THE IMPACT OF LACTOBACILLUS AND BACTERIOPHAGE ON GROUP B STREPTOCOCCUS AND THE PLACENTAL MEMBRANES By Megan Shiroda A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Microbiology and Molecular Genetics – Doctor of Philosophy 2019 ABSTRACT By Megan Shiroda THE IMPACT OF LACTOBACILLUS AND BACTERIOPHAGE ON GROUP B STREPTOCOCCUS AND THE PLACENTAL MEMBRANES The human microbiota encompasses the microbes that live on and in the human body. While some body sites have been studied extensively for their role in human health, other body sites including the extraplacental membranes (EPM) have been historically considered sterile and are less studied. The EPM surrounds the fetus during pregnancy and serves as an important protective barrier during pregnancy. Several studies have established which bacteria are found in this site, but few studies have been conducted to characterize their impact on the EPM in vitro or their potential to impact pathogens that invade the EPM during pregnancy. Further, our knowledge of the viral component of the microbiome in human health remains incomplete. In this dissertation, Lactobacillus, a well-studied probiotic in other body sites, was evaluated for its effect in the placental membranes and the opportunistic pathogen Group B Streptococcus (GBS). We confirmed the ability of four Lactobacillus strains representing three species to colonize a cell line model of the outermost layer of these membranes, the decidual cells. Further, Lactobacillus dampens a known cell signaling pathway, the Mitogen Activated Protein Kinase (MAPK) pathway, which is associated with inflammation and host cell death by reducing the total protein level of p38. Collectively, these data suggest that Lactobacillus could maintain a commensal interaction in the placental membranes as described in other body sites. The ability of the same Lactobacillus strains to impact two GBS strains was also examined as GBS can ascend from the vaginal tract to infect placental membranes, triggering premature birth or neonatal infection. We found live Lactobacillus does not affect GBS growth or biofilm production. L. gasseri increased association of both strains of GBS to the decidual cells but did not result in increased invasion of the cells. Instead, co-culture with Lactobacillus reduced host cell death. Secreted products of Lactobacillus drastically reduced growth in 35 GBS strains that broadly represent GBS diversity and could prevent biofilm formation; this inhibition was strain dependent. Unfortunately, increased GBS-induced host cell death with Lactobacillus supernatants was also observed. Collectively, these data suggest that both live Lactobacillus and its supernatant could impact GBS interactions with the placental membranes. Bacteriophage are one of the most abundant members of the microbiome but their impact on opportunistic pathogens such as GBS remains unknown. As GBS can be isolated from gastrointestinal tract, we hypothesized fecal phage communities (PC) would inhibit the growth of GBS in vitro. Approximately 6% of the tested communities inhibited the growth of GBS. To examine GBS host range, we examined capsule, sequence and clinical types of 35 strains with three inhibitory PCs. As no significant differences were found with these traits, we examined Clustered Regularly Interspaced Palindromic Repeats (CRISPR), which serve as an adaptive immune system against invading foreign DNA by the acquisition of spacer sequences. GBS strains with fewer than nine spacers were less likely to be lysed by a phage community than strains with more than sixteen spacers. Further, sensitive strains of GBS were significantly more likely to be inhibited by PCs with a lower abundance of GBS in the corresponding bacterial component of each PC. Collectively, these data suggest that the phage component of the intestinal microbiome could impact GBS colonization. The work described in this dissertation helps elucidate the impact of Lactobacillus and bacteriophage as members of the microbiota on GBS and the placental membranes and their impact on adverse pregnancy outcomes. For Maui and my boo. iv ACKNOWLEDGEMENTS This dissertation is a culmination of more than five years of work and the support of so many people during that time. I would like to thank Dr. Shannon Manning for welcoming me into her lab and supporting my exploration of what interested me both in her lab and outside of it. My committee has fostered my growth as a scientist and has always been available to me. I thank Dr. Arvidson for her expertise on Lactobacillus, Dr. Parent for welcoming me into her lab to do parts of my phage work, and Dr. Hammer for challenging me with his questions. I would also like to thank my fellow graduate students for have impressed me with their research, their journeys and their dedication. I look forward to following your careers. My family has supported me though a large number of interesting decisions, including moving to the Pacific and going back to graduate school. I thank my parents for their love and support, my sister for inspiring me and my bother for steadying me. Thank you to Dan Parrell, Bethany Hout, Michelle Korir and Robert Parker who were instrumental in me finding my feet when I started this program and making even the worst days better. My friends both in graduate school and my running groups have kept me a little closer to sanity and made my days so much happier. To my graduate school friends, I’m so happy I had you to go through this with. To my friends outside of graduate school, thank you for suffering through my babbling and giving me perspective. The Manning lab was a second family when I first joined, and I have valued my time with so many lab members, whether at lab lunch, Labsgiving or Party Barn. I have missed so many of you after you’ve moved on. I honestly don’t know what state the lab would be in v without Rebekah Mosci and Sam Carbonell, and I especially thank them for their patience while I was in various states of distraction while writing this dissertation. Part of graduate school is finding your calling, and I can honestly say that weekly coffee with Bethany pushed me to consider my long-term goals and find my love for teaching research. It’s because of Bethany that I found FAST, which was really the start of it all. The people of FAST, especially Dr. Rique Campa, Dr. Aseel Bala, Joelyn De Lima and Kellie Walters, have fostered my love of teaching and helped me recognize my interest in teaching research. Dr. John Merrill and Dr. Donna Koslowsky were incredible to teach for and they introduced me to the AACR group and a faculty learning community. Their guidance led me to the AACR group who served as another research home for me to explore the aspects of teaching research that really interested me. The entire group at CREATE for STEM has been so incredibly welcoming, but Kevin Haudek and Juli Uhl have been especially helpful in helping me recognize how to move forward as an education researcher. Lastly, Heather Blankenship and Brian Nohomovich have simply been the best support system I could ask for. I’m so honored to have graduated with both of you, and I can’t wait to see what we accomplish. Heather, I look forward to so many more miles with you and PRs (for me). Brian, I look forward to so many more years of challenging each other and tackling whatever our next steps throw at us. vi TABLE OF CONTENTS LIST OF TABLES .................................................................................................................... x LIST OF FIGURES ................................................................................................................. xi KEY TO ABBREVIATIONS .............................................................................................. xiii CHAPTER 1 LITERATURE REVIEW: THE HUMAN MICROBIOME AND ITS EFFECTS ON HEALTH AND GROUP B STREPTOCOCCUS .......................................... 1 THE HUMAN MICROBIOME AND ITS EFFECTS ON HUMAN HEALTH ............................... 2 The bacterial component .......................................................................................... 3 The viral component ................................................................................................ 4 THE MICROBIOTA OF THE REPRODUCTIVE TRACT ......................................................... 5 The bacterial microbiota of the lower reproductive tract and its impact on human health ........................................................................................................................ 5 The microbiota of the upper reproductive tract and its impacts .............................. 7 GROUP B STREPTOCOCCUS IS AN IMPORTANT OPPORTUNISTIC PATHOGEN IN THE REPRODUCTIVE TRACT ............................................................................................ 9 Group B Streptococcus presents a global human health burden .............................. 9 Group B Streptococcus strain variation and disease outcome ............................... 10 Group B Streptococcus has multiple routes of infection ....................................... 11 Current therapies for GBS and potential alternatives ............................................ 13 Current therapies and their success ............................................................... 13 Vaccination ..................................................................................................... 14 Phage therapy ................................................................................................. 14 Use of probiotic Lactobacillus against GBS .................................................. 17 REFERENCES .............................................................................................................. 20 CHAPTER 2 LACTOBACILLUS STRAINS VARY IN THEIR ABILITY TO INTERACT WITH HUMAN ENDOMETRIAL STROMAL CELLS.............................. 32 ABSTRACT ................................................................................................................... 33 INTRODUCTION ......................................................................................................... 34 MATERIALS AND METHODS .................................................................................. 36 Bacterial strains and growth conditions ................................................................. 36 Biofilm Production ................................................................................................. 36 Cell Culture ............................................................................................................ 37 Cytotoxicity assays ................................................................................................ 38 Detection of IL-10 by ELISA: ............................................................................... 38 Western Blotting for p38 ....................................................................................... 39 RESULTS ....................................................................................................................... 41 Lactobacillus strains vary in growth in various media types ................................. 41 vii L. reuteri strains form significantly better biofilms ............................................... 41 L. crispatus associates with decidual cells significantly better than other Lactobacillus strains .............................................................................................. 43 Lactobacillus does not produce an inflammatory response in dT-HESCs ............ 43 Lactobacillus does not trigger dT-HESC death ..................................................... 44 DISCUSSION ................................................................................................................. 45 ACKNOWLEDGEMENTS .......................................................................................... 49 APPENDIX .................................................................................................................... 50 REFERENCES .............................................................................................................. 56 CHAPTER 3 THE IMPACT OF LACTOBACILLUS ON GROUP B STREPTOCOCCAL INTERACTIONS WITH PLACENTAL MEMBRANES ............. 60 ABSTRACT ................................................................................................................... 61 INTRODUCTION ......................................................................................................... 63 MATERIALS AND METHODS .................................................................................. 65 Bacterial strains and growth conditions ................................................................. 65 Cell Culture ............................................................................................................ 66 Isolation of Lactobacillus supernatants ................................................................. 66 Bacterial growth curves ......................................................................................... 66 Biofilm assays ........................................................................................................ 67 Association with and invasion of decidual cells .................................................... 68 Cytotoxicity assays ................................................................................................ 69 RESULTS ....................................................................................................................... 70 Lactobacillus does not impede GBS growth in T-HESC infection media ............ 70 Lactobacillus does not affect GBS biofilm formation ........................................... 70 Lactobacillus variably affects GBS association with dT-HESCs .......................... 71 Co-culture with Lactobacillus variably affects host cell death .............................. 72 Lactobacillus supernatants inhibit GBS growth .................................................... 73 Lactobacillus supernatants prevent GBS biofilm formation ................................. 74 Lactobacillus supernatants variably affect association of GBS, but do not affect invasion .................................................................................................................. 74 Lactobacillus supernatants increase host cell death in the invading GBS strain ... 76 Supernatant from Lactobacillus reuteri 6475 broadly inhibits GBS strains.......... 77 DISCUSSION ................................................................................................................. 78 ACKNOWLEDGEMENTS .......................................................................................... 82 APPENDIX .................................................................................................................... 83 REFERENCES .............................................................................................................. 98 CHAPTER 4 THE EFFECT OF PHAGE COMMNUNITIES ISOLATED FROM HUMAN FECAL SAMPLES ON GROUP B STREPTOCOCCUS .................................. 102 ABSTRACT ................................................................................................................. 103 INTRODUCTION ....................................................................................................... 105 MATERIALS AND METHODS ................................................................................ 109 Bacterial strains and growth conditions ............................................................... 109 GBS genome sequencing and extraction of GBS CRISPR spacers ..................... 109 viii Isolation of phage communities from human fecal samples ................................ 110 Metagenomics of fecal samples and estimation of GBS abundance in the microbiome .......................................................................................................... 110 GBS growth inhibition by Phage Communities (PCs) ........................................ 111 Assessment of infectivity by spot plating ............................................................ 112 Plaque assays ....................................................................................................... 112 Enrichment of GBS-specific phage ..................................................................... 112 Microscopy of enriched phage samples ............................................................... 