Salmonella Enterica Serovar Typhimurium Infection in Mouse Pregnancy

SALMONELLA ENTERICA SEROVAR TYPHIMURIUM INFECTION IN MOUSE PREGNANCY

Shuhiba Mohammad

A thesis submitted to the Department of Biomedical and Molecular Sciences

In conformity with the requirements for

the degree of Master of Science

Queen’s University

Kingston, Ontario, Canada

December, 2016

Salmonella are Gram-negative, intracellular food-borne pathogens that cause pregnancy complications. In pregnant mice, Salmonella enterica serovar Typhimurium (S.Tm) infection results in placental bacterial replication, inflammation, necrosis, and fetal loss by unknown mechanisms.

Necroptosis, or programmed necrosis mediated by RIPK3 (receptor-interacting protein kinase 3), an inflammatory cell death pathway, is implicated in the pathogenesis of S.Tm in non-pregnant mice. This goal of this thesis was to investigate the role of necroptosis in the pathogenesis of S.Tm infection during mouse pregnancy. I hypothesized that elimination of the key necroptotic cell death protein RIPK3 would decrease placental inflammation and trophoblast cell death, and increase conceptus survival compared to controls.

Mice expressing a functional Slc11a1 (encodes the natural resistance-associated macrophage protein 1, NRAMP1) gene with or without RIPK3 function (Ripk3-/-Slc11a1+/+ compared to Slc11a1+/+) were infected with 103 S.Tm by tail vein injection on gestational day (GD) 12. Mice were euthanized on

GD 14 (48h post-infection) or GD 15 (72h post-infection) and implantation sites (IS) and maternal serum were harvested for analyses.

In nearly all challenged mice (except one outlier), S.Tm were detected in most IS within a litter but there was limited immune cell infiltration, placental damage or cell death in Slc11a1 competent mice regardless of Ripk3 gene deletion. Maternal serum cytokine analyses confirmed lack of maternal immune responses to S.Tm infection. IS amongst the litter of a single dam (Ripk3-/-Slc11a1+/+ at 72h post- infection) displayed heavy but not universal placental S.Tm infection of decidual tissues and spongiotrophoblast, associated with elevated maternal serum pro-inflammatory cytokines. S.Tm infection of the fetal yolk sac (YS) was observed in 54.5% of IS from this dam. YS infection was confirmed in archival samples in mice expressing Ripk3 with intact Slc11a1 and in mice lacking functional Slc11a1. In

Slc11a1 incompetent mice, S.Tm were detected in placental labyrinthine trophoblast. Based on the available data, this thesis suggests that Ripk3 and necroptosis have no significant roles in either promotion ii or prevention of progressive Salmonella infection during mouse pregnancy. It also provides pilot data that

NRAMP1 controls placental localization and lethality due to YS infection.

iii

Acknowledgements

This thesis has come to fruition through the unrelenting support of so many. Firstly, from my advisor, Dr. Anne Croy, for which her mentorship is and always will be, unmatched. Her undying support and counsel have had a massive role in my success. The effect of her guidance and support will impact my work for a lifetime. I will continue to exemplify her attitudes towards the pursuit of science by responding to emails within minutes of receipt (no matter the time), attending as many scientific meetings as humanely and financially possible, and to educate and discuss with everyone the wonders of this world. I also hope to emulate her impeccable style and fashion-sense, by which I am constantly inspired. Dr. Croy will always be my role model and she is the best advisor I could have ever hoped for. I would also like to acknowledge Dr. Neil MacLusky, former Chair of Biomedical Sciences at the University of Guelph for believing in me, especially when it was difficult to believe in myself. He supported and guided my passion for science by connecting me with Queen's University and Dr. Croy, and for that I am eternally grateful. Thank you to my collaborators at the University of Rochester, Dr. Shawn Murphy and Ian Perry. I will miss our lengthy discussions and the beautiful images of the microscopic monster we've been chasing for forever in "Where's Waldo"-like fashion. I would also like to thank Dr. Lakshmi Krishnan and Kristina Wachholz at the University of Ottawa for sample preparation. Thank you, Dr. Nancy Martin, for all of your expertise and help as part of my advisory committee. Thank you to all the mice sacrificed for this endeavor. I would like to thank my Croy lab mates Ashley Martin, Allison Felker, Mackenzie Redhead, Vanessa Kay, Nathália Portilho, Rayana Luna, Ernesto Figueiro-Filho, Matt Rätsep and countless others for their guidance and friendship. Thank you to all my colleagues in the Department of Biomedical and Molecular Sciences at Queen's University for which there are far too many to name here. A special thank you to my undergraduate volunteers, Zain Saleem and Olivia Jeon for their many hours spent running intricate histology protocols to my exact specifications. Many of the images included in this thesis were a direct result of their work. Their tireless efforts will not be forgotten; my wrist and sanity thanks them. Thank you to my enormous network of roommates, friends and my incredible family for their constant support throughout this thesis. Madar and Padar, thank you for supporting me through this journey. I know I will always be your curious baby that stuck keys in sockets, and would never believe in monsters lurking behind curtains. Thank you for supporting me in the pursuit of my dreams.

"Somewhere, something incredible is waiting to be known" - Carl Sagan

Table of Contents

Abstract ...... ii Acknowledgements ...... iv List of Figures ...... viii List of Tables ...... x List of Abbreviations ...... xi Chapter 1 Introduction and Literature Review ...... 1 1.1 Salmonella ...... 1 1.1.1 Natural course of Salmonella infection ...... 1 1.1.2 Salmonella infection in human pregnancy ...... 3 1.2 Pregnancy-enhanced pathogenesis of Salmonella ...... 3 1.2.1 Immunological shifts during pregnancy ...... 3 1.2.2 Placental Salmonella tropism ...... 5 1.3 Mouse Pregnancy ...... 9 1.3.1 Mouse placental anatomy ...... 9 1.3.2 Immune cells at the mouse maternal-fetal interface ...... 13 1.4 Salmonella enterica serovar Typhimurium infection during murine pregnancy ...... 17 1.5 Salmonella-induced cell death mechanisms ...... 22 1.5.1 Necroptosis ...... 23 1.6 Rationale, hypothesis and main objectives ...... 26 Chapter 2 Materials and Methods ...... 27 2.1 Mice ...... 27 2.2 Genotyping ...... 28 2.3 Mating strategy and pregnancy timing ...... 29 2.4 Bacterial preparation and in vivo infection ...... 29 2.5 Sample collection ...... 29 2.6 Tissue processing for histology ...... 30 2.7 Salmonella enterica serovar Typhimurium immunohistochemistry ...... 32 2.7.1 S.Tm Immunofluorescence ...... 34 2.8 Histopathology & morphometric measurements ...... 34 2.8.1 Routine histopathology with hematoxylin and eosin (H&E) staining ...... 34 2.8.2 Active-caspase-3 immunohistochemistry for detection of cell death ...... 35 2.8.3 Morphometric measurements of implantation sites ...... 35

2.9 Immunohistochemistry for leukocytes ...... 36 2.9.1 CD45 ...... 36 2.9.2 Neutrophil Ly6G ...... 36 2.9.3 Macrophage F4/80 ...... 37 2.10 Serum Analyses ...... 38 2.10.1 Lipopolysaccharide endotoxin quantification ...... 38 2.10.2 Progesterone ...... 38 2.10.3 Maternal cytokines, chemokines, and growth factors ...... 38 2.11 Statistical Analysis ...... 39 Chapter 3 Results ...... 40 3.1 Fetal Resorption ...... 40 3.2 Study Animals ...... 41 3.3 Fetal impacts of Salmonella enterica serovar Typhimurium infection during pregnancy ...... 44 3.3.1 Fetal:placental weight ratios ...... 44 3.3.2 Placental morphometric measurements ...... 47 3.4 Placental Salmonella enterica serovar Typhimurium Localization by Immunohistochemistry ...... 50 3.4.1 Placental localization of Salmonella enterica serovar Typhimurium foci and bacterial burden estimates ...... 50 3.4.2 Frequency of S.Tm+ placentas and distribution ...... 55 3.5 Histopathology ...... 61 3.5.1 Histopathology using Hematoxylin and Eosin staining ...... 61 3.5.2 Assessment of cell death by active caspase-3 immunohistochemistry ...... 70 3.6 Assessment of immune cell infiltration of implantation sites after intravenous maternal S.Tm challenge ...... 75 3.6.1 CD45 Immunohistochemistry ...... 75 3.6.2 Immunohistochemistry for neutrophil (anti-Ly6G) ...... 83 3.6.3 Immunohistochemistry for macrophages (anti-F4/80) ...... 89 3.7 Serum Analyses ...... 95 3.7.1 Quantification of maternal serum lipopolysaccharide endotoxin and progesterone ...... 95 3.7.2 Maternal serum cytokine analyses ...... 97 3.8 Focus on outlier Ripk3-/-Slc11a1+/+ dam #3715 ...... 102 Chapter 4 Discussion ...... 117 4.1.1 The outlier: a possible explanation for rapid fetal loss? ...... 121 4.1.2 Mechanisms of vertical S.Tm infection in mice ...... 124 vi

4.2 Study Limitations ...... 128 4.3 Future Directions ...... 131 4.4 Conclusion ...... 132 References ...... 133 Appendix A Sample Collection Log ...... 151 Appendix B Salmonella enterica serovar Typhimurium (S.Tm) immunohistochemistry & immunofluorescence protocol ...... 152 Appendix C Hematoxylin and Eosin staining protocl ...... 155 Appendix D Active caspase-3 immunohistochemistry staining protocol ...... 156 Appendix E Periodici-acid Schiff staining protocol ...... 158 Appendix F CD45 immunohistochemistry staining protocol ...... 160 Appendix G Neutrophil (Ly6G) immunohistochemistry staining protocol ...... 162 Appendix H Macrophage (F4/80) immunohistochemistry staining protocol ...... 164 Appendix I Placental morphometric measurements analyzed per litter ...... 166 Appendix J Anti-S.TM LPS positive control (liver) ...... 167 Appendix K Anti-active caspase-3 positive control ...... 169 Appendix L CD45+ leukocyte counts pooled by litter ...... 170

vii

List of Figures

Figure 1. Human placental structure and the maternal-fetal interface...... 8 Figure 2. Mouse placental structure and the maternal-fetal interface...... 17 Figure 3. Schematic diagram of necroptosis signaling in macrophages during infection with S.Tm...... 25 Figure 4. Methodology for placental and fetal weight determinations...... 31 Figure 5. Ripk3 deletion does not alter fetal resorption rates in the first 72h after S.Tm infection...... 41 Figure 6. Fetal:placental weight ratios were not significantly altered at 72h post-infection but infected mice with intact Ripk3 displayed heterogeneous developmental delay...... 46 Figure 7. Placental morphometric measurements analyzed per implantation site...... 49 Figure 8. Placental S.Tm immunohistochemistry and bacterial burden scores...... 54 Figure 9. Percentages of implantation sites infected and bacterial localization were not significantly altered by presence or absence of Ripk3...... 56 Figure 10. Placental S.Tm+ foci in Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice...... 58 Figure 11. S.Tm+ foci were not found in placentas of gestational-age matched naïve mice...... 60 Figure 12. Histopathological examination of implantation sites from the two less infected Ripk3-/- Slc11a1+/+ revealed no differences compared to naïve Ripk3-/-Slc11a1+/+...... 63 Figure 13. Histopathological examination of implantation sites from one infected Ripk3-/-Slc11a1+/+ (dam #3715) revealed major differences compared to placentas from naïve Ripk3-/-Slc11a1+/+...... 66 Figure 14. Photomicrographs of implantation sites from infected Slc11a1+/+ that resembled placentas from naïve Slc11a1+/+...... 68 Figure 15. Labyrinthine vascular congestion was not observed in the other two infected Ripk3-/-Slc11a1+/+ or Slc11a1+/+ implantation sites...... 69 Figure 16. Active caspase-3 immunohistochemistry of implantation sites from infected Ripk3-/-Slc11a1+/+ versus naive Ripk3-/-Slc11a1+/+ mice at GD 15...... 72 Figure 17. Active caspase-3 immunohistochemistry of implantation sites from infected Slc11a1+/+ versus naive Slc11a1+/+ mice at GD 15...... 74 Figure 18. CD45+ immunohistochemical cell counts in placental subregions and in the entire placenta. .76 Figure 19. CD45 immunohistochemistry in Ripk3-/-Slc11a1+/+ implantation sites comparing tissues from two of the three infected mice with limited pathology to tissues from naive controls...... 79 Figure 20. CD45 immunohistochemistry of placental tissues from heavily infected Ripk3-/-Slc11a1+/+ dam #3715...... 80 Figure 21. Implantation site CD45 immunohistochemistry from infected and naïve Slc11a1+/+ mice...... 82

viii

Figure 22. Ly6G neutrophil immunohistochemistry of implantation sites from infected and naïve Ripk3-/- Slc11a1+/+ mice...... 85 Figure 23. Ly6G neutrophil immunohistochemistry of implantation sites from heavily infected Ripk3-/- Slc11a1+/+ dam #3715...... 86 Figure 24. Ly6G neutrophil immunohistochemistry in implantation sites from infected and naïve Slc11a1+/+ mice...... 88 Figure 25. F4/80 macrophage immunohistochemistry of decidua basalis from implantation sites recovered from infected and naïve Ripk3-/-Slc11a1+/+ mice...... 90 Figure 26. F4/80 macrophage immunohistochemistry of labyrinth from implantation sites recovered from infected and naïve Ripk3-/-Slc11a1+/+ mice...... 92 Figure 27. F4/80 macrophage immunohistochemistry showed similar distributions of F4/80+ cells in implantation site tissues from infected and naïve Slc11a1+/+ mice...... 94 Figure 28. Serum lipopolysaccharide (LPS) endotoxin and progesterone (P4) quantification...... 96 Figure 29. Maternal serum cytokine values illustrated for molecules reported by others as altered in maternal serum of Slc11a1 competent mice infected with S.Tm during pregnancy...... 99 Figure 30. Maternal serum cytokine values illustrated for molecules generally implicated in bacterial infections...... 101 Figure 31. S.Tm immunohistochemistry of extraembryonic tissues including p-TGCs and yolk sac tissues from implantation sites of heavily infected Ripk3-/-Slc11a1+/+ dam #3715...... 108 Figure 32. S.Tm immunohistochemistry of chorionic plate and amnion tissues recovered from heavily infected implantation sites of Ripk3-/-Slc11a1+/+ dam #3715...... 109 Figure 33. Localization of S.Tm within highly immune infiltrated extraembryonic tissues recovered from a heavily infected implantation site of Ripk3-/-Slc11a1+/+ dam #3715...... 111 Figure 34. S.Tm immunofluorescence of extraembryonic tissues including p-TGCs and yolk sac tissues recovered from heavily infected implantation sites of Ripk3-/-Slc11a1+/+ dam #3715...... 112 Figure 35. S.Tm infection of extraembryonic tissues and anti-mesometrial decidua is associated with massive tissue damage and infiltration of CD45+ leukocytes in heavily infected implantation sites from Ripk3-/-Slc11a1+/+ dam #3715 compared to naïve controls...... 114 Figure 36. Functional Slc11a1 status determines placental localization of S.Tm, but fetal lethality likely results from S.Tm infection of fetal yolk sac...... 116 Figure 37. Vertical transmission of S.Tm in pregnant mice is dictated by functional Slc11a1 status...... 127

List of Tables

Table 1. Toll-like receptors involved in the recognition of Salmonella enterica serovar Typhimurium ...... 7 Table 2. Mouse strains used in this study...... 28 Table 3. Primers used for determining the lack of Ripk3 gene of presence of Slc11a1 gene...... 28 Table 4. Summary of all histological stains employed on implantation sites...... 32 Table 5. S.Tm bacterial burden scoring system...... 34 Table 6. Litter sizes were similar amongst Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice at mid-gestation and birth...... 40 Table 7. List study animals showing numbers of implantation sites (IS) processed, infection rates within litters, total number ...... 42 Table 8. S.Tm bacterial burden scoring system...... 52 Table 9. Ripk3-/-Slc11a1+/+ dam #3715 characteristics and implantation site outcomes...... 103

List of Abbreviations

AC; Amniotic cavity

ANOVA; analysis of variance

BHI; Brain heart infusion

BM; Basement membrane

BSA; Bovine serum albumin

CCL; CC-chemokine ligand

CFU; Colony-forming unit cIAP1/2; Cellular inhibition of apoptosis 1/2

CSF1; Colony stimulating factor 1

Ctrl; Control

CXCL; C-X-C motif ligand

CYLD; Ubiquitin carboxyl-terminal hydrolase (cylindromatosis)

DAB; 3,3' Diaminobenzidine

DAMPs; Damage-associated molecular patterns

Dapi; 4',6-diamidino-2-phenylindole

DB; Decidua basalis ddH2O; Double distilled water

DOB; Date of birth

EtOH; Ethanol

EVT; Extravillous trophoblast

FB; Fetal blood vessels

FC; Fetal circulation

FE; Fetal endothelium

GC; glycogen trophoblast cell xi

G-CSF; Granulocyte-colony stimulating factor

GD; Gestation day

GEO; Gene Expression Omnibus

GM-CSF; Granulocyte-macrophage colony-stimulating factor

H2O2; Hydrogen peroxide

H&E; Hematoxylin and eosin

HIER; Heat-induced antigen retrieval

HIV; Human immunodeficiency virus

IF; Immunofluorescence

IgG; Immunoglobulin G

IFN; Interferon

IFNAR1; type 1 IFN receptor

IHC; Immunohistochemistry

IL; Interleukin

IL-1Ra; Interleukin-1 receptor antagonist

Inf.; Infected

IS; Implantation site i.v.; Intravenous

JZ; Junctional zone

LIF; Leukemia inhibitory factor

LPS; Lipopolysachharide

MBS; Maternal blood space/sinusoid

M cells; Microfold cells

MFI; Maternal-fetal interface

MGI; Mouse Genome Informatics

xii

MLAp; Mesometrial lymphoid aggregate of pregnancy

MLKL; mixed-lineage kinase domain-like mRNA; Messenger RNA

NK; Natural killer

PAMPs; Pathogen-associated molecular patterns

PAS; Periodic-acid Schiff

PBS; Phosphate-buffered saline

PCD; Programmed-cell death

PCR; Polymerase chain reaction

PFA; Paraformaldehyde

PMN; Polymorphonuclear leukocyte

PRR; Pattern-recognition receptor p-TGC; Parietal trophoblast giant cell

PYS; Parietal endodermal yolk sac

NADPH; Nicotinamide adenine dinucleotide 3-phosphate [reduced form]

NLR; NOD-like receptor

NRAMP1; Natural-resistance associated macrophage protein 1

NRC; National Research Council

NTS; Non-typhoidal Salmonella

P4; Progesterone

RES; Reticuloendothelial system

RIPK; Receptor-interacting protein kinase

RM; Reichert's membrane

ROS; Reactive oxygen species

RNS; Reactive nitrogen species

xiii

SD; standard deviation

Slc11a1; solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1

SPI1; Salmonella pathogenicity island 1

SpT; Spongiotrophoblast s-TGC; Sinusoidal TGC

S.Tm; Salmonella enterica serovar Typhimurium

SynT; syncytiotrophoblast layer

TBC; Trophoblast cell

TBS; Tris-buffered saline

TGC; Trophoblast giant cell

TLR; Toll-like receptor

TNF; Tumour necrosis factor

TNFR1; Tumour necrosis factor receptor 1

TRAIL; Tumour necrosis factor-related apoptosis-inducing ligand

TRADD; Tumour necrosis factor receptor-associated death domain

TRAF2/5; Tumour necrosis factor receptor-associated factor 2/5

UC; Umbilical cord

UE; Uterine epithelium uNK; Uterine natural killer

UT; Uterine wall

VEGF; Vascular endothelial growth factor

VMYS; Visceral mesodermal yolk sac

VYS; Visceral endodermal yolk sac

WM-IF; Whole-mount immunofluorescence

WT; Wild-type

xiv

YS; Yolk sac

YSC; Yolk sac cavity

Chapter 1

Introduction and Literature Review

1.1 Salmonella

Salmonella species are Gram-negative, motile, facultative intracellular food-borne pathogens that contribute to significant morbidity and mortality in a variety of mammalian hosts. Salmonella species belong to the Enterobacteriaceae family; the genus is subdivided into two species, Salmonella bongori and Salmonella enterica (S.enterica) (1). Nearly all Salmonella organisms that cause disease in humans and animals belong to S.enterica, and are classified into over 2500 serovars (serological variants) based on the composition of their lipopolysaccharide (O) and flagellar (H) antigens (1). Clinically, disease presentation depends upon host susceptibility and S.enterica serovar. S.enterica serovar Typhi and

Paratyphi cause typhoid fever exclusively in humans while S.enterica serovar Enteritidis and

Typhimurium (often referred to as non-typhoidal Salmonella or NTS) cause self-limiting gastroenteritis in humans and a wide range of animal hosts (1). NTS are the second-leading cause of food-related illnesses

(after Norovirus) in the United States (2) and are estimated to cause approximately one million illnesses,

19,000 hospitalizations, and 380 deaths per year in USA

(http://www.cdc.gov/foodborneburden/PDFs/pathogens-complete-list-01-12.pdf). Salmonella enterica serovar Typhimurium (referred to as S.Tm) is the most commonly isolated serovar in clinical practice in

North America (1), and accounts for the majority of multi-drug resistant NTS bloodstream infections in children (3,4) and adults in Sub-Saharan Africa (5-7). Vaccines against Salmonella infection remain ineffective (8) and Salmonella species are becoming more resistant to antibiotic treatment (9-12).

1.1.1 Natural course of Salmonella infection

Salmonella are typically acquired orally through the ingestion of contaminated food or water, but can also be acquired through handling of pet reptiles and amphibians (13-15). Among characterized contaminated food sources are raw unpasteurized milk and milk products (16,17), egg products (18), 1 peanut butter and peanut products (19,20), meat (21) and vegetable products (22). After surviving the low pH of the stomach and entering the small intestine, Salmonella evade multiple innate host defenses

(23,24), pass through the intestinal mucous and colonize the underlying intestinal lumen epithelium by three known mechanisms. Firstly, Salmonella can invade non-phagocytic epithelial cells by bacterial- mediated endocytosis (25). Bacteria disrupt the epithelial cell brush border and induce membrane ruffling by actin rearrangement at the apical surface of these cells, which promotes bacterial engulfment (26,27).

Salmonella also enter specialized epithelial cells called microfold (M) cells, that function to sample antigens by pinocytosis to be relayed to lymphoid cells in underlying Peyer's patches (collections of gut- associated lymphoid tissue) (28). Through this process, Salmonella destroy M cells allowing microbial invasion of adjacent enterocytes (28). Both of the aforementioned processes require the function of invasion genes located in the Salmonella pathogenicity island 1 (SPI1), which encodes a protein secretion system essential for Salmonella virulence and invasion (29,30). Finally, through use of SPI1-deficient

S.Tm strains, it was determined that CD18-expressing phagocytes can uptake Salmonella and cross the intestinal barrier, potentially leading to systemically disseminated infection (31).

Once Salmonella have crossed the epithelial barrier, they enter intestinal macrophages and disseminate through the reticuloendothelial system (RES). The RES is a meshwork of connective tissue that contains immune cells involved in antigen surveillance. These cells elicit immune responses to foreign antigens (13). Macrophage dissemination typically occurs with infection by typhoidal serovars that are associated with systemic disease. Infection of immunocompetent hosts with NTS results in an early, local inflammatory response characterized by infiltration of polymorphonuclear leukocytes (PMNs) into the intestinal lumen, followed by self-limiting diarrhea (13). Symptoms last about 5-7 days and resolve spontaneously without therapeutic intervention except in patients where fluid loss is substantial or in patients who are severely ill and present with symptoms of disseminated systemic disease (1). Those most susceptible to disseminated systemic disease due to Salmonella infection include the young, elderly, and immunocompromised individuals, as well as women who are pregnant (32).

1.1.2 Salmonella infection in human pregnancy

Infection of pregnant women with Salmonella occurs at rates similar to those of the general population, with an incidence of 0.2% - 1.8% of positive rectal cultures at time of delivery (33-35).

Numerous case reports indicate Salmonella cause a variety of adverse pregnancy complications and severe illnesses such as chorioamnionitis, transplacental infections, abortions, and neonatal and maternal septicemia (36-42). The severity of outcomes from human gestational infection appears to be determined by the stage of pregnancy at the time of exposure. Fetal or neonatal complications are more limited the closer exposure is to the date of delivery (where term singleton pregnancy is defined as 280 days after the first day of the last menstrual period) (39). Pregnant women with acute Salmonella infection typically present with fever and malaise, classic signs of systemic Salmonella infection (36,37,39-41). Despite prompt antibiotic and supportive therapies, fetuses from mothers infected early in pregnancy have poor prognoses and show the severe illnesses listed above (36,40,41). Although maternal health improves after pregnancy, long-term secondary complications in the mother include endocarditis (43), osteomyelitis

(44), and long-term renal and hepatic dysfunction (45).

1.2 Pregnancy-enhanced pathogenesis of Salmonella

The physiological status of pregnancy enhances pathogenesis of Salmonella infection due to altered immunity during pregnancy and propensity of Salmonella species to infect placental tissues, as discussed below.

1.2.1 Immunological shifts during pregnancy

The increase in susceptibility to severe systemic Salmonella infection during pregnancy can be explained in part by the unique, physiological shifts in immune reactivity during pregnancy that are thought to promote tolerance of the semi-allogeneic fetus (46,47). It is a common misconception that immunological tolerance of the conceptus requires immunosuppression; instead it is appropriate to consider maternal immunity as being differentially modulated during pregnancy. In healthy individuals who are not pregnant and early in pregnancy (first and early second trimester), a pro-inflammatory state is 3 reflected in cytokine arrays of plasma. From studies in pregnant mice, it has been theorized that the maternal immune response is shifted towards anti-inflammatory type 2 (humoural) immunity, and shifted away from pro-inflammatory type 1 cell-mediated immune responses for the majority of pregnancy (46).

Experiments by Wegmann and his colleagues showed a propensity for tissues from the murine feto- placental unit to secrete the type 2 cytokines IL (interleukin)-4, IL-5, and IL-10 during all three trimesters of pregnancy (46). The same experiments revealed that murine abortion-prone pregnancies (ie. DBA/2- mated CBA/J females) induce mainly type 1 cytokines such as IL-2, TNF (tumour necrosis factor) and interferon (IFN)-γ (46). These studies and others suggested type 2 cytokine responses are protective for pregnancy, while type 1 cytokine responses compromise later pregnancy in mice (48-50). The sex hormone progesterone is believed to be involved in this immunosuppressive shift during mouse pregnancy (51,52).

During the type 2 cytokine-dominant interval of pregnancy, resistance to certain intracellular pathogens whose clearance requires a robust type 1 cytokine-mediated immune response is impaired. Foot pad inoculation of Leishmania major in non-mated C57BL/6 mice, a strain normally resistant to L.major infection elicits a strong type 1 immune response that eradicates the parasite (53). L.major challenge during C57BL/6 pregnancy results in a type 1 response that increases fetal resorptions, and is associated with decreased IL-10, and increased IFN-γ and TNF secretion by placental cells in vivo (53). It is important to note that these outcomes were observed in the absence of local placental infection; indicating that a systemic shift from a type 2 dominant immune response during pregnancy to a type 1 response lead to pregnancy compromise. In pregnant, BALB/c mice, a strain that generally mounts a type 2 anti-

L.major response, fetal loss was not observed. These studies highlight the importance of mouse genetic background in responsiveness to infection and that elevated type 1 immune responses are detrimental for placental function and pregnancy maintenance.

The shift towards type 2 immunity has also been observed in human pregnancy (54,55). This shift may account for the observations of gestational improvements in cell-mediated inflammatory disorders

4 including rheumatoid arthritis (56), and worsening of antibody-mediated inflammatory disorders such a systemic lupus erythematosus (57). Robust type 2 antibody-mediated immune responses efficiently eliminate extracellular pathogens during pregnancy (58). However, certain intracellular infections

(parasitic, viral and bacterial origin, including S.Tm) require robust type 1 responses for clearance and, during pregnancy, these type 1 responses appear to be deleterious to both mother and fetus.

1.2.2 Placental Salmonella tropism

In vitro studies have shown a propensity for S.Tm to infect human placental tissues. The placenta is a vital organ made of maternal and conceptus-derived tissues. It serves as the interface between the maternal and fetal environments (referred to as the maternal-fetal interface, MFI), and is essential for facilitating the exchange of gases, nutrients and waste products between mother and conceptus.

Additionally, the placenta is a significant source of pregnancy-associated hormones and growth factors, and is involved in protection of the fetus from both pathogens and maternal immune recognition. The mature human placenta is composed of two major components: the outer maternal layer composed of decidualized endometrial cells, maternal decidual leukocytes, and maternal vasculature that is responsible for bringing blood to the placenta and intervillous space where highly branched chorionic villi are bathed in a sea of maternal blood (Fig. 1A) (59). The chorionic villi are composed of layers of trophoblast cells

(TBCs); an outer continuous multinucleated layer called the syncytiotrophoblast, and the underlying mononuclear cytotrophoblast layer, and an inner core of fetal stromal cells and fetal blood vessels. It is important to note that TBCs, like intestinal epithelium, are an epithelial cell lineage and can be infected by S.Tm in culture. My collaborators at the University of Ottawa and the University of Rochester have shown the tropism and permissiveness of TBCs to S.Tm infection through a series of in vitro studies as described below.

Chattopadhyay and colleagues showed that S.Tm doubled in approximately 1 hour (h) in infected human choriocarcinoma cells (JEG-3 cell line) compared to 4h in HeLa cells that are also of epithelial origin (60). This finding was replicated in other trophoblast-derived cell lines including BeWo (human

5 choriocarcinoma cells) and HTR-8Neo (immortalized, first trimester human extravillous trophoblast cells)

(61). S.Tm proliferated more profoundly within trophoblast-derived cell lines than in macrophages (61).

Like macrophages, trophoblast-cell line-derived cells were able to phagocytize S.Tm, but were unable to clear S.Tm-containing phagosomes due to the production of IL-10 (61). This facilitated the uncontrolled proliferation of S.Tm in these trophoblast cell line cells. S.Tm has also been found to infect primary human cytotrophoblast cells, inducing secretion of pro-inflammatory cytokines IL-1β and IL-18 and subsequent cell death (62). Following bacterial challenge in culture, whole-mount immunofluorescence

(WM-IF) studies of human placental explants of chorionic villi from all three trimesters of pregnancy demonstrated S.Tm positive staining in syncytiotrophoblast as well as in cytotrophoblast cells (63).

However, bacterial burdens were not observed to increase over time suggesting that S.Tm could infect human trophoblasts, but not replicate in them to any significant extent (63). These results demonstrate that S.Tm is able to infect trophoblast-derived and primary TBCs in monolayer cultures, and in explant cultured villi that maintained an intact anatomical structure.

