4±143.3 a 82.3±113.2 a IL-2 0.6±0.7 a’ 0.6±0.6 a’ 0.4±0.5 a’ 0.4±0.6 a 1.2±2.8 a 0.6±0.6 a IL-4 0.6±0.4 a’ 0.8±0.4 b’ 0.7±0.4 a’, b’ 0.8±0.4 a 1.3±1.2 a 0.7±0.4 a IL-5 0.3±0.3 a’ 0.4±0.5 a’ 0.4±0.3 a’ 0.6±1.2 a 1.1±2.1 a 0.5±0.3 a IL-6 15.1±30. 5 a’ 4.1±5.6 a’ 3.4±5.0 a’ 36.0±71.1 a 6.3±10.7 a 5.3±8.0 a IL-7 56.0±121.4 a’ 34.6±33.7 a’ 29.9±37.1 a’ 55.4±117.5 a 90.4±269.4 a 29.3±36.2 a IL-8 1453.9±2230.1 a’ 1155.8±2716.0 a’ 604.4±985.1 a’ 2068.0±4658.2 a 418.2±442.6 a 855.0±1258.5 a IL-9 7.9±16.2 a’ 4.6±4.1 a’ 4.5±6.7 a’ 6.4±12.1 a 16.5±57.3 a 3.8±3.8 a IL-10 8.4±2.9 a’ 10.0±2.7 b’ 9.0±3.6 a’,b’ 8.4±3.2 a 11.0±4.6 b 9.9±2.4 a,b IL-12p70 57.8±89.8 a’ 55.0±47.0 a’ 44.6±38.1 a’, 50.3±72.7 a 90.0±217.5 a 41.4±22.1 a IL-13 4.7±8.8 a’ 3.4±2.1 a’ 3.3±3.0 a’ 5.1±9.8 a 9.7±27.8 a 3.1±2.8 a IL-15 1.0±1.1 a’ 1.3±1.5 a’ 0.8±0.9 a’ 1.0±1.0 a 1.6±1.6 a 0.8±0.9 a IL-17 4.3±2.7 a’ 5.0±2.6 a’ 3.7±1.8 a’ 4.3±2.6 a 7.7±7.1 a 3.8±1.8 a CCL2 10.7±17.4 a’ 8.4±4.3 a’ 6.8±4.6 a’ 10.6±13.3 a 11.6±10.2 a 9.5±8.5 a CCL3 1.9±1.5 a’ 1.5±1.0 a’ 1.6±1.8 a’ 3.7±6.1 a 1.7±1.4 a 1.6±0.9 a CCL4 9.1±9.5 a’ 5.6±8.8 a’ 5.0±13.4 a’ 21.1±48.8 a 3.8±3.3 a 6.3±7.6 a CCL5 32.8±134.4 a’ 3.8±1.5 a’ 2.9±1.4 a’ 10.0±33.8 a 4.4±3.1 a 3.3±1.5 a CCL11 11.2±19.4 a’ 14.6±36.7 a’ 7.2±9.6 a’ 6.6±11.8 a 24.8±78.1 a 11.0±12.7 a CSF2 15.1±13.2 a’ 12.1±7.0 a’ 9.4±7.1 a’ 13.0±12.0 a 21.9±45.6 a 9.2±6.4 a CSF3 131.9±156.2 a’ 58.1±129.8 a’, b’ 44.2±81.1 b’ 204.6±253.5 a 60.5±109.5 b 73.4±122.0 a CXCL10 1346.5±3779.0 a’ 639.9±762.5 a’ 309.7±354.0 a’ 512.7±1064.9 a 682.6±1371.7 a 581.2±866.2 a TNF-α 31.4±55.3 a’ 33.1±41.7 a’ 24.7±30.2 a’ 31.6±30.6 a 56.8±84.7 a 31.6±27.9 a IFN-Υ 49.0±45.9 a’ 73.5±46.4 a’ 62.0±55.1 a’ 80.4±68.4 a 127.4±110.9 a 67.3±54.4 a PDGF-bb 74.4±133.1 a’ 45.6±60.0 a’ 33.2±41.6 a’ 64.7±136.2 a 90.8±249.4 a 31.6±34.5 a bFGF 5.1±7.0 a’ 4.2±2.5 a’ 3.6±2.5 a’ 4.4±3.0 a 6.9±12.5 a 3.4±1.3 a VEGF 2982.6±4491.4 a’ 3666.5±4247.7 a’ 3541.9±4860.3 a’ 3883.9±8847.3 a 2856.4±2362.8 a 2623.1±2070.7 a

Chapter Six

General Discussion

140 141

Chapter 6

6. General Discussion

A disruption to the balance between pro-inflammatory cytokines and anti-inflammatory cytokines that favours an inflammatory milieu underlies the pathogenesis of infection/ inflammation associated preterm birth (PTB) (MacIntyre et al., 2012). A disturbance of the vaginal microbiota such as that observed in bacterial vaginosis (BV) also contributes to an increased risk of PTB (Donders et al., 2009). Limited knowledge is available regarding the use of probiotic lactobacilli as a prophylactic treatment for PTB. In this thesis, I assessed the effect of probiotic lactobacilli and its supernatant on the incidence of PTB and the immune- regulatory role of lactobacilli using pregnant mice. I also examined the effect of oral lactobacilli on the cervico-vaginal cytokines in pregnant women with an abnormal Nugent score. The effect of lactobacilli on the vaginal microbiota in both mice and pregnant women was also investigated.

I specifically studied 1) the effect of Lactobacillus rhamnosus GR-1 (GR-1) live bacteria and its supernatant (GR-1 SN) on the prevention of LPS-induced PTB in pregnant CD-1 mice; 2) the effect of GR-1 and GR-1 SN on the systemic and intra-uterine cytokine and chemokine profiles in pregnant CD-1 mice; 3) the effect of L. rhamnosus GR-1 and L. reuteri RC-14 (GR-1 and RC-14) live bacteria on the cervico-vaginal concentrations of cytokines and chemokines in pregnant women with an abnormal Nugent score; 4) the potential of using GR-1 to modulate the mouse vaginal microbiota; and 5) the potential of using GR-1 and RC-14 to alter the vaginal microbiota of pregnant women with an abnormal Nugent score.

I found that pre-treatment with GR-1 SN, but not with GR-1 live bacteria, reduces the incidence of inflammation (LPS)-induced PTB in pregnant CD-1 mice. I also observed that GR-1 SN and GR-1 live bacteria differentially modulate the systemic and intrauterine murine immune responses (Figure 6-1). I then investigated whether GR-1 live bacteria itself

142 has an immune-regulatory role in pregnant mice. The effects of oral GR-1 live bacteria systemically, as reflected by changes in the maternal plasma and locally within the intrauterine tissues, are mainly pro-inflammatory. I observed elevations in the pro-inflammatory cytokines TNFα, IL-6, IL-12p70, IL-17 and IFN-γ, and chemokines CCL2, CCL3, CCL4 and CCL5 with live bacteria administration. In contrast, GR-1 SN alone did not have any effect on pro-inflammatory cytokines or chemokines. The effect of GR-1 SN is primarily anti-inflammatory, with GR-1 SN alone increasing the placental anti-inflammatory cytokines IL-10 and IL-4 in pregnant CD-1 mice. This is consistent with previous in vitro studies, in which GR-1 SN increased the production of IL-10 in cultured human placental trophoblast cells (Yeganegi et al., 2010) and decidual cells (Li et al., 2014). Furthermore, I found that GR-1 SN dampens LPS-induced increases in pro-inflammatory cytokines and chemokines in pregnant mice (Figure 6-2). This differential effect on inflammatory mediators is in keeping with observations in previous studies, which have shown that the cell-free culture supernatant (CFS) of Bifidobacterium breve CNCM I-4035 is more effective than its live bacteria counterpart at suppressing the secretion of pro-inflammatory cytokines and chemokines in human dendritic cells (DCs) challenged (Bermudez-Brito et al., 2013). It has been shown that the maternal DCs surface expressions of co-stimulatory molecules CD86 and CD80 and antigen presenting molecule (HLA-DR) are reduced during pregnancy, suggesting DCs may be important in the immune tolerance of a semi-allogeneic fetus (Bachy et al., 2008). DCs treated with CNCM I-4036 live bacteria alone secrete inflammatory cytokines IL-1β, IL-6, IL-8, IL-12 and TNFα while CFS alone decreased the secretion of IL-8 and IL-12p40 (Bermudez-Brito et al., 2014). B. breve live bacteria alone are more potent stimulators of the pro-inflammatory cytokines and chemokines than its supernatant (Bermudez-Brito et al., 2013). Furthermore, B. breve CNCM I-4035 supernatant dampens the secretion of pro-inflammatory cytokines IL-1β, IL-6, IL-12p40, and chemokines MCP-1, MIP-1α and RANTES, while B. breve live bacteria increase the production of these chemokines in response to a challenge with Salmonella (Bermudez-Brito et al., 2013).

GR-1 live bacteria increased the concentration of CCL2 and IFN-γ in pregnant CD-1 mice both of which promote pathogen elimination. CCL2 is also responsible for the recruitment

143 of monocytes and their differentiation into macrophages (Mak, 2006). Furthermore, CCL2 enhances the phagocytic activity of macrophages. And IFN-γ possesses anti-pathogenic and anti-proliferative properties (Mak, 2006). IFN-γ has also been shown to reduce the expression of COX-2 and PGE2 in term and preterm placenta (Hanna et al., 2004). Taken together with the findings that GR-1 SN has anti-inflammatory properties, I speculate that the secreted metabolites in the GR-1 SN limit the inflammatory mediators produced by its live bacteria counterpart; while at the same time, the anti-infective properties of GR-1 live bacteria are maintained.

Lipoteichoic Acid (LTA), which is present on the cell surface of gram-positive lactobacilli, is immune-stimulatory through activation of the Toll-like receptor (TLR) 2 pathway in a murine model of colitis (Grangette et al., 2005). Enhanced anti-inflammatory activity has been found in a murine model of colitis when LTA is substituted (D-alanylation) or removed (Grangette et al,. 2005; Claes et al., 2010; Mohamadzadeh et al., 2011). It has been suggested that the active moiet(ies) responsible for the anti-inflammatory properties of B. breve CNCM I-4035 supernatant are likely proteins (Bermudez-Brito et al., 2013). These results suggest that soluble active metabolites, produced by GR-1 live bacteria and released into the supernatant, have anti-inflammatory properties, and GR-1 live bacteria and its supernatant exert their immune-regulatory effects via activation of signaling pathways.