113 RESULTS ..................................................................................................................... 114 Screening phage communities for GBS-specific phage isolation ........................ 114 Host range of select communities ........................................................................ 114 CRISPR spacer regions differ across GBS strains ............................................... 115 The number of CRISPR spacers significantly impacts the ability of phage communities to inhibit growth ............................................................................. 117 GBS presence in the microbiome is significantly correlated with GBS inhibition .............................................................................................................. 117 Recovery of phage from PCs capable of inhibiting GBS .................................... 118 Enriching GBS-specific phage ............................................................................. 121 Cesium Chloride Gradient ................................................................................... 122 DISCUSSION ............................................................................................................... 123 ACKNOWLEDGEMENTS ........................................................................................ 127 APPENDIX .................................................................................................................. 128 REFERENCES ............................................................................................................ 152 CHAPTER 5 CONCLUSIONS AND FUTURE DIRECTIONS .................................... 157 ix 84 129 131 134 LIST OF TABLES Extended GBS strain list Table 3.1. Table 4.1. GBS Strain List Table 4.2. Table 4.3. PC 801 and 895 have broad host range within Streptococcus CRISPR spacers can be annotated using NCBI x LIST OF FIGURES Figure 2.1. Figure 2.2. Figure 2.3. Figure 2.4. Figure 2.5. Figure 3.1. Figure 3.2. Figure 3.3. Figure 3.4. Figure 3.5. Figure 3.6. Figure 3.7. Figure 3.8. Lactobacillus growth varies by media type Lactobacillus species differ in the ability to form biofilms depending on media type Lactobacillus associates with decidualized Human Endometrial Stromal Cells (dT-HESC) Lactobacillus affects total p38 and phosphorylated levels of p38 (Pp38) Lactobacillus does not induce dT-HESC death Lactobacillus does not impact GBS growth Lactobacillus does not affect GBS biofilm formation Lactobacillus variably affects GBS association with dT-HESCs Lactobacillus variably affects host cell death Lactobacillus supernatants inhibit GBS growth Lactobacillus supernatants prevent GBS biofilm formation Lactobacillus supernatants variably affect host cell permeability dT-HESCs were infected with GBS at a MOI of 10 for two hours with or without 10% v/v Lactobacillus supernatant Lactobacillus supernatants increase GBS-induced host cell death Figure 3.9. Figure 3.10. Supernatant from Lr6475 broadly inhibits GBS growth Figure 4.1. Figure 4.2. CRISPR spacer regions differ across GBS strains Figure 4.3. The number of CRISPR spacers varies by capsule type Phage communities (PCs) can affect the growth of GBS 51 52 53 54 55 85 86 88 89 91 93 94 95 96 97 143 144 145 xi Figure 4.4. Figure 4.5. The number of CRISPR spacers and the percent GBS affect likelihood of lysis Phage titer experiments result in clearing or a confluent lawn between dilutions Figure 4.6. Use of BHI media resulted in clearer spot plates with clear lysogenic rings 148 Figure 4.7. Figure 4.8. Enrichment is not sufficient to isolate phage Fractions from a cesium chloride gradient maintained infectivity 149 151 146 147 xii KEY TO ABBREVIATIONS Absorbance 595nm Bovine serum albumin cyclic AMP Clonal complex Bicinchroninic acid Brain heart infusion Centers for Disease Control and Prevention A595 ANOVA Analysis of variance BCA BHI BLAST Basic Local Alignment Search Tool BSA cAMP CC CDC CFU CI CPS CRISPR Clustered Regularly Interspaced Short Palindromic Repeats CV dT-HESCs decidualized human endometrial stromal cells EPM GBS HESCs HMP IAP Intrapartum antibiotic prophylaxis Colony forming units Confidence interval Extraplacental membrane Group B Streptococcus Human endometrial stromal cells Human microbiome project Capsular polysaccharide Crystal violet xiii Multiplicity of infection National Center for Biotechnology Information Optical density Irritable bowel disease Immunoglobulin Lactobacillus cirspatus Lactobacillus gasseri Lipopolysaccharide Lactobacillus reuteri deMan, Rosoa and Sharpe Multilocus sequence types IBD IgA Lc Lg LPS Lr mTHB Modified Todd Hewitt broth MAPK Mitogen activated protein kinase MLST MOI MRS NCBI OD PBS PC PEG PTB PVDF RVC ST STI TBS TSA Sexually transmitted infection Phage community Polyetheleneglycol Tris-buffered saline Tryptic soy agar RVC Sequence type Phosphate buffered saline Preterm birth Polyvinyldene difluoride xiv TSB TSBd THA UTI Tryptic soy broth Tryptic soy broth with 1% dextrose Todd Hewitt agar Urinary tract infection xv CHAPTER 1 LITERATURE REVIEW: THE HUMAN MICROBIOME AND ITS EFFECTS ON HEALTH AND GROUP B STREPTOCOCCUS 1 The microbiota, or the microorganisms that live on and in the body, have been shown to play large roles in human health. These organisms contribute fundamentally to health by aiding in the extraction of nutrients from food, education of the immune system and as a barrier to invading pathogens. Studies including the Human Microbiome Project have begun to establish the members of the microbiota and have even pushed into examining the role the microbiome may play in complex disease states. Despite these advances in our understanding, some components of the microbiome have been less examined. First, the colonization of certain body sites, including the upper reproductive tract, remains controversial, and our understanding of the impact of individual organisms within the microbiota on health in these body sites remains incomplete. Specifically, the impact of the microbiota on opportunistic pathogens of the reproductive tract, including Group B Streptococcus, remains unknown. Finally, viruses are a generally under-appreciated residents of the human body that play large roles in the ecology of the microbiota but remain understudied in the reproductive tract. Herein, current knowledge on these topics will be reviewed and gaps in current understanding will be assessed. The Human Microbiome and Its Effects on Human Health Microorganisms, including bacteria, eukaryotes, archaea and viruses, live on and in the human body and are collectively called the microbiota. The genetic component of the microbiota is called the microbiome. Collectively, these organisms have a large impact on human health. While much work was initially done in culture-based methods, it was found that 80-95% of the microbiota were unculturable,1 leading to increased interest in culture independent methods, such as metagenomic sequencing. Both approaches have contributed to our understanding of the impacts of the microbiome in different ways. Metagenomic sequencing has given a better 2 understanding of which organisms are present, but it is limited because the presence of DNA does not always denote that the organism is living or active in the community. While this limitation is being actively combated by the use of other non-culture based methods, including proteomics and transcriptomics, culturing bacteria from a given body site provides insight into individual contributions of bacteria through in vitro assays and therefore the mechanisms behind the correlations observed in non-culture based methods. Continued efforts to culture complex communities through techniques including cultureomics, or the application of massive parallel culturing methods, will continue to increase our understanding of how the microbial communities that live on and in us impact our health. The bacterial component The Human Microbiome Project (HMP) has pushed our understanding of the bacterial component of the microbiome, identifying the unculturable component through the use of metagenomics. This project found that body sites contain distinct microbiomes, or genetic signatures of the microbiota, but that the potential function of that signature was still maintained between sites.2 Other work has found that stability of the microbiome is also determined by the site.2For example, the gut microbiome is stable over time, while the vaginal microbiome changes with menstrual cycle and pregnancy. Our microbiomes are acquired early in life and differ with age, geographical location, ethnicity, and health.3-7 To date, the microbiome has been associated with a wide variety of health topics including but not limited to obesity, cancer, vitamin production, pathogen invasion and education of the immune system and complex disease states including Irritable Bowel Disease (IBD), diabetes and preterm birth (PTB).8–11 In the first phase of HMP, the study largely focused on the bacterial component of the microbiome as 16S rRNA 3 sequencing technology was used to assess the composition of the bacterial community at different sites, but shotgun metagenomics was also used to determine the microbial pathways that were present in each site.12 Other research, including the second phase of the HMP (iHMP or HMP2), widened the scope of the research to include gene expression and protein profiles of the microbiome and host- specific properties such as genetic, epigenomics, and antibody and cytokine profiles, which collectively allow for better understanding of how the microbiome is affecting the host.13 Another often overlooked portion of the microbiome – the viral community – was also included in iHMP. The viral component The viral community is composed of both eukaryotic viruses that can themselves directly impact the host and bacteriophage that can impact the host through the bacterial component of the microbiome. The genetic signature of viruses is referred to as the virome. Similar to the bacterial communities, the viral community is seeded at an early and develops with over time.14 Additionally, the viral component also appears to be vertically transferred from mother to infant as the mode of birth impacts the composition of the fecal virome at one year of life.15 In contrast to the bacterial microbiome, the viral community is highly individualized.16 While research has found that the viral community can directly impact the immune system,17 many studies focus on the impact of bacteriophage – viruses that infect bacteria – on the bacterial component of the microbiota. Bacteriophage (phage) fall into two broad categories: lytic phage and lysogenic (temperate) phage. Lytic phages infect a bacterial host, replicate and obligately lyse the bacteria, 4 releasing viral particles and killing the host. Lysogenic phage can follow two lifestyles. First, a lysogenic phage will infect a host and integrate into the host genome. Its genome will replicate with the host until a trigger – typically a stress that causes DNA damage such as reactive oxygen species, antibiotics or ultraviolet light – causes the phage to enter a lytic life cycle, releasing viral progeny back into the environment. As bacteria are the natural host of bacteriophage, it is easy to see how the composition of the bacteriophage could affect the composition of the bacteria of the microbiota. Indeed, metagenomic research has found inverse correlations between bacteriophage and bacteria in the microbiome.18–20 For example, in IBD patients, increases in Caudovirales phage inversely correlated to its host Bacteroidaceae bacterial families in the disease state. The stability of bacteriophage in the human body is thought to be the result of ecological dynamics, in which the resident bacteriophages use resident bacteria to maintain numbers. This has been partially determined in gnotobiotic, or sterile, mouse models that have examined the bacteriophage-host dynamics in the gut.21 This stability is also maintained by the ability of phage to attach to mucosal surfaces.22 Barr et al. found that bacteriophage accumulate in the mucosa 10x more than bacteria.22 This research also found that these bacteriophages serve as a protective barrier to pathogen invasion. Most of the research done on the human virome has been conducted in the gut using fecal samples; therefore, more research is needed to fully examine if these dynamics apply to other body sites including the reproductive tract. The Microbiota of the Reproductive Tract The bacterial microbiota of the lower reproductive tract and its impact on human health The lower reproductive tract, or the vaginal tract, is the best studied region of the female urogenital tract. It maintains a high abundance, low diversity microbiome and is considered one 5 of the simplest in the human body.23 Traditionally, Lactobacillus has been considered the dominant member of the vaginal microbiome;23 however, variation has been observed across races as non-white women were shown to have lower levels of Lactobacillus.24 Despite differences between races, the presence of Lactobacillus is considered a marker of health, while a reduction in Lactobacillus abundance is a marker of dysbiosis, or a perturbed state in the microbiome, which is often linked to bacterial vaginosis.25,26 This dysbiosis is correlated to an increased risk of a variety of infections including sexually transmitted infections (STI),27 urinary tract infections2829 and yeast infections.29 This is in part due to Lactobacillus’s contribution to the acidification of the vaginal tract. A high abundance of Lactobacillus is also correlated to better pregnancy outcomes. Women with Lactobacillus-dominated vaginal microbiomes are less likely to deliver prematurely.30–33 Indeed, vaginal dysbiosis, which is characterized by depletion of Lactobacillus spp., was significantly increased prior to premature rupture of membranes.33 While certain species including, Group B Streptococcus and Sneathia spp., have been found to be increased, a general increase in diversity has mostly been observed.24,33 The importance of the vaginal microbiome extends beyond the mother’s health and pregnancy outcomes. The infant is seeded with the vaginal microbiome during vaginal birth, and this transfer has been shown to have short- term and long-term effects on outcomes in the child.34–3738 Indeed, the use of antibiotics during birth to prevent vertical transfer of pathogens such as Group B Streptococcus, and disruption of this vertical transfer by cesarean delivery can disrupt the gut microbiome of the infant for an extended period of time. 6 The microbiota of the upper reproductive tract and its impacts While the microbiota of the vaginal tract has been well studied, the upper reproductive tract, consisting of the uterus, fallopian tubes, ovaries, and placenta, is traditionally considered sterile. This sterility was thought to be maintained by the cervix, especially during pregnancy. Cervical cells not only secrete a thick mucus that physically restricts access to the upper reproductive tract, but they also secrete other products including defensins, lysozyme and nitric oxide.39 In spite of this barrier, research has demonstrated that bacteria are still capable of passing through the cervix to the uterus.40 Studies establishing the microbiota are limited in part because of the increased difficulty of sampling these areas compared to other body sites. However, current studies have reported that the microbiota is distinct from the vaginal tract and contains higher diversity than the Lactobacillus-dominated vaginal tract41 and a lower total abundance of bacterial DNA.42 Reported genera include Acinetobacter, Escherichia, Enterobacter and Lactobacillus. These studies remain controversial for several reasons. First, sampling of healthy women is mostly done by sampling through the vaginal tract, which may contaminate the sample with the vaginal microbiota. While some studies have examined women undergoing surgeries, these studies are severely limited and do not give an accurate picture of a “healthy” microbiome since the women have various health conditions requiring surgical intervention.41,43 Another common argument against studies that claim a low abundance microbiome is the existence of a “kit-ome,” or the DNA present in reagents used in metagenomic processing.44 In respect of these considerations, studies in these sites have provided negative controls45 or