Like cells of the innate immune system, human TBCs express surface toll-like receptors (TLRs) that recognize "danger" signals. TLRs are pattern-recognition receptors (PRRs) (10 are described in humans, TLR1-TLR10 (64)), that recognize and bind to sequences known as pathogen-associated molecular patterns (PAMPs), expressed by microorganisms. Upon binding of a microorganism-expressed surface ligand, trophoblast or immune cells experience an intracellular signaling cascade that results in the production of antimicrobial peptides and cytokines (65). Recognition of S.Tm is chiefly mediated by

TLR2, TLR4, and TLR5, which recognize peptidoglycan (bacterial cell wall component) and other lipoproteins including lipoteichoic acid; lipopolysaccharide on the surface of Gram-negative bacteria, and flagellin proteins involved in bacteria motility, respectively (66-72). Table 1 summarizes TLRs and their ligands pertaining to recognition of S.Tm.

Table 1. Toll-like receptors involved in the recognition of Salmonella enterica serovar Typhimurium Toll-like receptor (TLR) Ligand

TLR2 Peptidoglycan (bacterial cell wall component)

TLR4 Lipopolysaccharide (LPS) endotoxin

TLR5 Flagellin protein

Expression of all 10 TLRs have been described in human placental trophoblasts (73,74). Ligation of TLR2 or TLR4 has been found to induce trophoblast apoptosis or cytokine release, respectively in first trimester explant cultures (75), and in term placental explants (76,77). Therefore, TBCs should be able to respond to intracellular S.Tm infection using immune response-like mechanisms.

A schematic illustration of the MFI showing possible pathways of S.Tm placental infection and transmission to fetal circulation and tissues is presented as Figure 1B. Multiple layers of trophoblast

(EVTs, syncytiotrophoblast and cytotrophoblast) and the underlying basement membrane (BM) of the chorionic villi act as the physical and immunological barriers to infection. Damage to the continuous syncytiotrophoblast layer may allow S.Tm circulating within maternal blood or carried as phagocytosed viable agents by extravasating leukocytes to breach the chorionic villus layers and enter fetal tissues via hematogenous transmission. It is also theorized that S.Tm within decidual leukocytes may be able to infect EVTs that are in direct contact with maternal decidua and traverse the villous layers to reach fetal tissues, by non-hematogenous transmission. Both pathways cause fetal septicemia by transmission of bacteria directly into fetal circulation (fetal blood vessels, FB). Therefore, understanding of the human

MFI is needed as well as its differences to the MFI of mouse, a commonly used study surrogate for infectious challenge studies during pregnancy.

Figure 1. Human placental structure and the maternal-fetal interface. Schematic diagrams of the human placenta and maternal-fetal interface (MFI) show the hemochorial placenta consisting of chorionic villi bathed in maternal blood anchored into the maternal decidua through the invasive extravillous cytotrophoblasts (EVTs) (A). (B) shows a schematic illustration of the MFI showing possible pathways of S.Tm placental infection and transmission to fetal circulation and tissues. Barriers to infection consist of the multiple layers of trophoblast including EVTs, syncytiotrophoblast and cytotrophoblast layers, and the underlying basement membrane (BM). Damage to the syncytiotrophoblast layer may allow S.Tm bacteria circulating within maternal blood or carried by infected leukocytes (infected macrophage depicted here) to breach the chorionic villus layers and cross into fetal circulation by hematogenous transmission (1). Also, S.Tm (contained within infected decidual leukocytes) may be able to infect EVTs that are in direct contact with maternal decidua and traverse the villus layers to reach fetal circulation, by non-hematogenous transmission (2). Both pathways cause fetal septicemia (3) by transmission of bacteria directly into fetal circulation (fetal blood vessels, FB; fetal endothelium, FE). It is important to note that as pregnancy progresses, the villous cytotrophoblast layer disappears resulting in one continuous layer of syncytiotrophoblast separating maternal and fetal blood. Therefore, the mature human placenta is classified as hemomonochorial, while early placenta could be classified as hemodichorial. Figures redrawn from Robbins and Bakardjiev, Curr Opin Microbiol. 2012;15(1):36-43.

1.3 Mouse Pregnancy

Mouse pregnancy is a convenient tool to study the MFI during S.Tm infection due to many similarities with human pregnancy including implantation depth, decidual formation, placentation type and immune cell recruitment to the MFI. Gestation in mice is approximately 19 days when the morning of copulation plug is defined as gestational day (GD) 0. Implantation of the blastocyst into the anti- mesomestrial pole of the mouse uterus occurs between GD 4-4.5. Post-implantation development of the mouse embryo involves a series of rapid differentiation and fusion events, involving cells of extraembryonic lineages resulting in the development of the mature placenta while derivatives of the inner cell mass cells develop into the fetus. These lineage commitment events are reviewed in great detail elsewhere

(59,78). Here, discussion is focused on the mature mouse placenta, the topic of this thesis. The mature mouse placenta is established between GD 10-12, coinciding with the opening of placental circulation

(79). The opening of placental circulation refers to the process of remodeling of the maternal vasculature supplying the uterus. Maternal blood enters the mouse uterus through radial arteries that branch into approximately 5-10 spiral arteries (79). The spiral arteries become dilated and modified, losing elastin and smooth muscle in their walls and ultimately supply blood to the MFI (79). Spiral artery modification is an endocrine-mediated process that involves the local production of estrogen within the uterus, along with progesterone produced by the mouse ovary (specifically the corpus luteum) (80-82). Remodeling of the maternal vasculature to meet the demands of the developing conceptus is also mediated by uterine natural killer (uNK) cells, details of which are reviewed elsewhere (83). The anatomical features of the mature mouse placenta are discussed below.

1.3.1 Mouse placental anatomy

Although some gross anatomical and physiological differences exist between mouse and human placenta, they share notable cellular and molecular features (78,84). Both mice and humans have an invasive hemochorial placenta, resulting in the close opposition of maternal blood and fetal circulation

(78). Figure 2 shows a schematic diagram and matched hematoxylin and eosin (H&E)-stained histological

9 section illustrating the structure of the mature mouse placenta and MFI at GD 15, the gestational age focus for my studies. The mature mouse placenta is comprised of 3 major layers: the maternal decidua basalis, a middle "junctional" zone composed conceptus-derived trophoblast cells and maternal blood sinusoids, and the placental labyrinth, the inner-most layer composed of highly branched vessels of maternal and fetal origins.

At GD 15, the maternal decidua basalis is composed primarily of transformed endometrial stromal cells called decidual cells. Decidual cells are large, epithelial-like polygonal cells organized in closely packed cords. They are characterized histologically by their pale basophilic nuclei with narrow rims of eosinophlic cytoplasm, and indistinct cell borders (85). Decidual cells secrete hormones, cytokines and growth factors, and also act as phagocytes in placental remodeling (86). Highly permeable vascular channels lined by thin, flat endothelial cells run in the spaces between the decidual cords (85).

This area also contains invading trophoblast cells, maternal vasculature including spiral arteries and venous sinuses, and maternal immune cells, the latter of which are described in greater detail in section

1.3.2. The cells and structures of this area serve many essential functions; they (1) secrete molecules that regulate maternal-fetal cross-talk, (2) supply nutrition to the conceptus through an extensive vascular network, and (3) contribute to immune regulation at the MFI (82).

Bordering the maternal decidua is the junctional zone, an area composed of two distinct trophoblast layers; (1) trophoblast giant cells (TGCs), and (2) spongiotrophoblast. TGCs are analogous to the invasive human extravillous cytotrophoblast cells (78). Mouse TGCs are diverse in phenotype, since at least four TGC subtypes have been identified and characterized based on morphological and molecular criteria (87). These include: (1) parietal-TGCs (p-TGCs) that line the implantation site (IS) and are in direct contact with decidual and immune cells in the uterus, (2) spiral-artery associated TGCs responsible for eroding away the maternal endothelium and remodeling the maternal vasculature into low resistance vessels, (3) maternal-blood associated canal TGCs that line the central arterial channels in the placental labyrinth, and (4) sinusoidal TGCs (s-TGCs) that line the maternal sinusoidal blood spaces of the

10 labyrinth (87,88). P-TGCs form a thin layer of enormous, invasive, polyploid cells lining the entire IS between the overlying maternal decidua and underlying spongiotrophoblast, in addition to extending around the entire conceptus (Fig. 2). They are responsible for mediating the early invasion of the decidualized endometrium, and secrete a wide array of hormones important for maternal adaptations to pregnancy (87-89). TGCs also share many functional similarities with macrophages, implicating them as a critical immunological barrier to pathogens (90). These include their invasive quality, their expression of cell surface proteins and cytokines involved in regulating innate immune responses (90-92), and their ability to phagocytize microbial pathogens (93). Once TGCs phagocytose, they release reactive oxygen

(ROS) and nitrogen species (RNS) including superoxide anion, hydrogen peroxide and nitric oxide, further implicating them as having roles in defense against microbial pathogens at the MFI (94-96).

Indeed, the expression of enzymes such as NADPH (nicotinamide adenine dinucleotide 3-phosphate

[reduced form]) oxidase, capable of producing ROS in immune cells has been identified in TGCs (97,98).

Therefore, the TGCs represent a crucial immunological barrier to pathogens at the MFI.

The sub-layer underlying the p-TGCs is collectively referred to as the spongiotrophoblast layer.

This layer corresponds to the column cytotrophoblast of the human placenta, lying between the invasive human EVTs and the villous tree (78). The mouse spongiotrophoblast sub-layer is comprised of two distinct cell populations: (1) spongiotrophoblast cells, and (2) glycogen trophoblast cells (GCs). The cells of this layer are interspersed between trophoblast-lined, maternal arterial canals and venous sinusoids

(Fig. 2B) (79). Histologically, spongiotrophoblast cells are small, oval cells with basophilic nuclei, identified by their dense, compact morphology (85). GCs appear vacuolated or "foamy" in appearance and are commonly arranged in cords (99). They are known to migrate out of the junctional zone and into decidual tissues, to congregate near maternal spiral arteries by GD 14.5, and they may be involved in the initiation of parturition late in gestation (79,100,101). While exact functions of these cell populations have yet to be elucidated, they are known to include endocrine activities (102-105), and are critical for fetal survival (106,107). Some of these endocrine activities include the production of prolactin-like

11 hormones, lactogens and cytokines that feedback onto the corpus luteum to maintain progesterone secretion during pregnancy (108-110). The spongiotrophoblast layer is also involved in mediating the proper structural development of the underlying placental labyrinth region (106,111).

The labyrinth is the innermost placental region and is composed of two separate, highly branched, vascular networks, one of maternal and one of fetal origin. This highly tortuous, intertwined network becomes fully developed by GD 14.5, and is responsible for facilitating the maternal-fetal exchange of gases, nutrients and waste products (79). The mouse placental labyrinth area is analogous to the intervillous space/chorionic villi interface of the human placenta (78). The surface of the murine maternal-fetal exchange consists of three cell layers and is called the interhemal membrane. Starting from the maternal blood sinusoid and moving towards fetal circulation, the layers are: (1) mononuclear sinusoidal-TGCs (s-TGCs), (2) two continuous layers of polyploidal syncytiotrophoblasts called SynT-I and SynT-II, and (3) the fetal endothelium (see Fig. 2A, inset panel) (112). The mouse placenta is traditionally classified as hemotrichorial, as three layers separate the maternal blood from fetal blood

(112). However, ultrastructural studies of the interhemal membrane have shown that s-TGCs become perforated at GD 12.5, allowing direct contact between maternal and syncytiotrophoblast tissues (113).

Proper development of the labyrinth is critical for fetal growth and development, because defects in this area result in fetal growth restriction or, in severe cases, fetal death (59).

Just bordering the distal surface of the labyrinth, are extra-embryonic tissues including the chorionic plate, Reichert's membrane, yolk sac, and amnion. The chorionic plate serves as the attachment point for the umbilical cord containing the umbilical artery and veins. Underlying the chorionic plate is

Reichert's membrane; a tough, rodent-specific acellular basement membrane separating the trophoblast and underlying yolk sac (114). This membrane is suggested to act as a filter allowing free access of nutrients to the fetus while excluding maternal cells (115). The mouse yolk sac (YS) lies directly against

Reichert's membrane and consists of 3 cell layers: (1) parietal endodermal YS cells, (2) visceral endodermal YS cells, and (3) mesodermal YS cells (Fig. 2B) (116). Early in mouse pregnancy, the yolk

12 sac is the primary site of hematopoiesis (116) and plays important roles in nutrient transfer essential for proper brain development from GD 8.5 - 10.5 (117). Later in pregnancy, yolk sac continues to mediate critical nutrient transport functions between the mother and fetus, ensuring proper organogenesis and fetal survival from GD14.5 to term (117). The last cell layer closest-to and encapsulating the entire fetus is the amnion, a protective bilayered membrane (Fig. 2B) (85).

Unlike human studies of TLR expression in placenta, little is known regarding the basal expression levels of TLRs in mouse placenta. Some groups have observed low basal levels of TLR2

(recognizes peptidoglycan) messenger RNA (mRNA) expression in mouse placenta from GD 10 to term

(118). However, Gonzalez and colleagues detected modest expression of TLR2 from mid-pregnancy to term (119). TLR4 (recognizes LPS) is moderately expressed in mouse placenta, and decreases over the course of gestation but is expressed more abundantly in fetal membrane tissues (118-120). Others have confirmed TLR4 mRNA transcript and protein expression in the context of placental infection with periodontal, Gram-negative bacteria (121,122), or in mouse models of intrauterine inflammation (123).

Expression of TLR5 (recognizes flagellin) in mouse placenta has yet to be reported. A gene expression profile database search using the Gene Expression Omnibus (GEO) repository (National Centre for

Biotechnology Information (NCBI), U.S. National Library of Medicine) indicates TLR5 is expressed by mouse trophoblast stem cells. Together, these results indicate that there may be differential expression of

TLRs in a temporal and spatial manner during mouse pregnancy but this requires further investigation.

1.3.2 Immune cells at the mouse maternal-fetal interface

Immune cells at the mouse MFI are primarily contained within maternal uterine tissues particularly the decidualized mesometrial endometrium (decidua basalis). Fetal immune cell populations arise in the 2nd week of gestation but only circulate after completion of the cardiovascular system and are considered to have relatively little impact on uterine immune cell populations (124). Elegant experiments using flow cytometry, WM-IF and histological techniques have identified recruitment of large numbers of leukocytes to the MFI before the GD 10 opening of placental circulation (125,126). Decidual leukocytes

13 account for as much as 30-40% of decidual cells during early pregnancy (124). Decidual leukocytes have major roles in support of decidualization and decidual angiogenesis, facilitate invasion of trophoblasts, and ensure proper development of the mature placenta (124,126-128). The dominant leukocyte subset in early decidua basalis is the uNK cell. Other significant decidual leukocyte populations include monocytes, macrophages and dendritic cells (125,129). After mid-pregnancy, uNK cells diminish in number (130,131). Since the S.Tm challenge was given to my pregnant mice on GD 12, when uNK cells are already in decline, this introduction will now focus on decidual leukocyte populations found in the 2nd half of mouse pregnancy.

Immunohistochemical (IHC) localization of leukocytes in placental tissues using an antibody for

CD45 (a pan-leukocyte marker) has been essential for understanding leukocyte dynamics over the course of mouse gestation. CD45 IHC shows great numbers of CD45+ leukocytes in uterine and decidua basalis tissues up until approximately GD 13.5, after which numbers decrease to negligible levels just prior to parturition (132). In decidual tissues, leukocytes can be found both in the stroma, indicating infiltration into the extravascular environment, and in the intravascular environment contained within blood vessels.

Extravascular decidual leukocytes include regressing uNK cells, monocytes and neutrophils (133).

Intravascular decidual leukocytes consisting of granulocytes (40-60%), monocytes (25-30%), T cells (15-

20%), and B cells (5-8%) appeared adhered to vessel lumens (133). Decidual macrophages are scarce after GD 12.5 and thus do not contribute to a significant proportion of leukocytes within decidual tissues in the 2nd half of mouse pregnancy. Macrophages instead are mainly localized to the overlying myometrium (124). The functions of decidual leukocytes during mid-gestation remain largely unknown, but are thought to include immunoregulation at the MFI.

In the junctional zone and labyrinth placental regions, CD45+ leukocytes are rarely seen outside of maternal vessels or blood spaces (132,133). The leukocyte populations contained within these maternal blood spaces include neutrophils (40%–50%), monocytes (20%), B cells (10%–20%) and T cells (10%–

20%) (133). Although fetally-derived immune cells are postulated to have little impact of uterine immune

14 cell populations, fetal macrophages called Hofbauer cells, must be considered during the 2nd half of mouse pregnancy. Hofbauer cells develop within mouse fetal YS at GD 10 (134), and then circulate within vessels of the placental chorionic villi by GD 12 (135). At this stage, Hofbauer cells display reactivity for the macrophage surface marker, F4/80, and are capable of phagocytosis and lysosomal digestion (135). Hofbauer cells are believed to play roles in the remodeling of chorionic plate structures

(135) and in response to infection during pregnancy (136). Hofbauer cells have been implicated as the replicative niche of some pathogens including Zika virus, and may play a role in transplacental transmission of pathogens from mother to fetus (137-140). In the context of other pathogens such as human immunodeficiency virus (HIV), Hofbauer cells are thought to limit viral replication and transplacental transmission (141,142).

Towards the end of normal mouse pregnancy, uterine leukocyte recruitment and inflammation dramatically increases in preparation for parturition. Just prior to initiation of parturition in mice, progesterone secretion is abolished through luteolysis of ovarian corpora lutea, resulting in a dramatic decrease in plasma progesterone levels (143,144). This is considered a critical initiating event in murine parturition and is associated with the shift from the late gestational type 2-cytokine dominant immune response towards the non-pregnant type 1 pro-inflammatory immune environment. It is theorized that myometrial and decidual tissues secrete chemokines that recruit peripheral neutrophils and macrophages to infiltrate uterine tissues including the myometrium, decidua basalis, chorioamniontic membranes, and the cervix (145-147). This results in uterine inflammation and subsequent synthesis of prostaglandins, key promoters of uterine contractility (147). The link between intrauterine infection, inflammation and preterm birth is well known, giving support to the role of inflammation in initiating parturition.

Figure 2. Mouse placental structure and the maternal-fetal interface. Shown is both a schematic diagram (A) and matching hematoxylin and eosin (H&E)-stained mid-sagittal histological section (scale bar, 500 µm) (B) of a GD 15 Ripk3-/-Slc11a1+/+ implantation site (IS). The major placental regions are emphasized in dark blue text. The mesometrial lymphoid aggregrate of pregnancy (MLAp) is a transient structure forming in mid-gestation around the maternal vasculature entering the IS. It lies between the circular and longitudinal layers of the myometrium. Underlying the MLAp and myometrium is the decidua basalis, the outer-most placental region of maternal origin. The middle placental area is the junctional zone (JZ). It consists of a thin layer of parietal-trophoblast giant cells (p-TGCs) and spongiotrophoblast sub-layer. The junctional zone is characterized by dense, compact trophoblast cells and large maternal blood spaces. The labyrinth area is the inner-most placental region composed of loosely packed, tortuous labyrinthine vessels of maternal and fetal origin. Fetal vessels are schematically represented here, arranged in villi. Inset panel in (A) shows the interhemal membrane, the primary area of the maternal-fetal exchange of gases, nutrients, and waste products. The mouse placenta is considered to be hemotrichorial because 3 layers separate the maternal blood in maternal blood sinusoids (MBS), from the fetal blood in the fetal circulation (FC). The layers are the mononuclear sinusoidal-TGC (s-TGC), two continuous layers of syncytiotrophoblast (SynT-I and SynT-II), and finally the fetal endothelium (FE). The labyrinth rests the chorionic plate, of extraembryonic origin. This area serves as the attachment site for the umbilical cord containing the umbilical artery and veins. A thick, acellular basement membrane lies just distal to the chorionic plate called Reichert’s membrane (RM). Abutting RM is the yolk sac (YS), composed of 3 distinct cell layers: parietal endodermal YS cells (PYS), visceral endodermal YS cells (VYS), and visceral mesodermal YS cells (VMYS). Finally, the entire embryo is encapsulated by a bilayered membrane called the amnion. Panel A redrawn from The Guide to Investigation of Mouse Pregnancy (p. 145 & 179), Ed. Croy et al., 2014; London, UK: Academic Press/Elsevier.

1.4 Salmonella enterica serovar Typhimurium infection during murine pregnancy

Susceptibility of inbred mouse strains to infections is largely determined by genetics and in some instances, may be controlled by a single gene (148). Such is the case for mouse susceptibility to S.Tm infection that is determined by the presence or absence of a functional solute carrier family 11 (proton- coupled divalent metal ion transporters), member 1 (SLC11A1) protein, formerly known as the natural resistance-associated macrophage protein1 (NRAMP1) encoded by the Slc11a1 gene. For the purpose of this thesis, the gene name will be referred to as Slc11a1 following official Mouse Genome Informatics

(MGI) nomenclature, and its protein will be called by its preferred name, NRAMP1. In mice, the Slc11a1 gene is located on chromosome 1, and encodes an integral membrane phosphoglycoprotein found in phagosomes of macrophages and in granules of neutrophils (149,150). TBCs also possess phagosomes but Slc11a1 expression has yet to be reported in trophoblast. A GEO repository search indicates that 17

Slc11a1 is expressed in trophoblast stem cells. NRAMP1 is a divalent transition metal ion transporter involved in iron metabolism and host susceptibility/resistance to certain intracellular pathogens including

Mycobacteria, Salmonella and Leishmania (151). Resistance to pathogens is achieved through the active removal of iron and manganese from phagosomes, ions that are essential cofactors of prokaryotic and eukaryotic metallo-enzymes required for pathogen survival (151). Susceptible mouse strains such as

C57BL/6 and BALB/c carry a missense mutation resulting from a glycine to aspartate substitution at position 169 (often referred to in the literature as Slc11a1G169D or Slc11a1D169), affecting proper protein maturation and/or trafficking leading to no mature functional NRAMP1 (149). Resistant inbred mouse strains with wild-type (WT) Slc11a1 alleles express functional protein and include 129×1/SvJ, A/J, and

C3H (152,153). Non-pregnant susceptible mice are unable to control systemic S.Tm replication and reach their humane endpoint within 7 days while non-pregnant resistant mice control systemic replication and although initially ill after challenge, eventually clear the infection within 60 days (154,155). Therefore,

NRAMP1 plays a key role in regulating resistance to intracellular S.Tm infection.

Pregnancy enhances S.Tm-mediated pathogenesis in mice as in women. However, S.Tm infection during murine pregnancy leads to distinct outcomes based on NRAMP1 status. In resistant 1291/SvJ mice, mid-pregnancy (GD 10-12) intravenous (i.v.) infection with 103 S.Tm colony-forming units (CFUs) resulted in 100% fetal loss and greater than 60% maternal death within 6 days of bacterial challenge

(155). At 3 days after infection, splenic bacterial burdens were increased ~1000-fold when compared to burdens in non-pregnant challenged controls. Profound bacterial burdens were reported in the range of

107 bacteria in placental tissues pooled by litter as determined by CFU-analysis (155). This was accompanied by a mean fetal resorption rate of ~75%, a significant elevation compared to gestational-age matched, naïve (i.e. non-challenged) controls (155). In studies conducted only 14h after challenge, bacteria were detected in fetal livers pooled by litter, indicating rapid transplacental infection and subsequent fetal septicemia preceded fetal loss (155).

Since adaptive immune responses to S.Tm infection are delayed even in resistant hosts, a robust innate immune response is required for the clearance of S.Tm bacteria (156). Examination of innate immune cell populations in the study discussed above showed dampened innate maternal immune responses during pregnancy (155). In non-pregnant infected controls, splenic recruitment of macrophages, dendritic cells, neutrophils, and peripheral natural killer (NK) cell was significantly increased compared to non-infected, non-pregnant controls (155). This was accompanied by elevated plasma IL-12, a cytokine classically released by neutrophils, macrophages and dendritic cells to influence the functions of NK cells and T cells (155). Innate immune cell recruitment to spleen and plasma IL-12 elevation were not observed in pregnant, S.Tm infected 1291/SvJ mice, indicating pregnancy perturbs maternal innate immune responses to infection by intracellular pathogens (155). Serum IL-6 levels were elevated in pregnant infected mice 3 days after infection, but other pro-inflammatory cytokines such as IFN-γ and TNF were not altered (155). Pro-inflammatory cytokine transcripts were significantly increased in placental tissues 3 days after infection compared to placental transcripts from non-infected controls. These transcriptional changes included IL-6 (~40-fold increase), TNF-α (~8-fold increase), IL-18 (~3.5-fold increase) and paradoxically, a ~10-fold increase in IL-10 expression, a classic anti-inflammatory cytokine (155). This study highlighted the following important observations regarding S.Tm pathogenesis in murine pregnancy:

(1) resistance to severe systemic S.Tm infection is abolished by pregnancy in mice normally resistant in the non-pregnant state,

(2) placental S.Tm colonization greatly surpasses bacterial burdens in maternal organs indicating tropism of S.Tm for placenta,

(3) transplacental infection and subsequent fetal loss is rapid and precedes maternal death, and

(4) maternal innate immune responses to S.Tm are altered during pregnancy, with a strong type 1 pro- inflammatory response specifically in placenta.

In contrast to the above findings in S.Tm resistant, pregnant 1291/SvJ mice, C57BL/6J mice, an NRAMP-mutated and susceptible strain, reach humane endpoints 2-3 days after challenge with identical S.Tm dose (103 CFUs), route (i.v.) and timing of exposure (157). Mice were therefore bred to display "hybrid" resistance for further studies by mating resistant 1291/SvJ females by susceptible

C57BL/6J males to produce 129.B6F1 offspring (60). Non-pregnant 129.B6F1 mice (heterozygous for functional NRAMP1) developed chronic, non-lethal but unresolved S.Tm infection indicating partial

NRAMP1 function (60). Replicating the earlier studies in 1291/SvJ mice, a mid-pregnancy (GD 11-13) challenge of 103 wild-type (WT) S.Tm i.v. resulted in nearly 100% fetal loss 3 days post-infection, and

~80% maternal death within the following day. These lethal outcomes were correlated with massively increased splenic bacterial burdens relative to non-pregnant infected controls (105 vs. 104 bacterial CFUs, respectively), and ~108 bacterial CFUs in litter-pooled placentas (60). Thus, S.Tm appears to preferentially expand within placental tissues, with an estimated in vivo doubling time of less than 2h.

Infection of GD 11-13 pregnant 129.B6F1 mice with an avirulent auxotrophic mutant S.Tm strain

(∆aroA SL3261) that requires extraneous aromatic amino acids for its replication (158) resulted in productive expansion of placental S.Tm without fetal loss or marked placental inflammation or pathology

(60). In contrast, lethal infections with WT S.Tm in pregnant 129.B6F1 mice induced an overt serum inflammatory response with 5 to 30-fold increases in pro-inflammatory cytokines and chemokines including IL-6, CXCL13, granulocyte colony-stimulating factor (G-CSF), IL-1 receptor antagonist (IL-

1Ra), CCL1, CCL12, CXCL1, and CXCL2 and overt placental inflammation (elevations in mRNA transcript levels of G-CSF (~400-fold), TNF-α (~1000-fold), IL-6 (~200-fold), and IFN-γ compared to pregnant naïve mice) (60). Massive extravascular infiltration of PMNs was seen in placental histopathology from mice infected with WT S.Tm compared to either ∆aroA-infected pregnant mice or non-infected pregnant controls (60). The WT S.Tm-infected placentas had reduced decidua basalis area, and highly necrotic labyrinth. Using green fluorescent protein-tagged WT S.Tm, bacterial foci were localized throughout the necrotic labyrinth trophoblast, however the precise cellular localization of S.Tm

20 was not determined. These studies in hybrid resistant pregnant mice provided the following important advances:

(1) heterozygotes for NRAMP1 expression are resistant to severe S.Tm infection only in the non-pregnant state,

(2) bacterial burden does not correlate with placental pathology and fetal/maternal death, and

(3) inflammation in response to preferential and rapid S.Tm infection in placenta leads to adverse consequences including placental necrosis and pathology.

The timing of S.Tm challenge during murine pregnancy is an important variable. Mid-pregnancy i.v. WT S.Tm infections as discussed above result in acute fetal loss, while late-pregnancy (GD 14-15) i.v. infections with an identical dose induce preterm labour (155). Similar outcomes were observed with late-pregnancy infection by Salmonella enterica serovar Enteritidis (S.Enteritidis) using susceptible

BALB/c mice (159). In non-pregnant BALB/c mice, ingestion of a low dose of S.Enteritidis (3-4103

CFUs) normally results in self-limiting enterocolitis. When BALB/c mice were orally infected in late- pregnancy (GD 15), 40% of dams experienced preterm delivery, with a 33% fetal loss rate, and fetal- growth restriction was apparent in the surviving pups (159). There was no significant maternal mortality; all dams survived and appeared healthy post-challenge. Bacterial colonization of placentas from BALB/c dams that delivered preterm correlated with pup colonization detected by bacterial culture (159).

Histopathology 3 days post-infection (GD 18) revealed significantly higher bacterial burdens in placenta compared to other maternal tissues including Peyer's patches, intestine, and spleen (159). Late pregnancy challenge of BALB/c mice with S.Enteritidis also induced significant increases in pro-inflammatory cytokines levels in placental tissue, amniotic fluid and maternal plasma. Significant placental infiltration by PMNs was observed, with enhanced apoptosis in placental tissues of S.Enteritidis-infected mice compared to non-infected pregnant controls (159).

Although not addressed in any of the above studies, it is also important to consider the effects of

LPS endotoxin produced by S.Tm, a well-known abortifacient, in these in vivo models. It is currently unclear what roles LPS may play in mediating S.Tm-induced pregnancy pathology.

1.5 Salmonella-induced cell death mechanisms

Cells infected by S.Tm undergo programmed-cell death (PCD) as an immune defense mechanism triggered by the host to limit and control the spread of infection. Interestingly, S.Tm possess virulence strategies to manipulate and control PCD-mechanisms of host cells that facilitate its pathogenesis and survival (160). S.Tm can induce cell death by a variety of mechanisms, depending on the cell type infected. For example, S.Tm induces death of intestinal epithelial cells by apoptosis, a PCD pathway with a non-inflammatory outcome characterized by DNA fragmentation, cellular shrinkage and membrane blebbing (161,162). This discussion will focus on S.Tm-induced cell death of macrophages since it is the dominant replicative niche of S.Tm in vivo (163-165), and due to its similarities with trophoblasts for which mechanisms of S.Tm-induced cell death are unknown. S.Tm induces pro-inflammatory cell death in macrophages by pyroptosis and necroptosis, for which the latter is the focus of this thesis.