Previous studies have shown that the oral administration of Lactobacillus rhamnosus can influence body sites distant to the gut, such as the respiratory tract (Villena et al., 2012), the skin (Tanaka et al., 2009), and the heart of murine animals (Gan et al., 2014). Oral probiotics can modulate murine intestinal mucosal immune responses (Ogita et al, 2015) as well as systemic immune responses (Forsythe et al., 2012). Furthermore, it has been shown that the serum and intestinal fluid cytokine profiles are similar to each other after the oral administration of L. rhamnosus CRL 1505 in mice (Villena et al., 2012). Although orally administered L. rhamnosus GR-1 to pregnant CD-1 mice did not change the cecal microbiota, I did observe a change in the systemic and intrauterine production of cytokines and chemokines, as well as a change in the vaginal microbiota. It is possible that GR-1 passing through the mouse gut induces the intestinal mucosa to secrete signaling molecules

144 into the systemic circulation, which then exert immune-regulatory effects within the intrauterine tissues and amniotic fluid. I did not detect a change in Lactobacillus rhamnosus abundance in the mouse vaginal microbiota with oral administration of GR-1 live bacteria. Therefore it is plausible that the secreted signaling mediators, rather than GR-1 live bacteria travel directly to the mouse vagina, that causes changes in the vaginal environment and results in an altered vaginal microbiota in pregnant CD-1 mice. It is also possible that the Ion Torrent sequencing method that I utilized in these experiments is not sufficiently sensitive enough to detect small changes, if present, in Lactobacillus rhamnosus abundance.

The oral probiotic combination L. rhamnosus GR-1 and L. reuteri RC-14 (GR-1 and RC- 14) did not alter the cervico-vaginal cytokine concentrations in low risk pregnant women with an abnormal Nugent score. This is in contrast to the findings in pregnant mice, in which I found oral GR-1 live bacteria induced both systemic and intrauterine inflammatory cytokines. There are a number of possible explanations for these differences. Firstly, the effects of lactobacilli could be species specific. It is also possible that unlike in the mouse study, in which I used lipopolysaccharide to induce inflammation, pregnant women in our randomized controlled trial had low risk pregnancies with no evidence of clinical or subclinical inflammatory processes. A combination of probiotic strains (GR-1 and RC-14) was used in the studies with pregnant women, whereas a single strain (GR-1) was used in pregnant mice. Further investigations in pregnant mice using GR-1 and RC-14 could provide additional insights in to the potential efficacy of multi-strain probiotic preparations, and whether a multi-strain (GR-1 and RC-14) or a single strain (GR-1) would be more beneficial in pregnant women.

I have also found that oral GR-1 and RC-14, at the dose given in these experiments, did not alter the vaginal microbiota in pregnant women with an intermediate or BV Nugent score. This finding is in contrast to previous studies in non-pregnant women that reported oral GR- 1 and RC-14 reduce BV recurrence by restoring the indigenous lactobacilli (Reid et al., 2003a). This difference could be due to differences in the hormonal environment and/or the vaginal microbial stability between pregnant and non-pregnant women. High levels of estrogen during pregnancy likely accounts for a higher abundance of Lactobacillus spp.

145 observed in pregnant women when compared to non-pregnant women (Romero et al., 2014a), since higher estrogen leads to an increase in mucosal glycogen production, whose metabolized substrates support vaginal lactobacilli colonization (Spear et al., 2014). Furthermore, it has been shown that the vaginal microbiota composition of pregnant women is more stable than non-pregnant women (Romero et al., 2014a). The vaginal microbiota of pregnant women may thus be more resilient to changes caused by additional exogenous oral lactobacilli. The probiotic dosage chosen for our study (5 X 109 colony-forming units/ cfu) was based on previous studies in non-pregnant women (Reid at al., 2003). It is possible that a higher dose is needed to colonize the vagina of pregnant women.

The administration of GR-1 live bacteria or GR-1 SN alone does not change the normal gestational length, fetal weight nor litter size in pregnant CD-1 mice. In women, there were no adverse reactions reported following ingestion of GR-1 and RC-14. Gestational age at delivery, birth weight, and cord blood pH were not different between neonates born to placebo and probiotics GR-1 and RC-14 treated mothers. These findings are in agreement with a previous study that reviewed 37 studies of prenatal probiotics, and found no evidence of adverse maternal or neonatal outcomes (VandeVusse et al., 2013). We screened women for an abnormal Nugent score prior to randomization since the presence of an abnormal vaginal biota such as BV is associated with a 1.4-fold increased risk of PTB. This study is the first to our knowledge to investigate the effect of probiotic lactobacilli on the vaginal microbiota and cervico-vaginal cytokine profiles across gestation in pregnant women who had an abnormal Nugent score initially.

There are a few limitations to consider when interpreting the findings presented in this thesis. In the mouse studies, I observed that different batches/ bottles of LPS with the same catalogue number can have different potency; thus, giving variable preterm delivery (PTD) rate. For instance, 125 µg of LPS was required to result in 100% PTB in the mouse study that evaluated the GR-1 SN effect alone (Chapter 3); whereas 50 µg of LPS was sufficient to cause 100% PTB in the mouse study that evaluated the effect of GR-1 live bacteria (Chapter 4). Although the same batch of LPS was used within each set of experiments, two different batches of LPS were used overall. In addition, separate control experiments were used for

146 each set. I noted a difference in the baseline concentration of progesterone between the two series of mouse experiments. In the GR-1 SN alone study, baseline progesterone concentrations were 68 ± 4.6 ng/mL whereas in the GR-1 live bacteria study it was 40 ± 4.4 ng/mL There are a number of factors that could contribute to this difference, including the fact that the time of day that the mice were sacrificed was different; mice in the GR-1 SN study were sacrificed half a day earlier than the mice in the GR-1 live bacteria study. The mice in the GR-1 SN study received saline via intra-peritoneal injection, whereas those mice in the GR-1 live bacteria study received saline through oral gavage. Different types of procedure might place different levels of stress on the animals, which may alter the baseline hormonal concentrations. In the human study, compliance to the treatment protocol was determined by counting the numbers of pills remaining in the bottle returned at the end of the study. The study could be strengthened if stool samples were also collected and subjected to quantitative PCR amplification to quantify the amount of GR-1 and RC-14 present.

Future experiments are needed to identify the active moiety(ies) responsible for the anti- inflammatory properties of Lactobacillus rhamnosus GR-1 supernatant. The supernatant could first be fractionated into lipid, proteins and LTA components and tested in pregnant mice to evaluate which component(s) is associated with the inflammatory dampening effect. If it were the protein component, further fractionation based on the molecular weights of the proteins could be performed using Fast Protein Liquid Chromatography. Fractions that have similar inflammatory dampening effects as the crude GR-1 SN in pregnant mice could then be subjected to mass spectrometry to identify the active moiety(ies). This would allow concentration of the fraction, which could potentially enhance the anti-inflammatory properties.

Other experiments could include identifying the differential underlying mechanisms by which L. rhamnosus GR-1 live bacteria and its supernatant exert their effects in pregnant mice. The LTA component of GR-1 live bacteria could also be removed to evaluate whether it is responsible for the inflammatory stimulating effect of GR-1 live bacteria. Furthermore, knockout pregnant mice lacking the gene(s) for various TLRs could be used to

147 identify the signaling pathways responsible for the actions of GR-1 live bacteria and its supernatant. Further mechanistic pathways downstream of the TLRs could also be evaluated given that previous studies have demonstrated that GR-1 SN increases the production of anti-inflammatory cytokine IL-10 through the JAK/STAT and MAPK pathways in cultured human trophoblast cells (Yeganegi et al., 2010).

Since GR-1 SN has also previously been shown in vitro to reduce the synthesis of prostaglandins (PGs), future experiments could be performed to investigate the effect of GR- 1 SN in vivo on other mediators of parturition including PGs, PTGS, PGDH and MMPs in pregnant mice. Myometrial concentrations of pro-inflammatory cytokines increase in both infection (LPS)-induced PTL and non-infection (RU-486, a progesterone antagonist) associated PTL in pregnant CD-1 mice (Shynlova et al., 2013). LPS induces an increase in the mRNA levels of various pro-inflammatory cytokines in the myometrium of pregnant CD-1 mice as early as 2 hours after intrauterine LPS administration (Shynlova et al., 2014). This initial outburst of pro-inflammatory cytokines may contribute to luteolysis and cause progesterone withdrawal via activation of the NF-κB pathway (Vrachnis et al., 2012). In this study, the anti-inflammatory effect of GR-1 SN was independent of circulating maternal progesterone concentrations. Future experiments giving GR-1 SN to RU486-treated pregnant mice would provide evidence whether GR-1 SN dampens the initial pro- inflammatory cytokine outburst or targets progesterone withdrawal and its associated increase in pro-inflammatory cytokines.

Given that the litter size and the fetal weight of pups did not change with GR-1 SN treatment, future experiments could be performed to evaluate the health of mouse neonates born to mothers that received GR-1 live bacteria and/or its supernatant. Intra-uterine infection/ inflammation has been associated with an increased prevalence of adverse neurobehavioral outcomes such as cerebral palsy in exposed offspring in the human (Yoon et al., 2000; Wu, 2002) as well as fetal neuronal abnormalities in mice (Burd et al., 2010). In this thesis, I found that GR-1 SN dampens LPS-induced systemic and intra-uterine inflammation in pregnant mice; future studies could be performed to evaluate the potential of GR-1 SN at reducing inflammation (LPS)-induced fetal brain injury in pregnant mice.

148

In future clinical trials, pregnant women with high-risk pregnancies based on a previous PTB or short cervix in conjunction with bacterial vaginosis could be recruited to investigate the effect of oral GR-1 and RC-14. The vaginal microbial profile of pregnant women lacking Lactobacillus spp could also be used in conjunction with the Nugent score to identify women most likely to benefit from probiotics. Supplementation with probiotics is also known to improve intestinal dysbiosis (de Moreno de Blanc and LeBlanc, 2014) and mucosal immunity (Wan et al., 2015), and probiotics are widely used for non-pregnancy related conditions. Alterations of the intestinal biota in turn may be important in the pathogenesis of other pregnancy complications such as preeclampsia, intrauterine growth restrictions or miscarriage (Zhang et al., 2015). In order to confirm that oral probiotics colonize the gut of pregnant women, stool samples from the mothers could be collected to evaluate the gut microbiome. Compared to other body sites (skin, nose, vagina and gut), the human placenta is most similar to the oral microbiome, which suggests a hematogenous route of pathogenic transmission to the intrauterine cavity may be important (Aagaard et al., 2014) It has been previously observed that the relative abundance of Actinomycetales and Alphaproteobacteria are increased in the preterm placenta compared to the term placenta (Aagaard et al., 2014). Furthermore, the commensal bacterial species of the human oral microbiome, F. nucleatum, has been associated with intrauterine infections (Han et al., 2009). Therefore, crosstalk may exist between multiple bacterial communities in pregnant women and it is important to take into consideration other microbiome sites in future clinical studies.

The research findings in this doctoral thesis provide evidence to the efficacy of Lactobacillus rhamnosus GR-1 supernatant, but not the live bacteria, to reduce LPS-induced PTB and inflammation in pregnant mice. I have also shown that GR-1 live bacteria can modulate both systemic and intrauterine cytokines as well as the vaginal microbiota of pregnant mice. These findings provide further support for the potential benefit of lactobacilli supernatant in the prevention of inflammation-associated conditions during pregnancy including PTB.