assessed the sampling method itself.41 7 Recent papers on the microbiota of the upper reproductive tract have been reviewed by Peric et al.; therefore, this chapter will focus on publications involving the placenta and placental membranes.39 The placental microbiome was initially described by Aagaard et al. in 2014.46 Containing Bacteroidetes, Firmicutes, Fusobacteria, Proteobacteria and Tenericutes phyla, it was found to more closely resemble the oral microbiome than that of the vagina.46 While Lactobacillus was identified, it was at a lower abundance compared to other organisms, which is in direct contrast to what has been observed in the vaginal tract.47 Since this initial publication, some metagenomic studies have also reported bacteria in the placental membranes,48–50 while others assert that no microbial sequences were present after comparison to the kit-ome.51–53With this disagreement, it is important to note that observational and culture-based studies have claimed the presence of bacteria in the placenta and its membranes without visible damage to the tissue as early as 1982.54 Indeed, this study cultured vaginally associated bacteria including Lactobacillus from the placenta after cesarean delivery, which counters the argument of contamination from the vaginal tract. Confirmation of a placental microbiome would greatly benefit from the use of additional techniques beyond sequencing. Seferovic et al. utilized in situ hybridization with 16SrRNA probes to support the initial report of a low load of bacterial colonization without signs of tissue damage.55 This report has the benefit of visual evidence of a live bacterial component and contributes to structural knowledge of the microbiota. In addition to the assertion that a microbiome exists in the placental tissues, some studies have also found variation in the placental microbiome composition based on traits of the mother such as maternal obesity and gestational diabetes, the health of the pregnancy and neonatal outcomes.56–59 For example, Acinetobacter, Escherichia coli, Enterobacter and Lactobacillus 8 were increased in healthy, term pregnancies in comparison to preterm births with chorioamnionitis, or inflammation of the placental membranes.50 Higher abundance of Lactobacillus spp., Propionibacterium spp. and members of the Enterobacteriaceae family in healthy states were also confirmed in other studies.30,60,61 The weight of full-term infants also varied depending on the composition of the placental microbiome, with normal birth weight being associated with the presence of Lactobacillus.59 Group B Streptococcus is an important opportunistic pathogen in the reproductive tract The microbiota can also be a source of infection. One in every three preterm infants can be associated with intra-amniotic infection.62 The bacteria isolated from placental membranes are in many cases those that normally maintain an asymptomatic colonization of the vaginal tract.63,64 Such organisms that cause disease in only specific circumstances are called opportunistic pathogens. One such organism capable of ascending from the vaginal tract is Group B Streptococcus. Group B Streptococcus presents a global human health burden Group B Streptococcus (Streptococcus agalactiae, GBS) is an asymptomatic colonizer of the rectal vaginal tract of 18% of women globally, with regional variation ranging from 11- 35%.65 It also contributes to maternal infections, stillbirth, preterm birth and neonatal infections. In the United States, GBS is a leading cause of neonatal sepsis. GBS infections are lethal for 4- 6% of neonates and can result in other long-lasting effects, including deafness and developmental disabilities.66 GBS caused a total of 31,850 cases of invasive disease in 2017 in the United Stated, resulting in 2,030 deaths.67 GBS neonatal infections can be divided into two types based on time of onset. Late onset disease occurs between one week of life and three 9 months, and the route of infection remains unknown. Early onset disease occurs within the first week of life and is thought to occur when GBS is aspirated by the infant while passing through the vaginal tract or by ascending infection and invasion of the extraplacental membranes (EPM). Ascending infection is also thought to contribute to GBS’s ability to trigger premature birth and stillbirth. GBS can ascend from a commensal state in the vaginal tract, through the cervix and associate with the EPM. Group B Streptococcus strain variation and disease outcome GBS diversity can be examined in multiple ways and examining this diversity can reveal correlations between GBS types and disease outcome. First, GBS can be grouped phenotypically according to serotype based on the salicylic acid capsular polysaccharide (CPS).68–70 This capsule has also been found to be important in immune evasion, with serotype III being most commonly associated with disease.71,72 As CPS is not closely linked to phylogenetic lineage, capsule typing cannot be used to determine the relatedness of strains. A genotypic method, called multilocus sequence typing (MLST), examines sequence variation in seven conserved genes to group strains into different sequence types (STs).71 MLST is useful for examining genetic relatedness and evolutionary relationships among strains from different sources and serotypes. In GBS, MLST uses variation within seven conserved genes to group strains. Epidemiological studies have found that ST-17 strains are highly associated with neonatal disease;68,69,73 however, there is even strain variation within these groupings in phenotypes that may impact virulence. For example, our research group has found that ST-17 strains vary in: ability to attach to several human cell types, mechanisms of phagosomal stress survival, biofilm production, provocation of cytokine responses in macrophages, and induction 10 of host responses from decidual cells.74–79 We and others have also shown that ST-17 strains also have unique genetic characteristics that may impact pathogenesis, and that ST-17 strains are more likely to persist in pregnant women following antibiotic prophylaxis during childbirth.74,75,79–81 Enhanced persistence of ST-17 GBS following pregnancy likely contributes to the high incidence of late onset disease in infants; little is known, however, about other host factors that contribute to persistence. GBS strains can also be grouped according to where they were isolated from. In this way, a strain can be noted as invasive (isolated from an active infection) or colonizing (isolated from a mother who never had disease symptoms). Similarly, a strain’s resistance to antibiotics given during pregnancy can also be noted by grouping strains that were lost after antibiotic treatment together versus those that persisted. The CDC recently reported that while GBS remains sensitive to penicillin, the most commonly used antibiotic used for preventative measures, it has developed resistance to clindamycin and erythromycin. Very recently, rare instances of vancomycin resistance have also been detected.82 Group B Streptococcus has multiple routes of infection GBS infections begin with the ability to asymptomatically colonize the vaginal tract. The first step of this colonization is attachment. GBS cell surface proteins, ScpB, FbsA and laminin binding protein (Lmb), can bind to host extracellular matrix proteins fibronectin, fibrinogen and laminin, respectively.83–86 Distinct pili structures and serine-rich repeat proteins (srr-1/2) also interact with the epithelial surface of different cell types in the vagina, decidual layer of the placental membranes, and lung tissue.87–90 After establishing colonization, GBS must persist using multiple methods to avoid immunological clearing. First, host mimicry allows GBS to hide 11 from the immune system. The capsule (cps) of GBS is coated in sialic acid, which mimics the residues found on host cells.91,92 Another evasion and mimicry mechanism is the use of CspA to break down fibrinogen to fibrin, which coats and “hides” the cell.93 Secondly, GBS can also actively dampen the immune response using a serine protease (ScpB) that cleaves C5a, which is an important factor in the complement cascade of the immune system for recruitment of neutrophils.94–96 GBS can also avoid phagocytic uptake through a variety of mechanisms including the capsule, β-protein, Complement Interfering Protein (CIP) and BibA.97 This ability to persist in the presence of immune stressors is crucial to GBS survival and can contribute to disease during pregnancy. GBS can also cause disease by ascending from the vaginal tract into the upper reproductive tract, but little is known about what triggers ascending infections. Most studies focus on epidemiology of ascending infection in cohort studies, which only provide correlations, not molecular mechanisms. As described above, such studies link disruptions in the vaginal microbiome and increased risk of ascending infections. One molecular mechanism study in GBS did find that bacterial strains isolated from premature birth cases maintained active hyaluronidases, which could contribute to disease by cleaving hyaluronic acid in the cervix where it contributes to epithelial barrier function.98 This role was confirmed by reduced rates of ascending infection of GBS hyaluronidase (hylB) mutants in a murine model.98 Once ascending to the uterus, GBS can attach to the outermost layer of the placental membranes, the decidual cells, using the same attachment proteins (pili, ScpB, LMB and FbsA) described above. This association is thought to trigger an inflammatory cascade that leads to premature rupture of the membranes and premature birth.97 GBS is also capable of invading the placental membranes to access the infant during pregnancy. This can occur through at least three mechanisms. First, 12 GBS can secrete a cytolysin (ß-hemolysin) that lyses host cells, breaking down the host barriers.99–101 Secondly, it can invade transcellularly by hijacking host cell machinery to provoke cytoskeleton rearrangement to enter host cells.102 Finally, GBS can interact with gap junctions to utilize a paracellular route that does not result in tissue damage.103 After gaining access to the amniotic cavity, GBS can actively replicate in amniotic fluid,104 thereby allowing it to infect the fetus in utero via the lungs or gut.105,106 Once GBS enters the fetus, it can effectively enter the bloodstream, causing sepsis, and continue from the blood across the blood-brain barrier, resulting in meningitis. Current therapies for GBS and potential alternatives Current therapies and their success Public health policies in regard to GBS preventative therapies vary from country to country.107 While some countries utilize risk factors to guide risk assessment, the consensus guidelines in the United States test for GBS colonization of each mother using a rectal vaginal swab at 37 weeks of pregnancy.108 It is recommended that women with GBS colonization or who have tested positive in previous pregnancies receive intrapartum antibiotic prophylaxis (IAP). Successful IAP requires at least four hours of intravenous antibiotic treatment prior to the birth of the infant.108 IAP has successfully reduced early onset neonatal infections, but the timing of IAP during labor restricts its usefulness to early onset disease, which is only one of the many ways GBS can affect maternal and fetal health. Unfortunately, IAP has no effect on premature birth, stillbirth, maternal infections, or late onset disease. In addition to its limited use, IAP has known drawbacks, including disruption of the microbiome. As described above, the vaginal microbiome of the mother is determinant of infant microbiome and alterations in it can have 13 adverse effects on the mother and the infant.34–36 Additionally, the reduction of GBS-associated neonatal sepsis has occurred with a simultaneous increase in antibiotic resistant E. coli sepsis cases in very low birth weight infants, resulting in no overall reduction of neonatal sepsis cases.109,110 These shortcomings of IAP and the continued disease burden highlight the need for alternative therapies and continued research of this pathogen. Vaccination Alternative therapies for GBS should continue to focus on reducing maternal colonization but also need to be safe for earlier, more frequent use during pregnancy. During development of IAP, studies have examined the success of early and repeated use of antibiotics, but these treatments were associated with increased risk of infection and adverse pregnancy outcomes, resulting in the recommendation of a single dose during labor.108 One of the earliest suggested alternatives was maternal vaccination. This could be used to reduce colonization of the mother and pass protective antibodies to infants through breastmilk, thereby reducing late-onset disease.68,111 GBS vaccine development began after the observation of protective antibodies to GBS in mice.112 Initial rounds of vaccines used purified CPS as the antigen, but have also included conjugate polysaccharides.113 Rounds of vaccine testing have each had variable success in producing immune responses, but clinical trials with the most up to date vaccines are still on- going and have not been published.113 Phage therapy Another alternative therapy that has gained traction is the use of virulent bacteriophage to infect and kill unwanted bacteria, which is termed phage therapy. Phage therapy was examined soon after their discovery in the early 1900’s but was ignored in the United States and much of 14 Europe with the advent of antibiotics.114,115 With the rise of antibiotic resistance and the realization of the negative effects of antibiotics on the microbiome, phage therapy has been revisited for a number of different applications in human health and the food industry.116 In spite of early application of phage therapy and its continued use in Eastern Europe, few large trials that meet current standards have been performed to validate phage therapy or to ensure its safety.117 Instead, bacteriophage therapy is typically applied in single cases when antibiotics fail and patients are left with no other option.118–120 Phagoburn, which examined the use of bacteriophage on burn wound infections, became the first completed randomized, controlled and double-blinded study to pass all requirements for good clinical practices.121 In addition to testing the efficacy of the treatment, Jault et al. revealed previous unrecognized difficulties with production and administration of bacteriophage that will contribute to future studies.121 While the bacteriophage treatment was successful in reducing infective bacteria in the burn wounds, the trial was terminated early because the standard practice of antibiotic treatment was more successful. This outcome could have been due to the difficulties with instability of the phage cocktail, resulting in a drastically lower multiplicity of infection (MOI) than intended. In addition to this study, more studies with similar levels of controls have recently finished, are currently being executed or are recruiting patients.122 Use of bacteriophage during pregnancy is even less studied as little research has been conducted with pregnant women; however, the potential importance of phage therapy during pregnancy has been discussed.115 It is of particular interest to apply phage therapy to opportunistic pathogens that asymptomatically colonize the vaginal tract but cause issues during pregnancy, as this would allow for their selective removal without disrupting the microbiome.115 15 Bacterial targets include Ureaplasma spp. and GBS, which are associated with preterm birth, and Pseudomonas aeruginosa and E. coli, which are associated with urinary tract infections (UTIs). Indeed, UTIs are common during pregnancy and are typically treated with antibiotics, which affect the entire microbiome potentially leading to dysbiosis. As dysbiosis itself is associated with preterm birth and could potentially alter the transfer of a beneficial microbiome to the fetus, a more targeted approach is preferred. Sybesma et al. documented the success of a cocktail of bacteriophage in clearing UTIs from 50 patients with either E. coli or Klebsiella pneumoniae infections.123 In another study using a mouse model, a single oral dose of a phage cocktail was able to clear a UTI caused by uropathogenic E. coli.124 Studies on Ureaplasma spp. are limited as traditional phage isolation techniques are difficult; therefore, there