Since pyroptosis is not the main cell-death mechanism addressed in this thesis, it is only briefly described here. Pyroptosis is a pro-inflammatory, caspase-1-dependent cell death pathway characterized by the activation of a mulitmeric protein complex called the inflammasome (166). The inflammasome consists of cytosolic PRRs such as the NOD-like receptor (NLR) family of proteins which recognize bacterial components (PAMPs) and associated damage signals called DAMPs (damage-associated molecular patterns) (167). Upon inflammasome assembly, caspase-1 is activated, resulting in the activation and release of pro-inflammatory, pyrogenic cytokines IL-1β and IL-18 and subsequent pyroptotic-cell death (167). Pyroptotic-cell death is characterized by DNA fragmentation, rapid cell lysis, membrane rupture and release of pro-inflammatory intracellular contents into the extracellular environment (168,169). Although S.Tm is known to rapidly induce pyroptosis in macrophages in the

22 initial stages of infection, evidence also suggests that as infection progresses, pyroptosis may become a potent host immune defense mechanism to limit S.Tm pathogenesis (166,167,170). Activation of pyroptosis during S.Tm infection has been referred to as a "double-edged sword" from the perspectives of both pathogen and host survival.

1.5.1 Necroptosis

S.Tm also induces pro-inflammatory PCD of macrophages by necroptosis, or programmed necrosis. Historically, necrosis has been viewed as an accidental or unregulated type of cell death resulting from nonspecific and nonphysiological stress (171). Within the last decade, evidence has emerged indicating that necrosis can be regulated and induced under certain pathological and physiological circumstances (172-176). Necroptosis is a caspase-8 independent cell death pathway mediated by the receptor-interacting protein kinase (RIPK) family (schematically represented in Fig. 3).

Necroptosis can be induced through the ligation of various death receptors of the TNF super family (Fas ligand, TNF receptor 1 (TNFR1) and TNF-related apoptosis-inducing ligand (TRAIL)), type 1 IFN receptor (IFNAR1), and through TLR stimulation (173,177-179). The resulting intracellular signaling cascade is complex and is just beginning to be characterized. The most extensively characterized pathway is initiated through TNF. Upon receptor stimulation, the membrane-associated TNFR1 signaling complex

(called complex I) is formed which includes TNFR-associated death domain (TRADD), RIPK1, cellular inhibition of apoptosis 1/2 (cIAP1/2), TNFR-associated factor 2/5 (TRAF2/5), and ubiquitin carboxyl- terminal hydrolase CYLD (cylindromatosis) protein (180). This results in the phosphorylation of RIPK1, which then assembles with and phosphorylates RIPK3 forming the necrotic death complex or necrosome

(181). Phosphorylated RIPK3 then activates its downstream substrate mixed-lineage kinase domain-like

(MLKL) protein, which is believed to facilitate the disruption of cellular membranes (182-184). The activity of RIPK3 is critical for initiating necroptosis (180,185). Morphologically, necroptosis (like necrosis) is characterized by nuclear shrinkage, organelle and cell swelling, and rupture of the plasma

23 membrane (171). Intracellular contents including DAMPs are released into the extracellular environment, triggering inflammation.

S.Tm is known to induce expression of type 1 IFNs by macrophages (186). Building on this observation, Robinson and colleagues determined that S.Tm induces death of infected macrophages by necroptosis through IFN signaling, a potentially critical immunoevasive strategy (177). Survival was enhanced in C57BL/6J mice lacking IFNAR1 signaling and RIPK3 activity when challenged with 102

S.Tm by i.v. or intraperitoneal inoculation (177). Although the exact signaling mechanisms have yet to be elucidated, RIPK1 appears to associate with IFNAR1 (177). IFN signaling may also contribute to necroptosis in macrophages by promoting and sustaining the necrosome complex (187). All investigations to date of S.Tm-induced necroptosis have been conducted in non-pregnant mice. Key necroptotic signaling mediators however are expressed by cultured human trophoblasts (62,188,189). Little is known about the role of necroptosis in mouse placenta but a GEO repository search suggests both Ripk3 and its downstream mediator Mlkl are expressed in mouse trophoblast. Ripk3 is currently the only necroptosis related gene for which a viable mouse gene knockout is available (190). Studies of S.Tm infection in pregnant Ripk3 null mice are expected to identify the importance of necroptotic cell death in the pregnancy complications associated with gestational infection by this pathogen.

Figure 3. Schematic diagram of necroptosis signaling in macrophages during infection with S.Tm. S.Tm infection induces type I interferon (IFN) signaling through the production of type I IFNs (IFN-α/β). Interaction of IFNs with the type I IFN receptor (IFNAR1) results in recruitment of RIPK1. Ligation of the tumour necrosis factor (TNF) receptor 1 (TNFR1) leads to the formation of the complex I group of adaptor proteins and activation of RIPK1. TLR signaling through receptor stimulation by pathogen- associated molecular patterns (PAMPs) also induces assembly of the necrosome, which includes phosphorylated RIPK1 and RIPK3. RIPK3 phosphorylates its downstream effector mixed-lineage kinase- like protein (MLKL). MLKL disrupts cellular membranes and induces cation influxes resulting in cell swelling. This leads to membrane rupture and release of intracellular contents including damage- associated molecular patterns (DAMPs) triggering inflammation. cIAP 1/2, cellular inhibitor of apoptosis protein 1/2; CYLD, ubiquitin carboxyl-terminal hydrolase cylindramatosis protein; LPS, lipopolysaccharide; RIPK, receptor-interacting protein kinase; TLR, toll-like receptor; TRADD, TNF receptor-associated death domain; TRAF 2/5, TNF receptor-associated factor 2/5. Figure redrawn and adapted from Du et al., Cell Mol Immunol. 2013;10(1):4-6.

1.6 Rationale, hypothesis and main objectives

S.Tm preferentially replicates in placental tissues and is associated with deleterious, pro- inflammatory immune responses at the MFI. This ultimately leads to trophoblast cell death, placental compromise, pregnancy loss and maternal death in mice. The mechanisms for trophoblast cell death are unknown but may overlap with pro-inflammatory cell death of macrophages, given the striking functional similarities of these two cell types. I hypothesize that elimination of the key necroptotic cell death protein

RIPK3 involved in S.Tm-mediated death of infected trophoblast will decrease placental inflammation and trophoblast cell death, and increase conceptus survival compared to their controls. To test this hypothesis, congenic C57BL6/J mice expressing functional NRAMP1 (denoted Slc11a1+/+) with moderate susceptibility to severe S.Tm infection were used as the background strain, either with or without Ripk3 expression.

My main objectives are:

1. To assess if the interval between S.Tm infection and fetal death is extended by blockade of

RIPK3 signaling by gene deletion.

2. To characterize and quantify the extent of bacterial localization, TBC death and placental

damage, and immune cell infiltration in tissues from S.Tm-infected, pregnant gene-modified and

control mice (i.e. Ripk3-/-Slc11a1+/+ vs. Slc11a1+/+, respectively).

Chapter 2

Materials and Methods

2.1 Mice

Collaborators Dr. Lakshmi Krishnan and her trainees conducted all animal handling experiments and sample collections at the University of Ottawa (UOttawa). All animal use was conducted in accordance with guidelines the Canadian Council on Animal Care and under approval from the University of Ottawa

Animal Care Committee. Mice were housed in the animal facility at the National Research Council

Canada (NRC - Human Health Therapeutics, Ottawa, ON, Canada). Ripk3-/- (B6.129-Ripk3tm1Vmd) mice were provided as breeding pairs by Dr. V. Dixit (Genentech, San Francisco, CA, USA) (190) and

Slc11a1+/+ (B6J.129S1-Slc11a1r/GbrtJ) mice were provided as breeding pairs by Dr. G. Barton

(University of California, Berkeley, Berkeley, CA, USA) (191). C57BL/6J mice were purchased from

The Jackson Laboratory (Bar Harbor, ME, USA).

All mice used were bred on or had been backcrossed at least 10 generations to the C57BL/6J genetic background. Ripk3-/- mice were additionally crossed with Slc11a1+/+ mice in Dr. Krishnan's laboratory. F2 progeny were genotyped to determine the lack of Ripk3, and presence of the Slc11a1 gene. Suitable F2 progeny were then paired and bred until the 4th filial generation, generating the Ripk3-/-Slc11a1+/+ strain that was also used in my research. Mouse strains used in my studies are summarized in Table 2.

Archival specimens prepared by my Ottawa collaborators and histologically prepared and studied by an undergraduate thesis student, Annie Peng, prior to my arrival at Queen's University were also used in addition to mouse samples received and prepared by myself.

Table 2. Mouse strains used in this study. Genotype Genetic Background Outcome

Ripk3-/-Slc11a1+/+ B6.129-Ripk3tm1Vmd backcrossed to Deleted for Ripk3, Slc11a1 competent

B6J.129S1-Slc11a1r/GbrtJ

Slc11a1+/+ B6J.129S1-Slc11a1r/GbrtJ Functional Ripk3, Slc11a1 competent

2.2 Genotyping

All genotyping was conducted at the NRC, Ottawa. Ripk3 deficiency was validated by Polymerase

Chain Reaction (PCR) using genomic DNA obtained from mouse ear punches at NRC by K.Wachholz

(UOttawa) using the Proteinase K (Sigma-Aldrich Canada Co.; Oakville, ON) method. The amplified

DNA product was separated on a 1 to 1.5% agarose gel and visualized using ethidium bromide staining.

The expected product size for wildtype and null Ripk3 were ~1kb and ~1.2kb, respectively. Presence of a functional Slc11a1 gene was determined by sequencing of the PCR product to screen for a single point mutation (guanine to alanine) at position 169. This substitution results in a glycine to aspartic acid substitution within the gene product formerly known as the NRAMP1 protein. The DNA primers used are listed in Table 3.

Table 3. Primers used for determining the lack of Ripk3 gene of presence of Slc11a1 gene. Mouse Strain Primers used for genotyping

Ripk3-/- Ripk3 forward common 5’ - AATCGTTCCTGGATGGTGAG - 3’

Ripk3 reverse wild-type 5’ - GGAGCCATTCTCCATGAATC - 3’

Ripk3 reverse mutant 5’ - GATCCTGATCCTGACCCTGA - 3’

Ripk3 neomycin 5’ - ATCGACAAGACCGGCTTCCATCCGA - 3’

Slc11a1+/+ Slc11a1 forward 5’ - TCATCGGGACGGCTATCTCCTTCAA - 3’

Slc11a1 reverse 5’ TTGCGCAAACCTAGGGGTACAGGGA - 3’

2.3 Mating strategy and pregnancy timing

Two age-matched, eight to twelve week old female mice were placed in a cage for syngeneic mating overnight with one syngeneic male. The morning of copulation plug detection was taken as gestational day (GD) 0 of pregnancy. Bacterial challenge was conducted at GD 12.

2.4 Bacterial preparation and in vivo infection

Salmonella enterica serovar Typhimurium strain SL1344 (denoted as S.Tm) were grown at NRC in liquid culture in brain heart infusion (BHI) medium (Difco Laboratories, BD Biosciences; Franklin

Lakes, NJ, USA) under constant shaking at 37°C. At mid log phase (OD600 = 0.8), S.Tm were harvested, divided into aliquots and frozen at -80°C in 20% glycerol. One aliquot was used to prepare serial dilutions made in 0.9% saline that were spread on BHI agar plates containing streptomycin (Sigma-Aldrich Canada

Co.) to determine bacterial concentration in the frozen stock.

For in vivo infections, a frozen S.Tm stock was thawed and diluted in 0.9% saline to a final concentration of 1x103 bacteria colony-forming units (CFUs)/200µl and GD 12 pregnant mice were inoculated by K. Gurnani (NRC, Ottawa) via the lateral tail vein i.e. intravenous (i.v.) infection. GD- matched controls were inoculated with 0.9% saline i.v. (referred to as naïve). Post-injection CFU analyses on inocula were conducted for each study by K. Wachholz to confirm dosages received by each mouse.

Pregnant, inoculated mice were monitored and ranked for clinical condition twice per day by K.

Wachholz. Clinical condition of each animal was scored at these times as: A - healthy; B - ruffled fur, lively; C - ruffled fur, activity slowing, sick; D - ruffled fur, hunched, very little activity, eyes squeezed shut, very sick; E - moribund; F - deceased. Mice were humanely euthanized when a ranking of C or D was observed. For animals assigned to my research, no rankings of B to F were utilized.

2.5 Sample collection

Mice were anesthetized at NRC by isofluorane inhalation at 48 and 72 hours after infection (GD

14 and GD 15 of pregnancy, respectively) and terminal submandibular cheek bleeds were performed to collect between 200-900µl of blood in serum separator tubes (BD Biosciences; San Jose, CA, USA). 29

Tubes were centrifuged at 3000 rpm for 8 minutes at 4°C, and serum was collected and frozen at -80°C until shipment on dry ice to Queen's University.

Immediately following blood collection, mice were euthanized by cervical dislocation. Spleen, liver, ovaries, and uteri were dissected and immediately immersion-fixed immediately in 4% paraformaldehyde either overnight at 4°C, or for six hours at room temperature under slow rocking. Then, samples were washed thrice with phosphate buffered saline (PBS), followed by one wash with double distilled water (ddH2O), and then placed in 70% ethanol in individually-labeled glass scintillation vials until shipment on ice to Queen's University. Pertinent details such as date of birth, copulation plug detection date, euthanasia date, gross post-mortem pathological findings, and IS placement along the uterine horns were communicated at the time of shipment through a sample collection log (see Appendix

A for an example).

2.6 Tissue processing for histology

Upon receipt of tissues, weights of IS and fetuses were recorded as indicated in Figure 4. Briefly, the entire feto-placental unit was weighed (including uterine wall and fetal membranes), and then the fetus was removed and weighed. The final placental weight was determined by subtracting the fetal weight from the feto-placental unit weight. Fetal membranes were preserved only at the chorionic plate.

Placenta Feto-placental unit Fetus

Placenta weight = feto-placental unit weight - fetal weight

Figure 4. Methodology for placental and fetal weight determinations. Upon receipt of tissues, feto-placental units (including uterine wall and fetal membranes) were weighed, after which the fetus was removed and weighed. The final placental weight was determined according to the calculation, by subtracting the fetal weight from the feto-placental unit weight. The feto-placental unit and uterine wall component of the attachment site are included in this image from a GD 14 naïve Slc11a1+/+ mouse. Membranes are present but have been folded away from the fetus. Photo is original.

IS and fetuses were bisected sagittally, and processed immediately into paraffin using standard, vacuum processing methodology. IS from all litter-members (including resorbed implantation sites) of

S.Tm-infected dams (three pregnant females per genotype per time point) were serially sectioned (6µm), since preliminary data suggested variable S.Tm infection rates between individual IS from a single dam

(refer to Chapter 3, Table 7) for total numbers of slides sectioned. Two paraffin-embedded IS from each naïve GD-matched control (three pregnant females per genotype per time point) were serially sectioned for comparisons. For fetal histology, one fetus per Slc11a1+/+ dam at 72h post-infection were serially sectioned following similar procedures for IS. Summaries of all histological stains employed are presented in Table 4.

Table 4. Summary of all histological stains employed on implantation sites. Histological Stain Purpose Protocol Reference

Salmonella enterica serovar Typhimurium Identification of S.Tm bacteria Appendix B

(S.Tm) immunohistochemistry

Hematoxylin & eosin (H&E) Routine histopathology Appendix C

Caspase-3 immunohistochemistry Identification of cells undergoing Appendix D

apoptosis by extrinsic or intrinsic

pathways

Periodic-acid Schiff (PAS) Placental morphometric Appendix E

measurements

CD45 immunohistochemistry Identification of leukocytes Appendix F

Neutrophil (Ly6G) immunohistochemistry Identification of neutrophils Appendix G

Macrophage (F4/80) immunohistochemistry Identification of macrophages Appendix H

2.7 Salmonella enterica serovar Typhimurium immunohistochemistry

Serially sectioned IS, and positive control tissue (liver tissue sections from S.Tm- infected animals) slides were deparaffinized by immersion in three changes of xylene and rehydrated in 100%,

95% and 70% ethanol for 5 minutes each. Slides were then rinsed with ddH2O for two minutes after which sodium citrate (pH 6.0) heat-induced antigen retrieval (HIER) was performed. Sections were washed twice with Tris-buffered saline (TBS) with 0.1% Triton-X 100 (TBS-Triton-X 100) and blocked for endogenous peroxidase activity with 3% hydrogen peroxide (H2O2) at room temperature for 30 minutes. Sections were washed twice with TBS-Triton-X 100 and incubated with 10% normal goat serum

(Jackson Immuno Research Labs; West Grove, PA, USA - cat. no. 005-000-121) for 60 minutes at room temperature. Sections were then washed twice with TBS-Triton-X 100 and incubated with an

32 unconjugated, AffiniPure Fab Fragment goat anti-mouse IgG (H+L) (Jackson Immuno Research Labs - cat. no. 115-007-003) at a concentration of 20µg/ml in TBS with 0.05% Tween-20 (TBST) at room temperature for 60 minutes, to block endogenous mouse immunoglobulin G (IgG) interactions. Sections were washed twice with TBS-Triton-X 100 then incubated overnight at 4°C with monoclonal Salmonella

Typhimurium LPS antibody 1E6 (Thermo Fisher Scientific Inc.; Mississauga, ON – cat. no. MA1-83451) used at a dilution of 1:500 in TSBT, or purified mouse IgG1 isotypic control in TBST (Cedarlane Labs;

Burlington, ON - cat. no. CLX500AP) matching the concentration of the primary antibody, or just TBST for the negative control. At this step, an additional positive control consisting of an S.Tm culture smear on a glass slide (generously provided by Dr. Nancy Martin's laboratory, Queen's University) was added and incubated with antibody, similar to tissue section slides. The following morning, sections were washed thrice with TBST and incubated with peroxidase-conjugated AffiniPure goat anti-mouse IgG

(H+L) (Jackson Immuno Research Labs – cat. no. 115-035-062) at a dilution of 1:1000 in TBST at room temperature for 60 minutes. Sections were washed thrice with TBST then incubated with 3,3'

Diaminobenzidine (DAB) chromogen (Abcam Inc.; Toronto, ON – cat. no. ab64238) for three minutes.

Slides were rinsed with two changes of ddH2O and immersed in freshly filtered Harris's hematoxylin

(Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) for two minutes. Slides were immediately rinsed under running tap water for five minutes, then rehydrated in 70%, 95% and 100% ethanol for five minutes each, and finally cleared with two changes of xylene for three minutes each.

Slides were mounted with glass coverslips and Clarion™ Mounting Medium (Sigma-Aldrich Canada Co.;

Oakville, ON - cat. no. C0487). All protocol details are included in Appendix B.

For S.Tm bacteria identification, two mid-placental sections/IS were scanned at 400X magnification using a Zeiss M1 Imager microscope (Zeiss; Toronto, ON) equipped with Axiovision 4.8 software. Immunoreactive bacterial numbers and their regions of placental localization were scored for each IS from all S.Tm-infected dams at 48h and 72h post-infection. I developed and used the following

33 semi-quantitative S.Tm bacterial burden scoring system to obtain an ordinal value to describe progression in severity:

Table 5. S.Tm bacterial burden scoring system. Score # S.Tm+ foci per implantation site Description 0 0 Non-infected 1 1-10 Minimally infected 2 11-20 Mildly infected 3 21-50 Moderately infected 4 51+ Severely infected

2.7.1 S.Tm Immunofluorescence

S.Tm immunofluorescence staining was conducted similarly to the S.Tm IHC protocol as described above for the first day, with omission of the peroxidase-blocking step. After overnight incubation with primary antibodies, slides were washed thrice with TBST and incubated with Alexa

Fluor® 594 conjugated, goat anti-mouse IgG (H+L) secondary antibody (Life Technologies Inc.,

Invitrogen; Burlington, ON - cat. no. A-11005) at a dilution of 1:200 in TBST at room temperature for 60 minutes. Slides were washed thrice with TBST then mounted with glass coverslips and ProLong® Gold

Antifade Mountant with DAPI (Thermo Fisher Scientific Inc., Molecular Probes™ - cat. no. P36931). All protocol details are included in Appendix B.

2.8 Histopathology & morphometric measurements

2.8.1 Routine histopathology with Hematoxylin and Eosin (H&E) staining

Some slides were stained with hematoxylin (nucleic acids stained blue) and eosin (proteins stained pink) (H&E, see Appendix C) and were examined by light microscopy for histological differences. More specifically, tissues were examined for classic signs of cell death and tissue damage such as presence of apoptotic bodies, nuclear fragmentation (pyknosis), necrotic regions and cellular 34 debris. Tissues were also examined for cardinal signs of acute inflammation such as vessel vasodilation, edema, and leukocyte infiltration.

2.8.2 Active caspase-3 immunohistochemistry for detection of cell death

Immunostaining for the apoptotic cell death pathway protein activated caspase-3 was used for localization of dying cells. Serially sectioned slides that included internal positive controls (dying uterine natural killer cells in GD 15 IS sections (130,131)) were deparaffinized, rehydrated, and underwent sodium citrate (pH 6.0) HIER as described in section 2.7. Peroxidase blocking was conducted as in section 1.7 after which tissue sections were washed twice with TBST and incubated with CAS-Block™ histochemical reagent (Thermo Fisher Scientific Inc - cat. no. 008120) for 10 minutes at room temperature. Sections were incubated overnight at room temperature with polyclonal rabbit anti-activated caspase-3 antibody (Bioss Inc.; Woburn, MA, USA - cat. no. bs-2095R) at a dilution of 1:200 in CAS-

Block™, or rabbit IgG isotype control (Bioss Inc. – cat. no bs-0295P) matching the concentration of the primary antibody in CAS-Block™, or just CAS-Block™ for the negative control. The following morning, sections were washed thrice with TBST and incubated with biotinylated polyclonal goat anti-rabbit immunoglobulins (Dako Canada ULC; Mississauga, ON – cat. no. E0432) at a dilution of 1:500 in TBST at room temperature for 60 minutes. Sections were washed thrice with TBST and incubated with

ExtrAvidin®-Peroxidase (Sigma-Aldrich Canada Co. - cat. no. E2886) at a dilution of 1:100 in TBST at room temperature for 30 minutes. Sections were washed thrice with TBST then incubated with DAB chromogen for three minutes. Slides were rinsed and then stained using freshly filtered Harris's hematoxylin as described in section 2.7. Slides were rinsed, rehydrated, cleared, and coverslipped as described in section 2.7. All protocol details are included in Appendix D.

2.8.3 Morphometric measurements of implantation sites

To evaluate potential changes in the cross-sectional areas of IS or their major subcomponents following S.Tm infection, mid-sagittal placental sections from apparently viable IS were stained with

Periodic acid-Schiff (PAS, see Appendix E). Blinded cross-sectional area measurements for decidua 35 basalis, junctional zone (spongiotrophoblast), and placental labyrinth were made on images photographed at 25X magnification using one mid-placental section/IS and ImageJ (version 1.48) software (U. S.

National Institutes of Health; Bethesda, MD, USA).

2.9 Immunohistochemistry for leukocytes

2.9.1 CD45

An antibody for CD45, a pan-leukocyte marker, was used to enumerate and compare numbers of

CD45+ leukocytes in IS. Serially sectioned and positive control (sections of normal mouse spleen) slides were deparaffinized, rehydrated, and underwent sodium citrate (pH 6.0) HIER as described in section 2.7.

Peroxidase blocking was conducted as in section 2.7 after which tissue sections were washed twice with

TBST and incubated with 5% bovine serum albumin (BSA) in TBST at room temperature for 30 minutes.

Sections were incubated overnight at 4°C with monoclonal, biotin rat anti-mouse CD45 antibody

(BioLegend; San Diego, CA, USA – cat. no. 103104) at a dilution of 1:100 in TBST, or biotin rat IgG2b kappa isotype control (BioLegend – cat. no. 400603) matching the concentration of the primary antibody in TBST, or just TBST for the negative control. The following morning, sections were washed, and then incubated with ExtrAvidin®-Peroxidase as described in section 2.8.2. Sections were washed thrice with

TBST then incubated with DAB chromogen for five minutes. Slides were rinsed and then stained using freshly filtered Harris's hematoxylin as described in section 2.7. Slides were rinsed, rehydrated, cleared, and coverslipped as described in section 2.7. All protocol details are included in Appendix F. IHC- positive CD45 cells were counted using blind-encoded slides from 6-8 fields of view per 1 section/placenta and are presented as mean numbers of CD45+ cells per 400X field, in each placental area and total placental tissue.

2.9.2 Neutrophil Ly6G

An anti-Ly6G antibody was used to localize neutrophils, and their locations were compared descriptively. Serially sectioned and positive control (spleen and liver tissue sections from S.Tm-infected

36 animals) slides were deparaffinized and rehydrated as described in section 2.7. Slides were then rinsed with ddH2O for two minutes after which enzymatic antigen retrieval was performed using Proteinase K

(Sigma-Aldrich Canada Co. - cat. no. P6656) at room temperature for five minutes. Sections were washed twice with TBST, blocked for endogenous peroxidase activity, and serum blocked as per section 2.8.2.

Sections were incubated overnight at 4°C with monoclonal, purified rat anti-mouse Ly-6G monoclonal antibody (BD Pharmingen™, BD Biosciences – cat.no. 551459) at a dilution of 1:200 in CAS-Block™, or purified rat IgG2a kappa isotype control (BioLegend – cat. no. 400502) matching the concentration of the primary antibody in CAS-Block™, or just CAS-Block™ for the negative control. The following morning, sections were washed thrice with TBST and incubated with biotinylated polyclonal rabbit anti- rat immunoglobulins (Dako Canada ULC – cat. no. E0468) at a dilution of 1:1000 in TBST at room temperature for 60 minutes. Sections were washed, and then incubated with ExtrAvidin®-Peroxidase and

DAB chromogen as described in section 2.9.1. Slides were rinsed and then stained using freshly filtered

Harris's hematoxylin as described in section 2.7. Slides were rinsed, rehydrated, cleared, and coverslipped as per section 2.7. All protocol details are included in Appendix G.

2.9.3 Macrophage F4/80

An anti-F4/80 antibody was used to identify macrophages, and their regional distributions. These observations are reported descriptively. Serially sectioned and positive control (liver tissue sections) slides were deparaffinized, rehydrated, and underwent enzymatic antigen retrieval as described in section

2.9.2. Sections were washed twice with TBST, blocked for endogenous peroxidase activity, and serum blocked as described in section 2.8.2. Sections were incubated overnight at 4°C with monoclonal, biotin anti-mouse macrophage (F4/80) antibody (Cedarlane Labs – cat. no. CL8940B-3) at a dilution of 1:200 in

CAS-Block™, or biotin rat IgG2b kappa isotype control matching the concentration of the primary antibody in CAS-Block™, or just CAS-Block™ for the negative control. The following morning, sections were washed, and then incubated with ExtrAvidin®-Peroxidase and DAB chromogen as described in section 2.9.1. Slides were then rinsed and stained using freshly filtered Harris's hematoxylin as described

37 in section 2.7. Slides were rinsed, rehydrated, cleared, and coverslipped as per section 2.7. All protocol details are included in Appendix H.

2.10 Serum Analyses

2.10.1 Lipopolysaccharide endotoxin quantification

Serum lipopolysaccharide (LPS) endotoxin concentrations were estimated using a lipopolysaccharide enzyme-linked immunosorbent assay (ELISA) (Abbexa Ltd., Cambridge, UK - cat.no. abx150357) according to manufacturer's instructions. Note: the Pierce™ LAL (Limulus Amebocyte

Lysate) Chromogenic Endotoxin Quantitation Kit (Thermo Fisher Scientific Inc. - cat. no. 88282) was also employed and found to be unsuitable for mouse biological samples, despite advertisement for this purpose.

2.10.2 Progesterone

Serum progesterone concentrations were estimated using a mouse/rat progesterone ELISA

(ALPCO Diagnostics; Salem, NH, USA - cat.no. 55-PROMS-E01) according to manufacturer's instructions.

2.10.3 Maternal cytokines, chemokines, and growth factors

Maternal serum was thawed, diluted in PBS at Queen's University and shipped on dry ice to Eve

Technologies (Eve Technologies Corp, Calgary, AB, Canada) for quantification of cytokine, chemokine, or growth factor biomarkers using the Mouse Cytokine Array/Chemokine Array 32-Plex Discovery

Assay®. This multiplex assay uses the Bio-Plex™ 200 system (Bio-Rad Laboratories, Inc., Hercules, CA,

USA), and a Milliplex Mouse Cytokine/Chemokine kit (Millipore, St. Charles, MO, USA). The analytes evaluated were Eotaxin, G-CSF (granulocyte-colony stimulating factor), GM-CSF (granulocyte- macrophage colony-stimulating factor), IFN-γ (interferon gamma), IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5,

IL-6, IL-7, IL-9, IL-10, IL-12 (p40), IL-12 (p70), IL-13, IL-15, IL-17, LIF (leukemia inhibitory factor),

CSF1 (colony stimulating factor 1), CXCL1, CXCL2, CXCL5, CXCL9, CXCL10, CCL2, CCL3, CCL4,

CCL5, TNF-α (tumor necrosis factor alpha), and VEGF (vascular endothelial growth factor). The assay sensitivities for these markers ranged from 0.1 – 33.3 pg/mL.

2.11 Statistical Analysis

One-way ANOVAs followed by Tukey's multiple comparisons test, and the non-parametric Mann

Whitney U test were used as appropriate (stated in corresponding figure legends). Statistical significance is specified as: *: p<0.05, **: p<0.01, ***: p<0.005, ****: p<0.001.

Chapter 3

Results

3.1 Fetal Resorption

Mean numbers of implantation sites (IS) in naïve pregnant mice at GD 14 and GD 15 (N=3 per group) did not differ between Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice (Table 6). These litter sizes also did not differ from the mean number of viable pups born to naïve mothers (Table 6; data extracted from historical breeding colony records; N=33 litters from nine Ripk3-/-Slc11a1+/+dams; N=64 litters from 18

Slc11a1+/+ dams). Further, there were no differences between the genotypes in the percentage of viable litters at birth, mean number of litters per dam, or percentage of pups lost before weaning (data not shown).

Table 6. Litter sizes were similar amongst Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice at mid-gestation and birth. GD 14 GD 15 Birth

Ripk3-/-Slc11a1+/+ 8.7 ± 1.5* 9.3 ± 0.6* 6.6 ± 2.8æ

Slc11a1+/+ 8.7 ± 1.2* 9.3 ± 0.6* 6.6 ± 2.5í

Ripk3-/-Slc11a1+/+ 8.5 ± 7.5%* 14.8 ± 16.9%* - Mean Resorption rate (%) Slc11a1+/+ 3.3 ± 5.8%* 7.0 ± 6.1%*

Data are presented as mean ± SD. *N=3; æN=33 litters, n=9 dams; íN=64 litters, n=18 dams.

Deletion of Ripk3 did not alter fetal resorption rates in Slc11a1+/+ dams at either 48h or 72h after S.Tm infection (Fig. 5).