149

GR-1 SN GR-1 live bacteria

Maternal Plasma Maternal Plasma

No change Amniotic IL-12p40 Fluid TNFα

Fetal Membranes

CCL2, 3, CCL5 CCL5 !! 4, 5, 11

IL-4 IL-10 IL-10 IL-10 !! IL-4 !! IL-4 Placenta IL-17 TNFα IL-1α IL-6 Myometrium IFNγ IL-12p70 !!

Figure 6-1 Changes in sytemic and intrauterine cytokines after treatment with Lactobacillus rhamnosus GR-1 supernatant or live bacteria. Cytokines with a downward arrow decreased significantly following GR-1 treatment, when compared to mice that received saline. All other cytokines increased significantly following GR-1 treatment.

150

LPS LPS + GR-1 SN Maternal Plasma Maternal Plasma IL-1β CCL3 IL-1β !! CCL3 IL-6 CCL4 IL-6 !! CCL4 !! IL-12p40 CCL5 IL-12p40 !! CCL5 !! IL-12p70 Amniotic IL-12p70 Fluid TNFα TNFα !! IL-17 IL-17

Fetal Membranes

IL-1β IL-1β IL-6 IL-6 !! IL-12p40 IL-12p40 IL-6 IL-12p70 IL-6 !! IL-12p70 !! TNFα TNFα TNFα !! TNFα CCL3 IL-17 Placenta CCL3 IL-17 CCL4 CCL3 CCL4 !! CCL3 CCL5 CCL4 CCL5 CCL4 CCL5 CCL5 Myometrium IL-1β IL-1β IL-6 IL-6 !! IL-12p40 IL-12p40 IL-12p70 IL-12p70 !! TNFα TNFα !! IL-17 IL-17 !! CCL3,4,5 CCL3,4,5

Figure 6-2 LPS-induced sytemic and intrauterine cytokines that were dampened with GR-1 supernatant pretreatment.

Cytokines increased with LPS alone were shown on the left. On the right side of the figure, cytokines with a downward arrow decreased significantly with GR-1 supernatant pretreatment following subsequent LPS challenge, when compared to mice that received LPS alone. All other cytokines were not different between LPS group and LPS+GR-1 group.

References

151 152

List of References

Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, et al. (2014) The placenta harbors a unique microbiome. Sci Transl Med 6: 237ra265.

Aagaard K, Riehle K, Ma J, Segata N, Mistretta TA, et al. (2012) A metagenomic approach to characterization of the vaginal microbiome signature in pregnancy. PLoS One 7: e36466.

Aitchison J (1986) The Statistical Analysis of Compositional Data. New York: Chapman and Hall.

Ain R, Canham LN, Soares MJ (2003) Gestation stage-dependent intrauterine trophoblast cell invasion in the rat and mouse: novel endocrine phenotype and regulation. Dev Biol 260: 176-190.

Anteby EY, Natanson-Yaron S, Hamani Y, Sciaki Y, Goldman-Wohl D, et al. (2005) Fibroblast growth factor-10 and fibroblast growth factor receptors 1-4: expression and peptide localization in human decidua and placenta. Eur J Obstet Gynecol Reprod Biol 119: 27-35.

Anukam K, Osazuwa E, Ahonkhai I, Ngwu M, Osemene G, et al. (2006) Augmentation of antimicrobial metronidazole therapy of bacterial vaginosis with oral probiotic Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14: randomized, double-blind, placebo controlled trial. Microbes Infect 8: 1450-1454.

Arend WP, Malyak M, Guthridge CJ, Gabay C (1998) Interleukin-1 receptor antagonist: role in biology. Annu Rev Immunol 16: 27-55.

Aroutcheva A, Ling Z, Faro S (2008) Prevotella bivia as a source of lipopolysaccharide in the vagina. Anaerobe 14: 256-260.

Astle S, Newton R, Thornton S, Vatish M, Slater DM (2007) Expression and regulation of prostaglandin E synthase isoforms in human myometrium with labor. Mol Hum Reprod 13: 69-75.

Bachy V, Williams DJ, Ibrahim MA (2008) Altered dendritic cell function in normal pregnancy. J Reprod Immunol 78: 11-21.

153

Balkus J, Agnew K, Lawler R, Mitchell C, Hitti J (2010) Effects of pregnancy and bacterial vaginosis on proinflammatory cytokine and secretory leukocyte protease inhibitor concentrations in vaginal secretions. J Pregnancy 2010: 385981.

Baranao RI, Piazza A, Rumi LS, Polak de Fried E (1997) Determination of IL-1 and IL-6 levels in human embryo culture-conditioned media. Am J Reprod Immunol 37: 191-194.

Barfod KK, Roggenbuck M, Hansen LH, Schjorring S, Larsen ST, et al. (2013) The murine lung microbiome in relation to the intestinal and vaginal bacterial communities. BMC Microbiol 13: 303.

Behrman RE BA (2007) Preterm Birth: Causes, Consequences, and Prevention. Washington DC: National Academy of Sciences.

Beigi RH, Austin MN, Meyn LA, Krohn MA, Hillier SL (2004) Antimicrobial resistance associated with the treatment of bacterial vaginosis. Am J Obstet Gynecol 191: 1124-1129.

Bermudez-Brito M, Munoz-Quezada S, Gomez-Llorente C, Matencio E, Bernal MJ, et al. (2013) Cell-free culture supernatant of Bifidobacterium breve CNCM I-4035 decreases pro- inflammatory cytokines in human dendritic cells challenged with Salmonella typhi through TLR activation. PLoS One 8: e59370.

Bermudez-Brito M, Munoz-Quezada S, Gomez-Llorente C, Romero F, Gil A (2014) Lactobacillus rhamnosus and its cell-free culture supernatant differentially modulate inflammatory biomarkers in Escherichia coli-challenged human dendritic cells. Br J Nutr 111: 1727-1737.

Bhat G, Williams SM, Saade GR, Menon R (2014) Biomarker interactions are better predictors of spontaneous preterm birth. Reprod Sci 21: 340-350.

Blackburn S (2012) Maternal, Fetal & Neonatal Physiology. USA: Saunders.

Blencowe H, Cousens S, Oestergaard MZ, Chou D, Moller AB, et al. (2012) National, regional, and worldwide estimates of preterm birth rates in the year 2010 with time trends since 1990 for selected countries: a systematic analysis and implications. Lancet 379: 2162- 2172.

Bodean O, Munteanu O, Cirstoiu C, Secara D, Cirstoiu M (2013) Probiotics - a helpful additional therapy for bacterial vaginosis. J Med Life 6: 434-436.

154

Borges S, Silva J, Teixeira P (2014) The role of lactobacilli and probiotics in maintaining vaginal health. Arch Gynecol Obstet 289: 479-489.

Brace RA, Wolf EJ (1989) Normal amniotic fluid volume changes throughout pregnancy. Am J Obstet Gynecol 161: 382-388.

Brocklehurst P, Gordon A, Heatley E, Milan SJ (2013) Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev 1: CD000262.

Brostrom EB, Akre O, Katz-Salamon M, Jaraj D, Kaijser M (2013) Obstructive pulmonary disease in old age among individuals born preterm. Eur J Epidemiol 28: 79-85.

Buhimschi CS, Dulay AT, Abdel-Razeq S, Zhao G, Lee S, et al. (2009) Fetal inflammatory response in women with proteomic biomarkers characteristic of intra-amniotic inflammation and preterm birth. Bjog 116: 257-267.

Burd I, Bentz AI, Chai J, Gonzalez J, Monnerie H, et al. (2010) Inflammation-induced preterm birth alters neuronal morphology in the mouse fetal brain. J Neurosci Res 88: 1872- 1881.

Burd ID, Ness A, DiMuzio P, Ren GY, Tulenko TN (2009) Clostridium botulinum toxin A inhibits contractility in pregnant human myometrium in vitro. Reprod Sci 16: 1001-1004.

Chaban B, Links MG, Jayaprakash TP, Wagner EC, Bourque DK, et al. (2014) Characterization of the vaginal microbiota of healthy Canadian women through the menstrual cycle. Microbiome 2: 23.

Chaiworapongsa T, Romero R, Kim JC, Kim YM, Blackwell SC, et al. (2002) Evidence for fetal involvement in the pathologic process of clinical chorioamnionitis. Am J Obstet Gynecol 186: 1178-1182.

Challis J, Newnham J, Petraglia F, Yeganegi M, Bocking A (2013) Fetal sex and preterm birth. Placenta 34: 95-99.

Challis JR, Lockwood CJ, Myatt L, Norman JE, Strauss JF, 3rd, et al. (2009) Inflammation and pregnancy. Reprod Sci 16: 206-215.

Challis JRG, Matthews SG, Gibb W, Lye SJ (2000) Endocrine and paracrine regulation of birth at term and preterm. Endocr Rev 21: 514-550.

155

Chandiramani M, Seed PT, Orsi NM, Ekbote UV, Bennett PR, et al. (2012) Limited relationship between cervico-vaginal fluid cytokine profiles and cervical shortening in women at high risk of spontaneous preterm birth. PLoS One 7: e52412.

Chatterjee P, Chiasson VL, Bounds KR, Mitchell BM (2014) Regulation of the anti- inflammatory cytokines interleukin-4 and interleukin-10 during pregnancy. Front Immunol 5: 253.

Chau SE, Murthi P, Wong MH, Whitley GS, Brennecke SP, et al. (2013) Control of extravillous trophoblast function by the eotaxins CCL11, CCL24 and CCL26. Hum Reprod 28: 1497-1507.

Cheng SB, Sharma S (2014) Interleukin-10: A Pleiotropic Regulator in Pregnancy. Am J Reprod Immunol.

Christiaens I, Zaragoza DB, Guilbert L, Robertson SA, Mitchell BF, et al. (2008) Inflammatory processes in preterm and term parturition. J Reprod Immunol 79: 50-57.

Claes IJ, Lebeer S, Shen C, Verhoeven TL, Dilissen E, et al. (2010) Impact of lipoteichoic acid modification on the performance of the probiotic Lactobacillus rhamnosus GG in experimental colitis. Clin Exp Immunol 162: 306-314.

Clapcote SJ, Roder JC (2005) Simplex PCR assay for sex determination in mice. Biotechniques 38: 702, 704, 706.

Coad J MD (2011) Anatomy and Physiology for Midwives: Churchill Livingstone.

Coulam CH, Edwin SS, LaMarche S, Mitchell MD (1993a) Actions of interleukin-2 on amnion prostaglandin biosynthesis. Prostaglandins Leukot Essent Fatty Acids 49: 959-961.

Coulam CH, Edwin SS, LaMarche S, Mitchell MD (1993b) Actions of interleukin-2 on chorio-decidual prostaglandin biosynthesis. Prostaglandins 46: 145-156.