is no available information on phages capable of targeting this species. Currently, no clinical trial have been conducted examining the ability of a live phage to reduce colonization of GBS; however, in 2005, Cheng et al. examined the ability of a phage lysin to reduce GBS colonization in a mouse model.125 This phage treatment successfully targeted multiple strains of varying serotypes in vitro and eradicated GBS colonization in the vaginal tract with a single dose. Although several in vitro studies have examined the relationship between human GBS isolates and phage, these phages were mostly lysogenic, meaning they incorporate into the genome of the bacteria instead of obligately lysing the bacteria, as is the goal with phage therapy.115 Recently, the first virulent bacteriophage, HN48, that targeted GBS was isolated from pond water, but the GBS was associated with tilapia, not human infection.126 GBS colonizes a wide variety of animals including camels, crocodiles, cows and tilapia, but the host range of GBS 16 strains is species-specific.127 This host specificity raises the question of whether findings for one type of GBS are extendable to GBS infection in humans, but the discovery of the lytic HN48 is still exciting as all isolated phage for human associated GBS are lysogenic or temperate. Characterization of this Caudovirales phage demonstrated that it had a relatively wide host range (67%) when evaluated against GBS from a variety of fish sources. Moreover, HN48 was highly specific to GBS, suggesting it would be useful in controlling disease in tilapia.126 Additional experiments demonstrated that HN48 prevented disease in tilapia when co-infected with GBS; bacterial load was significantly decreased 12 hours post-infection and survival rate increased from 33% to 66%. It is also interesting to note that HN48 was sensitive to low pH (3-5), which would be important if applied in the vaginal tract.126 These finding underscore the importance of characterizing individual phage to examine traits that will help it be effective in the system of interest. Use of probiotic Lactobacillus against GBS The use of probiotics – or “live organisms which when administered in adequate amounts confer a health benefit on the host” – has also been suggested to reduce GBS colonization. Because Lactobacillus is a major contributor to health in the vaginal tract, and it has been found in the upper reproductive tract, it represents a useful bacterial population to study for probiotic effects. Metagenomic studies, for instance, have shown that GBS colonization is negatively correlated with Lactobacillus colonization in the vagianl tract.27,128 Studies on both Lactobacillus and GBS are mostly limited to the vaginal tract. DeGregorio et al. demonstrated that L. reuteri could modulate the immune response and reduce GBS colonization in a murine model.129,130 Recently, another group showed a reduction in rectal GBS colonization of 72% of women and in vaginal colonization of 68% of women.131 Additionally, studies testing oral dosing of murine 17 models have also demonstrated the ability of Lactobacillus to significantly reduce GBS colonization in the vaginal tract.129 By contrast, some studies have had less success in reducing GBS colonization, suggesting that the strain of Lactobacillus, the dose or timeframe may be important.132,133 Additional in vitro studies have examined the interactions between Lactobacillus and GBS, however, no human trials have been conducted to determine the effects of Lactobacillus beyond rectal vaginal colonization. In vitro studies have examined phenotypic effects such as GBS growth inhibition,134 reduction in attachment,135,136 effects of Lactobacillus bactericidins,137,138 and biosurfactants.139,140 Many different Lactobacillus traits could be beneficial for use as a probiotic, though significant strain variation has been observed. Lactobacillus is thought to establish association with the vaginal tract by binding fibronectin using fibronectin binding proteins. Lactobacillus strains with surface layers (S-layer) can also attach to host fibronectin using S-layer proteins, such as SlpA.141 As fibronectin is also commonly used by invading pathogens to establish infections, the presence of Lactobacillus at those sites could serve as a barrier to preventing pathogen colonization.142 Beyond barrier function, different strains of Lactobacillus also vary in the ability to directly inhibit pathogens. This inhibition could be due to the production of hydrogen peroxide, lactic acid and bactericidins. Lactic acid is one of the end products of the metabolism of all lactic acid bacteria, including Lactobacillus, and results in a general reduction of pH that can inhibit a range of Gram-positive and -negative bacteria. Indeed, production of this byproduct by Lactobacillus is credited with the acidic pH of the vaginal tract. Hydrogen peroxide is a reactive oxygen species that can lead to peroxidation of membrane lipids and increased membrane 18 permeability of other bacteria.143,144 Studies in the vaginal tract of women have found variation in the ability of vaginal Lactobacillus strains to produce hydrogen peroxide, which was correlated with differential risk of bacterial vaginosis.143,144 Bactericidin production vary between strains of Lactobacillus and the range of susceptible organisms for each bactericidin also varies.145,146 The considerable differences in host attachment and pathogen inhibition demonstrate the importance of examining a variety of strains for probiotic use, but also the importance of not generalizing the effect of one strain to other strains. 19 REFERENCES 20 REFERENCES 1. P.B. Eckburg, E.M. Bik, C.N. Bernstein, E. Purdom, L. Dethlefsen, M. Sargent, S.R. Gill, K.E. Nelson, D. A. R. Diversity of the human intestinal microbial flora. Science 308, 1635–1638 (2005). 2. Huttenhower, C. et al. Structure, function and diversity of the healthy human microbiome. Nature 486, 207–214 (2012). 3. Dominguez-Bello, M. G. et al. Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proceedings of the National Academy of Sciences of the United States of America 107, 11971–11975 (2010). 4. Yatsunenko, T. et al. Human gut microbiome viewed across age and geography. Nature vol. 486 222–227 (2012). 5. Vatanen, T. et al. The human gut microbiome in early-onset type 1 diabetes from the TEDDY study. Nature vol. 562 589–594 (2018). 6. Dominguez-Bello et al. Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc. Natl Acad. Sci. 107, 11971– 11975 (2010). 7. Franzosa et al. Identifying personal microbiomes using metagenomic codes. Proc Natl Acad Sci 112, E2930–E2938 (2015). 8. Jefferson, K. K. et al. Relationship between vitamin D status and the vaginal microbiome during pregnancy. Journal of Perinatology 39, 824–836 (2019). 9. Lloyd-Price, J. et al. Multi-omics of the gut microbial ecosystem in inflammatory bowel diseases. Nature 569, 655–662 (2019). 10. Fettweis, J. M. et al. The vaginal microbiome and preterm birth. Nature Medicine 25, 1012–1021 (2019). 11. Zhou, W. et al. Longitudinal multi-omics of host–microbe dynamics in prediabetes. Nature 569, 663–671 (2019). 12. Turnbaugh, P. J. et al. The Human Microbiome Project. Nature vol. 449 804–810 (2007). 13. The integrative human microbiome project: Dynamic analysis of microbiome-host omics profiles during periods of human health and disease corresponding author. Cell Host and Microbe vol. 16 276–289 (2014). 21 14. Breitbart, M. et al. Viral diversity and dynamics in an infant gut. Research in Microbiology 159, 367–373 (2008). 15. McCann, A. et al. Viromes of one year old infants reveal the impact of birth mode on microbiome diversity. PeerJ 2018, (2018). 16. Reyes, A. et al. Viruses in the faecal microbiota of monozygotic twins and their mothers. Nature 466, 334–338 (2010). 17. Górski, A. et al. Phages and immunomodulation. Future Microbiology vol. 12 905–914 (2017). 18. Monaco, C. L. et al. Altered Virome and Bacterial Microbiome in Human Immunodeficiency Virus-Associated Acquired Immunodeficiency Syndrome. Cell Host and Microbe 19, 311–322 (2016). 19. Norman, J. M. et al. Disease-specific alterations in the enteric virome in inflammatory bowel disease. Cell 160, 447–460 (2015). 20. Nakatsu, G. et al. Alterations in Enteric Virome Are Associated With Colorectal Cancer and Survival Outcomes. Gastroenterology 155, 529-541.e5 (2018). 21. Reyes, A., Wu, M., McNulty, N. P., Rohwer, F. L. & Gordon, J. I. Gnotobiotic mouse model of phage-bacterial host dynamics in the human gut. Proceedings of the National Academy of Sciences of the United States of America 110, 20236–20241 (2013). 22. Barr, J. J. et al. Bacteriophage adhering to mucus provide a non-host-derived immunity. Proceedings of the National Academy of Sciences of the United States of America 110, 10771–10776 (2013). 23. Ravel, J. et al. Vaginal microbiome of reproductive-age women. Proceedings of the National Academy of Sciences of the United States of America 108, 4680–4687 (2011). 24. Anahtar, M. N., Gootenberg, D. B., Mitchell, C. M. & Kwon, D. S. Cervicovaginal Microbiota and Reproductive Health: The Virtue of Simplicity. Cell Host and Microbe vol. 23 159–168 (2018). 25. Petrova, M. I., Lievens, E., Malik, S., Imholz, N. & Lebeer, S. Lactobacillus species as biomarkers and agents that can promote various aspects of vaginal health. Frontiers in Physiology vol. 6 (2015). 26. Serrano, M. G. et al. Racioethnic diversity in the dynamics of the vaginal microbiome during pregnancy. Nature Medicine (2019) doi:10.1038/s41591-019-0465-8. 27. van de Wijgert, J. H. H. M. The vaginal microbiome and sexually transmitted infections are interlinked: Consequences for treatment and prevention. PLoS Medicine 14, (2017). 22 28. A H Sumati and N K Saritha. Association of Urinary Tract Infection in Women with Bacterial Vaginosis. J Glob Infect Dis. 1, 151–152 (2009). 29. Zhou, X. et al. Characterization of vaginal microbial communities in adult healthy women using cultivation-independent methods. Microbiology 150, 2565–2573 (2004). 30. DiGiulio, D. B. et al. Temporal and spatial variation of the human microbiota during pregnancy. Proceedings of the National Academy of Sciences of the United States of America 112, 11060–11065 (2015). 31. Callahan, B. J. et al. Replication and refinement of a vaginal microbial signature of preterm birth in two racially distinct cohorts of US women. Proceedings of the National Academy of Sciences of the United States of America 114, 9966–9971 (2017). 32. Kindinger, L. M. et al. The interaction between vaginal microbiota, cervical length, and vaginal progesterone treatment for preterm birth risk. Microbiome 5, (2017). 33. Brown, R. G. et al. Vaginal dysbiosis increases risk of preterm fetal membrane rupture, neonatal sepsis and is exacerbated by erythromycin. BMC Medicine 16, (2018). 34. Corvaglia, L. et al. Influence of intrapartum antibiotic prophylaxis for group B streptococcus on gut microbiota in the first month of life. Journal of Pediatric Gastroenterology and Nutrition 62, 304–308 (2016). 35. Aloisio, I. et al. Influence of intrapartum antibiotic prophylaxis against group B Streptococcus on the early newborn gut composition and evaluation of the anti- Streptococcus activity of Bifidobacterium strains. Applied Microbiology and Biotechnology 98, 6051–6060 (2014). 36. Aloisio, I. et al. Evaluation of the effects of intrapartum antibiotic prophylaxis on newborn intestinal microbiota using a sequencing approach targeted to multi hypervariable 16S rDNA regions. Applied Microbiology and Biotechnology 100, 5537– 5546 (2016). 37. Duranti, S. et al. Maternal inheritance of bifidobacterial communities and bifidophages in infants through vertical transmission. Microbiome 5, (2017). 38. Wang, S. et al. Maternal Vertical Transmission Affecting Early-life Microbiota Development. Trends in Microbiology (2019) doi:10.1016/j.tim.2019.07.010. 39. Peric et al. Bacterial Colonization of the Female Reproductive Tract. Int. J. Mol. Sci. 20, 3405 (2019). 40. Hansen, L. K. et al. The cervical mucus plug inhibits, but does not block, the passage of ascending bacteria from the vagina during pregnancy. Acta Obstetricia et Gynecologica Scandinavica 93, 102–108 (2014). 23 41. Chen C et al. The microbiota continuum along the female reproductive tract and its relation to uterine-related diseases. Nat Commun 8, 875 (2017). 42. Mitchell, C. M. et al. Colonization of the upper genital tract by vaginal bacterial species in nonpregnant women. American Journal of Obstetrics and Gynecology 212, 611.e1-611.e9 (2015). 43. Miles, S. M., Hardy, B. L. & Merrell, D. S. Investigation of the microbiota of the reproductive tract in women undergoing a total hysterectomy and bilateral salpingo- oopherectomy. Fertility and Sterility 107, 813-820.e1 (2017). 44. de Goffau, M. C. et al. Recognizing the reagent microbiome. Nature Microbiology vol. 3 851–853 (2018). 45. Kim, D. et al. Optimizing methods and dodging pitfalls in microbiome research. Microbiome vol. 5 (2017). 46. Aagaard, K. et al. The Placenta Harbors a Unique Microbiome. Sci Transl Med. 6(237): 237 (2014). 47. Parnell, L. A. et al. Microbial communities in placentas from term normal pregnancy exhibit spatially variable profiles. Scientific Reports 7, (2017). 48. Tomlinson, M. S. et al. Microorganisms in the human placenta are associated with altered CpG methylation of immune and inflammation-related genes. PLoS ONE 12, (2017). 49. Prince, A. L. et al. The placental membrane microbiome is altered among subjects with spontaneous preterm birth with and without chorioamnionitis. American Journal of Obstetrics and Gynecology 214, 627e1-627e16 (2016). 50. Lauder, A. P. et al. Comparison of placenta samples with contamination controls does not provide evidence for a distinct placenta microbiota. Microbiome 4, (2016). 51. Leiby, J. S. et al. Lack of detection of a human placenta microbiome in samples from preterm and term deliveries. Microbiome 6, 196 (2018). 52. Bushman, F. D. De-Discovery of the Placenta Microbiome. American Journal of Obstetrics and Gynecology vol. 220 213–214 (2019). 53. Kovalovszki L, Villányi Z, Pataki I, Veszelowvsky I, N. Z. Isolation of aerobic bacteria from the placenta. Acta Paediatr Acad Sci Hung 23, 357–60 (1982). 54. Seferovic, M. D. et al. Visualization of microbes by 16S in situ hybridization in term and preterm placentas without intraamniotic infection. American Journal of Obstetrics and Gynecology 221, 146.e1-146.e23 (2019). 55. Bassols, J. et al. Gestational diabetes is associated with changes in placental microbiota 24 and microbiome. Pediatric Research 80, 777–784 (2016). 56. Zheng, J. et al. Correlation of placental microbiota with fetal macrosomia and clinical characteristics in mothers and newborns. www.impactjournals.com/oncotarget (2017). 57. Zheng, J. et al. The placental microbiota is altered among subjects with gestational diabetes mellitus: A spilot study. Frontiers in Physiology 8, (2017). 58. Zheng, J. et al. The placental microbiome varies in association with low birth weight in full-term neonates. Nutrients 7, 6924–6937 (2015). 59. Onderdonk, A. B., Delaney, M. L., DuBois, A. M., Allred, E. N. & Leviton, A. Detection of bacteria in placental tissues obtained from extremely low gestational age neonates. American Journal of Obstetrics and Gynecology 198, (2008). 60. Collado, M. C., Rautava, S., Aakko, J., Isolauri, E. & Salminen, S. Human gut colonization may be initiated in utero by distinct microbial communities in the placenta and amniotic fluid. Scientific Reports 6, (2016). 61. Yoon, B. H. et al. Clinical significance of intra-amniotic inflammation in patients with preterm labor and intact membranes. in American Journal of Obstetrics and Gynecology vol. 185 1130–1136 (Mosby Inc., 2001). 62. Romero, R., Dey, S. K. & Fisher, S. J. Preterm labor: One syndrome, many causes. Science vol. 345 760–765 (2014). 63. Romero, R. et al. The role of infection in preterm labour and delivery. Paediatr Perinat Epidemiol. 2:41-56 (2001). 64. Russell, N. J. et al. Maternal Colonization with Group B Streptococcus and Serotype Distribution Worldwide: Systematic Review and Meta-analyses. Clinical Infectious Diseases vol. 65 S100–S111 (2017). 65. Kohli-Lynch, M. et al. Neurodevelopmental Impairment in Children after Group B Streptococcal Disease Worldwide: Systematic Review and Meta-analyses. Clinical Infectious Diseases vol. 65 S190–S199 (2017). 66. Center for Disease Control. Active Bacterial Core Surveillance (ABCs) Report Emerging Infections Program Network Group B Streptococcus, 2014. (2015). 