60 Infected(Ripk3&/&Slc11a1⁺/⁺. Naive(Ripk3&/&Slc11a1⁺/⁺. 40 Infected(Slc11a1⁺/⁺. Naive(Slc11a1⁺/⁺. 20

%,Fetal,resorp5on/mouse 0

48h 72h Time,post1infec5on

Figure 5. Ripk3 deletion does not alter fetal resorption rates in the first 72h after S.Tm infection. Fetal resorption rates were not significantly different between infected and naïve mice regardless of Ripk3 deletion. N=3 dams per group/per time point. Data are presented as mean ± SD.

3.2 Study Animals

All animals used in this study are included in Table 7. A unique Dam ID# was given to each dam in this thesis. Details include numbers of viable and resorbed IS, percent of infected IS/dam and number of slides cut for analyses.

Table 7. List study animals showing numbers of implantation sites (IS) processed, infection rates within litters, total number IS sites and number of slides cut/dam.

Genotype Gestational Infection Dam ID # # Viable # Resorbed % Fetal % Infected IS # IS cut Total # slides cut [10(V) + Day Status IS (V) IS (R) resorption (S.Tm+) 5(R) for infected mice]

Ripk3-/-Slc11a1+/+ 14 Infected 3532 8 0 0 87.5 8 80

3533 2 2 50 50 4 30

3567 8 2 25 50 10 90

Naïve 3450 6 1 14.3 - 2 20

3472 8 1 11.1 - 2 20

3505 10 0 0 - 2 20

Slc11a1+/+ 14 Infected 3480 3 1 25 50 4 35

3481 9 0 0 44.4 9 90

3484 9 1 10 50 9 95

Naïve 3449 8 0 0 - 2 20

3463 8 0 0 - 2 20

3471 10 1 9.1 - 2 20

Ripk3-/-Slc11a1+/+ 15 Infected 3714 8 1 11.1 55.6 9 85

3715* 11 0 0 81.8 11 110

3799 9 0 0 88.9 9 90

Naïve 3698 6 3 33.3 - 2 20

3733 8 1 11.1 - 2 20

3735 10 0 0 - 2 20

Slc11a1+/+ 15 Infected 3588 8 2 20 70 10 90

3589 5 0 0 60 5 50

3590 5 0 0 100 5 50

Naïve 3667 9 1 10 - 2 20

3668 8 1 11.1 - 2 20

3669 9 0 0 - 2 20

*Denotes the heavily infected Ripk3-/-Slc11a1+/+ (dam #3715) referred to in the results section. Total number of slides cut for this study: 1135 (3 sections/slide therefore total of 3405 sections) IS, implantation site; V, viable; R, resorbed.

3.3 Fetal impacts of Salmonella enterica serovar Typhimurium infection during pregnancy

3.3.1 Fetal:placental weight ratios

Since my collaborators reported rapid transplacental infection and fetal liver colonization within

14h of S.Tm infection i.v. (155), I sought to examine whether fetal weights were affected. Analysis of fetal:placental weight ratios of apparently viable fetuses showed no significant differences with regards to infection or Ripk3 gene status at 72h post-infection (Fig. 6A). There were no differences when fetal weight only was considered, including the fetuses from Ripk3-/-Slc11a1+/+ dam #3715 who was subsequently found to be more heavily infected than all other infected mice provided to Queen’s

University that had received S.Tm (see Fig. 8D & E, 31 - 34). For infected Slc11a1+/+ dams only, variability in fetal size was striking upon gross post-mortem examination of fixed, shipped specimens

(Fig. 6B). Although slight variability in fetal weight and size occurred within litters, large variability in fetal weights occurred in genetically-matched, replicate dams. Preliminary histological analysis of H&E stained sagittal fetal sections (2 fetuses/Slc11a1+/+ dam at GD 15) showed profound differences in size when compared to representative fetuses from naïve and other infected dams (Fig. 6C). Fetuses from

Slc11a1+/+ dam #3588 at 72h post-infection were one day developmentally delayed at GD 15. A representative fetus from dam #3588 illustrated is at Theiler stage 23, equivalent to GD 14 (Fig. 6Cii).

Fetuses from another genetically-matched replicate infected dam (#3589) were two days developmentally delayed, at Theiler stage 22, equivalent to GD 13 (Fig. 6Ciii). Preliminary analyses suggest that fetal growth and subsequently size may be affected in infected Slc11a1+/+ pregnancies, since there were no indications of significant cell death in these sections compared to fetuses from naïve mice. Fetal weights and fetal:placental weight ratios were not different at 48h post-infection in mice with or without Ripk3 compared to their respective naïve controls (data not shown). Because there were no striking gross post- mortem differences between infected Ripk3-/-Slc11a1+/+ mice at 72h post-infection and their GD 15 naïve controls, these fetuses were not examined histologically for delays in development.

Figure 6. Fetal:placental weight ratios were not significantly altered at 72h post-infection but infected mice with intact Ripk3 displayed heterogeneous developmental delay. Ratios of fetal weight to placental weight did not differ statistically based on infection or Ripk3 gene status (A). Each data point represents an individual placenta (n numbers specified per group) of N=3 dams/genotype/group. Considerable variability was exhibited amongst infected Slc11a1+/+ mice noted upon gross post-mortem examination. Panel (B) illustrates a representative from each replicate dam (refer to Table 7 for dam number-specific details). Hematoxylin and eosin stained-sagittal histological sections of fetuses from these litters revealed not only striking differences in sizes but also in developmental stages of fetuses from infected compared to naïve dams (2 fetuses sectioned/Slc11a1+/+ dam) (C). (Ci) shows a fetus from a naïve GD 15 Slc11a1+/+ dam, corresponding to Theiler stage 24 (192), equivalent to GD 15 typical for normal development. (Cii) shows a Theiler stage 23 fetus from infected Slc11a1+/+ dam #3588, delayed developmentally by one day i.e. equivalent to GD 14. (Ciii) shows a Theiler Stage 22 fetus from infected Slc11a1+/+ dam #3589 delayed two days developmentally i.e. equivalent to GD 13. Although fetuses appeared smaller in size, preliminary histopathological analysis of revealed limited cell death (absence of apoptotic bodies or pyknotic cells and anti-active caspase-3 immunostaining) in these fetuses (data not shown). All data presented as mean ± SD. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. Inf., infected.

3.3.2 Placental morphometric measurements

My collaborators reported that TBCs are damaged as a result of S.Tm infection at mid-pregnancy in mice. Therefore, detailed assessment of the placenta and decidua of IS was undertaken. The initial studies involved morphometric estimation of the mid-sagittal surface areas of the decidua basalis (DB), junctional zone (JZ) and labyrinth of IS collected from naïve and infected dams of both genotypes. These studies revealed that JZ surface areas were significantly decreased in infected Ripk3-/-Slc11a1+/+ versus infected Slc11a1+/+ mice at 72h post-infection (Fig. 7C, p=0.0002). JZ surface areas of Ripk3-/-Slc11a1+/+ did not differ to naïve mice whether or not naïve mice were deleted for Ripk3. Infected Ripk3-/-Slc11a1+/+ had labyrinth surface areas significantly larger than infected Slc11a1+/+ mice (p=<0.0001) but there were no differences in labyrinth surface areas compared to naïve labyrinth surface areas of either genotype

(Fig. 7D). DB and total placental surface areas were not altered in mice having differing Ripk3 gene status or differing infectious challenge status (Fig. 7B & E). Interestingly, when morphometric measurements were pooled and averaged per litter, no significant differences were found with regards to infection or

Ripk3 gene status (Appendix I).

Figure 7. Placental morphometric measurements analyzed per implantation site. (A) Photomicrograph of a representative GD 15 naïve Ripk3-/-Slc11a1+/+ Periodic-acid Schiff-stained placental section. Yellow lines demarcate the areas of decidua basalis (DB), junctional zone (JZ), and labyrinth (Lab) measured using ImageJ. No differences in surface areas of DB or total placental area were found amongst all four groups (B & E). JZ surface area was significantly decreased (p=0.0002) in infected Ripk3-/-Slc11a1+/+ versus infected Slc11a1+/+ and surface areas did not differ compared to naïve mice (C). Labyrinth areas were significantly decreased (p=<0.0001) in infected Slc11a1+/+ mice versus Ripk3-/-Slc11a1+/+ mice, and there were no differences compared to naïve mice (D). Each data point represents an individual placenta (n=28 for infected Ripk3-/-Slc11a1+/+, n=18 for infected Slc11a1+/+, n=6 for both naïve groups) from N=3 dams/genotype/group. One-way ANOVA followed by Tukey’s post-test for multiple comparisons was used for statistical comparisons. All data are presented as mean ± SD. Inf, infected. Appendix I presents the same data set grouped by litter means, which masked the statistical differences illustrated in Figure 7.

3.4 Placental Salmonella enterica serovar Typhimurium Localization by Immunohistochemistry

3.4.1 Placental localization of Salmonella enterica serovar Typhimurium foci and bacterial burden estimates

To identify infected IS, an immunohistochemical approach was chosen to study all IS within each litter. A number of commercially available antibodies to bacterial cell wall components

(lipopolysaccharide, LPS) were assessed for their usefulness on hepatic sections from mice with known infections and for staining of glass microscope slides streaked with cultured S.Tm. The rabbit polyclonal anti-Salmonella antibody from Abcam Inc. (cat.no. ab156656) was found to be unsuitable because it stained organisms present in IS of naïve or non-challenged mice. This reactivity was considered to be recognition of normal microbial flora. The mouse monoclonal Salmonella Typhimurium LPS antibody

IE6 from Thermo Fisher Scientific Inc. (cat. no. MA1-83451) was selected for use in this thesis because of its specificity for the recognition S.Tm LPS, and lack of cross-reactivity to endotoxin of other

Enterobacteriaceae species. Currently, this is the only commercially available antibody specific to S.Tm appropriate for immunohistochemistry using formalin-fixed, paraffin-embedded tissues. Since this antibody is murine in origin, additional optimization and blocking steps were undertaken to eliminate background or non-specific secondary antibody staining to endogenous mouse IgG1 and Fc receptors.

S.Tm+ signals were only detected in tissue sections from S.Tm-challenged mice and not naïve mice.

Further, in infected mouse sections antibody only stained positive for S.Tm while isotype and secondary antibody alone did not. Representative images of positive control liver tissue sections are shown in

Appendix J.

Using the Salmonella Typhimurium LPS antibody IE6 (Thermo Fisher Scientific Inc., cat. no.

MA1-83451), S.Tm were detected in placental tissues of challenged and not naïve Ripk3-/-Slc11a1+/+and

Slc11a1+/+ mice using immunohistochemistry. Therefore, S.Tm+ signals represented true endotoxin signal from S.Tm and not endotoxin from commensal bacteria. Immunoreactivity was detected over structures with heterogeneous morphologies that were present in placental tissues in differing amounts. 50

S.Tm+ structures were small (~1-2 µm diameter), round to longish in structure and not uniformly stained.

Staining intensities ranged from brown to dark brown (Fig. 8A & B). S.Tm+ signals appeared to reflect either single bacterium (Fig. 8A) or small-to-large sized bacterial clusters (Fig. 8B & C). Clusters of

S.Tm+ foci were infrequent within decidual tissues of IS with mild bacterial burdens, while frequent in IS with the heaviest bacterial burdens.

A single bacterium or a bacterial cluster was enumerated as one S.Tm+ focus. The majority of

S.Tm+ foci were extracellular, being located within blood vessels or spaces of tissues that had a healthy appearance at 48h and 72h post-infection (Fig. 8B & C). Rarely, S.Tm+ foci were found intracellularly within leukocytes that morphologically resembled neutrophils, decidual cells, or spongiotrophoblast. Only within the most heavily burdened placentas from an individual infected Ripk3-/-Slc11a1+/+ mouse (dam

#3715; 72h post-infection) was IS pathology apparent (Fig. 8E). Intracellular S.Tm+ foci were more common in placental tissue sections from infected Ripk3-/-Slc11a1+/+ than infected Slc11a1+/+ mice.

S.Tm+ foci were not identified in resorbed IS, but these sites were included in all S.Tm identification analyses except for placental area localization because placental regions were indistinguishable in the sections of resorbing tissue.

I developed a scoring system for placental burden (Table 8) that ranged from zero (no detectable bacteria) to four (>50 bacteria up to too numerous to count). Only IS from infected Ripk3-/-Slc11a1+/+ mice received the highest bacterial burden score of four (severely infected) using my burden scoring system (Fig. 8G). Bacterial burden scores increased significantly within a genotype over time (48h versus

72h post-infection, p=0.0437 for placentas from Ripk3-/-Slc11a1+/+ and p=0.0190 for placentas from

Slc11a1+/+ mice). However, bacterial burden scores did not differ significantly between genotypes at 48h and 72h post-infection.

Table 8. S.Tm bacterial burden scoring system. Score # S.Tm+ foci per implantation site Description 0 0 Non-infected 1 1-10 Minimally infected 2 11-20 Mildly infected 3 21-50 Moderately infected 4 51+ Severely infected

Figure 8. Placental S.Tm immunohistochemistry and bacterial burden scores. Using an anti-S.Tm LPS antibody, S.Tm+ foci appeared as roundish to longish, brown to dark-brown stained structures either as single bacterium (A) or small to large clusters (B & C). Inset panel scale bar, 25µm. The majority of S.Tm+ foci were identified within vessels or other spaces of morphologically healthy tissues (B & C). S.Tm+ intracellular foci appeared within leukocytes (morphologically resembling neutrophils) and decidual cells (D), or in areas of the junctional zone containing cells of spongiotrophoblast (while devoid in trophoblast giant cells) in the most heaviest burdened placentas from a single infected Ripk3-/-Slc11a1+/+ mouse (dam #3715) at 72h post-infection, (F) is a photomicrograph of an isotype control corresponding to (E). S.Tm bacterial burden scores (0 – no foci, 1 – 1-10 foci, 2 – 11- 20 foci, 3 – 21-50 foci, 4 – 51+ foci) increased significantly between 48h and 72h post-infection within both genotypes (p=0.0437 for Ripk3-/-Slc11a1+/+ and p=0.0190 for Slc11a1+/+ using the Mann-Whitney test for comparisons), but were not significantly different between genotypes at either time point (G). Each data point represents an individual placenta (numbers specified for each group, includes resorbed IS) from n=3 dams/genotype/time point. Sections were counterstained with hematoxylin. All images are from tissues collected at GD 15, 72h post-infection. All scale bars indicated on individual images. DB, decidua basalis; JZ, junctional zone.

3.4.2 Frequency of S.Tm+ placentas and distribution

In all challenged mice at least one IS was S.Tm+ (see Table 7). IS-infected litter members ranged from 4/9 (44.4%) to 5/5 (100%). It was exceptionally rare to find every IS within a litter S.Tm+.

Percentages of IS with identifiable S.Tm+ foci did not differ between Ripk3-/-Slc11a1+/+ versus Slc11a1+/+ litters at either at 48h or 72h post-infection (Fig. 9A & B). S.Tm+ foci were predominately localized to the DB and placental labyrinth tissues. Relatively fewer signals were identified in the placental JZ (Fig.

9C & D). The sites of bacterial localization did not differ significantly over the course of infection, or with respect to Ripk3 gene status. Representative photomicrographs of S.Tm+ foci placental localization are presented in Figure 10. No S.Tm+ foci were identified in placentas from gestational age-matched naïve dams of either genotype (Fig. 11). One placenta out of 11 from Ripk3-/-Slc11a1+/+ dam #3715 was excluded from this analysis because the number S.Tm+ foci was too numerous to count (Fig. 8E).

IS from mice sacrificed at GD 14, 48h after S.Tm infection showed some antibody reactive S.Tm placental infection, but S.Tm were less common and not accompanied by any apparent placental pathology. Therefore, the subsequent studies undertaken for this thesis used 72h post-infection.

A B Ripk3&/&Slc11a1⁺/⁺ Slc11a1⁺/⁺

100 100

80 80

60 60

40 40 %"infected"placentas 20 %"infected"placentas 20

0 0

48h 72h 48h 72h

C D Slc11a1⁺/⁺ &/& ⁺/⁺ 100 ⁺/⁺ Ripk3 Slc11a1 Slc11a1 Labyrinth 43.10 36.71 100 80 100 Junc.onal(Zone Labyrinth Labyrinth Decidua(Basalis 80 33.02 41.00 60 43.10 36.71 Junc.onal(Zone80 17.87 Junc.onal(Zone Decidua(Basalis20.69 Decidua(Basalis 60 40 60 28.30 16.34 17.87 %%S.Tm+%foci/area 20.69 40 20 40 36.21 45.41 %%S.Tm+%foci/area 0 %%S.Tm+%foci/area 20 38.68 42.66 20 36.21 45.41 48h 72h 0 0 Time%post*infec.on

48h 72h 48h 72h Time%post*infec.on Time%post*infec.on

Figure 9. Percentages of implantation sites infected and bacterial localization were not significantly altered by presence or absence of Ripk3. The percentage of IS with S.Tm+ foci detection by IHC was not significantly increased from 48h to 72h post-infection in either Ripk3-/-Slc11a1+/+ (A) or Slc11a1+/+ (B) mice, and were similar between both genotypes at both time points after infection (each data point represents an individual animal from N=3 dams/genotype/group). (C & D) Foci were identified predominately within decidua and the placental labyrinth. Fewer foci were detected in the junctional zone at either 48h or 72h post-infection. Ripk3 deletion did not impact the sites of S.Tm localization. These analyses excluded resorbed IS in which tissue regions were no longer distinguishable.

Figure 10. Placental S.Tm+ foci in Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice. Representative photomicrographs of placental tissues immunostained with an anti-S.Tm LPS antibody show S.Tm+ foci in placental areas (circled in red) including decidua basalis (A & B), junctional zone (C & D), and labyrinth tissues (E & F). Representative isotype control images of labyrinth tissues (G & H). Left panel shows photomicrographs from Ripk3-/-Slc11a1+/+; right panel shows photomicrographs from Slc11a1+/+. All images are from tissues collected at GD 15, 72h post-infection. Sections were counterstained with hematoxylin. All scale bars, 50 µm.

Figure 11. S.Tm+ foci were not found in placentas of gestational-age matched naïve mice. Representative photomicrographs of placental tissues immunostained with an anti-S.Tm LPS antibody show placental areas including decidua basalis (A & B), junctional zone (C & D), and labyrinth tissues (E & F) with no detectable S.Tm+ foci. Representative isotype control images of labyrinth tissues (G & H). Left panel shows photomicrographs from Ripk3-/-Slc11a1+/+; right panel shows photomicrographs from Slc11a1+/+. All images are from tissues collected at GD 15, 72h post-infection. Sections were counterstained with hematoxylin. All images are from GD 15 mice. All scale bars, 50 µm.

3.5 Histopathology

3.5.1 Histopathology using Hematoxylin and Eosin staining

Histopathological examination of tissue sections from all IS of two out of the three available infected Ripk3-/-Slc11a1+/+ mice revealed limited placental damage at 72h post-infection (Fig. 12). The third replicate mouse (dam #3715) was distinct and is discussed below separately. In comparisons to the

IS from the three naïve Ripk3-/-Slc11a1+/+ females, low power scanning (25X magnification) of the IS from the two infected Ripk3-/-Slc11a1+/+ that appeared similar, identified limited differences confined to the JZ, which appeared to have more compacted cells and reduced blood sinusoidal space (inferred from less white space) (Fig. 12A & B). Higher power examination of the DB tissue revealed subtle microfoci of apoptosis. However, these apoptotic microfoci were also present in the DB of IS from naïve dams (Fig.

12C-F). Note: JZ spongiotrophoblast compaction can also be seen in these images). Subtle microfoci of apoptosis were also detected in the labyrinth of both infected and naïve dams (Fig. 12G & H).

Figure 12. Histopathological examination of implantation sites from the two less infected Ripk3-/- Slc11a1+/+ revealed no differences compared to naïve Ripk3-/-Slc11a1+/+. Representative low power photomicrographs of the entire placenta (A & B) show minimal placental differences except in the junctional zone (JZ) infected Ripk3-/-Slc11a1+/+. This area appeared more consolidated and less vascular (appearance of white space) than placentas from naïve animals. Inspection of infected Ripk3-/-Slc11a1+/+ decidua basalis (DB) at higher power magnifications visualized subtle microfoci of apoptosis with the presence of apoptotic bodies (indicated by black arrows and magnified in inset panels, scale bar = 50 µm). This however could also be detected in naïve DB, with more compact junctional zone tissue (C-F). Examination of labyrinthine (Lab) region revealed microfoci of apoptosis (black arrows) detectable in tissues from both infected and naïve mice (G & H). Necrosis was not observed in any part of the IS and the IS were judged to be healthy in appearance. All images are from GD 15 mice. Sections were stained with hematoxylin and eosin. Scale bars are indicated in individual images.

Profound placental damage and immune infiltration was observed in IS from one of the three infected Ripk3-/-Slc11a1+/+ dams (dam #3715; Fig. 13). Low power microscopic examination of the entire placenta (25X magnification) identified an immense leukocyte infiltration of the DB and JZ in 2/11 IS.

Heavy staining for hemosiderin (attributed to the presence of dilated and congested blood vessels) was observed throughout these 2 entire placentas, including in the labyrinthine maternal-fetal exchange area.

Higher power examination of DB and JZ areas showed profound tissue damage, and granulocyte infiltration (Fig. 13B-D). In these 2 IS, labyrinth TBCs had pyknotic nuclei and appeared to be degenerating. Labyrinths also had extensive vascular congestion and vasodilation with disrupted tissue architecture (Fig. 13E). Indeed, labryinthine vascular congestion was observed in 7/11 IS from this dam

(Fig. 15A). Vascular congestion was not observed in the labyrinthine tissue sections from the other two

Ripk3-/-Slc11a1+/+ mice (Fig. 15B).

Sections from infected and naïve Slc11a1+/+ mice (Fig. 14) did not differ from those of the two infected Ripk3-/-Slc11a1+/+ and three naïve Ripk3-/-Slc11a1+/+ presented in Figure 12.

Figure 13. Histopathological examination of implantation sites from one infected Ripk3-/-Slc11a1+/+ (dam #3715) revealed major differences compared to placentas from naïve Ripk3-/-Slc11a1+/+. Photomicrographs of H&E-stained sections from 2/11 IS showing the most severely inflamed and damaged phenotype. The other 9/11 IS from dam #3715 showed varying levels of placental damage less severe than those displayed here (data not shown). Lower power examination of entire placentas showed profound placental damage (A). Decidua basalis (DB) and junctional zone (JZ) were highly cellular due to an immense immune cell infiltration (indicated by black arrows). Heavy staining for hemosiderin indicative of vascular congestion could be seen in all areas of placenta, more specifically concentrated in the labyrinth (Lab) region (indicated by asterisks) (A-E). Higher power examinations of DB and JZ tissues showed profound necrosis and infiltration of granulocytes (black arrows) within these areas (B-D). Examination of labyrinthine tissues revealed intense vascular congestion with immense vasodilation of labyrinthine vessels, with disrupted architecture (H). Labyrinthine vascular congestion was noted in 7/11 IS from this individual dam. Photomicrograph in (H) was the most severely congested IS. Trophoblast cells in congested areas of the labyrinth had pyknotic nuclei, and appeared to be degenerating. From examination of H&E stained-section, tissue necrosis had not yet developed. All scale bars indicated in individual images.

Figure 14. Photomicrographs of implantation sites from infected Slc11a1+/+ that resembled placentas from naïve Slc11a1+/+. Representative low power photomicrographs of the entire placenta (A & B) show minimal placental differences except in the junctional zone (JZ) infected Slc11a1+/+. This area appeared more consolidated and less vascular (appearance of white space) than placentas from naïve animals. Inspection of infected Slc11a1+/+ decidua basalis (DB) at higher power magnifications visualized subtle microfoci of apoptosis with the presence of apoptotic bodies (indicated by black arrows and magnified in inset panels, scale bar = 50 µm). This however could also be detected in naïve DB, in regions with more compact junctional zone tissue (C-F). Examination of labyrinthine (Lab) regions detected microfoci of apoptosis (black arrows) detectable in tissues from both infected and naïve animals (G & H). Necrosis was not observed in any part of the IS and IS were judged to be healthy in appearance. Sections were stained with hematoxylin and eosin. Scale bars are indicated in individual images.

Figure 15. Labyrinthine vascular congestion was not observed in the other two infected Ripk3-/- Slc11a1+/+ or Slc11a1+/+ implantation sites. Panel (A) shows a representative photomicrograph of milder labyrinthine vascular congestion observed in 5/11 IS from infected Ripk3-/-Slc11a1+/+ dam #3715. Representative photomicrographs of hematoxylin and eosin stained placental labyrinthine sections of infected versus naïve Ripk3-/-Slc11a1+/+ showed no vascular congestion (B & C). Similarly, no labyrinthine vascular congestion was observed in tissue sections from infected versus naïve Slc11a1+/+ placentas (D & E). All scale bars, 50 µm. 69

3.5.2 Assessment of cell death by active caspase-3 immunohistochemistry

Because fetal loss and maternal humane endpoint are anticipated shortly after the 72h time-point examined by histopathology, I suspected infection-associated death of cells within the IS may have commenced but not reached levels detectable by routine histological screening. Therefore, immunostaining with anti-active caspase-3 was undertaken using all placentas from infected dams, and 2 placentas/naïve dam/genotype. In DB tissues, signals for active caspase-3 were observed in areas of subtle microfoci of apoptosis (identified previously by H&E staining), and were not different when comparing infection or Ripk3 gene status (Fig. 16A & B, Fig. 17A & B). In the labyrinth area, only trophoblast islands (isolated, round, dense networks of trophoblast cells) were active caspase-3 reactive but these did not differ when comparing infection or Ripk3 gene status (Fig. 16C & D, Fig. 17C & D).

Fetal membranes from infected Ripk3-/-Slc11a1+/+ mice were strongly reactive for active caspase-

3 staining compared to tissues from naïve mice (Fig. 16E & F), suggesting impending loss of fetal membrane integrity. Active caspase-3 staining was not observed in tissues from infected or naïve

Slc11a1+/+ mice, including fetal membranes (Fig. 17E & F). This suggests that Ripk3-associated gene products are protective in fetal membranes against S.Tm-associated caspase-3 activation.

Tissues from the severely infected and damaged placentas of Ripk3-/-Slc11a1+/+ dam #3715 were not reactive for active caspase-3 by IHC (data not shown), suggesting advanced cellular necrosis. Dying uNK cells served as internal positive controls (Appendix K) in each stained section.

Figure 16. Active caspase-3 immunohistochemistry of implantation sites from infected Ripk3-/- Slc11a1+/+ versus naive Ripk3-/-Slc11a1+/+ mice at GD 15. Representative photomicrographs show limited detection of active caspase-3. The IHC-positive signal (brown stain) in decidua basalis corresponded with areas of microfoci of apoptotic nuclei in both infected and naïve Ripk3-/-Slc11a1+/+ (A & B). In the labyrinth, weakly positive IHC signals were limited to trophoblast islands (infrequent dense, round structures containing trophoblast cells) in both infected and naïve Ripk3-/-Slc11a1+/+ mice (C & D). In fetal membranes of infected compared to naïve Ripk3-/- Slc11a1+/+ mice, significant active caspase-3 reactivity was detected (E & F). Representative isotype controls (G & H). Sections were counterstained with hematoxylin. All scale bars indicated in individual images.

Figure 17. Active caspase-3 immunohistochemistry of implantation sites from infected Slc11a1+/+ versus naive Slc11a1+/+ mice at GD 15. Representative photomicrographs show limited detection of active caspase-3 using an anti-active caspase- 3 antibody. The IHC-positive signal (brown stain) in decidua basalis corresponded with areas of microfoci of apoptotic nuclei in both infected and naïve Slc11a1+/+ (A & B). In the labyrinth, weakly positive IHC signals were limited to trophoblast islands (infrequent dense, round structures containing trophoblast cells) in both infected and naïve Slc11a1+/+ mice (C & D). Active caspase-3 reactivity did not differ in fetal membranes of infected compared to naïve Slc11a1+/+ mice (E & F). Representative isotype controls (G & H). Sections were counterstained with hematoxylin. All scale bars indicated in individual images.

3.6 Assessment of immune cell infiltration of implantation sites after intravenous maternal S.Tm challenge

3.6.1 CD45 Immunohistochemistry

To assess total immune cell infiltration into IS of dams challenged by S.Tm, immunohistochemical staining was undertaken using all IS from infected dams, and 2 placentas/naïve dam/genotype. CD45-reactive cells were then enumerated from 1 mid-sagittal section/IS by counting

CD45 IHC-positive cells from an area of approximately 1200 µm2/placental subregion (6 fields-of-view at

400X magnification). Blinded CD45+ cell counts revealed no significant differences in immune cell infiltration in the areas of DB, labyrinth, or the entire placenta (summed values of DB, JZ and labyrinthine areas) in infected versus naïve mice regardless of expression of the Ripk3 gene (Fig. 18A &

C-D). Very large variance was introduced into the decidual CD45+ cell counts in Ripk3-/-Slc11a1+/+ gene deleted mice due to inclusion of the IS from Ripk3-/-Slc11a1+/+ dam #3715 (Fig. 20) that were heavily leukocyte infiltrated. Two IS were identified as outliers (values of 179.33 and 45.83) using a ROUT

(Robust Regression and Outlier removal method offered by GraphPad Prism7 statistical software), a robust nonlinear regression method to identify outliers, but data was not re-analyzed to exclude these outliers.

Figure 18. CD45+ immunohistochemical cell counts in placental subregions and in the entire placenta. Mean CD45+ cell counts were not significantly different in the decidua basalis (A) and labyrinth areas (C), or total placenta (D) regardless of infection or Ripk3 gene status. Mean CD45+ cell counts were significantly higher in the infected Ripk3-/-Slc11a1+/+ junctional zone compared to infected Slc11a1+/+, p=0.0035 (B). However, CD45+ cell counts were not significantly altered in tissues from infected versus naïve animals regardless of Ripk3 deletion. Each data point represents an individual IS (n=28 for infected Ripk3-/-Slc11a1+/+, n=18 for infected Slc11a1+/+, while n=6 for both naïve groups) from N=3 animals/genotype/group. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. All data are presented as mean ± SD. Inf, infected.

The numbers of CD45+ cells in the JZs of infected Ripk3-/-Slc11a1+/+ mice were significantly higher (p=0.0035) than in the JZs of infected Slc11a1+/+ mice (Fig. 18B). This suggests Ripk3 gene product interactions normally dampen leukocyte recruitment and inflammatory responses within the placenta. CD45+ cell counts however, were not significantly higher in tissues from infected Ripk3-/-

Slc11a1+/+ compared to naïve Ripk3-/-Slc11a1+/+ or from infected Slc11a1+/+ compared to naïve

Slc11a1+/+ for the entire placenta or any placental subregions examined (Fig. 18). This indicates that a large sample size would be required to establish a role for Ripk3 gene product interactions during S.Tm IS infection. Representative photomicrographs further emphasize these finding in both Ripk3-/-Slc11a1+/+ and Slc11a1+/+ mice (Fig. 19 & 21, respectively), with exception to heavily infiltrated tissues of Ripk3-/-

Slc11a1+/+ dam #3715 (Fig. 16A & C). Labyrinth tissues from even the most severely infected placentas did not display significant immune cell infiltration, and CD45+ signals were predominately localized to leukocytes within maternal blood vessels (Fig. 20D).