Das C, Kumar VS, Gupta S, Kumar S (2002) Network of cytokines, integrins and hormones in human trophoblast cells. J Reprod Immunol 53: 257-268. de Moreno de LeBlanc A, LeBlanc JG (2014) Effect of probiotic administration on the intestinal microbiota, current knowledge and potential applications. World J Gastroenterol 20: 16518-16528.

156

Dekel N, Gnainsky Y, Granot I, Racicot K, Mor G (2014) The role of inflammation for a successful implantation. Am J Reprod Immunol 72: 141-147.

Deshpande G, Rao S, Patole S, Bulsara M (2010) Updated meta-analysis of probiotics for preventing necrotizing enterocolitis in preterm neonates. Pediatrics 125: 921-930.

Di Simone N, Caliandro D, Castellani R, Ferrazzani S, Caruso A (2000) Interleukin-3 and human trophoblast: in vitro explanations for the effect of interleukin in patients with antiphospholipid antibody syndrome. Fertil Steril 73: 1194-1200.

Donati L, Di Vico A, Nucci M, Quagliozzi L, Spagnuolo T, et al. (2010) Vaginal microbial biota and outcome of pregnancy. Arch Gynecol Obstet 281: 589-600.

Donders GG, Van Calsteren K, Bellen G, Reybrouck R, Van den Bosch T, et al. (2009) Predictive value for preterm birth of abnormal vaginal biota, bacterial vaginosis and aerobic vaginitis during the first trimester of pregnancy. BJOG 116: 1315-1324.

Douillard FP, Ribbera A, Kant R, Pietila TE, Jarvinen HM, et al. (2013) Comparative genomic and functional analysis of 100 Lactobacillus rhamnosus strains and their comparison with strain GG. PLoS Genet 9: e1003683.

Dudley DJ, Hunter C, Mitchell MD, Varner MW (1996) Elevations of amniotic fluid macrophage inflammatory protein-1 alpha concentrations in women during term and preterm labor. Obstet Gynecol 87: 94-98.

Dugoua JJ, Machado M, Zhu X, Chen X, Koren G, et al. (2009) Probiotic safety in pregnancy: a systematic review and meta-analysis of randomized controlled trials of Lactobacillus, Bifidobacterium, and Saccharomyces spp. J Obstet Gynaecol Can 31: 542- 552.

Ekelund CK, Vogel I, Skogstrand K, Thorsen P, Hougaard DM, et al. (2008) Interleukin-18 and interleukin-12 in maternal serum and spontaneous preterm delivery. J Reprod Immunol 77: 179-185.

El-Bastawissi AY, Williams MA, Riley DE, Hitti J, Krieger JN (2000) Amniotic fluid interleukin-6 and preterm delivery: a review. Obstet Gynecol 95: 1056-1064.

Elfayomy AK, Almasry SM (2014) Expression of tumor necrosis factor-alpha and vascular endothelial growth factor in different zones of fetal membranes: a possible relation to onset of labor. J Mol Histol 45: 243-257.

157

Elgert K (2009) Immunology: Understanding the immune system: Wiley blackwell.

Elliott CL, Loudon JA, Brown N, Slater DM, Bennett PR, et al. (2001) IL-1beta and IL-8 in human fetal membranes: changes with gestational age, labor, and culture conditions. Am J Reprod Immunol 46: 260-267.

Elliott CL, Slater DM, Dennes W, Poston L, Bennett PR (2000) Interleukin 8 expression in human myometrium: changes in relation to labor onset and with gestational age. Am J Reprod Immunol 43: 272-277.

Elovitz M, Wang Z (2004a) Medroxyprogesterone acetate, but not progesterone, protects against inflammation-induced parturition and intrauterine fetal demise. Am J Obstet Gynecol 190: 693-701.

Elovitz MA, Mrinalini C (2004b) Animal models of preterm birth. Trends Endocrinol Metab 15: 479-487.

Elovitz MA, Wang Z, Chien EK, Rychlik DF, Phillippe M (2003) A new model for inflammation-induced preterm birth: the role of platelet-activating factor and Toll-like receptor-4. Am J Pathol 163: 2103-2111.

Enomoto T, Sowa M, Nishimori K, Shimazu S, Yoshida A, et al. (2014) Effects of bifidobacterial supplementation to pregnant women and infants in the prevention of allergy development in infants and on fecal microbiota. Allergol Int 63: 575-585.

Esplin MS, Peltier MR, Hamblin S, Smith S, Fausett MB, et al. (2005) Monocyte chemotactic protein-1 expression is increased in human gestational tissues during term and preterm labor. Placenta 26: 661-671.

FAO/WHO (2001) Joint FAO/WHO expert consultation on evaluation of health and nutritional properties of probiotics in food.

Farage M, Maibach H (2006) Lifetime changes in the vulva and vagina. Arch Gynecol Obstet 273: 195-202.

Ferguson KK, McElrath TF, Chen YH, Mukherjee B, Meeker JD (2014) Longitudinal profiling of inflammatory cytokines and C-reactive protein during uncomplicated and preterm pregnancy. Am J Reprod Immunol 72: 326-336.

158

Fidel PI, Jr., Romero R, Maymon E, Hertelendy F (1998) Bacteria-induced or bacterial product-induced preterm parturition in mice and rabbits is preceded by a significant fall in serum progesterone concentrations. J Matern Fetal Med 7: 222-226.

Forsythe P, Wang B, Khambati I, Kunze WA (2012) Systemic effects of ingested Lactobacillus rhamnosus: inhibition of mast cell membrane potassium (IKCa) current and degranulation. PLoS One 7: e41234.

Fortunato SJ, Menon R, Lombardi SJ (1998) IL-15, a novel cytokine produced by human fetal membranes, is elevated in preterm labor. Am J Reprod Immunol 39: 16-23.

Fox MJ, Ahuja KD, Robertson IK, Ball MJ, Eri RD (2015) Can probiotic yogurt prevent diarrhoea in children on antibiotics? A double-blind, randomised, placebo-controlled study. BMJ Open 5: e006474.

Fredricks DN, Fiedler TL, Marrazzo JM (2005) Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 353: 1899-1911.

Freeman BE, Hammarlund E, Raue HP, Slifka MK (2012) Regulation of innate CD8+ T-cell activation mediated by cytokines. Proc Natl Acad Sci U S A 109: 9971-9976.

Furmento VA, Marino J, Blank VC, Roguin LP (2014) The granulocyte colony-stimulating factor (G-CSF) upregulates metalloproteinase-2 and VEGF through PI3K/Akt and Erk1/2 activation in human trophoblast Swan 71 cells. Placenta 35: 937-946.

Gan XT, Ettinger G, Huang CX, Burton JP, Haist JV, et al. (2014) Probiotic administration attenuates myocardial hypertrophy and heart failure after myocardial infarction in the rat. Circ Heart Fail 7: 491-499.

Gardiner GE, Heinemann C, Bruce AW, Beuerman D, Reid G (2002) Persistence of Lactobacillus fermentum RC-14 and Lactobacillus rhamnosus GR-1 but not L. rhamnosus GG in the human vagina as demonstrated by randomly amplified polymorphic DNA. Clin Diagn Lab Immunol 9: 92-96.

Geisert R, Fazleabas A, Lucy M, Mathew D (2012) Interaction of the conceptus and endometrium to establish pregnancy in mammals: role of interleukin 1beta. Cell Tissue Res 349: 825-838.

159

Giannoulias D, Patel FA, Holloway AC, Lye SJ, Tai HH, et al. (2002) Differential changes in 15-hydroxyprostaglandin dehydrogenase and prostaglandin H synthase (types I and II) in human pregnant myometrium. J Clin Endocrinol Metab 87: 1345-1352.

Gibb W, Challis JR (2002) Mechanisms of term and preterm birth. J Obstet Gynaecol Can 24: 874-883.

Gloor GB, Hummelen R, Macklaim JM, Dickson RJ, Fernandes AD, et al. (2010) Microbiome profiling by illumina sequencing of combinatorial sequence-tagged PCR products. PLoS One 5: e15406.

Goldenberg RL, Culhane JF, Iams JD, Romero R (2008a) Epidemiology and causes of preterm birth. Lancet 371: 75-84.

Goldenberg RL, Andrews WW, Goepfert AR, Faye-Petersen O, Cliver SP, et al. (2008b) The alabama preterm birth study: umbilical cord blood Ureaplasma urealyticum and Mycoplasma hominis cultures in very preterm newborn infants. Am J Obstet Gynecol 198: 43.e41-45.

Goldenberg RL, Hauth JC, Andrews WW (2000) Intrauterine infection and preterm delivery. N Engl J Med 342: 1500-1507.

Gotsch F, Romero R, Kusanovic JP, Mazaki-Tovi S, Pineles BL, et al. (2007) The fetal inflammatory response syndrome. Clin Obstet Gynecol 50: 652-683.

Grangette C, Nutten S, Palumbo E, Morath S, Hermann C, et al. (2005) Enhanced antiinflammatory capacity of a Lactobacillus plantarum mutant synthesizing modified teichoic acids. Proc Natl Acad Sci U S A 102: 10321-10326.

Gravett MG, Hitti J, Hess DL, Eschenbach DA (2000) Intrauterine infection and preterm delivery: evidence for activation of the fetal hypothalamic-pituitary-adrenal axis. Am J Obstet Gynecol 182: 1404-1413.

Grin PM, Kowalewska PM, Alhazzan W, Fox-Robichaud AE (2013) Lactobacillus for preventing recurrent urinary tract infections in women: meta-analysis. Can J Urol 20: 6607- 6614.

Gupta S, Kumar N, Singhal N, Kaur R, Manektala U (2006) Vaginal microbiota in postmenopausal women on hormone replacement therapy. Indian J Pathol Microbiol 49: 457-461.

160

Gutierrez-Orozco F, Thomas-Ahner JM, Galley JD, Bailey MT, Clinton SK, et al. (2015) Intestinal microbial dysbiosis and colonic epithelial cell hyperproliferation by dietary alpha- mangostin is independent of mouse strain. Nutrients 7: 764-784.

Hamilton S, Oomomian Y, Stephen G, Shynlova O, Tower CL, et al. (2012) Macrophages infiltrate the human and rat decidua during term and preterm labor: evidence that decidual inflammation precedes labor. Biol Reprod 86: 39.

Hamilton SA, Tower CL, Jones RL (2013) Identification of chemokines associated with the recruitment of decidual leukocytes in human labor: potential novel targets for preterm labor. PLoS One 8: e56946.

Han YW, Shen T, Chung P, Buhimschi IA, Buhimschi CS (2009) Uncultivated bacteria as etiologic agents of intra-amniotic inflammation leading to preterm birth. J Clin Microbiol 47: 38-47.

Hanna N, Bonifacio L, Reddy P, Hanna I, Weinberger B, et al. (2004) IFN-gamma-mediated inhibition of COX-2 expression in the placenta from term and preterm labor pregnancies. Am J Reprod Immunol 51: 311-318.