67. Johri, A., Grandi, G. & Rappuoli, R. Group B Stertococcus: global incidence and vaccine development. Nat Rev microbiol 4, 932–942 (2009). 68. Manning, S. D. et al. Multilocus sequence types associated with neonatal Group B streptococcal sepsis and meningitis in Canada. Journal of Clinical Microbiology 47, 1143–1148 (2009). 25 69. Slotved, H. C., Kong, F., Lambertsen, L., Sauer, S. & Gilbert, G. L. Serotype IX, a proposed new Streptococcus agalactiae serotype. Journal of Clinical Microbiology 45, 2929–2936 (2007). 70. Jones, N. et al. Multilocus Sequence Typing System for Group B Streptococcus. J. Clin. Microbiol. 41, 2530–2536 (2003). 71. Lin, F. Y. C. et al. Phylogenetic lineages of invasive and colonizing strains of serotype III group B streptococci from neonates: A multicenter prospective study. Journal of Clinical Microbiology 44, 1257–1261 (2006). 72. Luan, S. L. et al. Multilocus sequence typing of Swedish invasive group B streptococcus isolates indicates a neonatally associated genetic lineage and capsule switching. Journal of Clinical Microbiology 43, 3727–3733 (2005). 73. Korir, M. L. et al. Investigation of the role that NADH peroxidase plays in oxidative stress survival in Group B Streptococcus. Frontiers in Microbiology 9, (2018). 74. Korir, M. L. et al. Differing mechanisms of surviving phagosomal stress among Group B Streptococcus strains of varying genotypes. Virulence 8, 924–937 (2017). 75. Flaherty, R. A. et al. Modulation of death and inflammatory signaling in decidual stromal cells following exposure to Group B Streptococcus. Infection and Immunity (2019) doi:10.1128/iai.00729-19. 76. Flaherty, R. A. et al. Genetically distinct Group B Streptococcus strains induce varying macrophage cytokine responses. PLOS ONE 14, e0222910 (2019). 77. Parker, R. E. et al. Association between genotypic diversity and biofilm production in group B Streptococcus. BMC Microbiology 16, (2016). 78. Brochet, M. et al. Genomic diversity and evolution within the species Streptococcus agalactiae. Microbes and Infection 8, 1227–1243 (2006). 79. Springman, A. C. et al. Selection, recombination, and virulence gene diversity among group B streptococcal genotypes. Journal of Bacteriology 191, 5419–5427 (2009). 80. Springman, A. C. et al. Pilus distribution among lineages of Group B Streptococcus: An evolutionary and clinical perspective. BMC Microbiology 14, (2014). 81. CDC. ANTIBIOTIC RESISTANCE THREATS in the United States. (2013). https://www.cdc.gov/drugresistance/pdf/ar-threats-2013-508.pdf 82. Beckmann, C., Waggoner, J. D., Harris, T. O., Tamura, G. S. & Rubens, C. E. Identification of novel adhesins from group B streptococci by use of phage display reveals that C5a peptidase mediates fibronectin binding. Infection and Immunity 70, 2869–2876 (2002). 26 83. Schubert, A. et al. A fibrinogen receptor from group B Streptococcus interacts with fibrinogen by repetitive units with novel ligand binding sites. Molecular Microbiology 46, 557–569 (2002). 84. Spellerberg, B. et al. Lmb, a protein with similarities to the LraI adhesin family, mediates attachment of Streptococcus agalactiae to human laminin. Infection and Immunity 67, 871–878 (1999). 85. Tenenbaum, T. et al. Streptococcus agalactiae invasion of human brain microvascular endothelial cells is promoted by the laminin-binding protein Lmb. Microbes and Infection 9, 714–720 (2007). 86. Dramsi, S. et al. Assembly and role of pili in group B streptococci. Molecular Microbiology 60, 1401–1413 (2006). 87. Maisey, H. C., Hensler, M., Nizet, V. & Doran, K. S. Group B streptococcal pilus proteins contribute to adherence to and invasion of brain microvascular endothelial cells. Journal of Bacteriology 189, 1464–1467 (2007). 88. Sheen, T. R. et al. Serine-rich repeat proteins and pili promote Streptococcus agalactiae colonization of the vaginal tract. Journal of Bacteriology 193, 6834–6842 (2011). 89. Wang, N. Y. et al. Group B streptococcal serine-rich repeat proteins promote interaction with fibrinogen and vaginal colonization. Journal of Infectious Diseases 210, 982–991 (2014). 90. Marques, M. B., Kasper, D. L., Pangburn, M. K. & Wessels, M. R. Prevention of C3 deposition by capsular polysaccharide is a virulence mechanism of type III group B streptococci. Infection and Immunity 60, 3986–3993 (1992). 91. Campbell, J. R., Baker, C. J. & Edwards, M. S. Deposition and degradation of C3 on type III group B streptococci. Infection and Immunity 59, 1978–1983 (1991). 92. Harris, T. O., Shelver, D. W., Bohnsack, J. F. & Rubens, C. E. A novel streptococcal surface protease promotes virulence, resistance to opsonophagocytosis, and cleavage of human fibrinogen. Journal of Clinical Investigation 111, 61–70 (2003). 93. Takahashi, S. et al. Role of C5a-ase in group B streptococcal resistance to opsonophagocytic killing. Infection and Immunity 63, 4764–4769 (1995). 94. Bohnsack, J. F. et al. A role for C5 and C5a-ase in the acute neutrophile response to group B streptococcal infections. Journal of Infectious Diseases 175, 847–855 (1997). 95. Cheng, Q., Stafslien, D., Purushothaman, S. S. & Cleary, P. The Group B Streptococcal C5a Peptidase Is Both a Specific Protease and an Invasin The Group B Streptococcal C5a Peptidase Is Both a Specific Protease and an Invasin. Infection and immunity 70, 2408– 2413 (2002). 27 96. Korir, M. L., Manning, S. D. & Davies, H. D. Intrinsic Maturational Neonatal Immune Deficiencies and Susceptibility to Group B Streptococcus Infection. (2017) doi:10.1128/CMR. 97. Vornhagen, J. et al. Bacterial hyaluronidase promotes ascending GBS infection and preterm birth. mBio 7, 1–10 (2016). 98. Whidbey, C. et al. A hemolytic pigment of Group B Streptococcus allows bacterial penetration of human placenta. The Journal of experimental medicine 210, 1265–81 (2013). 99. Whidbey, C. et al. A streptococcal lipid toxin induces membrane permeabilization and pyroptosis leading to fetal injury. EMBO molecular medicine 7, 488–505 (2015). 100. Randis, T. M. et al. Group B Streptococcus β-hemolysin/cytolysin breaches maternal-fetal barriers to cause preterm birth and intrauterine fetal demise in vivo. Journal of Infectious Diseases 210, 265–273 (2014). 101. Baron, M. J., Bolduc, G. R., Goldberg, M. B., Aupérin, T. C. & Madoff, L. C. Alpha C protein of group B Streptococcus binds host cell surface glycosaminoglycan and enters cells by an actin-dependent mechanism. Journal of Biological Chemistry 279, 24714– 24723 (2004). 102. Soriani, M. et al. Group B Streptococcus crosses human epithelial cells by a paracellular route. The Journal of infectious diseases 193, 241–250 (2006). 103. Izabela Sitkiewicz, Nicole M. Green, Nina Guo, Ann Marie Bongiovanni, Steven S. Witkin, J. M. M. Transcriptome Adaptation of Group B Streptococcus to Growth in Human Amniotic Fluid. PLoS One 4, 6114 (2009). 104. Doran, K. S. & Nizet, V. Molecular pathogenesis of neonatal group B streptococcal infection: No longer in its infancy. Molecular Microbiology 54, 23–31 (2004). 105. Carl, M. A. et al. Sepsis from the gut: The enteric habitat of bacteria that cause late-onset neonatal bloodstream infections. Clinical Infectious Diseases 58, 1211–1218 (2014). 106. le Doare, K. et al. Intrapartum Antibiotic Chemoprophylaxis Policies for the Prevention of Group B Streptococcal Disease Worldwide: Systematic Review. Clinical Infectious Diseases vol. 65 S143–S151 (2017). 107. Verani JR, McGee L, Schrag SJ; Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases, C. for D. C. and P. (CDC). Prevention of perinatal group B streptococcal disease--revised guidelines from CDC, 2010. MMWR Recomm Rep. 59, 1–36 (2010). 108. Arbara, B. et al. Changes in pathogens causing early-onset sepsis in very-low-birth-weight infants. N Engl J Med vol. 347 (2002). 28 109. Stoll et al. Early Onset Neonatal Sepsis: The Burden of Group B Streptococcal and E. coli Disease Continues. Pediatrics 127(5): 817–826 (2011). 110. Leineweber B, Grote V, Schaad UB, H. U. Transplacentally acquired immunoglobulin G antibodies against measles, mumps, rubella and varicella-zoster virus in preterm and full- term newborns. Pediatr Infect Dis J. 23, 361–3 (2004). 111. Lancefield, R. C., Mccarty, M. & Everly, W. N. Multiple Mouse-protective antibodies directed against Group B Streptocococci. Special reference to antibodies effective against protein antigens. J Exp Med. 142(1): 165–179 (1975). 112. Chen VL, Avci FY, K. DL. A maternal vaccine against group B Streptococcus: past, present, and future. Vaccine 31, (2013). 113. Twort, F. W. L.R.C.P. LOND., M.R.C.S. An Investigation on the nature of ultra- microscopic viruses. The Lancet. 186 1241-1243 (1915). 114. Furfaro, L. L., Chang, B. J. & Payne, M. S. Applications for bacteriophage therapy during pregnancy and the perinatal period. Frontiers in Microbiology vol. 8 (2018). 115. Nobrega, F. L., Costa, A. R., Kluskens, L. D. & Azeredo, J. Revisiting phage therapy: New applications for old resources. Trends in Microbiology vol. 23 185–191 (2015). 116. Brüssow, H. What is needed for phage therapy to become a reality in Western medicine? Virology vol. 434 138–142 (2012). 117. Schooley, R. T. et al. Development and use of personalized bacteriophage-based therapeutic cocktails to treat a patient with a disseminated resistant Acinetobacter baumannii infection. Antimicrobial Agents and Chemotherapy 61, (2017). 118. Jennes, S. et al. Use of bacteriophages in the treatment of colistin-only-sensitive Pseudomonas aeruginosa septicaemia in a patient with acute kidney injury-a case report. Critical Care vol. 21 (2017). 119. Kutter, E. et al. Phage Therapy in Clinical Practice: Treatment of Human Infections. Current Pharmaceutical Biotechnology vol. 11 www.biochimpharm.ge (2010). 120. Jault, P. et al. Efficacy and tolerability of a cocktail of bacteriophages to treat burn wounds infected by Pseudomonas aeruginosa (PhagoBurn): a randomised, controlled, double-blind phase 1/2 trial. The Lancet Infectious Diseases 19, 35–45 (2019). 121. Ganeshan & Hosseinidoust. Phage Therapy with a focus on the Human Microbiota. Antibiotics 8, 131 (2019). 122. Sybesma, W. et al. Bacteriophages as potential treatment for urinary tract infections. Frontiers in Microbiology 7, (2016). 29 123. Galtier, M. et al. Bacteriophages to reduce gut carriage of antibiotic resistant uropathogens with low impact on microbiota composition. Environmental microbiology 18, 2237–2245 (2016). 124. Cheng, Q., Nelson, D., Zhu, S. & Fischetti, V. A. Removal of group B streptococci colonizing the vagina and oropharynx of mice with a bacteriophage lytic enzyme. Antimicrobial Agents and Chemotherapy 49, 111–117 (2005). 125. Luo, X. et al. Characterization of bacteriophage HN48 and its protective effects in nile tilapia oreochromis niloticus against streptococcus agalactiae infections. Journal of Fish Diseases 41, 1477–1484 (2018). 126. Chen SL. Genomic Insights Into the Distribution and Evolution of Group B Streptococcus. Front Microbiol. 10, 1447 (2019). 127. Rosen, G. H. et al. Group B Streptococcus and the Vaginal Microbiota. in Journal of Infectious Diseases vol. 216 744–751 (Oxford University Press, 2017). 128. de Gregorio, P. R., Juárez Tomás, M. S., Leccese Terraf, M. C. & Nader-Macías, M. E. F. Preventive effect of Lactobacillus reuteri CRL1324 on Group B Streptococcus vaginal colonization in an experimental mouse model. Journal of Applied Microbiology 118, 1034–1047 (2015). 129. de Gregorio, P. R., Juárez Tomás, M. S. & Nader-Macías, M. E. F. Immunomodulation of Lactobacillus reuteri CRL1324 on Group B Streptococcus Vaginal Colonization in a Murine Experimental Model. American Journal of Reproductive Immunology 75, 23–35 (2016). 130. Virginia Martín, Nivia Cárdenas, Sara Ocaña, María Marín, Rebeca Arroyo, David Beltrán, Carlos Badiola, L. F. and J. M. R. Rectal and Vaginal Eradication of Streptococcus agalactiae (GBS) in Pregnant Women by Using Lactobacillus salivarius CECT 9145, A Target-specific Probiotic Strain. Nutrients 11, 810 (2019). 131. Mary Sharpe. Effect of Probiotics on GBS Colonization Status During Pregnancy: A Pilot Randomized Controlled Trial. Clinical Trials.gov. 132. Olsen, P., Williamson, M., Traynor, V. & Georgiou, C. The impact of oral probiotics on vaginal Group B Streptococcal colonisation rates in pregnant women: A pilot randomised control study. Women and Birth 31, 31–37 (2018). 133. Marziali, G., Foschi, C., Parolin, C., Vitali, B. & Marangoni, A. In-vitro effect of vaginal lactobacilli against group B Streptococcus. Microbial Pathogenesis 136, 103692 (2019). 134. Ortiz, L., Ruiz, F., Pascual, L. & Barberis, L. Effect of two probiotic strains of Lactobacillus on in vitro adherence of listeria monocytogenes, Streptococcus agalactiae, and Staphylococcus aureus to vaginal epithelial cells. Current Microbiology 68, 679–684 (2014). 30 135. Zárate, G. & Nader-Macias, M. E. Influence of probiotic vaginal lactobacilli on in vitro adhesion of urogenital pathogens to vaginal epithelial cells. Letters in Applied Microbiology 43, 174–180 (2006). 136. Iseppi, R., Messi, P., Camellini, S. & Sabia, C. Bacteriocin activity of Lactobacillus brevis and Lactobacillus paracasei ssp. paracasei. Journal of Medical Microbiology 68, 1359– 1366 (2019). 137. Gaspar, C. et al. Bacteriocin production of the probiotic Lactobacillus acidophilus KS400. AMB Express 8, (2018). 138. Gudiña, E. J., Rocha, V., Teixeira, J. A. & Rodrigues, L. R. Antimicrobial and antiadhesive properties of a biosurfactant isolated from Lactobacillus paracasei ssp. paracasei A20. Letters in Applied Microbiology 50, 419–424 (2010). 139. Gudiña, E. J., Teixeira, J. A. & Rodrigues, L. R. Isolation and functional characterization of a biosurfactant produced by Lactobacillus paracasei. Colloids and Surfaces B: Biointerfaces 76, 298–304 (2010). 140. Hynönen, U., Westerlund-Wikström, B., Palva, A. & Korhonen, T. K. Identification by flagellum display of an epithelial cell- and fibronectin-binding function in the S1pA surface protein of Lactobacillus brevis. Journal of Bacteriology 184, 3360–3367 (2002). 141. Hymes, J. P. & Klaenhammer, T. R. Stuck in the middle: Fibronectin-binding proteins in gram-positive bacteria. Frontiers in Microbiology vol. 7 (2016). 142. Kong, S. & Davison2, A. J. The Role of Interactions between 02, HzOz, *OH, e-and 0; in Free Radical Damage to Biological Systems’. ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS vol. 204 (1980). 143. Dahl, T. A., Robert Midden, W. & Hartman, P. E. Comparison of Killing of Gram- Negative and Gram-Positive Bacteria by Pure Singlet Oxygent. Journal of Bacteriology vol. 171 (1989). 144. Mokoena, M. P. Lactic acid bacteria and their bacteriocins: Classification, biosynthesis and applications against uropathogens: A mini-review. Molecules vol. 22 (2017). 145. Naidu, A. S., Bidlack, W. R. & Clemens, R. A. Probiotic spectra of lactic acid bacteria (LAB). Critical Reviews in Food Science and Nutrition 39, 13–126 (1999). 31 LACTOBACILLUS STRAINS VARY IN THEIR ABILITY TO INTERACT WITH HUMAN ENDOMETRIAL STROMAL CELLS CHAPTER 2 32 ABSTRACT While previously thought to be sterile, the placental membranes that surround the fetus during pregnancy are now thought to contain a low abundance microbiome. Specifically, abundance of Lactobacillus, a probiotic and dominant member of the microbiome of the lower reproductive tract, has been shown to correlate to with healthy, term pregnancies. We sought to access if four different Lactobacillus strains are able to associate with a model of the outermost cells of the placental membranes (dT-HESCs). Further, we examined the outcomes of that interaction. Herein, we found that all four Lactobacillus strains were capable of interacting with dT-HESCs; however, L. crispatus was statistically more successful (p < 0.00005), with 10.6% of bacteria attaching to host cells, while only an average of 0.8% of the other strains associated. The four strains also varied in their ability to form biofilms. Dependent on media type, L. reuteri 6475 forms the strongest biofilms in vitro. To examine the potential impact of Lactobacillus association with these cells on immune responses, total and phosphorylated protein levels of p38 of the Mitogen Activated Protein Kinase (MAPK) pathway was examined. Total levels of p38 were reduced to an average of 44% that of the cells without Lactobacillus (p < 0.05). While we observed a trend towards reduction of the phosphorylated form of p38, this difference was not significant (p > 0.05). Further, we found that association with Lactobacillus did not result in increased host cell death. Collectively, these data suggest that Lactobacillus attaches to the outermost cells of the placental membranes and that this association would not induce the MAPK pathway which has been associated with inflammation and host cell death. 