The significant differences in leukocyte infiltration of IS regions when individual placentas were analyzed, disappeared upon pooling CD45+ cell counts per animal to normalize for differences in litter size (Appendix L).

Figure 19. CD45 immunohistochemistry in Ripk3-/-Slc11a1+/+ implantation sites comparing tissues from two of the three infected mice with limited pathology to tissues from naive controls. Images of representative photomicrographs taken at 400X magnification and used for CD45+ cell counting. CD45+ cells (identified using an anti-CD45 antibody; appearing as brown stain and indicated by black arrows) were scored in decidua basalis (A & B), in placental junctional zone (C & D), and in labyrinth (E & F). In all placental subregions, CD45+ leukocytes were intravascular and had not infiltrated into tissue. Representative image of an isotype control photomicrograph of decidua basalis from an infected dam (G). Isotype control-stained sections were used to validate cells included in enumeration for each region. Sections were counterstain with hematoxylin. All scale bars, 50 µm.

Figure 20. CD45 immunohistochemistry of placental tissues from heavily infected Ripk3-/-Slc11a1+/+ dam #3715. Representative photomicrographs taken at 400X magnification show the immune cell infiltration (CD45+ leukocytes identified using an anti-CD45 antibody; appearing as brown stain and indicated by black arrows in (C &D)) in tissues from the more severely infected Ripk3-/-Slc11a1+/+ dam (A & C). Decidua basalis (A) had a greater leukocyte burden than either placental region (C). The placental labyrinth had the fewest CD45+ cells (D) and these were not significantly elevated in number over CD45+ cells in the labyrinths of naïve controls of either genotype (compare with Fig. 15E & F). Most leukocytes were vessel-associated and appeared to be circulating (A, C, D). Representative image of an isotype control photomicrograph from decidua basalis of dam #3715 (B) corresponding to (A). Isotype control images were used to validate immunoreactive cells enumerated for each region (B). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

Figure 21. Implantation site CD45 immunohistochemistry from infected and naïve Slc11a1+/+ mice. Representative images of photomicrographs taken at 400x magnification that were used for CD45+ cell enumeration (CD45+ leukocytes identified using an anti-CD45 antibody; appearing as brown stain and indicated by black arrows) in decidua basalis (A & B), placental junctional zone (C & D), and labyrinth (E & F). In all placental subregions, CD45+ leukocytes were intravascular. Representative image of an isotype control photomicrograph from decidua basalis of a naïve GD 15 dam (G). Isotype controls were used to validate immunoreactive cells enumerated for each region. Sections were counterstained with hematoxylin. All scale bars, 50 µm.

3.6.2 Immunohistochemistry for neutrophil (anti-Ly6G)

Initial studies were undertaken to identify the lineages of CD45+ cells infiltrating IS 72h following maternal S.Tm challenge. To confirm the infiltration of neutrophils (recognized morphologically by their lobulated nuclei and size in dam #3715 in results sections 3.4, 3.5 and 3.6.1), immunostaining using anti-Ly6G was undertaken. Except for the single, heavily infected Ripk3-/-

Slc11a1+/+ dam #3715 (Fig. 23), Ly6G+ cell distribution in IS of infected Ripk3-/-Slc11a1+/+ and

Slc11a1+/+ did not differ to that from their naïve controls (Fig. 22 & 24). Ly6G+ neutrophils were found associated with blood vessel lumens (intravascular) in the DB, JZ and labyrinthine tissues. However, heavily infected placentas from Ripk3-/-Slc11a1+/+ dam #3715 showed profound neutrophil infiltration predominantly in the DB and JZ areas (Fig. 23A-D). These data suggest that the dominant leukocyte population recruited to severely infected IS is the Ly6G+ neutrophil.

Figure 22. Ly6G neutrophil immunohistochemistry of implantation sites from infected and naïve Ripk3-/-Slc11a1+/+ mice. Representative photomicrographs show Ly6G+ cells (identified using an anti-Ly6G antibody; appearing as brown stain over lymphocytes with polymorphonuclear appearance (black arrows)) in vessels of the decidua basalis (A & B), junctional zone (C & D), and labyrinth (E & F). Inset panel in (A) shows Ly6G+ cells at higher magnification. Representative image of an isotype control photomicrograph from the junctional zone of an infected dam (G). Sections counterstained with hematoxylin. All scale bars, 50 µm.

Figure 23. Ly6G neutrophil immunohistochemistry of implantation sites from heavily infected Ripk3-/-Slc11a1+/+ dam #3715. Representative photomicrographs show the massive infiltration of Ly6G+ cells (identified using an anti- Ly6G antibody; appearing as brown stain over mononuclear cells having a polymorphonuclear appearance (red circles or arrows)) in decidua basalis (A), junctional zone (C) and labyrinth (D). Representative isotype control image of decidua basalis (B). Sections were counterstained with hematoxylin. All scale bars indicated in individual images.

Figure 24. Ly6G neutrophil immunohistochemistry in implantation sites from infected and naïve Slc11a1+/+ mice. Representative photomicrographs show Ly6G+ cells (identified using an anti-Ly6G antibody; appearing as brown stain in mononuclear cells with a polymorphonuclear appearance, indicated by black arrows) in decidua basalis (A & B), junctional zone (C & D), and labyrinth (E & F) of infected and naïve Slc11a1+/+ IS. These cells were infrequent amongst all nucleated cells and did not appear to differ in frequency between infected and naïve Slc11a1+/+ IS. Inset panel in (A) shows Ly6G+ cells at higher magnification. Representative isotype control image of junctional zone tissue section from a naïve GD 15 dam (G). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

3.6.3 Immunohistochemistry for macrophages (anti-F4/80)

Because macrophages are known to support S.Tm, including macrophages within mouse IS, the second leukocyte lineage that was addressed was F4/80+ macrophages. Anti-F4/80 immunostaining was conducted using IS from all littermates of dams who were challenged and 2 IS/naïve dam/genotype.

F4/80+ macrophages that had similar morphology and distribution were present in placental tissues from

2 of 3 infected Ripk3-/-Slc11a1+/+ and in infected Slc11a1+/+ mice 72h post-infection (Fig. 25-27). The heterogeneous appearance of these F4/80+ cells and their distributions did not differ from those of F4/80+ cells in naïve Ripk3-/-Slc11a1+/+ or Slc11a1+/+(Fig. 25-27). In contrast to Ly6G+ neutrophils, F4/80+ macrophages were extravascular or perivascular in DB and labyrinthine tissues. In both tissues, shapes of

F4/80+ neutrophils ranged from flattened and spindle-like to round to stellate shaped with extending cytoplasmic processes.

A higher abundance of F4/80+ cells was found in tissues from heavily infected Ripk3-/-Slc11a1+/+ dam #3715. In only dam #3715, F4/80+ cells in the labyrinth appeared in large clusters (Fig. 25C & 26C

& D). In the other IS areas from this dam and in all other infected and naïve mice, F4/80+ cells appeared as single cells or as clusters of very few cells (Fig. 25 & 26). Although informative data was suggested by only the single, more heavily infected dam, these observations suggest that F4/80+ macrophages may utilize Ripk3 gene expression in the control of intracellular bacterial pathogens such as S.Tm.

Figure 25. F4/80 macrophage immunohistochemistry of decidua basalis from implantation sites recovered from infected and naïve Ripk3-/-Slc11a1+/+ mice. Representative photomicrographs show F4/80+ cells (identified using an anti-F4/80 antibody; appearing as brown stain over cells with variable morphologies, indicated by black arrows) were at similar abundance in decidua basalis of IS from two infected Ripk3-/-Slc11a1+/+ mice with limited histopathology compared to decidual basalis from naïve Ripk3-/-Slc11a1+/+ (A & B). Decidua basalis from IS of the more heavily infected Ripk3-/-Slc11a1+/+ dam #3715 displayed more F4/80+ cells (C). Representative isotype control specimen corresponding to image (C), (D). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

Figure 26. F4/80 macrophage immunohistochemistry of labyrinth from implantation sites recovered from infected and naïve Ripk3-/-Slc11a1+/+ mice. Representative photomicrographs show F4/80+ cells (identified using an anti-F4/80 antibody; appearing as brown stain in mononuclear leukocytes with variable appearance, indicated by black arrows) in similar abundance and location (single cells interspersed, closer to the chorionic plate) labyrinth tissues of most infected Ripk3-/-Slc11a1+/+ mice compared to tissues from naïve mice (A & B). Labyrinth tissues from Ripk3-/-Slc11a1+/+ dam #3715 displayed higher numbers of F4/80+ cells, presenting predominantly in clusters as imaged at 200X (C) and 400X (D) magnifications. Representative isotype control image corresponding to image (C), (E). Sections were counterstained with hematoxylin. All scale bars indicated in individual images.

Figure 27. F4/80 macrophage immunohistochemistry showed similar distributions of F4/80+ cells in implantation site tissues from infected and naïve Slc11a1+/+ mice. Representative photomicrographs show F4/80+ cells (identified using an anti-F4/80 antibody; appearing as brown stain in mononuclear leukocytes of variable morphological appearance, indicated by black arrows) in decidua basalis (A & B), and labyrinth tissues in similar proportions and presentation (single cells to small clusters localized closer to the chorionic plate) (C & D). Representative isotype control image of labyrinth tissue section from a GD 15 naïve dam (E). Sections were counterstained with hematoxylin. Scale bars indicated in individual images.

3.7 Serum Analyses

3.7.1 Quantification of maternal serum lipopolysaccharide endotoxin and progesterone

With the possible exception of Ripk3-/-Slc11a1+/+ dam #3715, I was unable to propose from my histopathological and immunohistochemical analyses a putative pathway for S.Tm-mediated pregnancy losses in mice. I was able to conclude and propose the hypothesis that RIPK3 had no major role in promotion or prevention of pregnancy pathology associated with S.Tm infection, at least for infectious agents delivered intravenously. It is well known that pregnancy loss in rodents can be acutely induced by endotoxins, products of Gram-negative bacteria, including Salmonella species (193). Furthermore, it is reported that LPS endotoxins act on the ovary to decrease progesterone (P4) levels (194). In mice, ovarian

P4 is essential for maintenance of pregnancy. Therefore, serum samples were collected at 72h post- infection or at GD 15 in naïve dams and quantified for LPS endotoxin and P4 concentrations. Because serum collections were initiated part way through my studies, not all mice were available for study. It is important to note that outlier dam Ripk3-/-Slc11a1+/+ dam #3715 was included in these studies, and data from the serum of this animal are distinguished from data from the other study animals by colour (red).

Unexpectedly, serum LPS levels were significantly lower in infected compared to naïve

Slc11a1+/+ mice, p=0.0225 (Fig. 28A). Infected Ripk3-/-Slc11a1+/+ dam #3715 had the lowest serum LPS concentration compared to the other two infected Ripk3-/-Slc11a1+/+ dams, and this concentration was comparable to that for infected Slc11a1+/+ mice. Serum progesterone (P4) levels did not differ significantly between the study groups regardless of infection or Ripk3 gene status (Fig. 28B). Thus, neither S.Tm endotoxin nor progesterone insufficiency appears to explain the impending loss of conceptus viability seen in this infection model.

Figure 28. Serum lipopolysaccharide (LPS) endotoxin and progesterone (P4) quantification. Serum LPS levels were significantly lower only in infected versus naïve Slc11a1+/+ mice, p=0.0225 (A). Serum P4 levels were not significantly different during infection, or with differing Ripk3 gene status (B). Each data point represents an individual animal of N=2 or 3 dams/genotype/group. All data are presented as mean ± SD. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. Conc., concentration; Inf., infected.

3.7.2 Maternal serum cytokine analyses

A classic inflammatory response is anticipated following bacterial infections. To address whether inflammatory cytokines in maternal plasma were elevated to levels that might affect conceptus survival, a

32 cytokine plasma analyte analysis was conducted. Serum from Ripk3-/-Slc11a1+/+ dam #3715 was included and is distinguished from other study animals by colour (red).

For all 32 cytokines, no significant differences were identified between infected and naïve dams of either genotype (Fig. 29 & 30), including cytokines previously reported (both as protein in serum and mRNA expression in infected placental tissues) as altered in response to S.Tm infection during pregnancy

(60,155) such as IL-6, G-CSF, CXCL1, CXCL2, and IL-12 (shown as IL-12 (p40) and IL-12 (p70)) (Fig.

29). Other cytokines known generally to be altered in response to infection with bacterial pathogens including IL-1α, IL-1β, CCL2, IFN-γ, and TNF-α are shown (Fig. 30). The means of the infected

Ripk3-/-Slc11a1+/+ group were skewed due to the presence of a consistent outlier, the heavily infected

Ripk3-/-Slc11a1+/+ dam #3715. Notably, G-CSF was increased >56-fold, IL-6 was increased >108-fold,

CXCL2 was increased >29-fold, and CCL2 was increased >81-fold compared to respective means of naïve Ripk3-/-Slc11a1+/+ mice.

Figure 29. Maternal serum cytokine values illustrated for molecules reported by others as altered in maternal serum of Slc11a1 competent mice infected with S.Tm during pregnancy. Cytokine bead array assays (Eve Technologies) revealed no significant difference in many cytokines previously identified as involved in response to S.Tm infection during pregnancy. These include IL-6 (A), G-CSF (B), CXCL1 (C), CXCL2 (D), and IL-12 (E & F). Means of the infected Ripk3-/-Slc11a1+/+ group were skewed due to the presence of a consistent outlier, the heavily infected Ripk3-/-Slc11a1+/+ dam #3715 (red data points). Each data point represents an individual animal of N=3 dams/genotype/group. All data presented as mean ± SD. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. Inf., infected.

100

Figure 30. Maternal serum cytokine values illustrated for molecules generally implicated in bacterial infections. Cytokine bead array assays (Eve Technologies) showed no significant differences for any analyte. Illustrated are the data for several other cytokines typically associated with responses to infection (195). Means of the infected Ripk3-/-Slc11a1+/+ group were skewed due to the heavily infected Ripk3-/-Slc11a1+/+ dam #3715 (red data points). Each data point represents an individual animal of N=3 dams/genotype/group. All data presented as mean ± SD. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. Inf., infected.

101

3.8 Focus on outlier Ripk3-/-Slc11a1+/+ dam #3715

I was unable to replicate my collaborator's previous findings in 129.B6F1 mice of massive placental S.Tm burden, inflammation or pathology in the majority of animals I studied. Therefore, attention was focused to the only animal that displayed an adverse phenotype by histology and by maternal serum cytokine profiling, Ripk3-/-Slc11a1+/+ dam #3715. I missed infection of tissues outside of the 3 major placental regions, DB, JZ and labyrinth areas because of heavy focus on these tissues central to my hypothesis. Therefore, these tissues were re-examined more closely. Dam-specific characteristics of dam #3715 and IS outcomes in terms of gross-post mortem observations, S.Tm bacterial burden scores, infection of yolk sac (YS) tissues, histopathological observations including placental damage and vascular congestion, and Ly6G+ neutrophil infiltration are included in Table 9. Ly6G+ infiltration was used as a representative marker to describe innate immune infiltration, because it was determined to be the predominant infiltrating mononuclear cell in heavily infected IS.

102

Table 9. Ripk3-/-Slc11a1+/+ dam #3715 characteristics and implantation site outcomes.

Parameter Clinical "health- A - healthy check" score prior to euthanasia

# Implantation sites 11

Fetal resorption rate 0.0%

% IS S.Tm+ 81.8%

Implantation Site Gross post-mortem # S.Tm+ foci/2 S.Tm burden Yolk sac Histopathology Vascular Ly6G+ neutrophil (lettered from left to observation mid-sagittal score** infection congestion infiltration of DB & right uterine horn) sections* JZ areas*** A Healthy size & colour 0 0 No Mild damage No Absent B Healthy size & colour 13 2 No Mild damage No Absent C Healthy size & colour 17 2 No Mild damage No Absent D Healthy size & colour 14 2 No Mild damage Yes - mild Absent E Healthy size & colour 8 1 Yes Mild damage Yes - mild Absent F Healthy size & colour 0 0 Yes Mild damage Yes - mild Absent G IS healthy size & colour, 200+ 4 Yes Severe damage Yes - severe Severe fetus brain hemorrhage H Healthy size & colour 81 4 Yes Severe damage Yes - severe Severe I Healthy size & colour 67 4 Yes Mild damage Yes - mild Absent J Healthy size & colour 26 3 Yes Mild damage Yes - mild Absent K Healthy size & colour 0 0 No Mild damage No Absent

* S.Tm+ foci were enumerated from decidua basalis (DB), junctional zone (JZ) and labyrinthine areas of placenta ** For descriptions of S.Tm bacterial burden scoring system I developed and used see Table 5) *** Ly6G+ neutrophil infiltration was noted here as a representative surrogate for immune infiltration because Ly6G+ neutrophils were the predominant mononuclear leukocyte in heavily infected IS

103

Upon re-examination of all IS from this specific animal, prominent chorio-deciduo-amnionitis was observed in 6/11 (54.5%) IS. Substantial S.Tm+ foci were detected in decidual tissues and spongiotrophoblast (Fig. 31A-C). S.Tm infection extended to the edge of the DB, which lies adjacent to the uterine myometrium and the thin layer of p-TGCs surrounding the extraembryonic membranes encapsulating the fetus (includes YS and amniotic sac) (Fig. 31A-C). Dense foci of S.Tm were detected by IHC in the parietal (PYS) and visceral (VYS) layers of the YS, but not in Reichert's membrane (RM) (Fig. 31B, C, E-F).

Immunoreactive S.Tm were also found in the heavily immune infiltrated p-TGC layer surrounding the amniotic sac containing the fetus, where the p-TGCs could no longer be distinguished (Fig. 31D). S.Tm+ foci were also detected in the amnion and within fetal vessels of the chorionic plate area (Fig. 32). This indicated fetal bacteremia in addition to infection of the amnion, the cell layer closest to the fetus.

Because S.Tm infection has yet to be described in extraembryonic tissues including fetal YS and amnion, immunofluorescence (IF) was conducted to validate these findings. S.Tm-IF staining replicated the

S.Tm-IHC staining pattern, confirming infection of the decidua and surrounding p-TGCs and extraembryonic tissues (Fig. 33 & 34). S.Tm+ foci were heavier and therefore brighter in areas of fetal yolk sac (Fig. 33 & 34). S.Tm-IF was successful since no positive staining was observed in isotype control sections of severely infected IS (Fig. 33B) or in IS from naïve Ripk3-/-Slc11a1+/+ mice (Fig. 34C).

Histopathology by H&E staining of these same heavily infected regions revealed profound tissue damage, necrosis and immune infiltration of mononuclear cells resembling neutrophils (Fig. 35). RM appeared thicker compared to RM from a gestational-age matched naïve Ripk3-/-Slc11a1+/+ control, indicating edema or response to tissue damage (Fig. 35A & C). The PYS was no longer a continuous and organized layer lining the fetal surface of the placenta. PYS cells were discontinuous and appeared to be degenerating and detaching from surface of RM (Fig. 35A & C). Immune infiltration was confirmed by

CD45 IHC (Fig. 35C), and the majority of immune cells in this area were Ly6G+ neutrophils (data not shown). Furthermore, the most severely infected, inflamed extraembryonic tissues were observed in

104

placentas from fetuses with gross post-mortem pathologies including a visible brain hemorrhage (data not shown). These studies indicate that there is preferential infection of decidua and extraembryonic tissues including the fetal YS rather than placenta during S.Tm infection of mid-gestation mice with intact Slc11a1 function. Destruction of early hematopoietic cells of the YS and fetal bacteremia could explain the rapid fetal loss (<24h) associated with S.Tm infection despite the healthy appearance and normal functions of most tissues at 72h post-infection.

Extraembyronic tissue infection was only observed in one dam gene-deleted for the key necroptosis protein RIPK3. I had hypothesized that elimination of necroptosis would result in a decrease of placental

S.Tm infection, inflammation and pathology compared to necroptosis-competent controls (Slc11a1+/+ mice).

To elucidate whether infection of extraembryonic tissues is a result of a lack of necroptotic-cell death I scanned my laboratory's archival specimen inventory for infected Slc11a1+/+ IS with observations of gross post-mortem pathology. These specimens had been prepared by my Ottawa collaborators and histologically prepared and studied by an undergraduate thesis student, Annie Peng, prior to my arrival at Queen's

University. One animal (Slc11a1+/+ dam #3293; GD 15 72h post-infection) was located with descriptions of pale IS. Tissue sections from one IS from dam #3293 were immunostained using my S.Tm-IHC protocol.

Localization of S.Tm+ foci was similar to that in Ripk3-/-Slc11a1+/+ dam #3715, and an S.Tm bacterial burden score of 4 (51+ S.Tm foci; refer to Table 5) was given (Fig. 36A & C). S.Tm+ foci were localized to

DB, JZ (in spongiotrophoblast and not TGCs), and extraembryonic tissues including fetal YS (Fig. 36A, C,

E). Thus, S.Tm is sporadic in its pathology in this model of i.v. S.Tm infection in mouse pregnancy and much larger sample sizes would be needed to estimate the frequency of lethal infections and to identify factors regulating this sporadic pathogenesis.

Since I was unable to find labyrinth trophoblast infection in any animal studied that had intact

Slc11a1 function, I analyzed IS from archival samples of infected C57BL/6J mice (lack functional

NRAMP1) with gross post-mortem pathology (smaller, pale IS). IS from infected GD 14 C57BL/6J at 48h

105

post-infection (dam #3377) showed an inverted staining pattern, with heavy S.Tm infection of labyrinthine tissues compared to other placental subregions including DB (Fig. 36B & D). S.Tm+ foci appeared within cells morphologically resembling trophoblasts (Fig. 36D). S.Tm were also found heavily concentrated in fetal yolk sac tissues (Fig. 36F). Thus, Slc11a1 appears to regulate the component of the placenta that becomes infected by S.Tm. In Slc11a1 competent mice, S.Tm is localized to the DB, spongiotrophoblast. In

Slc11a1 incompetent mice, S.Tm localizes to the TBCs of the placental labyrinth. However, the fetal lethality from S. Tm infection in Slc11a1 competent or Slc11a1 incompetent genotypes appears to be shared and independent of Slc11a1 status. Fetal lethality, I postulate is due to infection of yolk sac islands, hematopoietic failure and fetal bacteremia.

106

107

Figure 31. S.Tm immunohistochemistry of extraembryonic tissues including p-TGCs and yolk sac tissues from implantation sites of heavily infected Ripk3-/-Slc11a1+/+ dam #3715. Representative photomicrographs from 2/11 heavily infected IS shows heavy S.Tm+ IHC staining in the area of the anti-mesometrial decidua adjacent to p-TGCs (encircled in red), in addition to decidua basalis (DB) and spongiotrophoblast tissues (A & B). The parietal (PYS) and visceral (VYS) yolk sac layers are heavily infected with S.Tm, but Reichert's membrane (RM) is not (B-F). The anti-mesometrial decidua and p-TGC layer just adjacent to the uterine epithelium (UE) and uterine wall (UT) was also heavily infected with S.Tm, and p-TGCs could no longer be recognized due to the profound immune cell infiltrate (indicated by red arrows) (E). S.Tm appeared intracellularly within the endocytic vacuoles of VYS cells (F). These images show preferential infection of S.Tm in extraembryonic tissues including fetal yolk sac. Sections were counterstained with hematoxylin. Scale bars indicated in individual images. JZ, junctional zone; Lab, labyrinth; p-TGC, parietal-trophoblast giant cell.

108

Figure 32. S.Tm immunohistochemistry of chorionic plate and amnion tissues recovered from heavily infected implantation sites of Ripk3-/-Slc11a1+/+ dam #3715. Representative photomicrographs from the most heavily infected IS (2/11) from dam #3715 shows presence of S.Tm in chorionic plate tissues (identified using an anti-S.Tm LPS antibody), specifically within fetal vessels (distinguished by presence of nucleated erythrocytes) (A). Due to the presence of S.Tm within fetal vessels of the chorionic plate, these results indicate fetal bacteremia. Corresponding isotype control section to image in (A), (B). Cells of the amnion were positive for intracellular S.Tm (indicated by black arrow) (C). There was an abundance of CD45+ leukocytes intertwined within the layers of amnion (identified using an anti-CD45 antibody; cells stained brown), indicating amnionitis (D). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

109

110

Figure 33. Localization of S.Tm within highly immune infiltrated extraembryonic tissues recovered from a heavily infected implantation site of Ripk3-/-Slc11a1+/+ dam #3715. Photomicrographs show dense S.Tm+ staining (using an anti-S.Tm LPS antibody; red) in decidua basalis (DB) and junctional zone (JZ) tissues extending into the anti-mesometrial decidua adjacent to the uterine wall (UT) (A). Labyrinthine tissues appear completely devoid of any S.Tm+ staining. Dense S.Tm+ foci appear in parietal yolk sac (PYS) tissues on the fetal surface of the placenta (C). A dense infected immune infiltrate laid adjacent to the discontinuous and degenerating PYS. Visceral yolk sac (VYS) cells were also infected heavily by S.Tm (C). S.Tm+ foci did not appear to extend into the uterine wall or myometrium. S.Tm-IF staining confirms localization of S.Tm to yolk sac tissues. Isotype control image shows no S.Tm+ staining (B). Red, Alexa-fluor® 594 secondary antibody to S.Tm; Blue, dapi. All scale bars included individual images. Ctrl, control.

111

Figure 34. S.Tm immunofluorescence of extraembryonic tissues including p-TGCs and yolk sac tissues recovered from heavily infected implantation sites of Ripk3-/-Slc11a1+/+ dam #3715. Photomicrographs show intracellular S.Tm foci in p-TGCs adjacent to heavily infected, discontinuous and degenerating parietal yolk sac (PYS) cells (A). Visceral yolk sac (VYS) cells were also infected. Higher power image of VYS shows intracellular S.Tm within a cell resembling a heavily infected mononuclear cell (B). Tissues from naïve dams did not display any S.Tm+ staining (C). These images highlight the ability of S.Tm to infect YS tissues, and elicit an immune response (inferred through presence of mononuclear immune cells). Red, Alexa-fluor® 594 secondary antibody to S.Tm; Blue, dapi. All scale bars indicated in individual images. 112

113

Figure 35. S.Tm infection of extraembryonic tissues and anti-mesometrial decidua is associated with massive tissue damage and infiltration of CD45+ leukocytes in heavily infected implantation sites from Ripk3-/-Slc11a1+/+ dam #3715 compared to naïve controls. Histopathology by hematoxylin and eosin (H&E) staining showed massive tissue damage in the area of anti- mesometrial decidua (A). Reichert's membrane (RM) appears thick and discontinuous compared to tissues from a representative naïve control (B). The parietal yolk sac (PYS) cells are no longer an organized continuous layer lining the fetal side of the placenta (see in comparison to naïve tissue in (B)). P-TGCs adjacent to failing PYS, displayed fragmented nuclei and vacuolated cytoplasm, indicating of necrotic cell death. The visceral yolk sac (VYS) appears to be degenerating and the yolk sac blood islands appear congested (asterisks) compared to VYS of naïve controls (A & B). Profound infiltration of CD45+ leukocytes was observed in the necrotic anti-mesometrial decidua, which was not seen in tissues CD45- stained naïve control tissues (C &D). Destruction of the fetal yolk sac is apparent and the extent of damage is apparent upon comparison to tissues from gestation-age matched naïve controls. Anti-CD45 IHC sections in (C & D) were counterstained with hematoxylin. All scale bars, 100 µm.

114

115

Figure 36. Functional Slc11a1 status determines placental localization of S.Tm, but fetal lethality likely results from S.Tm infection of fetal yolk sac. Using an anti-S.Tm antibody, tissue sections from an infected Slc11a1+/+ dam (#3293) at 72h post-infection showed similar S.Tm staining patterns in placenta (decidua basalis (DB) and junctional zone (JZ)) to IS from Ripk3-/-Slc11a1+/+ dam #3715 (A). In JZ tissues, S.Tm+ foci were limited to spongiotrophoblasts (SpT), and absent in p-TGCs (C). In an infected Slc11a1 incompetent C57BL/6J dam (#3377), S.Tm+ foci were localized to the placental labyrinth in an inverted staining pattern compared to tissues from Slc11a1 competent dams (B). Examination of placental regions at higher power revealed the complete absence of S.Tm+ foci in SpT tissues, and infection of cells resembling placental labyrinthine trophoblasts (D). Intracellular S.Tm+ foci were detected in fetal yolk sac (YS) tissues in both mice regardless of Slc11a1 functional status (E & F). The progression of YS infection in (F) appears especially advanced, as YS cell types cannot be distinguished. Fetal YS destruction is likely the cause of rapid fetal loss in pregnant mice regardless of their susceptibility to S.Tm infection. Panel (G) is the corresponding isotype control to the image in (D). Sections were counterstained with hematoxylin. All scale bars indicated in individual images.

116

Chapter 4

Discussion

Genetic elimination of the key necroptotic cell death protein RIPK3 neither decreased placental inflammation or trophoblast cell death, nor altered conceptus survival compared to controls over the first 72h hours after mid-pregnancy i.v. challenge with 103 S.Tm. Additionally, no differences were observed with respect to fetal loss, placental damage, and immune cell infiltration of placental tissues when infected dams were compared to naïve, gestational-age matched controls, regardless of Ripk3 gene status, except for one outlier Ripk3 gene-deleted dam, discussed separately in section 4.1.1. A genetic knockout of the downstream mediator of RIPK3, MLKL, has also been noted to have no effect on polymicrobial septic shock-induced animal death (196). Therefore, elimination of necroptosis by gene deletion appears to be insufficient to prevent or delay death of animals following i.v. bacterial challenge leading to septicemia.

I was unable to replicate my collaborators’ previous findings of significant fetal loss, massive placental S.Tm infection, necrosis, and inflammation (60,155). Previously published work used 129×1/SvJ mice with intact Slc11a1 function. When pregnant 129×1/SvJ mice were challenged on GD 10-11 with 103

S.Tm i.v., a fetal resorption rate of ~75% was observed at 72h post-infection (155). The background mouse strain used in my studies would closely resemble the 129×1/SvJ inbred strain that was used to generate the congenic C57BL/6J mice with intact Slc11a1 function. The gestational day of S.Tm challenge also differed but structurally, IS developmental maturity would have been similar (i.e. placental layers, spiral arterial modification and placental circulation were established). My collaborators’ study reported elevated placental pro-inflammatory cytokine mRNA transcript levels, despite a dampened maternal innate immune response to

S.Tm infection. Further my collaborators’ studies using 129.B6F1 mice with hybrid resistance echoed the initial placental inflammatory phenotype found in resistant 129×1/SvJ mice. In these studies, use of an avirulent S.Tm mutant, showed that bacterial burden alone does not correlate with placental pathology and 117

subsequent fetal or maternal death (60). Instead, adverse, severe pro-inflammatory immune responses to heavy S.Tm-infection in placental tissues were interpreted as the probable etiology for conceptus-loss (60). I was unable to validate these previous findings using 103 S.Tm infection at GD 12 (24-48h later in development) using a novel mouse strain (B6. Slc11a1+/+) or a mutated variant of the novel strain (Ripk3-/-

Slc11a1+/+). The congenic Slc11a1+/+ mouse strain had not been examined in the context of S.Tm infection in pregnancy, but was expected to respond in a manner identical to the 129×1/SvJ strain.