Hanna N, Bonifacio L, Weinberger B, Reddy P, Murphy S, et al. (2006) Evidence for interleukin-10-mediated inhibition of cyclo- oxygenase-2 expression and prostaglandin production in preterm human placenta. Am J Reprod Immunol 55: 19-27.

Hannan NJ, Jones RL, White CA, Salamonsen LA (2006) The chemokines, CX3CL1, CCL14, and CCL4, promote human trophoblast migration at the feto-maternal interface. Biol Reprod 74: 896-904.

Hart PH, Cooper RL, Finlay-Jones JJ (1991) IL-4 suppresses IL-1 beta, TNF-alpha and PGE2 production by human peritoneal macrophages. Immunology 72: 344-349.

Hayashi M, Sohma R, Sumioka Y, Inaba N (2006) Granulocyte-macrophage colony- stimulating factor levels in amniotic fluid before the onset of labor and during labor do not differ in normal pregnancies. Am J Reprod Immunol 55: 69-75.

Hayashi M, Zhu K, Sagesaka T, Fukasawa I, Inaba N (2008) Amniotic fluid levels of tumor necrosis factor-alpha and soluble tumor necrosis factor receptor 1 before and after the onset of labor in normal pregnancies. Horm Metab Res 40: 251-256.

161

Heng YJ, Liong S, Permezel M, Rice GE, Di Quinzio MK, et al. (2014a) The interplay of the interleukin 1 system in pregnancy and labor. Reprod Sci 21: 122-130.

Heng YJ, Pennell CE, Chua HN, Perkins JE, Lye SJ (2014b) Whole blood gene expression profile associated with spontaneous preterm birth in women with threatened preterm labor. PLoS One 9: e96901.

Hirsch E, Filipovich Y, Mahendroo M (2006) Signaling via the type I IL-1 and TNF receptors is necessary for bacterially induced preterm labor in a murine model. Am J Obstet Gynecol 194: 1334-1340.

Hirsch E, Muhle R (2002) Intrauterine bacterial inoculation induces labor in the mouse by mechanisms other than progesterone withdrawal. Biol Reprod 67: 1337-1341.

Hitti J, Hillier SL, Agnew KJ, Krohn MA, Reisner DP, et al. (2001) Vaginal indicators of amniotic fluid infection in preterm labor. Obstet Gynecol 97: 211-219.

Holst RM, Hagberg H, Wennerholm UB, Skogstrand K, Thorsen P, et al. (2011) Prediction of microbial invasion of the amniotic cavity in women with preterm labor: analysis of multiple proteins in amniotic and cervical fluids. BJOG 118: 240-249.

Homayouni A, Bastani P, Ziyadi S, Mohammad-Alizadeh-Charandabi S, Ghalibaf M, et al. (2014) Effects of probiotics on the recurrence of bacterial vaginosis: a review. J Low Genit Tract Dis 18: 79-86.

Houser BL (2012) Decidual macrophages and their roles at the maternal-fetal interface. Yale J Biol Med 85: 105-118.

Hummelen R, Fernandes AD, Macklaim JM, Dickson RJ, Changalucha J, et al. (2010) Deep sequencing of the vaginal microbiota of women with HIV. PLoS One 5: e12078.

Hummelen R, Macklaim JM, Bisanz JE, Hammond JA, McMillan A, et al. (2011) Vaginal microbiome and epithelial gene array in post-menopausal women with moderate to severe dryness. PLoS One 6: e26602.

Huse SM, Dethlefsen L, Huber JA, Mark Welch D, Relman DA, et al. (2008) Exploring microbial diversity and taxonomy using SSU rRNA hypervariable tag sequencing. PLoS Genet 4: e1000255.

162

Hyman RW, Fukushima M, Jiang H, Fung E, Rand L, et al. (2014) Diversity of the vaginal microbiome correlates with preterm birth. Reprod Sci 21: 32-40.

Imai M, Tani A, Saito M, Saito K, Amano K, et al. (2001) Significance of fetal fibronectin and cytokine measurement in the cervicovaginal secretions of women at term in predicting term labor and post-term pregnancy. Eur J Obstet Gynecol Reprod Biol 97: 53-58.

Ishikawa H, Matsumoto S, Ohashi Y, Imaoka A, Setoyama H, et al. (2011) Beneficial effects of probiotic bifidobacterium and galacto-oligosaccharide in patients with ulcerative colitis: a randomized controlled study. Digestion 84: 128-133.

Ito M, Nakashima A, Hidaka T, Okabe M, Bac ND, et al. (2010) A role for IL-17 in induction of an inflammation at the fetomaternal interface in preterm labor. J Reprod Immunol 84: 75-85.

Jacobsson B, Mattsby-Baltzer I, Hagberg H (2005) Interleukin-6 and interleukin-8 in cervical and amniotic fluid: relationship to microbial invasion of the chorioamniotic membranes. BJOG 112: 719-724.

Jauniaux E, Jurkovic D (2012) Placenta accreta: pathogenesis of a 20th century iatrogenic uterine disease. Placenta 33: 244-251.

Jayasooriya GS, Lamont RF (2009) The use of progesterone and other progestational agents to prevent spontaneous preterm labor and preterm birth. Expert Opin Pharmacother 10: 1007-1016.

Jiang ZY, Guo YY, Ren HB, Zou YF, Fan MS, et al. (2012) Tumor necrosis factor (TNF)- alpha upregulates progesterone receptor-A by activating the NF-kappaB signaling pathway in human decidua after labor onset. Placenta 33: 1-7.

Jun JK, Yoon BH, Romero R, Kim M, Moon JB, et al. (2000) Interleukin 6 determinations in cervical fluid have diagnostic and prognostic value in preterm premature rupture of membranes. Am J Obstet Gynecol 183: 868-873.

Karmakar S, Dhar R, Das C (2004) Inhibition of cytotrophoblastic (JEG-3) cell invasion by interleukin 12 involves an interferon gamma-mediated pathway. J Biol Chem 279: 55297- 55307.

163

Keelan JA, Sato TA, Mitchell MD (1998) Comparative studies on the effects of interleukin- 4 and interleukin-13 on cytokine and prostaglandin E2 production by amnion-derived WISH cells. Am J Reprod Immunol 40: 332-338.

Keelan JA, Sato TA, Hansen WR, Gilmour JS, Gupta DK, et al. (1999) Interleukin-4 differentially regulates prostaglandin production in amnion-derived WISH cells stimulated with pro-inflammatory cytokines and epidermal growth factor. Prostaglandins Leukot Essent Fatty Acids 60: 255-262.

Keelan JA, Blumenstein M, Helliwell RJ, Sato TA, Marvin KW, et al. (2003) Cytokines, prostaglandins and parturition--a review. Placenta 24 Suppl A: S33-46.

Kemgang TS, Kapila S, Shanmugam VP, Kapila R (2014) Cross-talk between probiotic lactobacilli and host immune system. J Appl Microbiol 117: 303-319.

Keyes LE, Moore LG, Walchak SJ, Dempsey EC (1996) Pregnancy-stimulated growth of vascular smooth muscle cells: importance of protein kinase C-dependent synergy between estrogen and platelet-derived growth factor. J Cell Physiol 166: 22-32.

Kim SO, Sheikh HI, Ha SD, Martins A, Reid G (2006) G-CSF-mediated inhibition of JNK is a key mechanism for Lactobacillus rhamnosus-induced suppression of TNF production in macrophages. Cell Microbiol 8: 1958-1971.

Klimkiewicz-Blok D, Florjanski J, Zalewski J, Blok R (2012) Analysis of the concentrations of interleukin 15 in amniotic fluid in the second and the third trimesters of pregnancy. Adv Clin Exp Med 21: 75-79.

Koga K, Mor G (2010) Toll-like receptors at the maternal-fetal interface in normal pregnancy and pregnancy disorders. Am J Reprod Immunol 63: 587-600.

Koren O, Goodrich JK, Cullender TC, Spor A, Laitinen K, et al. (2012) Host remodeling of the gut microbiome and metabolic changes during pregnancy. Cell 150: 470-480.

Koscica KL, Ananth CV, Placido J, Reznik SE (2007) The effect of a matrix metalloproteinase inhibitor on inflammation-mediated preterm delivery. Am J Obstet Gynecol 196: 551.e551-553.

Kostic AD, Howitt MR, Garrett WS (2013) Exploring host-microbiota interactions in animal models and humans. Genes Dev 27: 701-718.

164

Kuczynski J, Lauber CL, Walters WA, Parfrey LW, Clemente JC, et al. (2012) Experimental and analytical tools for studying the human microbiome. Nat Rev Genet 13: 47-58.

La Sala GB, Ardizzoni A, Capodanno F, Manca L, Baschieri MC, et al. (2012) Protein microarrays on midtrimester amniotic fluids: a novel approach for the diagnosis of early intrauterine inflammation related to preterm delivery. Int J Immunopathol Pharmacol 25: 1029-1040.

Lamont RF, Sobel JD, Akins RA, Hassan SS, Chaiworapongsa T, et al. (2011) The vaginal microbiome: new information about genital tract biota using molecular based techniques. BJOG 118: 533-549.

Ley RE, Turnbaugh PJ, Klein S, Gordon JI (2006) Microbial ecology: human gut microbes associated with obesity. Nature 444: 1022-1023.

Li W, Challis JR (2005) Corticotropin-releasing hormone and urocortin induce secretion of matrix metalloproteinase-9 (MMP-9) without change in tissue inhibitors of MMP-1 by cultured cells from human placenta and fetal membranes. J Clin Endocrinol Metab 90: 6569- 6574.

Li W, Yang S, Kim SO, Reid G, Challis JR, et al. (2014) Lipopolysaccharide-induced profiles of cytokine, chemokine, and growth factors produced by human decidual Cells Are Altered by Lactobacillus rhamnosus GR-1 Supernatant. Reprod Sci 21: 939-947.

Lin YP, Thibodeaux CH, Pena JA, Ferry GD, Versalovic J (2008) Probiotic Lactobacillus reuteri suppress proinflammatory cytokines via c-Jun. Inflamm Bowel Dis 14: 1068-1083.

Lindsay KL, Brennan L, McAuliffe FM (2014) Acceptability of and compliance with a probiotic capsule intervention in pregnancy. Int J Gynaecol Obstet 125: 279-280.

Ling Z, Kong J, Liu F, Zhu H, Chen X, et al. (2010) Molecular analysis of the diversity of vaginal microbiota associated with bacterial vaginosis. BMC Genomics 11: 488.

Lockwood CJ, Basar M, Kayisli UA, Guzeloglu-Kayisli O, Murk W, et al. (2014) Interferon-gamma protects first-trimester decidual cells against aberrant matrix metalloproteinase 1, 3, and 9 expression in preeclampsia. Am J Pathol 184: 2549-2559.