33 INTRODUCTION The placental membranes surround the fetus during pregnancy and are made of two layers, the amnion and choriodecidua, held together by connective tissue and fibroblast cells. These layers serve as the final barrier between the fetus and ascending pathogens from the vaginal tract. Pathogens such as Group B Streptococcus that cause adverse pregnancy outcomes, including preterm birth and in utero infections, must interact with these membranes in order to cause infection. Infection of these membranes is thought to cause weakening of the membranes, leading to miscarriage, preterm birth or neonatal sepsis.1,2 While previously considered sterile, some metagenomic studies have found that the placenta and placental membranes may have a small load of commensal bacteria, including the genera Acinetobacter, Escherichia coli, Enterobacter and Lactobacillus.3 Some bacteria, including Lactobacillus species, have been detected in healthy, term pregnancies,3 yet few live culture studies have been performed to determine whether Lactobacillus can interact with the host cells in the placental membranes. Lactobacillus is a commensal bacterium that dominates the healthy vaginal tract of most women.4 Its role in human health has been heavily studied in the gut and urogenital tract. Mechanisms of health promotion include modulation of host immune responses and by direct or indirect inhibition of pathogens.5 Lactobacillus can indirectly inhibit pathogens by occupying host attachment sites or by competing for nutrients or other growth factors. Direct inhibition includes production of lactic acid or other secondary metabolites. Immune modulation by Lactobacillus can be localized or systematic. Mechanisms include increased production of immunoglobulins (IgA), stimulation of cytokines and increased phagocytosis by immune cells.6 As Lactobacillus is generally considered to be safe, its use as a probiotic during pregnancy in the urogenital tract is also being accessed. 34 In the urogenital tract, Lactobacillus has been found to inhibit a variety of individual pathogens including HIV, Uropathogenic Escherichia coli (UPEC), Group B Streptococcus (GBS), Neisseria gonorrhoeae and Garnella vaginalis.7 It has also been shown to counteract more complex disease states such as bacterial vaginosis.8 Modulation of the immune system has been more heavily studied in the gut, but recent studies suggest that Lactobacillus may also play a role in the urogenital tract. Specifically, supernatants from lactobacilli were shown affect the cytokine response to lipopolysacchadide (LPS) in the decidua, reducing a potentially harmful inflammatory response during pregnancy.9 To better understand the interactions between live Lactobacilli and the placental membranes, we sought to characterize the interactions of four Lactobacilli strains with a cell line model of the outermost layer of cells of the placental membranes: decidualized human endometrial stromal cells (dT-HESCs). Using this model, we assessed whether Lactobacillus maintains a beneficial interaction with these cells to provide similar protective functions as has been described in other body sites. Due to known variation between Lactobacilli strains, we also characterized growth and biofilm phenotypes of the strains. 35 MATERIALS AND METHODS Bacterial strains and growth conditions Lactobacillus strains were selected to represent species that have been found in both the vagina and extraplacental membranes via metagenomics.3,10 Additionally, previous work on Lactobacillus has focused on certain species including L. reuteri (Lr), L. gasseri (Lg) and L. crispatus (Lc); therefore, strains were selected to represent these species. Lr6475 was isolated from breast milk (MM4-1A PTA-6475), Lg33323 from the vagina (DSM 20243 [63 AM]), and Lc19390 from stool. Lr17938 is the daughter strain of L. reuteri ATCC 55730 which was isolated from breast milk but also carried potentially transferable resistance traits for tetracycline and lincomycin.11 Lr17938 no longer carries this plasmid. Lactobacillus strains were cultured in deMan, Rososa and Sharpe (MRS) broth (Difco 288130) or agar at 37°C with 5% CO2. Growth curves were performed by diluting an overnight culture of Lactobacillus 1:10 and observing growth by OD595 for 8 hours in a BioTek Cytation 3 Imager. Biofilm Production Over-night cultures of Lactobacillus were inoculated into fresh Tryptic Soy Broth (TSB) + 1% dextrose (TSBd) or MRS in a 96 well plate. Each biological replicate contains at least four technical replicates and media controls. Plates were incubated for 48 hours and subsequently washed twice with PBS to removal non-adherent cells. 100µl of 3% Crystal Violet was added to each well and incubated for 10 minutes. Wells were washed four times with PBS, and 200µl of ethanol was added to solubilize the crystal violet. 50µl was moved to a fresh plate to read by 595 absorbance. To analyze, the media control was subtracted and each resulting number was multiplied by four. Technical replicates were averaged within a plate. At least three biological 36 replicates were performed. Significance was determined using an unpaired analysis of variance (ANOVA). Cutoff values were established as preciously described.13 Briefly, an OD595 cut-off value (ODc) was assigned based on the negative control and its standard deviations across three biological replicates (ODc = negative control + 3 standard deviations = 0.25). Non-biofilm producers fall below the ODc, while weak biofilm formers fall below 2 x ODc (0.50), moderate biofilm formers are below 3 x ODc (.75), and strong biofilm formers are above 4 x ODc (1.0). Cell Culture Telomerase-immortalized human endometrial stromal cells (T-HESC, ATCC CRL- 4003)12 were cultured in DMEM/Nutrient Mixture Ham's F-12 with l-glutamine (Sigma) supplemented with 1% ITS+ Universal Culture Supplement Premix (BD), 1.5 g/liter sodium bicarbonate, 2% pen/strep, and 10% charcoal-treated FBS (HyClone), which is referred to as HESC medium herein. For the cell experiments, T-HESCs were decidualized (dT-HESCs) as previously described14,15 by incubating the cells with 0.5 mM 8-bromo-cyclic amp (cAMP) (Sigma Aldrich, St. Louis, MO) for three to six days. Completely confluent monolayers were used for all experiments to ensure that only decidual cell surfaces were available for bacterial interactions. Association of Lactobacillus strains with decidual cells. Association assays were performed as previously described.15,16 Briefly, monolayers of decidual cells were washed thrice with phosphate buffered saline (PBS) before infection with bacterial cultures. Overnight bacterial cultures of Lactobacillus were washed once with PBS and re-suspended in infection media (T- HESC medium without ITS+ or antibiotics and with 2% charcoal-treated FBS). dT-HESC cells were infected at a multiplicity of infection (MOI) of ten Lactobacillus cell per host cell 37 (MOI=10). Following a two-hour incubation at 37ᵒC in atmospheric conditions, samples were taken from each well to determine the final colony forming units (CFU) of Lactobacillus. To quantify the number of associated cells, wells were washed three times to remove unattached cells. Host cells were disrupted using Triton-X, and wells were scraped thoroughly. Repeated pipetting and vortexing ensured even resuspension before plating. The percent of associated cells was calculated by dividing the CFU of associated cells by the final CFU of each well. These experiments were performed in three to four biological replicates of technical triplicates. Significance was determined using an unpaired ANOVA. Cytotoxicity assays Monolayers of dT-HESCs were cultured in 24-well plates and cells were infected with GBS (MOI=10) using a protocol described by Korir, et al.17 After incubation for four hours, cells were washed twice with PBS, treated with 4µM ethidium homodimer 1 (Molecular Probes, Eugene, OR), suspended in PBS, and incubated at room temperature for 30 mins without light. Fluorescence was measured at 528-nm excitation and 617-nm emission using a plate reader (Beckman Coulter, Inc). The total number of cells in each well was calculated by adding 0.1% (wt/v) Saponin (Sigma) and incubating for at least 20 minutes before repeating the fluorescence reading. The percent permeability (cell death) was calculated by dividing the initial reading by the second and multiplying by 100. These experiments were performed in three to four biological replicates of technical triplicates. Significance was determined using an unpaired ANOVA. Detection of IL-10 by ELISA: dT-HESCs were infected with Lactobacillus (MOI=10). Cell supernatants were collected following a three-hour incubation, spun down at 28,000 rpm for 20 minutes to remove bacteria 38 and cellular debris, and stored at -20°C. Concentration of IL-10 was determined using the IL-10 Human ELISA kit: ab100549 (Abcam;Cambridge, MA) according to manufacturer’s instructions and comparisons were made relative to mock infection. Data from three biological replicates were pooled to quantify the average cytokine concentrations (pg/mL). Significance was determined by unpaired ANOVA and post-hoc Dunnett’s tests relative to mock infection. Western Blotting for p38 dT-HESC were infected with Lactobacillus (MOI=10) as described above and protein lysates were collected following a three-hour infection to quantify the abundance of p38. Cells were washed twice with PBS and lysed with 300µl of lysis buffer for ten minutes at 4°C. Each well was scraped, and the lysate was collected. Samples were spun at 28,000 rpm for 20 minutes to pellet cellular debris, transferred to a fresh tube, and stored at -20°C. Protein concentration was determined by bicinchoninic acid (BCA) assay (Pierce) using Bovine Serum Albumin (BSA) immediately prior to separation on a 4-15% polyacrylamide gel (BioRad). Samples were transferred to a polyvinylidene difluoride (PVDF) membrane and the membrane was blocked in 5% BSA (Fisher Scientific) + 0.1% Tween 20 (Sigma) in Tris-Buffered Saline (TBS) for at least two hours. The membranes were exposed to primary antibodies (1:1000) for phospho MAPK- p38 (T180+Y182; #4511S, Cell Signaling Technology), total MAPK p38 (1:1000; #8690S, Cell Signaling), or beta-tubulin (1:500; Santa Cruz Biotechnology) overnight at 4°C. Membranes were washed five times over an hour in TBS+0.1% Tween 20 and incubated with goat-anti- rabbit IgG-HRP secondary antibodies (1:5000, Life Technologies) at room temperature for 1.5 hours followed by another set of wash steps in TBS+0.1% Tween 20. The membranes were incubated with ECL chemiluminescence reagent (Pierce) prior to developing using an Amersham 39 Imager 600 (GE Life Sciences). Relative abundance of the proteins was determined using Image J and values from three independent replicates were pooled. Beta tubulin was used as a loading control. 40 RESULTS Lactobacillus strains vary in growth in various media types To determine growth phenotypes of these Lactobacillus strains in the various media types required for experiments, growth curves were performed. deMan, Rososa and Sharpe (MRS) broth is a standard growth broth for Lactobacillus. All the strains grew in MRS media and demonstrated similar lengths of lag phase and growth rates (Figure 2.1A). However, they did differ in the length of exponential phase and maximum OD595. Lr6475 entered stationary phase earliest at 10 hours and reached a maximum OD595 of 1.5 at 12 hours. Lg 33323 reached a similar maximum OD595 of 1.5 but entered stationary phase three hours later. Lr17938 reached the lowest OD595 of 1.14 after 14 hours of growth. Lc19390 reached the highest OD595 of 1.73 but also entered stationary phase later, about 20 hours into growth. TSB containing 1% dextrose (TSBd) added has been used previously to assess biofilms in GBS; therefore, we sought to characterize Lactobacillus biofilms in this media as well.15 All the Lactobacillus strains grew poorly in the media. Lc19390 and Lg33323 grew similarly, each reaching a maximum OD595 of 0.02. Lr6475 grew slightly better, reaching an OD595 of 0.04, while Lr17938 grew the best, reaching 0.06 in 48 hours (Figure 2.1B). Human Endometrial Stromal Cells (HESC) 2/0 media is a minimal media, which contains 2% per volume fetal bovine serum and 0% antibiotics and antimycotics. None of the Lactobacillus strains grew in the cell culture media, never reaching above 0.005 OD595 (Figure 2.1C). L. reuteri strains form significantly better biofilms Bacterial biofilms have been shown to play a role in adherence to surfaces and persistence in hosts; therefore, we assessed these strains differ for the ability to form biofilms. 41 Weak versus strong biofilms were differentiated as described in the methods. Biofilm formation was dependent on media type (Figure 2.2). L. reuteri 6475 and L. gasseri 33323 formed significantly better biofilms in MRS media (p < 0.05; p < 0.00005) compared to TSBd. Conversely, L. reuteri 17938 formed the strongest biofilm in TSBd compared to MRS (p < 0.05). While there was not a statistically significant difference in biofilm formation for L. crispatus between media types, it did form a higher biofilm in TSBd, increasing from 0.039 (non-biofilm former) to 0.37 (weak biofilm former). There were also differences between Lactobacillus strains within media types. In the MRS media, biofilms formed by L. reuteri 6475 were significantly higher than L. reuteri 17938 (p < 0.05) and L. crispatus (p< 0.005) but were not different from those formed by L. gasseri (p > 0.05). Biofilms formed by L. gasseri were also significantly higher than L. reuteri 17938 (p < 0.005) and L. crispatus (p< 0.005). Biofilm formation was not statistically different between L. reuteri 17938 and L. crispatus in the MRS media type (p > 0.05). In TSBd both L. reuteri strains formed strong biofilms, reaching ODs of 1.74 and 2.71, respectively (Figure 2.2B). L. crispatus formed a weak biofilm of 0.37. L. gasseri was classified as a non-biofilm former in TSBd, falling below the ODc at 0.18. In this media, biofilms formed by L. reuteri 17938 were significantly stronger than both L. gasseri and L. crispatus (p < 0.05, 0.005; unpaired ANOVA). L. reuteri 6475 was only significantly stronger than L. gasseri. (p < 0.05, unpaired ANOVA). There was no statistical difference in biofilm production between L. gasseri and L. crispatus (p > 0.05). 