Absence of an adverse inflammatory phenotype in both novel strains was not due to a lack of S.Tm infection of placental tissues. S.Tm+ foci were detected by IHC as early as 48h post-infection in IS from mice with intact Slc11a1 with or without Ripk3 (means of 48.1% versus 62.5% infected IS respectively).

However, even at 72h post-infection, S.Tm+ foci were sparse in placental tissues of both novel strains and foci could be counted individually. The majority of S.Tm+ foci were detected in extracellular environments; for example, contained within blood vessels or spaces in the most perfused placental sub-regions including the DB and labyrinth. S.Tm enumeration by IHC was performed using two mid-sagittal sections per IS, and therefore only represents bacteria localized to the center of the IS. Despite the lack of sensitivity for bacterial burden estimation by IHC, IS infection rates (or percentage of IS with S.Tm-reactive foci by IHC) were similar to those obtained in parallel studies through bacterial culture CFU analysis conducted by my collaborator Kristina Wachholz, UOttawa (157). At 72h post-infection, S.Tm+ foci were detectable in 75.4% of IS from infected Ripk3-/-Slc11a1+/+ dams, and 80% of IS from infected Slc11a1+/+ dams. Despite the presence of S.Tm in placental tissues, there was no associated inflammatory response to infection

(determined through leukocyte immunostaining) and no major placental damage (determined through histopathology and anti-active caspase 3 immunostaining). Similar observations have been reported in cytomegalovirus (CMV) infection of human placenta, where CMV could be detected by IHC without associated inflammation or pathology (197).

118

H&E-staining did reveal spongiotrophoblast cell compaction in the area of the JZ that was accompanied by reduced maternal blood sinusoidal space (qualitative observation) in IS from infected dams regardless of Ripk3 status. However, JZ surface area was not altered when IS from infected dams were compared to those from naïve dams regardless of Ripk3 status. Cells of the JZ are thought to have endocrine activities involved in maintenance of pregnancy (108-110). It is unknown how compaction of JZ cells may have affected the endocrine functions of these cells. The reduction of maternal blood space may have decreased blood flow to the underlying labyrinth area and perturbed maternal-fetal exchange of gases and nutrients essential for fetal growth and development. Use of micro-computed tomography of fetal labyrinthine vessels or maternal to fetal tissue transport of biotin would be techniques to determine possible functional impacts on vascular structure and molecular exchange parameters. Reduced blood flow to the placenta leads to intrauterine growth restriction. Some developmental delay was observed in fetuses only from infected Slc11a1+/+ dams at 72h post-infection, but this observation was inconsistent. Furthermore, fetal to placental weight ratios were not significantly altered in infected versus naïve Slc11a1+/+ dams regardless of Ripk3 status.

Because there were no major differences in placental pathology at 72h post-infection when pregnancy loss was imminent, I suspected infection-associated cell death may have been activated but not yet reached levels detectable by histopathology. To address this question, cell death detection was conducted by active caspase-3 immunostaining. Active caspase-3 is crucial mediator of apoptosis (171,198). Success of active caspase-3 immunostaining was confirmed with the convenient internal positive control, physiological death of uNK cells in late decidua. There were no significant differences in areas IHC-positive for caspase-3 activity between infected and naïve dams with or without Ripk3, with the exception of fetal membranes which were positive only in mice lacking Ripk3. Activation of necroptosis was not assessed in my studies because antibodies specific for the detection of key phosphorylated residues involved in activities of RIPK3 and MLKL are in the early stages of development and optimization for IHC analyses (199).

119

Alternative explanations to account for impending rapid fetal loss by 96h post-infection in Slc11a1 competent mice with or without Ripk3 were pursued through examination of LPS endotoxin and P4 levels in maternal serum. It is well known that pregnancy loss in rodents can be acutely induced by exposure to LPS

(193), a product of the cell walls of S.Tm and other organisms. LPS acts on the ovary to decrease P4 levels

(194). P4 is required for the maintenance of mouse pregnancy. Thus, it was important to consider how S.Tm and its endotoxin LPS might alter P4 levels. Furthermore, metabolomic studies involving non-pregnant mice orally infected with S.Tm identified important shifts in steroid and eicosanoid hormone biosynthesis and metabolism (200). Steroid and eicosanoid hormones play important roles not only in pregnancy maintenance and but also in normal immune functions. Maternal serum LPS levels were significantly reduced in only infected Slc11a1+/+ mice compared to their naïve controls. A reduction of LPS endotoxin in serum of S.Tm- infected dams initially seemed paradoxical, but could result from clearance of LPS or its binding to cells via

TLR4. The higher serum LPS levels obtained from naïve dams may be due to the presence of non-bioactive

LPS from other bacterial species, for example intestinal commensal bacteria (201). P4 levels were not altered between infected and naïve Slc11a1 competent mice regardless of Ripk3 status, suggesting that decreased P4 levels do not contribute to the pathogenesis of pregnancy loss in S.Tm infection.

Maternal serum cytokine profiling decisively showed the absence of a maternal immune response to

S.Tm infection. Pro-inflammatory cytokines indicative of activation of the innate immune system such as IL-

6, G-CSF, CXCL1, CXCL2, IFN-γ, TNF-α and IL-1β were not significantly different between infected versus naïve Slc11a1 competent dams regardless of Ripk3 status. This indicates that in the majority of dams included in my studies, S.Tm was present in placental tissues with no local or systemic maternal immune response to infection. Although I could not assess if the transcription of pro-inflammatory cytokines was elevated in placental tissues, I believe my histological studies support my conclusion regarding absence of local immune response to S.Tm infection in placenta. These non-remarkable histological and serological

120

findings were at odds with the imminent, rapid loss of fetuses by 96h post-infection. At 72h post-infection, the majority of tissues (both IS and fetal) appeared healthy and viable.

4.1.1 The outlier: a possible explanation for rapid fetal loss?

Efforts were shifted to the heavily infected, outlier dam #3715 (Ripk3-/-Slc11a1+/+), sacrificed at 72h post-infection. This dam received a clinical health-check score of "A" or "healthy" with no observable maternal distress (i.e. ruffled fur, hunched, slowed activity) prior to sacrifice. Dam #3715 had 11 viable IS

(fetal resorption rate of 0) and 9/11 (81.8%) IS displayed S.Tm+ foci in placental tissues. There was a graded

S.Tm infection amongst IS, receiving S.Tm bacterial burden scores ranging from 0 (2/11 IS), up to scores of

4 (indicating 51+ foci, detected only in 3/11 IS) (see Table 9 for further details). S.Tm+ foci could not be enumerated in 1/11 IS because of the overwhelming bacterial burden within the DB and spongiotrophoblast regions. Severe placental damage and necrosis of DB and JZ was detected in 2/11 IS, sites that corresponded with highest placental bacterial burdens (Table 9). These sites had a massive infiltration of Ly6G+ neutrophils, not F4/80+ macrophages. Active-caspase-3 staining showed no significant apoptosis even in the

2 most damaged IS, indicating S.Tm-mediated necrosis as the major of pathway of cell death within infected

DB and JZ tissues. Trophoblast cells of the placental labyrinth were not infected, but appeared to be degenerating in these 2 IS. Severe labyrinthine vascular congestion was observed in these 2 specific IS, with another 5/11 IS from this dam displaying mild but noticeable vascular congestion. The remaining 9/11 IS from this dam appeared to have mild placental damage and immune cell infiltration. Only a proportion (2/11 or 18.1%) of IS displayed substantial S.Tm infection, immune infiltration and placental damage, with the remaining IS having less infection and milder damage. The maternal serum cytokine analyses revealed large elevations of pro-inflammatory cytokines (IL-6, G-CSF, CXCL1, CCL2 and TNF-α) indicative of a strong innate maternal immune response to S.Tm infection. These results suggested that prior to any fetal resorption or pregnancy loss, pro-inflammatory cytokines are elevated in maternal plasma, in response to heavy S.Tm

121

infection and damage in selected IS. These results did not explain imminent pregnancy failure of all members of a litter within the next 24h.

Potential mechanisms for rapid fetal loss were only discovered through localization of S.Tm by IHC.

To my knowledge, this is the first investigation of S.Tm localization by IHC in mouse placenta. My IHC protocols were carefully vetted and optimized to ensure elimination of non-specific background staining. Re- examination of the entire IS including the uterine wall and myometrium, placenta and extraembryonic tissues revealed profound and consistent infection of fetal yolk sac (YS) membranes in 6/11 IS from Ripk3-/-

Slc11a1+/+ dam #3715. Because YS infection has not been previously described during S.Tm infection in mouse pregnancy, S.Tm-IHC staining results were further validated by S.Tm-IF staining. In mouse pregnancy, pathogen infection of YS has only been reported for Brucella abortus (B. abortus) infection. B. abortus is a facultative, intracellular, Gram-negative bacterium with tropism for the pregnant reproductive tract of ruminants (202). Using BALB/c mice infected on GD 9 with 105.7 B. abortus by intraperitoneal injection, Tobias and colleagues localized bacteria in tissues collected on GD 18 intracellularly, in neutrophils in the DB, both spongiotrophoblasts and TGCs of the JZ, and the fetal YS (203). These findings were observed in conjunction with severe necrosuppurative placentitis in the DB and JZ areas (203).

Infection was not localized to the placental labyrinth area. They also noted that these observations occurred prior to fetal loss and commented on the sporadic nature of the infection and resulting pathology (203). The findings of mid-pregnancy infection with B. abortus in mouse pregnancy bear striking resemblance with my observations in IS tissues from dam #3715, with exception to TGC infection. Unlike S.Tm, B. abortus is non-motile and is internalized by TGCs through interaction with cell surface heat shock cognate protein 70

(hsc70) (204,205).

Within the YS, S.Tm were specifically localized within parietal YS (PYS) cells and visceral YS

(VYS) cells. Both cell types are of the extraembryonic endoderm lineage, arising from embryonic hypoblast

(forms from the inner cell mass). The mouse yolk sac completely surrounds the embryo external to the

122

amnion (refer to Chapter 1, Fig. 2B). At the gestational age range of focus for my studies (GD 12-15), the mouse yolk sac functions to deliver nutrients to the fetus critical for proper fetal growth (117). VYS cells are capable of both receptor-mediated and bulk-endocytosis, which are involved in the histotrophic digestion of nutrients within the YS cavity (116). Nutrients are delivered to the fetus through YS (vitelline) blood circulation (206). Proper vitelline circulation is essential for embryo survival. Ligation of vitelline circulation at GD 12.5 or 13.5 resulted in 100% fetal death by GD 17.5 (117,207). Vitelline ligation at GD 14.5 resulted in 57% fetal death by GD 17.5 (117,207). Therefore, damage to YS tissues mediated by S.Tm infection would be sufficient to produce an occlusion in vitelline circulation leading to rapid fetal death.

In addition to nutrient digestion and delivery functions, the YS is an important hematopoietic organ for primitive and definitive erythropoiesis essential to the progression of fetal development (208,209).

Erythrogenesis requires high amounts of iron for the synthesis of heme. This provides a region of enhanced iron metabolism, which may explain the tropism of S.Tm for fetal yolk sac tissues. Iron is a critical nutrient for the growth and survival of most bacterial species including S.Tm (210,211). Bacteria must acquire iron from their host environment, and possess pathogenicity mechanisms to compete with host iron sequestration mechanisms (211). The virulence of S.Tm in mice is dependent on the presence of iron in serum and tissues, and experimentally induced hypoferremic mice exhibit enhanced resistance to S.Tm infection (212). It is logical that S.Tm display strong tropism for YS tissues in an area with an abundance of iron. Loss of a critical site of erythropoiesis would result in a detrimental oxidative shift in fetuses, potentially contributing to fetal distress and subsequent loss. This hypothesis has extra appeal since the IS assessed with YS infection appeared grossly pale. Together, destruction of YS with potential occlusion of vitelline circulation and inhibition of nutrient transfer, coupled with failure of erythrogenesis leads to fetal anemia and rapid fetal death.

YS infection was found in infected pregnant mice with functional Slc11a1 regardless of Ripk3 gene status at 72h post-infection. YS infection was also observed in a pregnant infected C57BL/6J dam (archival

123

samples studied) lacking functional Slc11a1 at 48h post-infection. Therefore, mice with differing susceptibility to S.Tm infection may share common mechanisms of fetal loss related to compromises in the functions of the YS erythropoiesis and vitelline circulation.

The translational potential for this novel hypothesis regarding fetal loss by S.Tm YS infection to human pregnancy may be restricted to early pregnancy exposure because of differences in the development, structure and function of mouse versus human YS. Mice have an inverted YS that envelops the entire embryo and is active from gastrulation until birth. Humans have a free, balloon-shaped yolk sac that connects to the embryo via the yolk stalk (213). The human YS is thought to play important roles is nutrient exchange before the vascularization of chorionic villi, i.e. up to the 9th week of pregnancy, after which it degenerates (213,214). However, fundamental roles of YS are preserved in mammalian evolution. Therefore

S.Tm infection of YS during early human pregnancy may be considered.

4.1.2 Mechanisms of vertical S.Tm infection in mice

Preliminary data in selected, heavily S.Tm-infected IS revealed unique placental localization of

S.Tm based on functional Slc11a1 status regardless of Ripk3 gene status. In mice with functional Slc11a1,

S.Tm+ foci were limited to the DB and occasionally spongiotrophoblast. Mice without functional Slc11a1 displayed S.Tm+ foci in placental labyrinth tissues, likely within TBCs. Therefore, Slc11a1 was found to dictate trophoblast susceptibility to S.Tm infection in mice in vivo, and thus determined placental localization of S.Tm. Based on these observations, I propose two potential mechanisms of vertical transmission of S.Tm in mouse, paraplacental transmission in Slc11a1 competent mice and transplacental transmission in Slc11a1 incompetent mice (Fig. 37). In Slc11a1 competent mice, hematogenous S.Tm challenge leads to mesometrial decidual infection with occasional spongiotrophoblast infection. Infection spreads non-hematogenously towards the thin anti-mesometrial decidua directly adjacent to the single continuous layer of p-TGC cells surrounding the fetal yolk and amniotic sacs. S.Tm within the YS cavity are endocytosed by PYS and VYS cells and enter vitelline circulation. S.Tm spreads to the fetal liver, causing 124

fetal bacteremia that can be seen in chorionic plate vessels of the placenta (Fig. 37A). In Slc11a1 incompetent mice, hematogenous S.Tm challenge leads to rapid labyrinthine TBC infection and hematogenous transmission to fetal blood circulation resulting in fetal bacteremia (Fig. 37B). At this point, the exact mechanisms of YS infection are unknown in Slc11a1 incompetent mice but S.Tm may reach the vitelline YS circulation through fetal liver. It is known that fetal liver colonization is rapid following i.v.

S.Tm infection in 129×1/SvJ mice (155), but this is unknown in the novel mutant strains used in this study

(Ripk3-/-Slc11a1+/+ and Slc11a1+/+).

At this stage, these results should be interpreted with caution until they can be reproduced in sample sizes greater than N=1 per genotype. The immediate next step should be to confirm Slc11a1 expression and localization by IHC or IF in mice.

125

126

Figure 37. Vertical transmission of S.Tm in pregnant mice is dictated by functional Slc11a1 status. Schematic diagrams of an entire implantation site including fetus (both represent GD 15) shows potential S.Tm infection pathways in utero-placental tissues of pregnant mice. In Slc11a1 competent mice (A), vertical transmission of S.Tm is likely paraplacental. Hematogenous S.Tm challenge (1) (shown as S.Tm entering the radial artery) results in colonization of the decidua basalis (DB) and junctional zone tissues (2), specifically spongiotrophoblast (SpT). S.Tm spreads non-hematogenously in DB and JZ tissues to infect the thin layer of anti-mesometrial decidua, and parietal-trophoblast giant cells (p- TGCs), spreading to the adjacent parietal yolk sac (PYS) and visceral yolk sac (VYS) tissues (3). S.Tm spreads through yolk sac vitelline circulation through the umbilical cord (UC) to fetal liver resulting in fetal bacteremia (4). This may explain the observation of S.Tm in fetal blood vessels of the chorionic plate without placental labyrinth infection. S.Tm are also found in the amniotic membrane. In Slc11a1 incompetent mice (B), vertical transmission of S.Tm is likely transplacental. Hematogenous S.Tm challenge (1) results in rapid colonization of placental labyrinth tissues (likely labyrinthine trophoblasts) (2). S.Tm crosses the interhemal membrane and enters fetal circulation (FC) (therefore hematogenous spread), leading to rapid colonization of fetal liver and fetal bacteremia (3). Fetal liver infection likely results in the spread of S.Tm to fetal yolk sac tissues through vitelline circulation (4). Schematic drawings are original and not to scale. AC, amniotic cavity; MLAp, mesometrial lymphoid aggregate of pregnancy; UT, uterine wall; YSC, yolk sac cavity.

127

4.2 Study Limitations

Low sample size was a major limitation in the studies reported in this thesis since inherent variability complicated all of the assays conducted. Variability began within each litter where some but not all IS became infected by S.Tm. Infected IS were not identical to each other, even amongst the infected sites in a single litter. This necessitated all IS from litter members (n= 4 to 11 IS) of each infected dam be sectioned and studied to enable my analyses. To study n=3 pregnancies per group at two time points post-infection, I sectioned a total of 1135 slides (3 sections/slide therefore 3405 sections) for the histological analyses incorporated into this thesis. This tedious technical component consumed a large amount of my research time. The essential need to study every IS/pregnancy and the time requirement for preparation of such material meant that replicate numbers could not be increased above 3 dams per group.

The second aspect of variability was inconsistency of outcomes between genetically-identical replicate dams despite careful experimental execution. Replicate dams within a specific group (i.e. GD 15

Slc11a1+/+ 72h post-infection, N=3) were maternal age-matched and gestational-age matched (often sisters from the same litter) and received identical S.Tm challenge (103 S.Tm i.v.) from a common bacterial stock.

All mice were housed on the same vent racks with identical environmental exposures (stress from typical mouse husbandry and handling, ambient temperature, noise etc.). All mice in this study (excluding archival specimens, data unknown) received a health-check clinical score of "A", indicating no visible maternal distress. Despite controlling for these variables, experimental replicates of S.Tm infection during pregnancy differed significantly. Some of this variability may be attributed to differences in bacterial seeding and initial organ colonization in vivo. It has been estimated that after 2h of infection with 103 S.Tm i.v. in pregnant

129.B6F1 mice ~10% of bacteria trafficked to various organs including spleen, liver and placenta (60).

Because in vivo S.Tm doubling time in placenta is rapid (<2h) (60), if >10% of bacteria trafficked to the placenta, this may account for the cases of mice with massive placental S.Tm burdens (as was observed in dam #3715 and archival tissue sample dams #3293 and #3376). Other explanations for the variable

128

pathological phenotypes between replicate animals could be based on genetic variability within the inbred mice used in my studies.

The use of genetically-defined, inbred mouse strains homozygous at all genetic loci is an essential tool to limit variability in biomedical research. Inbred strains are defined as having been created by more than 20 generations of filial or sibling matings. In reality, a strain will not be completely homozygous at all loci until at least the 40th inbreeding generation because of residual heterozygosity (215,216). According to the Jackson Laboratory Mouse Strain Data sheet for the congenic background strain used in my studies,

C57BL/6J with functional Slc11a1 (i.e. Slc11a1+/+), were only backcrossed 12 filial generations. Therefore, some of the post infection pathological differences that I observed may be attributed to mild genetic variability between the pregnant dams. The phenomena of genetic drift and the appearance of new mutations within stable, inbred mouse strains occur and must also be considered as an explanation contributing to the variability between my animals (217). Substantially increased numbers of replicate dams would have been needed to estimate the frequency of fetal YS infection and bacteremia. In addition to limits in technical/thesis time for this, such increased numbers were not justifiable economically or ethically in regards to appropriate utilization of research animals. The Canadian Council on Animal Care (CCAC) is responsible for the oversight of research animal usage in Canada. One of the major tenets of the CCAC is "Reduction", referring to any strategy that will result in fewer animals being used. Therefore, it is important to design animal studies to use the minimum number of animals necessary to achieve scientific objectives without wasting crucial resources.

The model that I used for studying S.Tm pathogenesis during mouse pregnancy (i.e. mid-pregnancy infection at GD 12 with 103 S.Tm i.v.) has limited value for further studies because of the low levels of detectable pathology and its sporadic nature. Tissues examined at 72h post-infection in Slc11a1 competent mice revealed either limited S.Tm+ foci and no maternal immune response to infection, or massive infiltration of bacteria in areas outside the placenta coinciding with inflammation and pathology in an

129

occasional animal. Histological examinations could not be conducted 24h later (96h post-infection), due to significant fetal loss and severe maternal illness (157). I believe increasing the bacterial dose or altering the route, or timing of S.Tm administration would resolve these issues.

A dose-response experiment should have been conducted in the novel mutant strains (Ripk3-/-

Slc11a1+/+ and Slc11a1+/+) used in this study to determine an appropriate i.v. S.Tm dose to achieve consistent observations of gross pathology. The selected dose was extrapolated from prior work in other mouse strains. During my studies, I received tissues from 3 infected Ripk3-/-Slc11a1+/+ dams at 72h post- infection that had received ~1350 S.Tm i.v. instead of the validated dose of 1000 S.Tm i.v. Since these dams had a 35% greater bacteria exposure than all other infected study animals, I excluded data from these specimens during my analyses to ensure precision in comparisons. Tissues from these dams displayed variable fetal loss (ranging from 20%-100%), massive S.Tm infection in all IS (resorbing and viable), and profound chorioamnionitis and fetal YS infection in viable IS (data not shown). Therefore, I postulate that the infectious dose used for i.v. inoculation of Slc11a1+/+ function must be higher and that this should be re- evaluated and reported in the novel mutant strains. Dose-response studies conducted in similar murine infection studies (mid-pregnancy infection with i.v. bacterial challenge) with other food-borne pathogens like

Listeria monocytogenes used infectious doses ranging from 5×103 - 5×105 CFUs (218). Based on my observations, a dose-response experiment should be conducted in the novel mouse strains using 2-fold to 10- fold higher doses than the 103 inoculum used in this study.

Increasing the dose of i.v. S.Tm may provoke a more acute, severe phenotype. Therefore, alternative exposure routes should be investigated and standardized using dose response curves in the novel mutant strains. Infection of 108 S.Tm by oral gavage in GD 10 C57BL/6J mice lacking Slc11a1 function gave somewhat consistent placental S.Tm infection at 5 days post-infection (157). Oral inoculation of S.Tm with higher bacterial doses (up to 5×108) was attempted in Slc11a1+/+ mice, but consistent placental S.Tm infection could not be established at 5 days post-infection (157).

130

Previously published work from my collaborators had shown that infection earlier in pregnancy (as early as GD 10) with identical S.Tm dose resulted in profound placental S.Tm infection and statistically significantly increases in expression of pro-inflammatory cytokine mRNA transcripts at 72h post-infection compared to naïve control tissues (60,155). Therefore, the timing of i.v. infection to the onset of placental circulation on GD 10 may provide a more robust model.

4.3 Future Directions

The mouse model used in this thesis was not found to be efficacious for the study of Salmonella pathogenesis during pregnancy. My observations in mouse placenta indicate that trophoblast susceptibility to

S.Tm infection depends on Slc11a1 function. Therefore, efforts should be focused on investigating the role of NRAMP1 as an innate immune defense mechanism in human placenta. Based upon my results, my collaborators at the University of Rochester are currently studying NRAMP1 expression in human placental explant cultures using WM-IF. NRAMP1 polymorphisms have been noted in human populations (219-222), and these may explain why Salmonella infection in human pregnancy does not always result in neonatal death. Variant populations may be at an increased risk for severe Salmonella infection during pregnancy, however, NRAMP1 polymorphisms have only been studied to date in the context of Mycobacterium tuberculosis infection in non-pregnant individuals. Further studies are required to evaluate associations between individuals with NRAMP1 polymorphisms and susceptibility to severe Salmonella infection.

I was the first to note specific infection of mouse PYS and VYS cells in vivo by S.Tm using IHC.

Ultrastructural examination of infected YS cells would be beneficial to confirm intracellular bacterial localization in YS cells. S.Tm infection and tropism for YS has yet to be described in the literature and warrants further investigation. In vitro mouse YS primary cell culture and other models are available and may be useful for the study of S.Tm permissiveness and pathogenesis in specific YS cell types. F9 teratocarcinoma cells or mouse embryonic stem cells can be induced to differentiate into PYS or VYS cells 131

(223-226). Although the structure and function of human YS differs from that of mouse, this avenue of research has potential application to investigations of Salmonella infection in the first trimester of pregnancy when the YS is a major source of fetal nutrition and erythrocytes. While it is known that Salmonella infection in early human pregnancy results in spontaneous abortion (41,227), it is unknown whether human

YS tissues are a potential reservoir or route for transplacental infection of Salmonella from mother to fetus.

4.4 Conclusion

Based on the available data, this thesis suggests that Ripk3 and necroptosis have no significant roles in either promotion or prevention of progressive Salmonella infection during mouse pregnancy. It also provides pilot data that NRAMP1 controls placental localization and lethality may be due to YS infection.

Thus, bacterial properties and not immune responsiveness may account for adverse pregnancy outcomes in

Salmonella-infected mice.

132

References

1. Fàbrega A, Vila J. Salmonella enterica serovar Typhimurium skills to succeed in the host: Virulence and regulation. Clinical Microbiology Reviews 2013;26(2):308-41.

2. Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson M-A, Roy SL, et al. Foodborne illness acquired in the United States major pathogens. Emerging Infectious Diseases 2011;17(1):7-15.

3. Kariuki S, Revathi G, Kariuki N, Kiiru J, Mwituria J, Muyodi J, et al. Invasive multidrug-resistant non- typhoidal Salmonella infections in Africa: Zoonotic or anthroponotic transmission? Journal of Medical Microbiology 2006;55(5):585-91.

4. Kariuki S, Revathi G, Kariuki N, Kiiru J, Mwituria J, Hart CA. Characterisation of community acquired non-typhoidal Salmonella from bacteraemia and diarrhoeal infections in children admitted to hospital in Nairobi, Kenya. BMC Microbiology 2006;6:101.

5. Gordon MA, Graham SM, Walsh AL, Wilson L, Phiri A, Molyneux E, et al. Epidemics of invasive Salmonella enterica serovar Enteritidis and S. Enterica serovar Typhimurium infection associated with multidrug resistance among adults and children in Malawi. Clinical Infectious Diseases 2008;46(7):963-9.

6. Reddy EA, Shaw AV, Crump JA. Community-acquired bloodstream infections in Africa: A systematic review and meta-analysis. Lancet Infectious Diseases 2010;10(6):417-32.

7. Feasey NA, Masesa C, Jassi C, Faragher EB, Mallewa J, Mallewa M, et al. Three epidemics of invasive multidrug-resistant Salmonella bloodstream infection in Blantyre, Malawi, 1998-2014. Clinical Infectious Diseases 2015;61 Suppl 4:S363-71.

8. McGregor AC, Waddington CS, Pollard AJ. Prospects for prevention of Salmonella infection in children through vaccination. Current Opinion in Infectious Diseases 2013;26(3):254-62.

9. Kaufmann SH, Raupach B, Finlay BB. Introduction: Microbiology and immunology: Lessons learned from Salmonella. Microbes and Infection 2001;3(14):1177-81.

10. Khumalo J, Saidi B, Mbanga J. Evolution of antimicrobial resistance of Salmonella Enteritidis (1972-- 2005). Onderstepoort Journal of Veterinary Research 2014;81(1):1-6.

11. Campos J, Cristino L, Peixe L, Antunes P. MCR-1 in multidrug-resistant and copper-tolerant clinically relevant Salmonella 1,4,[5],12:I:- and S. Rissen clones in portugal, 2011 to 2015. European Surveillance 2016;21(26):1-5.

12. Hu Y, Liu F, Lin IY, Gao GF, Zhu B. Dissemination of the mcr-1 colistin resistance gene. The Lancet Infectious Diseases 2016;16(2):146-7.

13. Haraga A, Ohlson MB, Miller SI. Salmonellae interplay with host cells. Nature Reviews Microbiology 2008;6(1):53-66. 133

14. Sanchez S, Hofacre CL, Lee MD, Maurer JJ, Doyle MP. Animal sources of Salmonellosis in humans. Journal of the American Veterinary Medical Association 2002;221(4):492-7.

15. Sanyal D, Douglas T, Roberts R. Salmonella infection acquired from reptilian pets. Archives of Disease in Childhood 1997;77(4):345-6.

16. Cody SH, Abbott SL, Marfin AA, Schulz B, Wagner P, Robbins K, et al. Two outbreaks of multidrug- resistant Salmonella serotype Typhimurium DT104 infections linked to raw-milk cheese in northern california. JAMA 1999;281(19):1805-10.

17. American Academy of Pediatrics. Consumption of raw or unpasteurized milk and milk products by pregnant women and children. Pediatrics 2014;133(1):175-9.

18. Hedberg CW, David MJ, White KE, MacDonald KL, Osterholm MT. Role of egg consumption in sporadic Salmonella Enteritidis and Salmonella Typhimurium infections in Minnesota. The Journal of Infectious Diseases 1993;167(1):107-11.

19. Centers for Disease Control and Prevention (CDC). Multistate outbreak of Salmonella serotype Tennessee infections associated with peanut butter--United States, 2006-2007. MMWR Morbidity and Mortality Weekly Report 2007;56(21):521-4.

20. Cavallaro E, Date K, Medus C, Meyer S, Miller B, Kim C, et al. Salmonella Typhimurium infections associated with peanut products. New England Journal of Medicine 2011;365(7):601-10.

21. Dechet AM, Scallan E, Gensheimer K, Hoekstra R, Gunderman-King J, Lockett J, et al. Outbreak of multidrug-resistant Salmonella enterica serotype Typhimurium definitive type 104 infection linked to commercial ground beef, northeastern United States, 2003-2004. Clinical Infectious Diseases 2006;42(6):747-52.

22. Behravesh CB, Blaney D, Medus C, Bidol SA, Phan Q, Soliva S, et al. Multistate outbreak of Salmonella serotype Typhimurium infections associated with consumption of restaurant tomatoes, USA, 2006: Hypothesis generation through case exposures in multiple restaurant clusters. Epidemiology & Infection 2012;140(11):2053-61.