Lopez Bernal A (2007) Overview. Preterm labor: mechanisms and management. BMC Pregnancy Childbirth 7 Suppl 1: S2.

165

MacIntyre DA, Sykes L, Teoh TG, Bennett PR (2012) Prevention of preterm labor via the modulation of inflammatory pathways. J Matern Fetal Neonatal Med 25 Suppl 1: 17-20.

Madsen G, Zakar T, Ku CY, Sanborn BM, Smith R, et al. (2004) Prostaglandins differentially modulate progesterone receptor-A and -B expression in human myometrial cells: evidence for prostaglandin-induced functional progesterone withdrawal. J Clin Endocrinol Metab 89: 1010-1013.

Magurran AE (2003) Measuring Biological Diversity. MA, USA: Wiley-Blackwell.

Mak T (2006) The Immune Response: Basic and Clinical Principles. San Diego: Academic Press.

Marciniak B, Patro-Malysza J, Poniedzialek-Czajkowska E, Kimber-Trojnar Z, Leszczynska-Gorzelak B, et al. (2011) Glucocorticoids in pregnancy. Curr Pharm Biotechnol 12: 750-757.

Markert UR, Morales-Prieto DM, Fitzgerald JS (2011) Understanding the link between the IL-6 cytokine family and pregnancy: implications for future therapeutics. Expert Rev Clin Immunol 7: 603-609.

Martinez-Garcia EA, Chavez-Robles B, Sanchez-Hernandez PE, Nunez-Atahualpa L, Martin-Maquez BT, et al. (2011) IL-17 increased in the third trimester in healthy women with term labor. Am J Reprod Immunol 65: 99-103.

Martins AJ, Spanton S, Sheikh HI, Kim SO (2011) The anti-inflammatory role of granulocyte colony-stimulating factor in macrophage-dendritic cell crosstalk after Lactobacillus rhamnosus GR-1 exposure. J Leukoc Biol 89: 907-915.

Mastromarino P, Vitali B, Mosca L (2013) Bacterial vaginosis: a review on clinical trials with probiotics. New Microbiol 36: 229-238.

Matoba N, Yu Y, Mestan K, Pearson C, Ortiz K, et al. (2009) Differential patterns of 27 cord blood immune biomarkers across gestational age. Pediatrics 123: 1320-1328.

Maymon E, Romero R, Pacora P, Gomez R, Mazor M, et al. (2001) A role for the 72 kDa gelatinase (MMP-2) and its inhibitor (TIMP-2) in human parturition, premature rupture of membranes and intraamniotic infection. J Perinat Med 29: 308-316.

166

McDonald HM, Brocklehurst P, Gordon A (2007) Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev: CD000262.

Meisser A, Cameo P, Islami D, Campana A, Bischof P (1999) Effects of interleukin-6 (IL-6) on cytotrophoblastic cells. Mol Hum Reprod 5: 1055-1058.

Menard JP, Fenollar F, Henry M, Bretelle F, Raoult D (2008) Molecular quantification of Gardnerella vaginalis and Atopobium vaginae loads to predict bacterial vaginosis. Clin Infect Dis 47: 33-43.

Menon R, Jones J, Gunst PR, Kacerovsky M, Fortunato SJ, et al. (2014) Amniotic fluid metabolomic analysis in spontaneous preterm birth. Reprod Sci 21: 791-803.

Mesiano S, Chan EC, Fitter JT, Kwek K, Yeo G, et al. (2002) Progesterone withdrawal and estrogen activation in human parturition are coordinated by progesterone receptor A expression in the myometrium. J Clin Endocrinol Metab 87: 2924-2930.

Meysick KC, Garber GE (1992) Interactions between Trichomonas vaginalis and vaginal biota in a mouse model. J Parasitol 78: 157-160.

Micallef A, Grech N, Farrugia F, Schembri-Wismayer P, Calleja-Agius J (2014) The role of interferons in early pregnancy. Gynecol Endocrinol 30: 1-6.

Mileti E, Matteoli G, Iliev ID, Rescigno M (2009) Comparison of the immunomodulatory properties of three probiotic strains of lactobacilli using complex culture systems: prediction for in vivo efficacy. PLoS One 4: e7056.

Mirmonsef P, Hotton AL, Gilbert D, Burgad D, Landay A, et al. (2014) Free glycogen in vaginal fluids is associated with Lactobacillus colonization and low vaginal pH. PLoS One 9: e102467.

Mitchell C, Marrazzo J (2014) Bacterial vaginosis and the cervicovaginal immune response. Am J Reprod Immunol 71: 555-563.

Mohamadzadeh M, Pfeiler EA, Brown JB, Zadeh M, Gramarossa M, et al. (2011) Regulation of induced colonic inflammation by Lactobacillus acidophilus deficient in lipoteichoic acid. Proc Natl Acad Sci U S A 108 Suppl 1: 4623-4630.

Morelli L, Zonenenschain D, Del Piano M, Cognein P (2004) Utilization of the intestinal tract as a delivery system for urogenital probiotics. J Clin Gastroenterol 38: S107-110.

167

Mukhopadhya I, Hansen R, El-Omar EM, Hold GL (2012) IBD-what role do Proteobacteria play? Nat Rev Gastroenterol Hepatol 9: 219-230.

Murphy SP, Tayade C, Ashkar AA, Hatta K, Zhang J, et al. (2009) Interferon gamma in successful pregnancies. Biol Reprod 80: 848-859.

Myatt L, Sun K (2010) Role of fetal membranes in signaling of fetal maturation and parturition. Int J Dev Biol 54: 545-553.

Nakashima A, Ito M, Yoneda S, Shiozaki A, Hidaka T, et al. (2010) Circulating and decidual Th17 cell levels in healthy pregnancy. Am J Reprod Immunol 63: 104-109.

Naruse K, Innes BA, Bulmer JN, Robson SC, Searle RF, et al. (2010) Secretion of cytokines by villous cytotrophoblast and extravillous trophoblast in the first trimester of human pregnancy. J Reprod Immunol 86: 148-150.

Nelson DB, Hanlon A, Nachamkin I, Haggerty C, Mastrogiannis DS, et al. (2014) Early pregnancy changes in bacterial vaginosis-associated bacteria and preterm delivery. Paediatr Perinat Epidemiol 28: 88-96.

Nenadic DB, Pavlovic MD (2008) Cervical fluid cytokines in pregnant women: Relation to vaginal wet mount findings and polymorphonuclear leukocyte counts. Eur J Obstet Gynecol Reprod Biol 140: 165-170.

Nugent RP, Krohn MA, Hillier SL (1991) Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 29: 297- 301.

O'Brien JM, Lewis DF (2009) Progestins for the prevention of spontaneous preterm birth: review and implications of recent studies. J Reprod Med 54: 73-87.

O'Hanlon DE, Moench TR, Cone RA (2013) Vaginal pH and microbicidal lactic acid when lactobacilli dominate the microbiota. PLoS One 8: e80074.

Ogita T, Bergamo P, Maurano F, D'Arienzo R, Mazzarella G, et al. (2015) Modulatory activity of Lactobacillus rhamnosus OLL2838 in a mouse model of intestinal immunopathology. Immunobiology.

168

Olgun NS, Reznik SE (2010) The matrix metalloproteases and endothelin-1 in infection- associated preterm birth. Obstet Gynecol Int 2010.

Oliver RS, Lamont RF (2013) Infection and antibiotics in the aetiology, prediction and prevention of preterm birth. J Obstet Gynaecol 33: 768-775.

Olson DM, Ammann C (2007) Role of the prostaglandins in labor and prostaglandin receptor inhibitors in the prevention of preterm labor. Front Biosci 12: 1329-1343.

Oreshkova T, Dimitrov R, Mourdjeva M (2012) A cross-talk of decidual stromal cells, trophoblast, and immune cells: a prerequisite for the success of pregnancy. Am J Reprod Immunol 68: 366-373.

Oskoui M, Coutinho F, Dykeman J, Jette N, Pringsheim T (2013) An update on the prevalence of cerebral palsy: a systematic review and meta-analysis. Dev Med Child Neurol 55: 509-519.

Othman M, Neilson JP, Alfirevic Z (2007) Probiotics for preventing preterm labor. Cochrane Database Syst Rev: CD005941.

Pabona JM, Zhang D, Ginsburg DS, Simmen FA, Simmen RC (2014) Prolonged pregnancy in women is associated with attenuated myometrial expression of progesterone receptor co- regulator kruppel-like factor 9. J Clin Endocrinol Metab: jc20142846.

Papatheodorou DC, Karagiannidis LK, Paltoglou G, Margeli A, Kaparos G, et al. (2013) Pulsatile interleukin-6 leads CRH secretion and is associated with myometrial contractility during the active phase of term human labor. J Clin Endocrinol Metab 98: 4105-4112.

Papatsonis DN, Flenady V, Liley HG (2013) Maintenance therapy with oxytocin antagonists for inhibiting preterm birth after threatened preterm labor. Cochrane Database Syst Rev 10: CD005938.

Parizek A, Koucky M, Duskova M (2014) Progesterone, inflammation and preterm labor. J Steroid Biochem Mol Biol 139: 159-165.

Park KH, Chaiworapongsa T, Kim YM, Espinoza J, Yoshimatsu J, et al. (2003) Matrix metalloproteinase 3 in parturition, premature rupture of the membranes, and microbial invasion of the amniotic cavity. J Perinat Med 31: 12-22.

169

Pavcnik-Arnol M, Lucovnik M, Kornhauser-Cerar L, Premru-Srsen T, Hojker S, et al. (2014) Lipopolysaccharide-binding protein as marker of fetal inflammatory response syndrome after preterm premature rupture of membranes. Neonatology 105: 121-127.

Perez Leiros C, Ramhorst R (2013) Tolerance induction at the early maternal-placental interface through selective cell recruitment and targeting by immune polypeptides. Am J Reprod Immunol 69: 359-368.

Petrou S (2005) The economic consequences of preterm birth during the first 10 years of life. BJOG 112 Suppl 1: 10-15.

Petrova MI, van den Broek M, Balzarini J, Vanderleyden J, Lebeer S (2013) Vaginal microbiota and its role in HIV transmission and infection. FEMS Microbiol Rev 37: 762- 792.

Pongcharoen S, Niumsup P, Sanguansermsri D, Supalap K, Butkhamchot P (2006) The effect of interleukin-17 on the proliferation and invasion of JEG-3 human choriocarcinoma cells. Am J Reprod Immunol 55: 291-300.

Pongcharoen S, Somran J, Sritippayawan S, Niumsup P, Chanchan P, et al. (2007) Interleukin-17 expression in the human placenta. Placenta 28: 59-63.

Puchner K, Iavazzo C, Gourgiotis D, Boutsikou M, Baka S, et al. (2011) Mid-trimester amniotic fluid interleukins (IL-1beta, IL-10 and IL-18) as possible predictors of preterm delivery. In Vivo 25: 141-148.

Ravel J, Gajer P, Abdo Z, Schneider GM, Koenig SS, et al. (2011) Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A 108 Suppl 1: 4680-4687.