42 L. crispatus associates with decidual cells significantly better than other Lactobacillus strains A key step to establishing an interaction with a host cell is attachment. To determine if different Lactobacillus strains can interact with dT-HESCs, association assays were performed. As this is the first assay with Lactobacillus using this cell type, we compared association levels with those published for an organism known to colonize decidual cells, Group B Streptococcus (GBS). Our lab has previously used this experimental design to establish GBS association from approximately 0.02 to 0.6% (Korir, 2014). This level of association is similar to what we observe with Lactobacillus, suggesting Lactobacillus is capable of associating with this cell type. Nonetheless, variation between strains was observed (Figure 2.3). Of the strains tested, L. crispatus associated significantly better than the other strains with 10.6% of the total bacteria in the well establishing a stable association with the cells (p < 0.0005, unpaired ANOVA). The other strains were not significantly different from each other in percent association (p > 0.05); 1.11% of L. reuteri 6475, 0.52% of L. gasseri and 0.77% of L. reuteri 17938 associated with the dT-HESCs. Lactobacillus does not produce an inflammatory response in dT-HESCs As Lactobacillus associated with dT-HESCs, we next sought to assess if they triggered an immune response in this cell line. p38 is a key player in the MAPK pathway that promotes cell death and inflammation.18 As previous studies have found that Lactobacillus can affect levels of this protein and protect against invading pathogens,19-23 we wanted to determine if these strains have a similar affect and if strains vary in that effect. Western blots were used to determine the levels of both the phosphorylated (active) form and total p38 after a three-hour infection with Lactobacillus. In comparison to a mock infection with cell culture media, each Lactobacillus 43 strain lowered the levels of total p38, but did not affect the phosphorylated form of p38 (Figure 2.4A). Compared to the mock infection, total p38 was significantly reduced to 42.7%, 44.3%, 46.6% and 41.5%, respectively, for L. reuteri 6475, L. gasseri, L. reuteri 17938 and L. crispatus (p < 0.05). There were no significant differences between the Lactobacillus strains (p > 0.05). While levels of phosphorylated p38 were also decreased with the addition of Lactobacillus in comparison to the mock infection, these results were not statistically significant (p> 0.05). L. reuteri 6475, L. gasseri, L. reuteri 17938 and L. crispatus reduced levels of phosphorylated p38 to 66.5%, 56.7%, 52.0% and 69.4%, respectively (Figure 2.4B). There were no differences between strains. To further assess if Lactobacillus had any anti-inflammatory effects, we also utilized an ELISA to detect changes in IL-10 secretion. We were unable to detect any changes in IL-10 production as very low production of this cytokine was observed in both the mock and Lactobacilli treatments (data not shown). Lactobacillus does not trigger dT-HESC death Reduction in total p38 and maintenance of the phosphorylated form, led us to assess downstream effects of the mitogen-activated protein kinase pathway to determine the effect of the reduction in total levels of p38. Because this pathway is associated with cell death,18 we compared host cell permeability during incubation with Lactobacillus to that of a mock infection as a marker of host cell death (Figure 2.5). No significant differences in the host cell permeability were observed with Lactobacilli infection (p> 0.05, unpaired ANOVA). The media control contributed to a 24.9% cell death rate, while L. reuteri 6475, L. gasseri, L. reuteri 17938 and L. crispatus caused 29.1%, 26.93%, 30.2% and 27.0% cell death, respectively. Further, we observed no differences between the individual strains (p> 0.05, unpaired ANOVA, Figure 2.5). 44 DISCUSSION Lactobacillus has been studied as a probiotic in the gut and urogenital tract, but its effects in the placental membranes remain uncharacterized. As this bacterium is being proposed as a probiotic during pregnancy,24,25 it is important to understand the potential effects of Lactobacillus colonization of the placental membranes should it reach them. Previous work has found vast differences in probiotic capacity between species of Lactobacillus and stains within the same species;26 therefore, we first sought to characterize traits that may affect their ability to interact with host cell membranes including growth and biofilm formation. Growth differences were observed between strains and between medias. As HESC 2/0 media lacks critical nutrients, it is not surprising that all four of the Lactobacillus strains were unable to grow. The inability to grow in this model system may affect some traits such as the production of secondary metabolites. To colonize the placental membranes, Lactobacillus must interact with the outermost layer of cells in the membranes, decidualized stromal cells. Indeed, we have demonstrated that each of the Lactobacillus strains was capable of attaching to dT-HESCs, though variation was observed in the percent association across strains. The overall level of association to dT-HESCs was similar to that described for GBS, an opportunistic pathogen that can attach to and invade dT-HESCs even though variation across GBS strains was also observed.14,15,21 Notably, the L. crispatus strain was significantly better at associating with the decidualized stromal cells, which provides support for the recovery of L. crispatus from the placental membranes of healthy, term pregnancies using metagenomics.3 As L. acidophilus was the other Lactobacilli species found in this study,3 future work should also examine if it is more capable of associating with decidual cells. The ability to attach to the placental membranes suggests that these strains may be able to 45 serve as a barrier to invading pathogens as their presence may remove potential attachment sites for those bacteria. Further work coculturing Lactobacillus with different pathogens known to colonize the placental membranes should be completed to assess this potential. The ability to form a biofilm has been associated with persistence in many environments, and strains of Lactobacillus have been found to persist in other host sites including the intestines for up to a week.27 We found that the Lactobacillus biofilms were dependent on growth media, suggesting nutrients are an important factor in the ability of Lactobacillus to form biofilms. Indeed, L. reuteri 6475 and L. gasseri 33323 formed significantly better biofilms in MRS media (p < 0.05; p < 0.00005), while L. reuteri 17938 formed a stronger biofilm in TSBd (p < 0.05). Further, we found differences in biofilm formation between individual Lactobacillus strains. L. crispatus did not form strong biofilms in either media, while L. reuteri 6475 formed strong biofilms in both medias. L. gasseri and L. reuteri 17938 were more drastically impacted by media type as the growth media determined if each was a low or high biofilm former. These differences in biofilm production could impact a given strain’s ability to persist in a given environment. Premature birth and preterm premature rupture of the membranes (pPROM) are associated with inflammation and host cell death in the placental membranes;28 therefore, it is important to assess whether Lactobacillus has an effect on immune signaling and host cell death. The MAPK pathway responds to stress and is involved in inflammatory signaling and cell death.18 We observed significantly reduced levels of total p38 among all four strains. The lower amount of total protein could reduce the overall output of the pathway, potentially stunting the ability of this pathway to promote host cell death. Though not significant, we also found a trend 46 toward reduced levels of phosphorylated or active form of p38 in response to Lactobacillus compared to the mock infection, suggesting that Lactobacillus is not inducing this pathway. While not previously studied in this cell type, other studies have found that Lactobacillus can induce the MAPK pathway in macrophages, while other studies observed reduced MAPK induction with Lactobacillus in the gut, intestinal epithelial cells and the liver.19–23 Still other work in a nematode model of E. coli sepsis, only certain strains of Lactobacillus further increased MAPK activation during infection, resulting in better survival outcomes.29 Further work will be needed to more fully understand the effect of Lactobacillus on this pathway and its downstream effect on pregnancy outcomes. The reduced activation of the MAPK pathway in addition to no changes in host cell death (Fig 2.4) suggests that colonization of Lactobacillus would not likely damage the placental membranes through this cell type. As there are other cells types in the placental membranes, including macrophages, it would also be important to access the effects of Lactobacillus in these other cell types as well or in a more complex model. Indeed, there were no significant differences in host cell permeability after four hours of incubation. In this same experimental design and conditions, known pathogen GBS can cause about 70% host cell death (Chapter 3, Figure 3.4), suggesting that Lactobacillus does not induce host death in these cells. In other systems, Lactobacillus modulates the immune system to reduce inflammation or prime immune cells for invading pathogens.21 Production of IL-10 is a marker of an anti- inflammatory response to Lactobacillus. This interleukin has previously been shown to be induced by Lactobacillus in other cell types including macrophage cell lines (RAW264.7 (mouse) and THP-1(human)), cervical tumor cells (HeLa), and colonic epithelial cells (Caco2).21,30,31 However, we observed very little production of IL-10 in mock or following 47 infection with all four strains of Lactobacillus in our study. Because it is possible that this cell line does not produce higher levels of cytokines in general, other models, such as whole decidua that produce higher cytokine levels, may serve as a better model of studying this interaction.9,32 Collectively, these data suggest that different strains of Lactobacillus sustain a commensal relationship with cells of the decidua. We found variation among strains of Lactobacillus in growth, biofilm production and association with dT-HESCs, suggesting that this relationship will be strain-dependent, as suggested in the literature for other body sites. Further research needs to be done to examine if these characteristics allow Lactobacillus to perform the barrier function against pathogen invasion as seen in the vaginal tract. These strains should also be studied in more complex models to examine the effect of the other cell types in the decidua, as they may alter the relationship. Since the placental membranes play such an important role in maintaining a healthy pregnancy, increasing our understanding of how traditional commensal bacteria affect these membranes could enhance knowledge of premature birth and other adverse pregnancy-related outcomes. 48 ACKNOWLEDGEMENTS We would like to thank Dr. Cindy Arvidson and Dr. Poorna Visvanathan for donation of Lactobacillus strains. We would also like to thank Dr. Neal Hammer and Dr Arvidson for their careful editing of the chapter. 49 APPENDIX 50 Figure 2.1. Lactobacillus growth varies by media type. Lactobacillus strains were grown overnight, washed in PBS, and inoculated into fresh A) MRS media; B) TSA + 1% Dextrose; or C) HESC 2/0 media. The Optical Density 595 (OD595) was measured every 30 minutes for 48 hours using a spectrophotometer. 51 Figure 2.2. Lactobacillus species differ in the ability to form biofilms depending on media type. Lactobacillus was grown overnight, washed in PBS and incubated in: A) TSB with 5% dextrose; or B) MRS media for 48 hours in a 96-well plate. Biofilms were assessed via crystal violet staining. Normalized absorbance is calculated by taking the OD595 subtracting that of the control and multiplying by four to account for dilutions. Experiments were performed in biological triplicate of technical quadruplicates. Error bars represent standard deviation between biological trials. Significance was determined by an unpaired ANOVA. 52 Figure 2.3. Lactobacillus associates with decidualized Human Endometrial Stromal Cells (dT-HESC). dT-HESCs were infected with Lactobacillus at a MOI of 10 for two hours. The percent of associated bacteria was calculated relative to the total number of bacteria in the well. Experiments were completed in biological quadruplets of technical triplicates. Error bars represent standard deviation between biological trials. Significance was determined by an unpaired ANOVA. d e t a i c o s s A t n e c r e P 2 0 1 5 1 0 5 2 1 0 p < 0.05 L r6 4 7 5 L g 3 3 3 2 3 L r1 7 9 3 8 L c 1 9 3 9 0 53 Figure 2.4. Lactobacillus affects total p38 and phosphorylated levels of p38 (Pp38). dHESCs were infected with Lactobacillus at a MOI of 10 and incubated for three hours. Protein lysates were collected and analyzed via Western Blotting using densitometry in Image J. A β-tubulin internal loading control was used to account for loading differences. Graphed densitometry data is presented as a percent of the mock control for each of three biological replicates. Error bars represent the standard deviation of the data. Significance was determined using an unpaired ANOVA. A. y r t e m o t i s n e D t n e c r e P n o i t c e f n I k c o M o t d e z i l a m r o N 1 2 5 1 0 0 7 5 5 0 2 5 0 T o ta l p 3 8 p < 0.05 M ock Lr6475 Lg33323 Lr17938 Lc19390 B. y r t e m o t i s n e D t n e c r e P n o i t c e f n I k c o M o t d e z i l a m r o N 1 2 5 1 0 0 7 5 5 0 2 5 0 54 P p 3 8 M ock Lr6475 Lg33323 Lr17938 Lc19390 Figure 2.5. Lactobacillus does not induce dT-HESC death. dT-HESCs were infected with Lactobacillus at a MOI of 10 and incubated for four hours. Cell permeability was detected using an ethidium homodimer assay, and percent permeability was calculated by lysing the remaining cell in each well. Graphed data represents three biological replicates, and the error bars represent the standard deviation of the data. Significance was determined using an unpaired ANOVA. y t i l i b a e m r e P t n e c r e P 5 0 4 0 3 0 2 0 1 0 0 Lr6475 Lg33323 Lr17938 Lc19390 M edia 55 REFERENCES 56 REFERENCES 1. Doran, K. S. & Nizet, V. Molecular pathogenesis of neonatal group B streptococcal infection: No longer in its infancy. Molecular Microbiology vol. 54 23–31 (2004). 2. Wennekamp, J. & Henneke, P. Induction and termination of inflammatory signaling in group B streptococcal sepsis. Immunological Reviews vol. 225 114–127 (2008). 3. Prince, A. L. et al. The placental membrane microbiome is altered among subjects with spontaneous preterm birth with and without chorioamnionitis. American Journal of Obstetrics and Gynecology 214, 627e1-627e16 (2016). 4. Ravel, J. et al. Vaginal microbiome of reproductive-age women. Proceedings of the National Academy of Sciences of the United States of America 108, 4680–4687 (2011). 5. Bryan Tungland. Microbiota and the Urogenital Tract, Pathogenesis, and Therapies. in Human Microbiota in Health and Disease From Pathogenesis to Therapy 605–647 (2018). 6. Erickson, K. L. & Hubbard, N. E. Probiotic Immunomodulation in Health and Disease. The Journal of Nutrition 130, 403S-409S (2000). 7. Spurbeck, R. R. & Arvidson, C. G. Lactobacilli at the front line of defense against vaginally acquired infections. Future Microbiology vol. 6 567–582 (2011). 8. Reid G, Burton J, D. E. The rationale for probiotics in female urogenital healthcare. MedGenMed 6, 49 (2004). 9. Li, W. et al. Lipopolysaccharide-induced profiles of cytokine, chemokine, and growth factors produced by human decidual cells are altered by lactobacillus rhamnosus gr-1 supernatant. Reproductive Sciences 21, 939–947 (2014). 10. Aagaard, K. et al. The Placenta Harbors a Unique Microbiome. Sci Transl Med. 6(237): 237 (2014). 11. Urbańska, M. & Szajewska, H. The efficacy of Lactobacillus reuteri DSM 17938 in infants and children: a review of the current evidence. European Journal of Pediatrics 173, 1327–1337 (2014). 12. Krikun, G. et al. A novel immortalized human endometrial stromal cell line with normal progestational response. Endocrinology 145, 2291–2296 (2004). 13. Stepanović et al. Quantification of biofilm in microtiter plates: overview of testing conditions and practical recommendations for assessment of biofilm production by staphylococci. APMIS 115(8):891-9 (2007). 14. Brosens, J. J. et al. Human Endometrial Fibroblasts Immortalized by Simian Virus 40 57 Large T Antigen Differentiate in Response to a Decidualization Stimulus. Endocrinology. 137(6): 2225-31 (1996). 15. Korir, M. L. et al. Association and virulence gene expression vary among serotype III group B Streptococcus isolates following exposure to decidual and lung epithelial cells. Infection and Immunity 82, 4587–4595 (2014). 16. Parker, R. E. et al. Association between genotypic diversity and biofilm production in group B Streptococcus. BMC Microbiology 16, (2016). 17. Korir, M. L. et al. Investigation of the role that NADH peroxidase plays in oxidative stress survival in group B Streptococcus. Frontiers in Microbiology 9, (2018). 18. Cargnello, M. & Roux, P. P. Activation and Function of the MAPKs and Their Substrates, the MAPK-Activated Protein Kinases. Microbiology and Molecular Biology Reviews 75, 50–83 (2011). 19. Ho, M. et al. Oral Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 to reduce Group B Streptococcus colonization in pregnant women: A randomized controlled trial. Taiwanese Journal of Obstetrics and Gynecology 55, 515–518 (2016). 20. Virginia Martín, Nivia Cárdenas, Sara Ocaña, María Marín, Rebeca Arroyo, David Beltrán, Carlos Badiola, L. F. and J. M. R. Rectal and Vaginal Eradication of Streptococcus agalactiae (GBS) in Pregnant Women by Using Lactobacillus salivarius CECT 9145, A Target-specific Probiotic Strain. Nutrients 11, 810 (2019). 21. Strus M1, Malinowska M, H. PB. In vitro antagonistic effect of Lactobacillus on organisms associated with bacterial vaginosis. J Reprod Med. 47, 41–6 (2002). 