23. Selsted ME, Miller SI, Henschen AH, Ouellette AJ. Enteric defensins: Antibiotic peptide components of intestinal host defense. The Journal of Cell Biology 1992;118(4):929-36.

24. Prouty AM, Brodsky IE, Falkow S, Gunn JS. Bile-salt-mediated induction of antimicrobial and bile resistance in Salmonella Typhimurium. Microbiology 2004;150(Pt 4):775-83.

25. Francis CL, Starnbach MN, Falkow S. Morphological and cytoskeletal changes in epithelial cells occur immediately upon interaction with Salmonella Typhimurium grown under low-oxygen conditions. Molecular Microbiology 1992;6(21):3077-87.

134

26. Takeuchi A. Electron microscope studies of experimental Salmonella infection. I. Penetration into the intestinal epithelium by Salmonella Typhimurium. The American Journal of Pathology 1967;50(1):109-36.

27. Finlay BB, Ruschkowski S, Dedhar S. Cytoskeletal rearrangements accompanying Salmonella entry into epithelial cells. Journal of Cell Science 1991;99(Pt 2):283-96.

28. Jones BD, Ghori N, Falkow S. Salmonella Typhimurium initiates murine infection by penetrating and destroying the specialized epithelial M cells of the peyer's patches. The Journal of Experimental Medicine 1994;180(1):15-23.

29. Galán JE, Curtiss R. Cloning and molecular characterization of genes whose products allow Salmonella Typhimurium to penetrate tissue culture cells. Proceedings of the National Academy of Sciences of the United States of America1989;86(16):6383-7.

30. Collazo CM, Galán JE. The invasion-associated type-iii protein secretion system in Salmonella--a review. Gene 1997;192(1):51-9.

31. Vazquez-Torres A, Jones-Carson J, Bäumler AJ, Falkow S, Valdivia R, Brown W, et al. Extraintestinal dissemination of Salmonella by cd18-expressing phagocytes. Nature 1999;401(6755):804-8.

32. Doffinger R, Patel S, Kumararatne DS. Human immunodeficiencies that predispose to intracellular bacterial infections. Current Opinion in Rheumatology 2005;17(4):440-6.

33. Shimizu T, Ezaki Y, Tamura H, Kobayashi H, Shibamoto T, Nakai Y, et al. [Studies on campylobacter and Salmonella in feces of pregnant women and newborn infants]. Nihon Sanka Fujinka Gakkai Zasshi 1986;38(4):493-8.

34. Roberts C, Wilkins EG. Salmonella screening of pregnant women. Journal of Hospital Infection 1987;10(1):67-72.

35. Citernesi A, Spinelli G, Piazzesi G, Innocenti S, Curiel P. [Screening for Salmonella in pregnancy]. Minerva Ginecologica 1994;46(12):681-6.

36. Scialli AR, Rarick TL. Salmonella sepsis and second-trimester pregnancy loss. Obstetrics & Gynecology 1992;79(5(Pt 2)):820-1.

37. Ault KA, Kennedy M, Seoud MA, Reiss R. Maternal and neonatal infection with Salmonella Heidelberg: A case report. Infectious Diseases in Obstetrics & Gynecology 1993;1(1):46-8.

38. Roll C, Schmid EN, Menken U, Hanssler L. Fatal Salmonella Enteritidis sepsis acquired prenatally in a premature infant. Obstetrics & Gynecology 1996;88(4):692-3.

39. Schloesser RL, Schaefer V, Groll AH. Fatal transplacental infection with non-typhoidal Salmonella. Scandinavian Journal of Infectious Diseases 2004;36(10):773-4.

135

40. Gyang A, Saunders M. Salmonella Mississippi: A rare cause of second trimester miscarriage. Archives of Gynecology & Obstetrics 2008;277(5):437-8.

41. Vigliani MB, Bakardjiev AI. First trimester typhoid fever with vertical transmission of Salmonella Typhi, an intracellular organism. Case Reports in Medicine 2013;2013(973297):1-5.

42. Rai B, Utekar T, Ray R. Preterm delivery and neonatal meningitis due to transplacental acquisition of non-typhoidal Salmonella serovar Montevideo. BMJ Case Reports 2014;2014:1-3.

43. Ozer O, Cebesoy FB, Sari I, Davutoglu V. A case of Salmonella Typhi endocarditis in pregnancy. The American Journal of Medical Sciences 2009;337(3):210-1.

44. Agustsson AI, Olafsson K, Thorisdottir AS. Salmonella osteomyelitis in pregnancy. Acta Obstetricia et Gynecologica Scandinavica 2009;88(10):1171-3.

45. Alvarez JA, Al-Khan A, Ganesh V, Apuzzio JJ. Salmonella as a causative organism of acute pyelonephritis during pregnancy. American Journal of Obstetrics & Gynecology 2004;190(5):1482- 3.

46. Wegmann TG, Lin H, Guilbert L, Mosmann TR. Bidirectional cytokine interactions in the maternal-fetal relationship: Is successful pregnancy a TH2 phenomenon? Immunology Today 1993;14(7):353-6.

47. Kourtis AP, Read JS, Jamieson DJ. Pregnancy and infection. New England Journal of Medicine 2014;370(23):2211-8.

48. Tezabwala BU, Johnson PM, Rees RC. Inhibition of pregnancy viability in mice following IL-2 administration. Immunology 1989;67(1):115-9.

49. Chaouat G, Menu E, Clark DA, Dy M, Minkowski M, Wegmann TG. Control of fetal survival in CBA x DBA/2 mice by lymphokine therapy. Journal of Reproduction & Fertility 1990;89(2):447-58.

50. Chaouat G, Assal Meliani A, Martal J, Raghupathy R, Elliott JF, Elliot J, et al. IL-10 prevents naturally occurring fetal loss in the CBA x DBA/2 mating combination, and local defect in IL-10 production in this abortion-prone combination is corrected by in vivo injection of IFN-tau. Journal of Immunology 1995;154(9):4261-8.

51. Szekeres-Bartho J, Wegmann TG. A progesterone-dependent immunomodulatory protein alters the Th1/Th2 balance. Journal of Reproductive Immunology 1996;31(1-2):81-95.

52. Szekeres-Bartho J, Par G, Szereday L, Smart CY, Achatz I. Progesterone and non-specific immunologic mechanisms in pregnancy. American Journal of Reproductive Immunology 1997;38(3):176-82.

53. Krishnan L, Guilbert LJ, Wegmann TG, Belosevic M, Mosmann TR. T helper 1 response against leishmania major in pregnant C57BL/6 mice increases implantation failure and fetal resorptions. Correlation with increased IFN-gamma and TNF and reduced IL-10 production by placental cells. The Journal of Immunology 1996;156(2):653-62. 136

54. Marzi M, Vigano A, Trabattoni D, Villa ML, Salvaggio A, Clerici E, Clerici M. Characterization of type 1 and type 2 cytokine production profile in physiologic and pathologic human pregnancy. Clinical & Experimental Immunology 1996;106(1):127-33.

55. Reinhard G, Noll A, Schlebusch H, Mallmann P, Ruecker AV. Shifts in the Th1/Th2 balance during human pregnancy correlate with apoptotic changes. Biochemical and Biophysical Research Communications 1998;245(3):933-8.

56. Klipple GL, Cecere FA. Rheumatoid arthritis and pregnancy. Rheumatic Diseases Clinics of North America 1989;15(2):213-39.

57. Cauldwell M, Nelson-Piercy C. Maternal and fetal complications of systemic lupus erythematosus. The Obstetrician & Gynaecologist 2012;14(3):167-74.

58. Jamieson DJ, Theiler RN, Rasmussen SA. Emerging infections and pregnancy. Emerging Infectious Diseases 2006;12(11):1638.

59. Watson ED, Cross JC. Development of structures and transport functions in the mouse placenta. Physiology (Bethesda) 2005;20:180-93.

60. Chattopadhyay A, Robinson N, Sandhu JK, Finlay BB, Sad S, Krishnan L. Salmonella enterica serovar Typhimurium-induced placental inflammation and not bacterial burden correlates with pathology and fatal maternal disease. Infection & Immunity 2010;78(5):2292-301.

61. Nguyen T, Robinson N, Allison SE, Coombes BK, Sad S, Krishnan L. IL-10 produced by trophoblast cells inhibits phagosome maturation leading to profound intracellular proliferation of Salmonella enterica Typhimurium. Placenta 2013;34(9):765-74.

62. Nguyen T, Perry I, Murphy S, Gruslin A, Krishnan L. Salmonella enterica serovar Typhimurium productively infects human primary trophoblast cells and induces inflammatory cell death. [abstract]. Placenta. 2014; 35(9):A62-A63.

63. Perry ID, Nguyen T, Miller RK, Krishnan L, Murphy SP. Salmonella Typhimurium infects placental explants from multiple gestation ages. [abstract]. American Journal of Reproductive Immunology 2015;73(Suppl)S2:34–89.

64. Takeda K, Akira S. Roles of toll-like receptors in innate immune responses. Genes to Cells 2001;6(9):733-42.

65. Akira S. Toll-like receptor signaling. Journal of Biological Chemistry 2003;278(40):38105-8.

66. Hapfelmeier S, Stecher B, Barthel M, Kremer M, Müller AJ, Heikenwalder M, et al. The Salmonella pathogenicity island (SPI)-2 and SPI-1 type III secretion systems allow salmonella serovar typhimurium to trigger colitis via myd88-dependent and myd88-independent mechanisms. The Journal of Immunology 2005;174(3):1675-85.

137

67. O'brien AD, Rosenstreich DL, Scher I, Campbell GH, MacDermott RP, Formal SB. Genetic control of susceptibility to Salmonella Typhimurium in mice: Role of the LPS gene. The Journal of Immunology 1980;124(1):20-4.

68. Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, Du X, et al. Defective LPS signaling in C3H/hej and C57BL/10sccr mice: Mutations in tlr4 gene. Science 1998;282(5396):2085-8.

69. Royle MC, Tötemeyer S, Alldridge LC, Maskell DJ, Bryant CE. Stimulation of toll-like receptor 4 by lipopolysaccharide during cellular invasion by live Salmonella Typhimurium is a critical but not exclusive event leading to macrophage responses. The Journal of Immunology 2003;170(11):5445- 54.

70. Vazquez-Torres A, Vallance BA, Bergman MA, Finlay BB, Cookson BT, Jones-Carson J, Fang FC. Toll-like receptor 4 dependence of innate and adaptive immunity to Salmonella: Importance of the kupffer cell network. The Journal of Immunology 2004;172(10):6202-8.

71. Smith KD, Andersen-Nissen E, Hayashi F, Strobe K, Bergman MA, Barrett SL, et al. Toll-like receptor 5 recognizes a conserved site on flagellin required for protofilament formation and bacterial motility. Nature Immunology 2003;4(12):1247-53.

72. Feuillet V, Medjane S, Mondor I, Demaria O, Pagni PP, Galán JE, et al. Involvement of toll-like receptor 5 in the recognition of flagellated bacteria. Proceedings of the National Academy of Sciences of the United States of America 2006;103(33):12487-92.

73. Klaffenbach D, Rascher W, Röllinghoff M, Dötsch J, Meissner U, Schnare M. Regulation and signal transduction of toll-like receptors in human chorioncarcinoma cell lines. American Journal of Reproductive Immunology 2005;53(2):77-84.

74. Mitsunari M, Yoshida S, Shoji T, Tsukihara S, Iwabe T, Harada T, Terakawa N. Macrophage-activating lipopeptide-2 induces cyclooxygenase-2 and prostaglandin E(2) via toll-like receptor 2 in human placental trophoblast cells. Journal Reproductive Immunology 2006;72(1-2):46-59.

75. Abrahams VM, Bole-Aldo P, Kim YM, Straszewski-Chavez SL, Chaiworapongsa T, Romero R, Mor G. Divergent trophoblast responses to bacterial products mediated by TLRs. Journal of Immunology 2004;173(7):4286-96.

76. Holmlund U, Cebers G, Dahlfors AR, Sandstedt B, Bremme K, Ekström ES, Scheynius A. Expression and regulation of the pattern recognition receptors toll-like receptor-2 and toll-like receptor-4 in the human placenta. Immunology 2002;107(1):145-51.

77. Kumazaki K, Nakayama M, Yanagihara I, Suehara N, Wada Y. Immunohistochemical distribution of toll-like receptor 4 in term and preterm human placentas from normal and complicated pregnancy including chorioamnionitis. Human Pathology 2004;35(1):47-54.

78. Rossant J, Cross JC. Placental development: Lessons from mouse mutants. Nature Review Genetics 2001;2(7):538-48.

138

79. Adamson SL, Lu Y, Whiteley KJ, Holmyard D, Hemberger M, Pfarrer C, Cross JC. Interactions between trophoblast cells and the maternal and fetal circulation in the mouse placenta. Developmental Biology 2002;250(2):358-73.

80. Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery CA, et al. Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes & Development 1995;9(18):2266-78.

81. Das A, Mantena SR, Kannan A, Evans DB, Bagchi MK, Bagchi IC. De novo synthesis of estrogen in pregnant uterus is critical for stromal decidualization and angiogenesis. Proceedings of the National Academy of Sciences of the United States of America 2009;106(30):12542-7.

82. Favaro R, Abrahamsohn PA, Zorn MT. 11 – Decidualization and Endometrial Extracellular Matrix Remodeling. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 125-142.

83. Rätsep MT, Felker AM, Kay VR, Tolusso L, Hofmann AP, Croy BA. Uterine natural killer cells: Supervisors of vasculature construction in early decidua basalis. Reproduction 2015;149(2):R91- 102.

84. Hemberger M, Cross JC. Genes governing placental development. Trends in Endocrinology & Metabolism 2001;12(4):162-8.

85. Bolon B. 14 – Pathology Analysis of the Placenta. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 175-188.

86. Zorn TM, Bijovsky AT, Bevilacqua EM, Abrahamsohn PA. Phagocytosis of collagen by mouse decidual cells. The Anatomical Record 1989;225(2):96-100.

87. Simmons DG, Fortier AL, Cross JC. Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta. Developmental Biology 2007;304(2):567-78.

88. Hu D, Cross JC. Development and function of trophoblast giant cells in the rodent placenta. The International Journal of Developmental Biology 2010;54(2-3):341-54.

89. Hemberger M. IFPA award in placentology lecture--characteristics and significance of trophoblast giant cells. Placenta 2008;29:4-9.

90. Guilbert L, Robertson SA, Wegmann TG. The trophoblast as an integral component of a macrophage- cytokine network. Immunology & Cell Biology 1993;71(Pt 1):49-57.

91. Arceci RJ, Shanahan F, Stanley ER, Pollard JW. Temporal expression and location of colony-stimulating factor 1 (CSF-1) and its receptor in the female reproductive tract are consistent with CSF-1-regulated placental development. Proceedings of the National Academy of Sciences of the United States of America 1989;86(22):8818-22.

139

92. Guleria I, Pollard JW. The trophoblast is a component of the innate immune system during pregnancy. Nature Medicine 2000;6(5):589-93.

93. Amarante-Paffaro A, Queiroz GS, Corrêa ST, Spira B, Bevilacqua E. Phagocytosis as a potential mechanism for microbial defense of mouse placental trophoblast cells. Reproduction 2004;128(2):207-18.

94. Gagioti S, Colepicolo P, Bevilacqua E. Post-implantation mouse embryos have the capability to generate and release reactive oxygen species. Reproduction, Fertility & Development 1995;7(5):1111-6.

95. Gagioti S, Colepicolo P, Bevilacqua E. Reactive oxygen species and the phagocytosis process of hemochorial trophoblast. Ciência e Cultura (Säo Paulo) 1996;48(1/2):37-42.

96. Gagioti S, Scavone C, Bevilacqua E. Participation of the mouse implanting trophoblast in nitric oxide production during pregnancy. Biology of Reproduction 2000, Feb;62(2):260-8.

97. Gomes SZ, Lorenzon AR, Vieira JS, Rocha CR, Bandeira C, Hoshida MS, et al. Expression of NADPH oxidase by trophoblast cells: Potential implications for the postimplanting mouse embryo. Biology of Reproduction 2012, Feb;86(2):56.

98. Bevilacqua E, Gomes SZ, Lorenzon AR, Hoshida MS, Amarante-Paffaro AM. NADPH oxidase as an important source of reactive oxygen species at the mouse maternal-fetal interface: Putative biological roles. Reproductive BioMedicine Online 2012;25(1):31-43.

99. Simmons DG, Cross JC. Determinants of trophoblast lineage and cell subtype specification in the mouse placenta. Developmental Biology 2005;284(1):12-24.

100. Coan PM, Conroy N, Burton GJ, Ferguson-Smith AC. Origin and characteristics of glycogen cells in the developing murine placenta. Developmental Dynamics 2006;235(12):3280-94.

101. Burdon C, Mann C, Cindrova-Davies T, Ferguson-Smith AC, Burton GJ. Oxidative stress and the induction of cyclooxygenase enzymes and apoptosis in the murine placenta. Placenta 2007;28(7):724-33.

102. Wiemers DO, Shao L-J, Ain R, Dai G, Soares MJ. The mouse prolactin gene family locus. Endocrinology 2003;144(1):313-25.

103. Wynne F, Ball M, McLellan AS, Dockery P, Zimmermann W, Moore T. Mouse pregnancy-specific glycoproteins: Tissue-specific expression and evidence of association with maternal vasculature. Reproduction 2006;131(4):721-32.

104. Simmons DG, Rawn S, Davies A, Hughes M, Cross JC. Spatial and temporal expression of the 23 murine prolactin/placental lactogen-related genes is not associated with their position in the locus. BMC Genomics 2008;9(1):1.

140

105. Knox K, Leuenberger D, Penn AA, Baker JC. Global hormone profiling of murine placenta reveals secretin expression. Placenta 2011;32(11):811-6.

106. Guillemot F, Caspary T, Tilghman SM, Copeland NG, Gilbert DJ, Jenkins NA, et al. Genomic imprinting of mash2, a mouse gene required for trophoblast development. Nature Genetics 1995;9(3):235-42.

107. Tanaka M, Gertsenstein M, Rossant J, Nagy A. Mash2 acts cell autonomously in mouse spongiotrophoblast development. Developmental Biology 1997;190(1):55-65.

108. Ain R, Canham LN, Soares MJ. Gestation stage-dependent intrauterine trophoblast cell invasion in the rat and mouse: Novel endocrine phenotype and regulation. Developmental Biology 2003;260(1):176-90.

109. Malassine A, Frendo J-L, Evain-Brion D. A comparison of placental development and endocrine functions between the human and mouse model. Human Reproduction Update 2003;9(6):531-9.

110. Soares MJ. The prolactin and growth hormone families: Pregnancy-specific hormones/cytokines at the maternal-fetal interface. Reproductive Biology and Endocrinology 2004;2(1):1.

111. Oh-McGinnis R, Bogutz AB, Lefebvre L. Partial loss of ascl2 function affects all three layers of the mature placenta and causes intrauterine growth restriction. Developmental Biology 2011;351(2):277-86.

112. Simmons DG. 12 – Postimplantation Development of the Chorioallantoic Placenta. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 143-161.

113. Coan PM, Ferguson-Smith AC, Burton GJ. Ultrastructural changes in the interhaemal membrane and junctional zone of the murine chorioallantoic placenta across gestation. Journal of Anatomy 2005;207(6):783-96.

114. Salamat M, Miosge N, Herken R. Development of Reichert's membrane in the early mouse embryo. Anatomy and Embryology 1995;192(3):275-81.

115. Gardner RL. Origin and differentiation of extraembryonic tissues in the mouse. International Review of Experimental Pathology 1982;24:63-133.

116. Jollie WP. Development, morphology, and function of the yolk-sac placenta of laboratory rodents. Teratology 1990;41(4):361-81.

117. Rennie MY, Mu J, Rahman A, Qu D, Whiteley KJ, Sled JG. 16 – The Uteroplacental, Fetoplacental, and Yolk Sac Circulations in the Mouse. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 201-210.

141

118. Harju K, Glumoff V, Hallman M. Ontogeny of toll-like receptors TLR2 and TLR4 in mice. Pediatric Research 2001;49(1):81-3.

119. Gonzalez JM, Xu H, Ofori E, Elovitz MA. Toll-like receptors in the uterus, cervix, and placenta: Is pregnancy an immunosuppressed state? American Journal of Obstetrics & Gynecology 2007;197(3):296.e1-6.

120. Harju K, Ojaniemi M, Rounioja S, Glumoff V, Paananen R, Vuolteenaho R, Hallman M. Expression of toll-like receptor 4 and endotoxin responsiveness in mice during perinatal period. Pediatric Research 2005;57(5 Pt 1):644-8.

121. Arce RM, Barros SP, Wacker B, Peters B, Moss K, Offenbacher S. Increased TLR4 expression in murine placentas after oral infection with periodontal pathogens. Placenta 2009;30(2):156-62.

122. Arce RM, Caron KM, Barros SP, Offenbacher S. Toll-like receptor 4 mediates intrauterine growth restriction after systemic Campylobacter rectus infection in mice. Molecular Oral Microbiology 2012;27(5):373-81.

123. Elovitz MA, Mrinalini C. Can medroxyprogesterone acetate alter toll-like receptor expression in a mouse model of intrauterine inflammation? American Journal of Obstetrics & Gynecology 2005;193(3):1149-55.

124. Erlebacher A. 19 – Leukocyte Population Dynamics and Functions at the Maternal–Fetal Interface. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 227-242.

125. Croy BA, Chen Z, Hofmann AP, Lord EM, Sedlacek AL, Gerber SA. Imaging of vascular development in early mouse decidua and its association with leukocytes and trophoblasts. Biology of Reproduction 2012;87(5):125.

126. Lima PD, Zhang J, Dunk C, Lye SJ, Croy BA. Leukocyte driven-decidual angiogenesis in early pregnancy. Cellular & Molecular Immunology 2014;11(6):522-37.

127. Blois SM, Klapp BF, Barrientos G. Decidualization and angiogenesis in early pregnancy: Unraveling the functions of DC and NK cells. Journal of Reproductive Immunology 2011;88(2):86-92.

128. Hofmann AP, Gerber SA, Croy BA. Uterine natural killer cells pace early development of mouse decidua basalis. Molecular Human Reproduction 2014;20(1):66-76.

129. Felker AM, Croy BA. Uterine natural killer cell partnerships in early mouse decidua basalis. Journal of Leukocyte Biology 2016;100(4):645-55.

130. Delgado SR, McBey BA, Yamashiro S, Fujita J, Kiso Y, Croy BA. Accounting for the peripartum loss of granulated metrial gland cells, a natural killer cell population, from the pregnant mouse uterus. Journal of Leukocyte Biology 1996;59(2):262-9.

142

131. Paffaro VA, Bizinotto MC, Joazeiro PP, Yamada AT. Subset classification of mouse uterine natural killer cells by DBA lectin reactivity. Placenta 2003;24(5):479-88.

132. Edwards AK, Janzen-Pang J, Peng A, Tayade C, Carniato A, Lima PDA, Tse D. 3 – Microscopic Anatomy of the Pregnant Mouse Uterus Throughout Gestation. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 43-67.

133. Kruse A, Martens N, Fernekorn U, Hallmann R, Butcher EC. Alterations in the expression of homing- associated molecules at the maternal/fetal interface during the course of pregnancy. Biology of Reproduction 2002;66(2):333-45.

134. Naito M, Yamamura F, Nishikawa S, Takahashi K. Development, differentiation, and maturation of fetal mouse yolk sac macrophages in cultures. Journal of Leukocyte Biology 1989;46(1):1-10.

135. Takahashi K, Naito M, Katabuchi H, Higashi K. Development, differentiation, and maturation of macrophages in the chorionic villi of mouse placenta with special reference to the origin of Hofbauer cells. Journal of Leukocyte Biology 1991;50(1):57-68.

136. Tang Z, Abrahams VM, Mor G, Guller S. Placental Hofbauer cells and complications of pregnancy. Annals of the New York Academy of Sciences 2011;1221(1):103-8.

137. Quicke KM, Bowen JR, Johnson EL, McDonald CE, Ma H, O'Neal JT, et al. Zika virus infects human placental macrophages. Cell Host & Microbe 2016;20(1):83-90.

138. Tabata T, Petitt M, Puerta-Guardo H, Michlmayr D, Wang C, Fang-Hoover J, et al. Zika virus targets different primary human placental cells, suggesting two routes for vertical transmission. Cell Host & Microbe 2016;20(2):155-66.

139. Rosenberg AZ, Yu W, Hill DA, Reyes CA, Schwartz DA. Placental pathology of Zika virus: Viral infection of the placenta induces villous stromal macrophage (Hofbauer cell) proliferation and hyperplasia. Archives of Pathology & Laboratory Medicine 2016, Sep 28. [Epub ahead of print]

140. Jurado KA, Simoni MK, Tang Z, Uraki R, Hwang J, Householder S, et al. Zika virus productively infects primary human placenta-specific macrophages. JCI Insight 2016;1(13): :e88461.

141. Johnson EL, Chakraborty R. Placental hofbauer cells limit HIV-1 replication and potentially offset mother to child transmission (MTCT) by induction of immunoregulatory cytokines. Retrovirology 2012;9:101.

142. Johnson EL, Chu H, Byrareddy SN, Spearman P, Chakraborty R. Placental Hofbauer cells assemble and sequester HIV-1 in tetraspanin-positive compartments that are accessible to broadly neutralizing antibodies. Journal of the International AIDS Society 2015;18:19385.

143. Murr SM, Stabenfeldt GH, Bradford GE, Geschwind II. Plasma progesterone during pregnancy in the mouse 1. Endocrinology 1974;94(4):1209-11. 143

144. Sugimoto Y, Yamasaki A, Segi E, Tsuboi K, Aze Y, Nishimura T, et al. Failure of parturition in mice lacking the prostaglandin F receptor. Science 1997;277(5326):681-3.

145. Shynlova O, Nedd-Roderique T, Li Y, Dorogin A, Nguyen T, Lye SJ. Infiltration of myeloid cells into decidua is a critical early event in the labour cascade and post-partum uterine remodeling. Journal of Cellular and Molecular Medicine 2013;17(2):311-24.

146. Shynlova O, Nedd-Roderique T, Li Y, Dorogin A, Lye SJ. Myometrial immune cells contribute to term parturition, preterm labour and post-partum involution in mice. Journal of Cellular and Molecular Medicine 2013;17(1):90-102.

147. Shynlova O, Lye SJ. 32 – Regulation of Parturition. In: Croy BA, Yamada AT, DeMayo FJ, Adamson SL, editors. The Guide to Investigation of Mouse Pregnancy. Boston: Academic Press; 2014. p. 373- 389.

148. Lam-Yuk-Tseung S, Gros P. Genetic control of susceptibility to bacterial infections in mouse models. Cellular Microbiology 2003;5(5):299-313.

149. Vidal SM, Pinner E, Lepage P, Gauthier S, Gros P. Natural resistance to intracellular infections: Nramp1 encodes a membrane phosphoglycoprotein absent in macrophages from susceptible (nramp1D169) mouse strains. Journal of Immunology 1996;157(8):3559-68.

150. Canonne-Hergaux F, Calafat J, Richer E, Cellier M, Grinstein S, Borregaard N, Gros P. Expression and subcellular localization of NRAMP1 in human neutrophil granules. Blood 2002;100(1):268-75.

151. Nevo Y, Nelson N. The NRAMP family of metal-ion transporters. Biochimica Et Biophysica Acta (BBA)-Molecular Cell Research 2006;1763(7):609-20.

152. Plant J, Glynn AA. Natural resistance to Salmonella infection, delayed hypersensitivity and ir genes in different strains of mice. Nature 1974, Mar 22;248(446):345-7.

153. Vidal S, Tremblay ML, Govoni G, Gauthier S, Sebastiani G, Malo D, et al. The ity/lsh/bcg locus: Natural resistance to infection with intracellular parasites is abrogated by disruption of the nramp1 gene. Journal of Experimental Medicine 1995;182(3):655-66.

154. Monack DM, Bouley DM, Falkow S. Salmonella Typhimurium persists within macrophages in the mesenteric lymph nodes of chronically infected Nramp1+/+ mice and can be reactivated by IFN- gamma neutralization. Journal of Experimental Medicine 2004;199(2):231-41.

155. Pejcic-Karapetrovic B, Gurnani K, Russell MS, Finlay BB, Sad S, Krishnan L. Pregnancy impairs the innate immune resistance to Salmonella Typhimurium leading to rapid fatal infection. Journal of Immunology 2007;179(9):6088-96.

156. Luu RA, Gurnani K, Dudani R, Kammara R, van Faassen H, Sirard JC, et al. Delayed expansion and contraction of CD8+ T cell response during infection with virulent Salmonella Typhimurium. Journal of Immunol 2006;177(3):1516-25. 144

157. Wachholz KLC. Placental infection by Salmonella Typhimurium in a murine model: the role of innate immune mediators in cell death at the fetal-maternal interface. MSc [thesis]. Ottawa (ON): University of Ottawa, 2016.

158. Hoiseth SK, Stocker BA. Aromatic-dependent Salmonella Typhimurium are non-virulent and effective as live vaccines. Nature 1981;291(5812):238-9.

159. Noto Llana M, Sarnacki SH, Aya Castañeda Mdel R, Pustovrh MC, Gartner AS, Buzzola FR, et al. Salmonella enterica serovar Enteritidis enterocolitis during late stages of gestation induces an adverse pregnancy outcome in the murine model. PLoS One 2014;9(11):e111282.

160. Lamkanfi M, Dixit VM. Manipulation of host cell death pathways during microbial infections. Cell Host & Microbe 2010;8(1):44-54.

161. Paesold G, Guiney DG, Eckmann L, Kagnoff MF. Genes in the Salmonella pathogenicity island 2 and the Salmonella virulence plasmid are essential for Salmonella-induced apoptosis in intestinal epithelial cells. Cellular Microbiology 2002;4(11):771-81.

162. Zeng H, Wu H, Sloane V, Jones R, Yu Y, Lin P, et al. Flagellin/TLR5 responses in epithelia reveal intertwined activation of inflammatory and apoptotic pathways. American Journal of Physiology & Gastrointestinal Liver Physiology 2006;290(1):G96-G108.

163. Fields PI, Swanson RV, Haidaris CG, Heffron F. Mutants of Salmonella Typhimurium that cannot survive within the macrophage are avirulent. Proceedings of the National Academy of Sciences of the United States of America 1986;83(14):5189-93.

164. Linehan SA, Holden DW. The interplay between Salmonella Typhimurium and its macrophage host-- what can it teach us about innate immunity? Immunology Letters 2003;85(2):183-92.

165. Worley MJ, Nieman GS, Geddes K, Heffron F. Salmonella Typhimurium disseminates within its host by manipulating the motility of infected cells. Proceedings of the National Academy of Sciences of the United States of America 2006;103(47):17915-20.