Reagan-Shaw S, Nihal M, Ahmad N (2008) Dose translation from animal to human studies revisited. FASEB J 22: 659-661.

Reid G, Brigidi P, Burton JP, Contractor N, Duncan S, et al. (2015) Microbes central to human reproduction. Am J Reprod Immunol 73: 1-11.

Reid G (2012) Probiotic and prebiotic applications for vaginal health. J AOAC Int 95: 31- 34.

Reid G, Charbonneau D, Erb J, Kochanowski B, Beuerman D, et al. (2003a) Oral use of Lactobacillus rhamnosus GR-1 and L. fermentum RC-14 significantly alters vaginal biota:

170 randomized, placebo-controlled trial in 64 healthy women. FEMS Immunol Med Microbiol 35: 131-134.

Reid G, Bocking A (2003b) The potential for probiotics to prevent bacterial vaginosis and preterm labor. Am J Obstet Gynecol 189: 1202-1208.

Reid G (2001a) Probiotic agents to protect the urogenital tract against infection. Am J Clin Nutr 73: 437S-443S.

Reid G, Bruce AW (2001b) Selection of lactobacillus strains for urogenital probiotic applications. J Infect Dis 183 Supp 1: S77-80.

Robertson SA, Christiaens I, Dorian CL, Zaragoza DB, Care AS, et al. (2010) Interleukin-6 is an essential determinant of on-time parturition in the mouse. Endocrinology 151: 3996- 4006.

Robertson SA, Care AS, Skinner RJ (2007a) Interleukin 10 regulates inflammatory cytokine synthesis to protect against lipopolysaccharide-induced abortion and fetal growth restriction in mice. Biol Reprod 76: 738-748.

Robertson SA (2007b) GM-CSF regulation of embryo development and pregnancy. Cytokine Growth Factor Rev 18: 287-298.

Robertson SA, Skinner RJ, Care AS (2006) Essential role for IL-10 in resistance to lipopolysaccharide-induced preterm labor in mice. J Immunol 177: 4888-4896.

Robertson SA, O’Connell, A., and A. Ramsey (2000) The effect of interleukin-6 deficiency on implantation, fetal development and parturition in mice. . Proc Aust Soc ReprodBiol 31: 97.

Romero R, Hassan SS, Gajer P, Tarca AL, Fadrosh DW, et al. (2014a) The composition and stability of the vaginal microbiota of normal pregnant women is different from that of non- pregnant women. Microbiome 2: 4.

Romero R, Hassan SS, Gajer P, Tarca AL, Fadrosh DW, et al. (2014b) The vaginal microbiota of pregnant women who subsequently have spontaneous preterm labor and delivery and those with a normal delivery at term. Microbiome 2: 18.

171

Romero R, Stanczyk FZ (2013) Progesterone is not the same as 17alpha- hydroxyprogesterone caproate: implications for obstetrical practice. Am J Obstet Gynecol 208: 421-426.

Ruiz FO, Gerbaldo G, Garcia MJ, Giordano W, Pascual L, et al. (2012) Synergistic effect between two bacteriocin-like inhibitory substances produced by lactobacilli Strains with inhibitory activity for Streptococcus agalactiae. Curr Microbiol 64: 349-356.

Schwebke JR, Muzny CA, Josey WE (2014) Role of Gardnerella vaginalis in the pathogenesis of bacterial vaginosis: a conceptual model. J Infect Dis 210: 338-343.

Schwenke M, Knofler M, Velicky P, Weimar CH, Kruse M, et al. (2013) Control of human endometrial stromal cell motility by PDGF-BB, HB-EGF and trophoblast-secreted factors. PLoS One 8: e54336.

Sennstrom MB, Ekman G, Westergren-Thorsson G, Malmstrom A, Bystrom B, et al. (2000) Human cervical ripening, an inflammatory process mediated by cytokines. Mol Hum Reprod 6: 375-381.

Shennan A, Crawshaw S, Briley A, Hawken J, Seed P, et al. (2006) A randomised controlled trial of metronidazole for the prevention of preterm birth in women positive for cervicovaginal fetal fibronectin: the PREMET Study. BJOG 113: 65-74.

Shynlova O, Dorogin A, Li Y, Lye S (2014) Inhibition of infection-mediated preterm birth by administration of broad spectrum chemokine inhibitor in mice. J Cell Mol Med 18: 1816- 1829.

Shynlova O, Nedd-Roderique T, Li Y, Dorogin A, Lye SJ (2013) Myometrial immune cells contribute to term parturition, preterm labor and post-partum involution in mice. J Cell Mol Med 17: 90-102.

Shynlova O, Tsui P, Jaffer S, Lye SJ (2009) Integration of endocrine and mechanical signals in the regulation of myometrial functions during pregnancy and labor. Eur J Obstet Gynecol Reprod Biol 144 Suppl 1: S2-10.

Slater D, Dennes W, Sawdy R, Allport V, Bennett P (1999a) Expression of cyclo-oxygenase types-1 and -2 in human fetal membranes throughout pregnancy. J Mol Endocrinol 22: 125- 130.

172

Slater DM, Dennes WJ, Campa JS, Poston L, Bennett PR (1999b) Expression of cyclo- oxygenase types-1 and -2 in human myometrium throughout pregnancy. Mol Hum Reprod 5: 880-884.

Smayevsky J, Canigia LF, Lanza A, Bianchini H (2001) Vaginal microbiota associated with bacterial vaginosis in nonpregnant women: reliability of sialidase detection. Infect Dis Obstet Gynecol 9: 17-22.

Smith R (2007) Parturition. N Engl J Med 356: 271-283.

Spear GT, French AL, Gilbert D, Zariffard MR, Mirmonsef P, et al. (2014) Human alpha- amylase present in lower-genital-tract mucosal fluid processes glycogen to support vaginal colonization by Lactobacillus. J Infect Dis. 210: 1019-28.

Srinivasan S, Hoffman NG, Morgan MT, Matsen FA, Fiedler TL, et al. (2012) Bacterial communities in women with bacterial vaginosis: high resolution phylogenetic analyses reveal relationships of microbiota to clinical criteria. PLoS One 7: e37818.

Strakova Z, Srisuparp S, Fazleabas AT (2000) Interleukin-1beta induces the expression of insulin-like growth factor binding protein-1 during decidualization in the primate. Endocrinology 141: 4664-4670.

Subramaniam A, Abramovici A, Andrews WW, Tita AT (2012) Antimicrobials for preterm birth prevention: an overview. Infect Dis Obstet Gynecol 2012: 157159.

Tan H, Yi L, Rote NS, Hurd WW, Mesiano S (2012) Progesterone receptor-A and -B have opposite effects on proinflammatory gene expression in human myometrial cells: implications for progesterone actions in human pregnancy and parturition. J Clin Endocrinol Metab 97: E719-730.

Tanaka A, Jung K, Benyacoub J, Prioult G, Okamoto N, et al. (2009) Oral supplementation with Lactobacillus rhamnosus CGMCC 1.3724 prevents development of atopic dermatitis in NC/NgaTnd mice possibly by modulating local production of IFN-gamma. Exp Dermatol 18: 1022-1027.

Timmerman HM, Niers LE, Ridwan BU, Koning CJ, Mulder L, et al. (2007) Design of a multispecies probiotic mixture to prevent infectious complications in critically ill patients. Clin Nutr 26: 450-459.

173

Timmons BC, Reese J, Socrate S, Ehinger N, Paria BC, et al. (2014) Prostaglandins are essential for cervical ripening in LPS-mediated preterm birth but not term or antiprogestin- driven preterm ripening. Endocrinology 155: 287-298.

Top KA, Buet A, Whittier S, Ratner AJ, Saiman L (2012) Predictors of rectovaginal colonization in pregnant women and risk for maternal and neonatal infections. J Pediatric Infect Dis Soc 1: 7-15.

Toth B, Haufe T, Scholz C, Kuhn C, Friese K, et al. (2010) Placental interleukin-15 expression in recurrent miscarriage. Am J Reprod Immunol 64: 402-410.

Ugwumadu AH (2002) Bacterial vaginosis in pregnancy. Curr Opin Obstet Gynecol 14: 115-118.

Unal ER, Cierny JT, Roedner C, Newman R, Goetzl L (2011) Maternal inflammation in spontaneous term labor. Am J Obstet Gynecol 204: 223 e221-225.

Vacca P, Mingari MC, Moretta L (2013) Natural killer cells in human pregnancy. J Reprod Immunol 97: 14-19.

Vallor AC, Antonio MA, Hawes SE, Hillier SL (2001) Factors associated with acquisition of, or persistent colonization by, vaginal lactobacilli: role of hydrogen peroxide production. J Infect Dis 184: 1431-1436.

Vanderhoof JA (2008) Probiotics in allergy management. J Pediatr Gastroenterol Nutr 47 Suppl 2: S38-40.

VandeVusse L, Hanson L, Safdar N (2013) Perinatal outcomes of prenatal probiotic and prebiotic administration: an integrative review. J Perinat Neonatal Nurs 27: 288-301; quiz E281-282.

Velez DR, Fortunato SJ, Morgan N, Edwards TL, Lombardi SJ, et al. (2008) Patterns of cytokine profiles differ with pregnancy outcome and ethnicity. Hum Reprod 23: 1902-1909.

Verstraelen H, Verhelst R, Claeys G, De Backer E, Temmerman M, et al. (2009) Longitudinal analysis of the vaginal microbiota in pregnancy suggests that L. crispatus promotes the stability of the normal vaginal microbiota and that L. gasseri and/or L. iners are more conducive to the occurrence of abnormal vaginal microbiota. BMC Microbiol 9: 116.

174

Verstraelen H, Verhelst R, Claeys G, Temmerman M, Vaneechoutte M (2004) Culture- independent analysis of vaginal microbiota: the unrecognized association of Atopobium vaginae with bacterial vaginosis. Am J Obstet Gynecol 191: 1130-1132.

Villena J, Chiba E, Tomosada Y, Salva S, Marranzino G, et al. (2012) Orally administered Lactobacillus rhamnosus modulates the respiratory immune response triggered by the viral pathogen-associated molecular pattern poly(I:C). BMC Immunol 13: 53.

Vogel I, Goepfert AR, Thorsen P, Skogstrand K, Hougaard DM, et al. (2007) Early second- trimester inflammatory markers and short cervical length and the risk of recurrent preterm birth. J Reprod Immunol 75: 133-140.

Voltolini C, Petraglia F (2014) Neuroendocrinology of pregnancy and parturition. Handb Clin Neurol 124: 17-36. von Minckwitz G, Grischke EM, Schwab S, Hettinger S, Loibl S, et al. (2000) Predictive value of serum interleukin-6 and -8 levels in preterm labor or rupture of the membranes. Acta Obstet Gynecol Scand 79: 667-672. von Wolff M, Thaler CJ, Strowitzki T, Broome J, Stolz W, et al. (2000) Regulated expression of cytokines in human endometrium throughout the menstrual cycle: dysregulation in habitual abortion. Mol Hum Reprod 6: 627-634.