166. Lamkanfi M, Dixit VM. Mechanisms and functions of inflammasomes. Cell 2014;157(5):1013-22.

167. Ashida H, Mimuro H, Ogawa M, Kobayashi T, Sanada T, Kim M, Sasakawa C. Cell death and infection: A double-edged sword for host and pathogen survival. Journal of Cell Biologyg 2011;195(6):931-42.

168. Fink SL, Cookson BT. Apoptosis, pyroptosis, and necrosis: Mechanistic description of dead and dying eukaryotic cells. Infection & Immunity 2005;73(4):1907-16.

169. Fink SL, Cookson BT. Caspase-1-dependent pore formation during pyroptosis leads to osmotic lysis of infected host macrophages. Cellular Microbiology 2006;8(11):1812-25.

145

170. Fink SL, Cookson BT. Pyroptosis and host cell death responses during Salmonella infection. Cellular Microbiology 2007;9(11):2562-70.

171. Galluzzi L, Vitale I, Abrams JM, Alnemri ES, Baehrecke EH, Blagosklonny MV, et al. Molecular definitions of cell death subroutines: Recommendations of the nomenclature committee on cell death 2012. Cell Death & Differentiation 2012;19(1):107-20.

172. Degterev A, Huang Z, Boyce M, Li Y, Jagtap P, Mizushima N, et al. Chemical inhibitor of nonapoptotic cell death with therapeutic potential for ischemic brain injury. Nature Chemical Biology 2005;1(2):112-9.

173. Hitomi J, Christofferson DE, Ng A, Yao J, Degterev A, Xavier RJ, Yuan J. Identification of a molecular signaling network that regulates a cellular necrotic cell death pathway. Cell 2008;135(7):1311-23.

174. Galluzzi L, Kroemer G. Necroptosis: A specialized pathway of programmed necrosis. Cell 2008;135(7):1161-3.

175. Zhou W, Yuan J. Necroptosis in health and diseases. Seminars in Cell & Developmental Biology 2014;35:14-23.

176. Vandenabeele P, Galluzzi L, Berghe TV, Kroemer G. Molecular mechanisms of necroptosis: An ordered cellular explosion. Nature Reviews Molecular Cell Biology 2010;11(10):700-14.

177. Robinson N, McComb S, Mulligan R, Dudani R, Krishnan L, Sad S. Type I interferon induces necroptosis in macrophages during infection with Salmonella enterica serovar Typhimurium. Nature Immunology 2012;13(10):954-62.

178. He S, Liang Y, Shao F, Wang X. Toll-like receptors activate programmed necrosis in macrophages through a receptor-interacting kinase-3-mediated pathway. Proceedings of the National Academy of Sciences of the United States of America 2011;108(50):20054-9.

179. Kaiser WJ, Sridharan H, Huang C, Mandal P, Upton JW, Gough PJ, et al. Toll-like receptor 3-mediated necrosis via TRIF, RIP3, and MLKL. Journal of Biological Chemistry 2013;288(43):31268-79.

180. Moriwaki K, Chan FK. RIP3: A molecular switch for necrosis and inflammation. Genes & Development 2013;27(15):1640-9.

181. Zhang J, Yang Y, He W, Sun L. Necrosome core machinery: MLKL. Cellular and Molecular Life Sciences 2016;73(11-12):2153-63.

182. Wang H, Sun L, Su L, Rizo J, Liu L, Wang LF, et al. Mixed lineage kinase domain-like protein MLKL causes necrotic membrane disruption upon phosphorylation by RIP3. Molecular Cell 2014;54(1):133-46.

146

183. Cai Z, Jitkaew S, Zhao J, Chiang HC, Choksi S, Liu J, et al. Plasma membrane translocation of trimerized MLKL protein is required for TNF-induced necroptosis. Nature Cell Biology 2014;16(1):55-65.

184. Chen X, Li W, Ren J, Huang D, He WT, Song Y, et al. Translocation of mixed lineage kinase domain- like protein to plasma membrane leads to necrotic cell death. Cell Research 2014;24(1):105-21.

185. Newton K, Dugger DL, Wickliffe KE, Kapoor N, de Almagro MC, Vucic D, et al. Activity of protein kinase RIPK3 determines whether cells die by necroptosis or apoptosis. Science 2014;343(6177):1357-60.

186. Sing A, Merlin T, Knopf H-P, Nielsen PJ, Loppnow H, Galanos C, Freudenberg MA. Bacterial induction of beta interferon in mice is a function of the lipopolysaccharide component. Infection & Immunity 2000;68(3):1600-7.

187. McComb S, Cessford E, Alturki NA, Joseph J, Shutinoski B, Startek JB, et al. Type-I interferon signaling through ISGF3 complex is required for sustained RIP3 activation and necroptosis in macrophages. Proceedings of the National Academy of Sciences of the United States of America 2014;111(31):E3206-13.

188. Bailey L, Caniggia I. Necroptosis: A novel contributor to placental cell death in preeclampsia. Placenta 2015;36(9):A4.

189. Chu X, Chen W, Li N, Hu XZ, Du CT, Yu SX, et al. Cytosolic double-stranded DNA induces nonnecroptotic programmed cell death in trophoblasts via IFI16. Journal of Infectious Diseases 2014;210(9):1476-86.

190. Newton K, Sun X, Dixit VM. Kinase RIP3 is dispensable for normal NF-kappa beta, signaling by the B- cell and T-cell receptors, tumor necrosis factor receptor 1, and toll-like receptors 2 and 4. Molecular and Cell Biology 2004;24(4):1464-9.

191. Arpaia N, Godec J, Lau L, Sivick KE, McLaughlin LM, Jones MB, et al. TLR signaling is required for virulence of an intracellular pathogen. Cell 2011;144(5):675.

192. Kaufman MH. The atlas of mouse development. 1st ed. Academic Press; 1992.

193. Zahl PA, Bjerknes C. Induction of decidua-placental hemorrhage in mice by the endotoxins of certain gram-negative bacteria. Experimental Biology and Medicine 1943;54(3):329-32.

194. Aisemberg J, Vercelli CA, Bariani MV, Billi SC, Wolfson ML, Franchi AM. Progesterone is essential for protecting against LPS-induced pregnancy loss. LIF as a potential mediator of the anti- inflammatory effect of progesterone. PLoS One 2013;8(2):e56161.

195. Shi C, Pamer EG. Monocyte recruitment during infection and inflammation. Nature Reviews Immunology 2011;11(11):762-74.

147

196. Wu J, Huang Z, Ren J, Zhang Z, He P, Li Y, et al. Mlkl knockout mice demonstrate the indispensable role of mlkl in necroptosis. Cell Research 2013;23(8):994-1006.

197. Mühlemann K, Miller RK, Metlay L, Menegus MA. Cytomegalovirus infection of the human placenta: An immunocytochemical study. Human Pathology 1992;23(11):1234-7.

198. Elmore S. Apoptosis: A review of programmed cell death. Toxicologic Pathology 2007;35(4):495-516.

199. He S, Huang S, Shen Z. Biomarkers for the detection of necroptosis. Cellular and Molecular Life Sciences 2016;73(11-12):2177-81.

200. Antunes LC, Arena ET, Menendez A, Han J, Ferreira RB, Buckner MM, et al. Impact of Salmonella infection on host hormone metabolism revealed by metabolomics. Infection & Immunity 2011;79(4):1759-69.

201. Munford RS. Endotoxemia menace, marker, or mistake? Journal of Leukocyte Biology 2016;100(4):687-98.

202. Anderson TD, Meador VP, Cheville NF. Pathogenesis of placentitis in the goat inoculated with Brucella abortus. I. Gross and histologic lesions. Veterinary Pathology 1986;23(3):219-26.

203. Tobias L, Cordes DO, Schurig GG. Placental pathology of the pregnant mouse inoculated with Brucella abortus strain 2308. Veterinary Pathology 1993;30(2):119-29.

204. Watanabe K, Tachibana M, Tanaka S, Furuoka H, Horiuchi M, Suzuki H, Watarai M. Heat shock cognate protein 70 contributes to Brucella invasion into trophoblast giant cells that cause infectious abortion. BMC Microbiology 2008;8:212.

205. Watanabe K, Tachibana M, Kim S, Watarai M. EEVD motif of heat shock cognate protein 70 contributes to bacterial uptake by trophoblast giant cells. Journal of Biomedical Science 2009;16:113.

206. Zohn IE, Sarkar AA. The visceral yolk sac endoderm provides for absorption of nutrients to the embryo during neuralation. Birth Defects Research Part A: Clinical and Molecular Teratology 2010;88(8):593-600.

207. Mu J, Adamson SL. Developmental changes in hemodynamics of uterine artery, utero- and umbilicoplacental, and vitelline circulations in mouse throughout gestation. American Journal of Physiology: Heart and Circulatory Physiology 2006;291(3):H1421-8.

208. Mikkola HK, Orkin SH. The journey of developing hematopoietic stem cells. Development 2006;133(19):3733-44.

209. McGrath KE, Palis J. Hematopoiesis in the yolk sac: More than meets the eye. Experimental Hematology 2005;33(9):1021-8.

148

210. Messenger AJ, Barclay R. Bacteria, iron and pathogenicity. Biochemical Education 1983;11(2):54-63.

211. Frawley ER, Fang FC. The ins and outs of bacterial iron metabolism. Molecular Microbiology 2014;93(4):609-16.

212. Gauthier Y, Isoard P. Increased resistance to Salmonella infection of hypoferremic mice fed a low- protein diet. Microbiology and Immunology 1986;30(5):425-35.

213. Freyer C, Renfree MB. The mammalian yolk sac placenta. Journal of Experimental Zoology Part B: Molecular and Developmental Evolution 2009;312(6):545-54.

214. Burton GJ, Watson AL, Hempstock J, Skepper JN, Jauniaux E. Uterine glands provide histiotrophic nutrition for the human fetus during the first trimester of pregnancy. Journal of Clinical Endocrinology and Metabolism 2002;87(6):2954-9.

215. Jackson Laboratory (Bar Harbor M), Fox RR, Witham BA. Handbook on genetically standardized JAX mice. The Laboratory; 1997.

216. Genetic and phenotypic definition of laboratory mice and rats/what constitutes an acceptable genetic- phenotypic definition; Microbial and phenotypic definition of rats and mice: Proceedings of the 1998 US/japan conference. 1999.

217. Stevens JC, Banks GT, Festing MF, Fisher EM. Quiet mutations in inbred strains of mice. Trends in Molecular Medicine 2007;13(12):512-9.

218. Le Monnier A, Join-Lambert OF, Jaubert F, Berche P, Kayal S. Invasion of the placenta during murine Listeriosis. Infection & Immunity 2006;74(1):663-72.

219. Bellamy R, Ruwende C, Corrah T, McAdam KP, Whittle HC, Hill AV. Variations in the NRAMP1 gene and susceptibility to tuberculosis in West Africans. New England Journal of Medicine 1998;338(10):640-4.

220. Stagas MK, Papaetis GS, Orphanidou D, Kostopoulos C, Syriou S, Reczko M, Drakoulis N. Polymorphisms of the NRAMP1 gene: Distribution and susceptibility to the development of pulmonary tuberculosis in the Greek population. Medical Science Monitor 2011;17(1):PH1-6.

221. Nugraha J, Anggraini R. NRAMP1 polymorphism and susceptibility to lung tuberculosis in Surabaya, Indonesia. Southeast Asian Journal of Tropical Medicin and Public Health 2011;r;42(2):338-41.

222. Li X, Yang Y, Zhou F, Zhang Y, Lu H, Jin Q, Gao L. SLC11A1 (NRAMP1) polymorphisms and tuberculosis susceptibility: Updated systematic review and meta-analysis. PLoS One 2011;6(1):e15831.

223. Strickland S, Smith KK, Marotti KR. Hormonal induction of differentiation in teratocarcinoma stem cells: Generation of parietal endoderm by retinoic acid and dibutyryl cAMP. Cell 1980;21(2):347- 55. 149

224. Hogan BL, Taylor A, Adamson E. Cell interactions modulate embryonal carcinoma cell differentiation into parietal or visceral endoderm. Nature 1981;291(5812):235-7.

225. Doetschman TC, Eistetter H, Katz M, Schmidt W, Kemler R. The in vitro development of blastocyst- derived embryonic stem cell lines: Formation of visceral yolk sac, blood islands and myocardium. Journal of Embryology and Experimental Morphology 1985;87:27-45.

226. Haumaitre C, Reber M, Cereghini S. Functions of HNF1 family members in differentiation of the visceral endoderm cell lineage. Journal of Biological Chemistry 2003;278(42):40933-42.

227. Coughlin LB, McGuigan J, Haddad NG, Mannion P. Salmonella sepsis and miscarriage. Clinical Microbiology and Infection 2003;9(8):866-8.

150

Appendix A

Sample Collection Log

151

Appendix B

Salmonella enterica serovar Typhimurium (S.Tm) immunohistochemistry & immunofluorescence staining protocols

Salmonella enterica serovar Typhimurium Immunohistochemistry Protocol

Positive Control - non-pregnant orally infected livers & S.Tm bacterial smears on glass slides Negative Control - purified mouse IgG1 isotypic control (unconjugated) or TBST (diluent)

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes EtOH 70% - 5 minutes

6. Quenching: 3% H2O2 – 30 minutes *Thaw normal goat serum* 7. Wash: 2x with TBS + 0.1% Triton X-100 8. Serum Block: 10% normal goat serum (Jackson Immuno Research Labs; West Grove, PA, USA - cat. no. 005-000-121) at room temperature – 60 minutes 9. Wash: 2x with TBS + 0.1% Triton X-100 10. IgG Block: AffiniPure Fab Fragment goat anti-mouse IgG (H+L) @ 20ug/ml (Jackson Immuno Research Labs; West Grove, PA, USA - cat. no. 115-007-003) in TBST at room temperature – 60 minutes 11. Wash: 3x with TBS + 0.1% Triton X-100 12. Primary Antibody – in TBST – Overnight at 4°C • Salmonella Typhimurium LPS antibody 1E6 (Thermo Fisher Scientific Inc; Mississauga, ON – cat. no. MA1-83451) @ 1mg/ml - 1:500 dilution 152

• Negative control – purified mouse IgG1 isotypic control (Cedarlane Labs; Burlington, ON - cat. no. CLX500AP) – matching concentration above or just TBST Day 2 13. Wash: 3x with TBST 14. Secondary Antibody – peroxidase-conjugated AffiniPure goat anti-mouse IgG (H+L) (Jackson Immuno Research Labs, West Grove, PA, USA – cat. no. 115-035-062) @ 0.8 mg/ml at 1:1000 in TBST at room temperature – 60 minutes 15. Wash: 3x with TBST 16. Apply 3,3' Diaminobenzidine (DAB) chromogen (Abcam Inc, Toronto, ON – cat. no. ab64238) – 3 minutes

17. Wash: 2x with ddH2O 18. Counterstain: filtered Harris' Hematoxylin (Electron Microscopy Sciences, Hatfield, PA, USA - cat no. 26754) – 2 minutes 19. Rinse: running tap water - 5 minutes 20. Dehydration: EtOH 70% - 5 minutes EtOH 95% - 5 minutes EtOH 100% - 5 minutes 21. Clearing: Xylene 4 – 3 minutes Xylene 5 – 3 minutes 22. Mounting & Coverslipping: with Clarion™ mounting medium (Sigma-Aldrich Canada Co., Oakville, ON - cat. no. C0487)

Salmonella enterica serovar Typhimurium immunofluorescence protocol

Positive Control - non-pregnant orally infected livers & S.Tm bacterial smears on glass slides Negative Control - purified mouse IgG1 isotypic control (unconjugated) or TBST (diluent)

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes

153

EtOH 70% - 5 minutes

3. Rinse: ddH2O – 2 minutes 4. Antigen Retrieval: Sodium citrate (pH 6.0) HIER using microwave – bring slides and buffer almost to boiling then zap for 15 seconds, every 30 seconds for total of 5 minutes. Let slides cool for 15 minutes. 5. Wash: 2x with TBS + 0.1% Triton X-100 *Mark samples with Dako pen* 6. Serum Block: 10% normal goat serum (Jackson Immuno Research Labs; West Grove, PA, USA - cat. no. 005-000-121) at room temperature – 60 minutes 7. Wash: 2x with TBS + 0.1% Triton X-100 8. IgG Block: AffiniPure Fab Fragment goat anti-mouse IgG (H+L) @ 20ug/ml (Jackson Immuno Research Labs; West Grove, PA, USA - cat. no. 115-007-003) in TBST at room temperature – 60 minutes 9. Wash: 3x with TBS + 0.1% Triton X-100 10. Primary Antibody – in TBST – Overnight at 4°C • Salmonella Typhimurium LPS antibody 1E6 (Thermo Fisher Scientific Inc; Mississauga, ON – cat. no. MA1-83451) @ 1mg/ml - 1:500 dilution • Negative control – purified mouse IgG1 isotypic control (Cedarlane Labs; Burlington, ON - cat. no. CLX500AP) – matching concentration above or just TBST

Day 2 11. Wash: 3x with TBST 12. Secondary Antibody – Alexa Fluor® 594 conjugated, goat anti-mouse IgG (H+L) secondary antibody (Life Technologies Inc., Invitrogen; Burlington, ON - cat. no. A-11005) @ 1 mg/ml - 1:200 dilution in TBST at room temperature – 60 minutes 13. Wash: 3x with TBST 14. Mounting & Coverslipping: with ProLong® Gold Antifade Mountant with DAPI (Thermo Fisher Scientific Inc., Molecular Probes™ - cat. no. P36931)

154

Appendix C

Hematoxylin and Eosin staining protocol

Deparaffinization Xylene 1 (5 minutes) Xylene 2 (5 minutes) Xylene 3 (5 minutes) Hydration 100% ethanol 1 (5 minutes) *Start filtering hematoxylin using P2 Fisherbrand Filter Paper* 95% ethanol 1 (5 minutes) 80% ethanol 1 (5 minutes) 70% ethanol 1 (5 minutes)

Rinse ddH2O (1 minutes) Staining (H) Harris Hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) (2 minutes) Rinse Running tap water (5 minutes) Staining (E) Eosin Y (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26051-10) (2.5 minutes) Dehydration 70% ethanol 2 (5 minutes) 80% ethanol 2 (5 minutes) 95% ethanol 2 (5 minutes) 100% ethanol 2 (5 minutes) 100% ethanol 3 (5 minutes) Clearing Xylene 4 (3 minutes) Xylene 5 (3 minutes) Mounting Coverslip using Clarion™ mounting medium (Sigma-Aldrich

Canada Co.; Oakville, ON - cat. no. C0487)

155

Appendix D

Active caspase-3 immunohistochemistry staining protocol

Positive control - GD 15 mouse decidua (dying uterine natural killer cells - uNK) Negative control - rabbit IgG isotype control or just diluent (Cas-Block™ histochemical reagent)

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes EtOH 70% - 5 minutes 3. Rinse: ddH2O – 2 minutes 4. Antigen Retrieval: Sodium citrate (pH 6.0) HIER using microwave – bring slides and buffer almost to boiling then zap for 15 seconds, every 30 seconds for total of 5 minutes. Let slides cool for 15 minutes. 5. Wash: 2x with TBST (Tris-buffered saline with 0.05% Tween 20)

6. Quenching: 3% H2O2 room temperature – 30 minutes 7. Wash: 2x with TBST 8. Serum Block: CAS-Block™ histochemical reagent (Thermo Fisher Scientific Inc; Mississauga, ON – cat. no. 008120) – 10 minutes 9. Primary Antibody – in CAS-Block™ - overnight at room temperature • Polyclonal Rabbit anti-caspase-3 (Bioss Inc; Woburn, MA, USA– cat. no. bs-2593R) @ 1ug/ul – 1:200 dilution (stored at -20°C) • Negative control – Rabbit IgG isotype ctrl (Bioss Inc; Woburn, MA, USA– cat. no bs-0295P) @ 1ug/ul – matching concentration above (stored at -20°C) or just diluent (CAS-Block™) Day 2 10. Wash: 3x with TBST 11. Secondary Antibody –in TBST - 60 minutes at room temperature • Polyclonal Goat Anti-Rabbit Immunoglobulins/Biotinylated (Dako Canada ULC; Mississauga, ON – cat. no. E0432) @ 0.82g/L – 1:500 dilution 12. Wash: 3x with TBST 156

13. Incubate with ExtrAvidin®-Peroxidase (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. E2886) 1:100 in TBST – 30 minutes 14. Wash: 3x with TBST 15. Apply 3,3' Diaminobenzidine (DAB) chromogen (Abcam Inc; Toronto, ON – cat. no. ab64238) – 4 minutes

16. Rinse: 2x ddH2O 17. Counterstain: filtered Harris hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) – 2 minutes 18. Rinse: running tap water - 5 minutes 19. Dehydration: EtOH 70% - 5 minutes EtOH 95% - 5 minutes EtOH 100% - 5 minutes 20. Clearing: Xylene 4 – 3 minutes Xylene 5 – 3 minutes 21. Mounting & Coverslipping: with Clarion™ mounting medium (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. C0487)

157

Appendix E

Periodic-acid Schiff staining protocol

Rinse ddH2O (1 minutes) Staining 1% periodic acid (Sigma-Aldrich Canada Co., Oakville, ON - cat. no. BCBB7710V) (7 minutes)

Rinse ddH2O (5 minutes) Staining Schiff's reagent (Sigma-Aldrich Canada Co., Oakville, ON - cat. no. 3952016) (10 minutes)

Rinse ddH2O (5 minutes) Staining Sulphurous acid solution (5 minutes)

Rinse ddH2O (5 minutes) Staining Harris Hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) (2 minutes) Rinse Running tap water (5 minutes) Dehydration 70% ethanol 2 (5 minutes) 80% ethanol 2 (5 minutes) 95% ethanol 2 (5 minutes) 100% ethanol 2 (5 minutes) 100% ethanol 3 (5 minutes) Clearing Xylene 4 (3 minutes) Xylene 5 (3 minutes)

158

Mounting Coverslip using Clarion™ mounting medium (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. C0487)

159

Appendix F

CD45 immunohistochemistry staining protocol

Positive control - normal mouse spleen Negative controls - biotin rat IgG2b kappa isotype control or TBST

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes EtOH 70% - 5 minutes

6. Quenching: 3% H2O2 at room temperature – 30 minutes 7. Wash: 2x with TBST 8. Serum Block: 5% bovine serum albumin (BSA) in TBST at room temperature – 30 minutes 9. Wash: 2x with TBST 10. Primary Antibody – in TBST - overnight at 4°C • Biotin anti-mouse CD45 antibody (BioLegend; San Diego, CA, USA – cat. no. 103104) @ 0.5 mg/ml – 1:100 dilution • Negative control – Biotin rat IgG2b kappa isotype control (BioLegend; San Diego, CA, USA – cat. no. 400603) @ 0.5 mg/ml -1:100 or just TBST Day 2 11. Wash: 3x with TBST 12. Incubate with ExtrAvidin®-Peroxidase (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. E2886) 1:100 dilution in TBST – 30 minutes 13. Wash: 3x with TBST

160

14. Apply 3,3' Diaminobenzidine (DAB) chromogen (Abcam Inc.; Toronto, ON – cat. no. ab64238) – 5 minutes

15. Rinse: 2x ddH2O 16. Counterstain: filtered Harris Hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) – 2 minutes 17. Rinse: running tap water - 5 minutes 18. Dehydration: EtOH 70% - 5 minutes EtOH 95% - 5 minutes EtOH 100% - 5 minutes 19. Clearing: Xylene 4 – 3 minutes Xylene 5 – 3 minutes 20. Mounting & Coverslipping: with Clarion™ mounting medium (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. C0487)

161

Appendix G

Neutrophil (Ly6G) immunohistochemistry staining protocol

Positive controls – Infected spleen & liver Negative controls - Rat IgG2a kappa isotype control or CAS-Block™

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes EtOH 70% - 5 minutes

3. Rinse: ddH2O – 2 minutes 4. Wash: 2x with TBST (TBS with Tween 20 0.05%) 5. Antigen Retrieval: Enzymatic digestion with Proteinase K (Sigma-Aldrich Canada Co., Oakville, ON - cat. no. P6656) at room temperature – 5 minutes 6. Wash: 2x with TBST

7. Quenching: 3% H2O2 at room temperature – 30 minutes 8. Wash: 2x with TBST 9. Serum Block: CAS-Block™ histochemical reagent (Thermo Fisher Scientific Inc; Mississauga, ON – cat. no. 008120) at room temperature – 10 minutes 10. Primary Antibody – in CAS-Block™ - overnight at 4°C • Purified rat anti-mouse Ly6G monoclonal antibody (BD Pharmingen™, BD Biosciences; San Jose, CA, USA – cat. no. 551459) @ 0.5 mg/ml - 1:200 dilution • Negative control – Rat IgG2a kappa isotype control (BioLegend; San Diego, CA, USA – cat. no. 400502) @ 0.5 mg/ml – matching concentration above or just CAS-Block™ Day 2 11. Wash: 3x with TBST 12. Secondary Antibody –in TBST - 60 minutes at room temperature • Polyclonal rabbit anti-rat immunoglobulins biotinylated (Dako Canada ULC; Mississauga, ON – cat. no.- E0468) @ 0.85 g/L - 1:1000 dilution 13. Wash: 3x with TBST 162

14. Incubate with ExtrAvidin®-Peroxidase (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. E2886) at room temperature - 30 minutes 15. Wash: 3x with TBST 16. Apply 3,3' Diaminobenzidine (DAB) chromogen (Abcam Inc; Toronto, ON – cat.no. ab64238) – 4 minutes

17. Rinse: 2x ddH2O 18. Counterstain: filtered Harris hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) – 2 minutes 19. Rinse: running tap water - 5 minutes 20. Dehydration: EtOH 70% - 5 minutes EtOH 95% - 5 minutes EtOH 100% - 5 minutes 21. Clearing: Xylene 4 – 3 minutes Xylene 5 – 3 minutes 22. Mounting & Coverslipping: with Clarion™ mounting medium (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. C0487)

163

Appendix H

Macrophage (F4/80) immunohistochemistry staining protocol

Positive controls – Infected mouse spleen or liver Negative controls - biotin rat IgG2b k isotype control or just Cas-Block™

Day 1 1. Deparaffinization: Xylene 1 – 5 minutes Xylene 2 – 5 minutes Xylene 3 – 5 minutes 2. Rehydration: EtOH 100% - 5 minutes EtOH 95% - 5 minutes EtOH 70% - 5 minutes

3. Rinse: ddH2O – 2 minutes 4. Wash: 2x with TBST (Tween 20 0.05%) *Circle slides for antigen retrieval with hydrophobic pen (PAP pen)* 5. Antigen Retrieval: Enzymatic digestion with Proteinase K (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. P6656) at room temperature – 5 minutes 6. Wash: 2x with TBST

7. Quenching: 3% H2O2 at room temperature – 30 minutes 8. Wash: 2x with TBST 9. Serum Block: CAS-Block™ histochemical reagent (Thermo Fisher Scientific Inc.; Mississauga, ON – cat. no. 008120) – 10 minutes 10. Primary Antibody – in in CAS-Block™ - overnight at 4°C • Biotin Anti-Mouse Macrophages (F4/80) monoclonal antibody (Cedarlane Labs; Burlington, ON – cat. no. CL8940B-3) @ 0.1 mg/ml - 1:200 dilution • Negative control –Biotin Rat IgG2b kappa isotype control (BioLegend; San Diego, CA, USA – cat. no. 400603) @ 0.5 mg/ml – matching concentration above (diluted to 0.1 mg/ml) Day 2 11. Wash: 3x with TBST 12. Incubate with ExtrAvidin®-Peroxidase (Sigma-Aldrich Canada Co.; Oakville, ON - cat. no. E2886) 1:100 in TBST at room temperature – 30 minutes

164

13. Wash: 3x with TBST 14. Apply 3,3' Diaminobenzidine (DAB) chromogen (Abcam Inc; Toronto, ON – cat. no. ab64238) - 3 minutes (shorter for spleen or liver - 1 minute)

15. Rinse: 2x ddH2O 16. Counterstain: filtered Harris hematoxylin (Electron Microscopy Sciences; Hatfield, PA, USA - cat no. 26754) – 2 minutes 17. Rinse: running tap water - 5 minutes 18. Dehydration: EtOH 70% - 5 minutes EtOH 95% - 5 minutes EtOH 100% - 5 minutes 19. Clearing: Xylene 4 – 3 minutes Xylene 5 – 3 minutes 20. Mounting & Coverslipping: with Clarion™ mounting medium (Sigma-Aldrich Canada Co.; Oakville,

ON - cat. no. C0487)

165

Appendix I

Placental morphometric measurements analyzed per litter

Appendix I. Placental morphometric measurements analyzed per litter. There were no statistical differences between mean surface areas of placental regions amongst all of the groups analyzed (A-D). Each data point represents the mean of all litter members for infected animals, and for naïve animals, each data point represents the mean of two placentas/animal from N=3 animals/genotype/group. Statistical analyses used were a one-way ANOVA followed by Tukey’s post-test for multiple comparisons. All data presented as mean ± SD. Inf, infected. Statistical differences were identified when placentas were considered individually (see Fig. 7). 166

Appendix J

Anti-S.Tm LPS positive control (liver)

167

Appendix J. Positive control liver tissue immunostaining shows S.Tm+ foci. Shown is a representative positive control photomicrograph of anti-S.Tm LPS immunostained liver tissue section obtained from a non- pregnant C57BL/6J female 5 days post oral inoculation with 108 CFUs S.Tm (A). S.Tm+ signals appear as discrete, brown intracellular foci (indicated by black arrows). Staining was validated in comparison to an isotype control section (B). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

168

Appendix K

Anti-active caspase-3 positive control

Appendix K. Dying uNK cells serve as an internal positive control for validation of active caspase-3 immunostaining in GD 15 implantation sites. Shown are representative photomicrographs of DB tissue from a GD 15 naïve Ripk3-/-Slc11a1+/+. UNK cells are regressing at this gestational time-point, and can be distinguished by their vacuolated, brown-staining cytoplasm (A). Staining was validated in comparison to an isotype control section (B). Sections were counterstained with hematoxylin. All scale bars, 50 µm.

169

Appendix L

CD45+ leukocyte counts pooled by litter

Appendix L. Pooled mean CD45+ cell counts were not significantly altered regardless of infection or Ripk3 gene status. CD45+ IHC cell counts were not significantly different in the decidua basalis (A) junctional zone, and labyrinth areas (C), or in all areas (total) of placenta (D) regardless of infection or Ripk3 gene status when values were pooled and averaged for all placentas per infected animals, or 2 placentas/naïve dam. Each data point represents an individual animal from N=3 dams/genotype/group. All data are presented as mean SD. Statistical tests employed were a one-way ANOVA followed by Tukey's post-test for multiple comparisons. See Figure 18 for data presented as counts from individual implantation sites.

170