Vos M, Quince C, Pijl AS, de Hollander M, Kowalchuk GA (2012) A comparison of rpoB and 16S rRNA as markers in pyrosequencing studies of bacterial diversity. PLoS One 7: e30600.

Wallace AE, Fraser R, Cartwright JE (2012) Extravillous trophoblast and decidual natural killer cells: a remodelling partnership. Hum Reprod Update 18: 458-471.

Walsh CJ, Guinane CM, O'Toole PW, Cotter PD (2014) Beneficial modulation of the gut microbiota. FEBS Lett.

Walther-Antonio MR, Jeraldo P, Berg Miller ME, Yeoman CJ, Nelson KE, et al. (2014) Pregnancy's stronghold on the vaginal microbiome. PLoS One 9: e98514.

Wan LY, Chen ZJ, Shah NP, El-Nezami H (2015) Modulation of intestinal epithelial defense responses by probiotic bacteria. Crit Rev Food Sci Nutr: 0.

175

Wei SQ, Fraser W, Luo ZC (2010) Inflammatory cytokines and spontaneous preterm birth in asymptomatic women: a systematic review. Obstet Gynecol 116: 393-401.

Weissenbacher T, Laubender RP, Witkin SS, Gingelmaier A, Schiessl B, et al. (2013) Diagnostic biomarkers of pro-inflammatory immune-mediated preterm birth. Arch Gynecol Obstet 287: 673-685.

Wen A, Srinivasan U, Goldberg D, Owen J, Marrs CF, et al. (2014) Selected vaginal bacteria and risk of preterm birth: an ecological perspective. J Infect Dis 209: 1087-1094.

Whitcomb BW, Schisterman EF, Luo X, Chegini N (2009) Maternal serum granulocyte colony-stimulating factor levels and spontaneous preterm birth. J Womens Health (Larchmt) 18: 73-78.

Wilczynski JR (2005) Th1/Th2 cytokines balance--yin and yang of reproductive immunology. Eur J Obstet Gynecol Reprod Biol 122: 136-143.

Wu YW (2002) Systematic review of chorioamnionitis and cerebral palsy. Ment Retard Dev Disabil Res Rev 8: 25-29.

Wynne S (2013) C. difficile in pregnancy: an emerging problem. Midwives 16: 54.

Yan F and Polk B (2012) Characterization of a probiotic-derived soluble protein which reveals a mechanism of preventive and treatment effects of probiotics on intestinal inflammatory diseases. Gut Microbes. 3(1):25-8.

Yang Q, El-Sayed Y, Rosenberg-Hasson Y, Hirschberg DL, Nayak NR, et al. (2009) Multiple cytokine profile in plasma and amniotic fluid in a mouse model of pre-term labor. Am J Reprod Immunol 62: 339-347.

Yang Y, Guo Y, Kan Q, Zhou XG, Zhou XY, et al. (2014a) A meta-analysis of probiotics for preventing necrotizing enterocolitis in preterm neonates. Braz J Med Biol Res 47: 804- 810.

Yang S, Li W, Challis JR, Reid G, Kim SO, et al. (2014b) Probiotic Lactobacillus rhamnosus GR-1 supernatant prevents lipopolysaccharide-induced preterm birth and reduces inflammation in pregnant CD-1 mice. Am J Obstet Gynecol 211: 44 e41-44 e12.

Yang S, Reid G, Challis JR, Kim SO, Gloor GB, Bocking AD (2015) Is there a role for probiotics in the prevention of preterm birth? Front Immunol 6:62.

176

Yao Y, Li W, Kaplan MH, Chang CH (2005) Interleukin (IL)-4 inhibits IL-10 to promote IL-12 production by dendritic cells. J Exp Med 201: 1899-1903.

Yeganegi M, Leung CG, Martins A, Kim SO, Reid G, et al. (2011) Lactobacillus rhamnosus GR-1 stimulates colony-stimulating factor 3 (granulocyte) (CSF3) output in placental trophoblast cells in a fetal sex-dependent manner. Biol Reprod 84: 18-25.

Yeganegi M, Leung CG, Martins A, Kim SO, Reid G, et al. (2010) Lactobacillus rhamnosus GR-1-induced IL-10 production in human placental trophoblast cells involves activation of JAK/STAT and MAPK pathways. Reprod Sci 17: 1043-1051.

Yeganegi M, Watson CS, Martins A, Kim SO, Reid G, et al. (2009) Effect of Lactobacillus rhamnosus GR-1 supernatant and fetal sex on lipopolysaccharide-induced cytokine and prostaglandin-regulating enzymes in human placental trophoblast cells: implications for treatment of bacterial vaginosis and prevention of preterm labor. Am J Obstet Gynecol 200: 532 e531-538.

Yoon BH, Oh SY, Romero R, Shim SS, Han SY, et al. (2001) An elevated amniotic fluid matrix metalloproteinase-8 level at the time of mid-trimester genetic amniocentesis is a risk factor for spontaneous preterm delivery. Am J Obstet Gynecol 185: 1162-1167.

Yoon BH, Romero R, Park JS, Kim CJ, Kim SH, et al. (2000) Fetal exposure to an intra- amniotic inflammation and the development of cerebral palsy at the age of three years. Am J Obstet Gynecol 182: 675-681.

Yoshimura K, Hirsch E (2005) Effect of stimulation and antagonism of interleukin-1 signaling on preterm delivery in mice. J Soc Gynecol Investig 12: 533-538.

Young A, Thomson AJ, Ledingham M, Jordan F, Greer IA, et al. (2002) Immunolocalization of proinflammatory cytokines in myometrium, cervix, and fetal membranes during human parturition at term. Biol Reprod 66: 445-449.

Youssef N, Sheik CS, Krumholz LR, Najar FZ, Roe BA, et al. (2009) Comparison of species richness estimates obtained using nearly complete fragments and simulated pyrosequencing- generated fragments in 16S rRNA gene-based environmental surveys. Appl Environ Microbiol 75: 5227-5236.

Zhang D, Huang Y, Ye D (2015) Intestinal dysbiosis: An emerging cause of pregnancy complications? Med Hypotheses 84: 223-226.

Appendices

177 178

List of Appendices

Appendix I

Cytokine and Chemokine Assay Protocol

1. Dilute 10x beads with assay buffer to 1x concentration. 2. Add 50 µL of 1x beads to each well of a 96 well plate. 3. Wash the plate twice with 100 µL of wash buffer. 4. Dilute samples with sample diluent in a 1:4 ratio for plasma samples and in a 1:1 ratio for all other samples. 5. Add 50 µL of standards, blank (diluent) or samples to each well in duplicate. 6. Cover the plate with a plastic film and then with aluminum foil and incubate on shaker (850 rpm) for 30 minutes at room temperature. 7. Wash the plate thrice with 100 µL of wash buffer. 8. Dilute 10x detection antibody with antibody diluent to 1x concentration. 9. Add 25 µL of 1x detection antibody to each well. 10. Cover the plate with a plastic film and then with aluminum foil and incubate on shaker (850 rpm) for 30 minutes at room temperature. 11. Wash the plate thrice with 100 µL of wash buffer. 12. Dilute 100x streptavidin-PE with assay buffer to 1x concentration. 13. Add 50 µL of 1x streptavidin-PE to each well. 14. Cover the plate with a plastic film and then with aluminum foil and incubate on shaker (850 rpm) for 10 minutes at room temperature. 15. Wash the plate thrice with 100 µL of wash buffer. 16. Re-suspend the beads in each well with 125 µL assay buffer. 17. Cover the plate with a plastic film and then with aluminum foil and incubate on shaker (850 rpm) for 30 seconds at room temperature. 18. Proceed to read the plate on Bioplex machine.

179

Appendix II

Progesterone EIA Assay Protocol

1. Reconstitute the progesterone AChE Tracer with 6 mL of EIA buffer. 2. Reconstitute the progesterone EIA antiserum with 6 mL EIA buffer. 3. Set-up the plate to include 2 wells for Blank, 2 wells for non specific binding (NSB),

3 wells for Maximum binding (B0), 1 well for Total activity (TA). 4. Add 100 µL of EIA buffer to NSB wells.

5. Add 50 µL of EIA buffer to B0 wells. 6. Add 50 µL of standards and samples to each well in duplicate. 7. Each sample is assayed at three dilutions and each dilution is assayed in duplicate. 8. Add 50 µL of diluted progesterone AChE to each well except the TA and the blank wells. 9. Add 50 µL of diluted progesterone EIA antiserum to each well except the TA, the NSB and the blank wells. 10. Cover the plate with a plastic film and incubate for 1 hour at room temperature on shaker at 300 rpm. 11. Empty the wells and rinse 5 times with 200 µL of wash buffer. 12. Reconstitute Ellman’s reagent with 20 mL of UltraPure water. 13. Add 200 µL of Ellman’s reagent to each well. 14. Add 5 µL of tracer to the TA wells. 15. Cover the plate with a plastic film and then with aluminum foil. The plate is left in a dark room to develop on a shaker (300 rpm) for 60 to 90 minutes. 16. Read the plate at a wavelength between 405 nm to 420 nm.

180

Appendix III

PowerSoil®DNA Isolation Kit Protocol

1. Add vaginal or cecal tissues into PowerBead Tubes and vortex to mix. 2. Add 60 µl of Solution C1 cell lysis buffer. Secure the tubes horizontally on a vortex pad with tape and vortex at maximum speed for 10 minutes. 3. Centrifuge the tubes at 10,000x g for 30 sec at 25oC and transfer 500 µl of the supernatant to a clean 2 ml tube. 4. Add 250 µl of Solution C2 inhibitor removal buffer, vortex for 5 sec and incubate at 4°C for 5 min. 5. Centrifuge the tubes at 25°C for 1 min at 10,000 x g and transfer 500 µl of the supernatant to a clean 2 ml tube. 6. Add 200 µl of Solution C3 inhibitor removal buffer, vortex for 5 sec and incubate at 4°C for 5 min. Repeat Step 5. 7. Add 1200 µl of Solution C4 containing high concentration of salt to bind DNA, and vortex the tube for 5 seconds. 8. Load 600 µl onto a Spin Filter, centrifuge at 25°C for 1 min at 10,000 x g, and discard the flow through. Repeat Step 9 twice. 9. Add 500 µl of Solution C5 ethanol containing buffer, centrifuge at 25°C for 30 sec at 10,000 x g and discard the flow through. 10. Centrifuge again at 25°C for 1 min at 10,000 x g. 11. Place the spin filter in a clean 2 ml tube and add 100 µl of Solution C6 sterile elution buffer and centrifuge at at 25°C for 30 sec at 10,000 x g. 12. Store the DNA in the tube at -80oC until further analysis.

181

Appendix IV

182

Appendix V

183

Appendix VI