22. Alander, M. et al. Persistence of Colonization of Human Colonic Mucosa by a Probiotic Strain, Lactobacillus rhamnosus GG, after Oral Consumption. Applied and Environmental Microbiology vol. 65 (1999). 23. Jeong, M. et al. Heat-Killed Lactobacillus plantarum KCTC 13314BP Enhances Phagocytic Activity and Immunomodulatory Effects Via Activation of MAPK and STAT3 Pathways. Journal of Microbiology and Biotechnology 29, 1248–1254 (2019). 24. Cui, Y. et al. Lactobacillus reuteri ZJ617 Culture Supernatant Attenuates Acute Liver Injury Induced in Mice by Lipopolysaccharide. The Journal of Nutrition (2019) doi:10.1093/jn/nxz088. 25. Kanmani, P. & Kim, H. Functional capabilities of probiotic strains on attenuation of intestinal epithelial cell inflammatory response induced by TLR4 stimuli. BioFactors 45, 223–235 (2019). 26. Li, H. et al. Lactobacillus acidophilus alleviates the inflammatory response to enterotoxigenic Escherichia coli K88 via inhibition of the NF-κB and p38 mitogen- activated protein kinase signaling pathways in piglets. BMC Microbiology 16, 1–8 (2016). 58 27. Kamaladevi, A. & Balamurugan, K. Lactobacillus casei triggers a TLR mediated RACK-1 dependent p38 MAPK pathway in Caenorhabditis elegans to resist Klebsiella pneumoniae infection. Food and Function 7, 3211–3223 (2016). 28. Menon R. and Papaconstantinou J. p38 Mitogen Activated Protein Kinase (MAPK): A New Therapeutic Target for Reducing the Risk of Adverse Pregnancy Outcomes. Expert Opin Ther Targets. 20(12): 1397–1412 (2016). 29. Zhou, M. et al. Cell signaling of Caenorhabditis elegans in response to enterotoxigenic Escherichia coli infection and Lactobacillus zeae protection. Frontiers in Immunology 9, (2018). 30. Garcia-Castillo, V. et al. Characterization of the immunomodulatory and anti- Helicobacter pylori properties of the human gastric isolate Lactobacillus rhamnosus UCO-25A. Biofouling 1–16 (2019). 31. Kook, S.-Y., Chung, E.-C., Lee, Y., Lee, D. W. & Kim, S. Isolation and characterization of five novel probiotic strains from Korean infant and children faeces. PLOS ONE 14, e0223913 (2019). 32. Boldenow, E. et al. Antimicrobial peptide response to Group B Streptococcus in human extraplacental membranes in culture. Placenta 34, 480–485 (2013). 59 THE IMPACT OF LACTOBACILLUS ON GROUP B STREPTOCOCCAL INTERACTIONS WITH PLACENTAL MEMBRANES CHAPTER 3 60 ABSTRACT Group B Streptococcus (Streptococcus agalactiae, GBS) contributes to the global human disease burden through adverse pregnancy outcomes and neonatal disease. Currently, the only preventative measure implemented in the United States is intrapartum antibiotic prophylaxis (IAP), which has only impacted one form of neonatal disease and does not improve pregnancy outcomes. Lactobacillus is the dominant member of the microbiota of the lower reproductive tract and has recently been identified in the upper reproductive tract including the placental membranes during healthy, term pregnancies. Previous work has shown that Lactobacillus reduces rectal vaginal colonization of GBS, but no studies have examined how Lactobacillus might impact the ability of GBS to cause adverse pregnancy outcomes by associating with and invading the placental membranes. Herein, we characterized interactions using two GBS strains (colonizing and invasive) and four Lactobacillus strains that had been previously characterized for their ability to interaction with a model of the outermost cells of the placental membranes, dT-HESCs. We found that live Lactobacillus does not affect growth or biofilm production of GBS in vitro, but that L. gasseri 33323 increases GBS association with dT-HESCs in both GBS strains by 4 - 6% (p < 0.005). Increased association did not result in increased invasion (p > 0.05) or increased host cell death. Indeed, a statistically significant reduction in host cell death was observed with certain combination of GBS and Lactobacillus (p < 0.05). As Lactobacillus is known to secrete many inhibitory compounds, we also sought to characterize the effect of Lactobacillus supernatants on GBS. We found that these supernatants were able to inhibit growth (p < 0.00005) and biofilm formation (p < 0.005) of GBS, though this was strain dependent. We also observed increases in GBS-induced host cell death with Lactobacillus supernatants in the invasive strain of GBS. Finally, we assessed whether the supernatant from one strain, L. reuteri 61 6475, could broadly inhibit growth of GBS. Indeed, we found that growth was reduced to an average of 46.6% of each GBS strain alone. Collectively, these data suggest that both live Lactobacillus and its supernatant could impact GBS interactions with the placental membranes. 62 INTRODUCTION Group B Streptococcus (Streptococcus agalactiae, GBS) is a global human health burden , contributing to neonatal infections and deaths and adverse pregnancy outcomes, including premature birth and stillbirth. In the United States alone, GBS is thought to cause 12% of stillbirths1 and is the leading cause of neonatal pneumonia, sepsis and meningitis.2,3 4-6% of infected newborns succumb to the infection, while others have long-lasting effects including deafness and developmental disabilities.3 GBS is thought to contribute to premature birth and stillbirth by accessing the fetus via an ascending infection. While GBS vaginal colonization is asymptomatic, GBS association or invasion of the placental membranes has been linked to inflammation and subsequent fetal infection.4–7 Currently, consensus guidelines in the United States recommend that mothers who are culture positive for GBS at 37-39 weeks of pregnancy should receive intrapartum antibiotic prophylaxis (IAP). Proper administration of IAP requires four hours of antibiotics during labor. This practice has reduced the rate of early onset neonatal disease number of early onset cases but has not affected late onset disease or adverse pregnancy outcomes.8 Additionally, the antibiotic treatment is known to affect both the maternal vaginal microbiome and the neonatal gut microbiome, which are important for neonatal development.9–13 These shortcomings of IAP combined with growing concern of antibiotic resistance have resulted in increased interest in alternative therapies for the prevention of GBS-associated neonatal disease and adverse pregnancy outcomes. As maternal colonization is the primary risk factor for neonatal infection, therapies should reduce maternal colonization.2 These therapies should be more specific than the 63 antibiotic course; however, so that multiple doses could be used to combat adverse outcomes in the mother and baby and the microbiomes of the mother and neonate could remain unaltered. Lactobacillus, a dominant member of the vaginal and cervical microbiota in most women, has been examined as a probiotic against a variety of bacteria including GBS. Human trials using Lactobacillus alone or in combination with other probiotic species, for instance, have shown variable, but promising success in reducing rectal-vaginal colonization in pregnant mothers.14–16 Reasons for variability include differences in the length of intervention, strains and dosage.14 There is less research available on how Lactobacillus may affect GBS and ascending infection; however, some studies have found that Lactobacillus can reduce the risk of premature birth triggered by inflammation.17–19 To our knowledge, no prior studies have been conducted to determine how Lactobacillus affects GBS ascension to the vagianl tract or virulence phenotypes. As Lactobacillus has been identified in the human placental membranes using metagenomic techniques,20 we aimed to assess how Lactobacillus alters the ability of GBS to associate with and invade this important barrier to the fetus. To further characterize the interaction, we examined changes in GBS growth and host cell death. These in vitro studies reveal the mechanisms behind Lactobacillus’s ability to affect colonization and premature birth, allowing for better understanding of its use as a probiotic and of its limitations. 64 MATERIALS AND METHODS Bacterial strains and growth conditions GBS strains were selected based on multilocus sequence type (ST) designation, capsular serotype, and source. Two strains belonged to ST-17, which has been most associated with severe disease in neonates.21,22 One strain, GB00112, was isolated using a rectal-vaginal swab from a colonized pregnant woman23 and the second from the blood of a septic newborn (GB00411)24 GBS was cultured in Todd-Hewitt broth (THB) or half concentrated THB with agar (THA) at 37°C with 5% CO2. The extended strain set that was used to determine the effect of Lactobacillus supernatant on GBS growth is detailed in Table 1. Lactobacillus strains were selected to represent species that have been found in both the vaginal tract and extraplacental membranes via metagenomic techniques.20 Since previous work on Lactobacillus has focused on certain species including, L. reuteri (Lr), L. gasseri (Lg) and L. crispatus (Lc), which commonly inhabit the genitourinary tract, strains representing these species were selected. Lr6475 was isolated from breast milk (MM4-1A PTA-6475), Lg33323 from the vagina (DSM 20243 [63 AM]), and Lc19390 from stool. Lr17938 is the daughter strain of L. reuteri ATCC 55730, which was isolated from breast milk but also carried potentially transferable resistance traits for tetracycline and lincomycin.25 Lr17938 no longer carries this plasmid. All strains were cultured in deMan, Rososa and Sharpe (MRS, Difco 288130) broth or agar at 37°C with 5% CO2. Plating of a co-culture of Lactobacillus and GBS was performed on sheep's blood agar plates (Tryptic Soy broth supplemented with 5% sheep’s blood, Northeast Lab Services) at 37°C and 5% CO2. Lactobacillus colonies were given 48 hours to grow as colonies were small. 65 Cell Culture Telomerase-immortalized human endometrial stromal cells (T-HESC; ATCC CRL- 4003)26 were cultured in DMEM/Nutrient Mixture Ham's F-12 with L-glutamine (Sigma) supplemented with 1% BD ITS+ universal culture supplement premix, 1.5 g/liter sodium bicarbonate, 2% penicillin / streptomycin, and 10% charcoal-treated FBS (HyClone), which is referred to as HESC medium herein. For all cell experiments, the HESC line was decidualized (dT-HESCs) as previously described27 by incubating with 0.5 mM 8-bromo-cyclic amp (cAMP) (Sigma) for three to six days. The resulting cells are referred to as decidual cells in this work. Assays were only performed when cells reached a 100% confluent monolayer so that only decidual cell surfaces were available for bacterial interactions. Isolation of Lactobacillus supernatants Lactobacillus was loop-inoculated into 10mL of MRS broth with or without 5mM glycerol in a 15mL conical tube and incubated at 37°C for 18-20 hours with the cap slightly loosened. After incubation, cultures were vortexed, centrifuged to pellet the bacteria, and filter sterilized with a 0.22 micron filter. Supernatants were used immediately. Bacterial growth curves To examine the effect of live Lactobacillus and its supernatants, growth curves were generated by serial plating or using a plate reader (Beckman Coulter, Inc). To examine the effect of co-culture on GBS growth, overnight cultures of GBS and Lactobacillus were washed once with PBS and diluted to an equivalent optical density (OD)600 of 0.1 in HESC infection media. Each culture was inoculated 1:10 to have a starting culture with a 1:1 ratio of each bacteria. Samples were collected hourly for six hours and differentially plated on tryptic soy agar (TSA) 66 supplemented with 5% sheep blood. To examine the effect of supernatants, 100µL of 0.1OD 600 culture in HESC infection media was added to a 96-well plate with 25µL of supernatant or additional infection media. Time points were collected by plate reader every 15 minutes for eight hours (BioTek Cytation 3 Imager). The Area Under the Curve (AUC) was calculated using GraphPad Prism 6. Significant differences were determined by comparing the AUC of three biological replicates by unpaired ANOVA. Biofilm assays Biofilm production was assessed as previously described in Chapter 2.28 Briefly, overnight cultures of GBS and Lactobacillus were diluted to an equivalent OD600 of 0.1 and resuspended in tryptic soy broth supplemented with 1% Dextrose (TSAd); 50µl of each culture was added to a 96- well plate for co-culture biofilms. Mono-culture wells contained 50µl of culture and 50µl of media. Lactobacillus supernatant was added to GBS in mono-culture at the beginning of the incubation period at 10% v/v of the total volume of the well. Plates were incubated for 48 hours at 37°C and 5% CO2. After incubation, wells were washed twice and 100µl of crystal violet was added. After a ten-minute incubation, crystal violet was removed, and the wells were washed four times with 150µl of PBS. Remaining crystal violet was solubilized with 100% ethanol; 50µl was taken from each well, and absorbance at OD595 was determined using a plate reader (Beckman Coulter, Inc). The total absorbance was calculated by subtracting the average of the media controls and multiplying by four. Significance was determined by unpaired ANOVA of at least three biological replicates of technical quadruplicates. 67 Association with and invasion of decidual cells Monolayers of decidual cells were washed thrice with phosphate buffered saline (PBS) before infection with bacterial cultures. Overnight bacterial cultures of GBS and Lactobacillus were washed once with PBS and re-suspended in HESC infection media. dT-HESC cells were infected at a multiplicity of infection (MOI) of ten bacterial (GBS) cell per host cell (MOI = 10) for both mono-culture and co-culture wells to assure the same number of GBS cells were available to affect decidual cells. Co-culture wells had an additional MOI = 10 of Lactobacillus. Experiments with Lactobacillus supernatant 10% v/v of supernatant was added to each well at the beginning of the two-hour incubation. An equivalent volume of HESC infection media was added to each control well to control for total volume between wells. Following a two-hour incubation at 37ᵒC in atmospheric conditions, samples were taken from each well to quantify the final colony forming units (CFU) of GBS. To calculate the number of associated cells (attached and invaded), wells were washed three times to remove unattached cells and host cells were disrupted using Triton-X as described previously.28 Wells were scraped and thoroughly re-suspended before plating for CFU. To enumerate intracellular bacteria, extracellular bacteria were killed with 100 μg/ml of gentamicin (Gibco) and 5 μg/ml of penicillin G (Sigma) for one hour prior to continuing the Triton-X treatment and enumeration steps described above. The percent associated cells were determined by dividing the associated cells by the final CFU of each well. The invasion frequencies were calculated by dividing each well by the average of the three technical replicates of the final CFU. All presented data represent the average of three biological replicates of three technical replicates. 68 Cytotoxicity assays Monolayers of dT-HESCs were cultured in 24-well plates. Cells were infected as described above and/or treated with 10% v/v of Lactobacillus supernatant. After incubation, cells were washed twice with PBS and treated with 4µM ethidium homodimer 1 (Molecular Probes, Eugene, OR) suspended in PBS as previously described.30 Plates were incubated at room temperature for 30 minutes without light, and fluorescence was measured at 528-nm excitation and 617-nm emission using a plate reader. The total number of cells in each well was calculated by adding 0.1% (wt/v) Saponin (Sigma) and incubating for at least 20 minutes before repeating the fluorescence reading. The percent permeability (cell death) was calculated by dividing the initial reading by the second and multiplying by 100. Significance was determined by unpaired ANOVA of at least three biological replicates of technical triplicates. 69 RESULTS Lactobacillus does not impede GBS growth in T-HESC infection media To evaluate growth of each strain in the media used for each experiment, growth curves were performed with bacteria in mono- and in co-culture. None of the Lactobacillus strains grew in the T-HESC infection media (shown previously in Figure 2.1) yet both GB112 and GB411 grew in the media (Figure 3.1A). To determine if there was a difference in growth, Area Under the Curve (AUC) was calculated. Co-culturing GBS with Lactobacillus did not affect the growth of GBS (p > 0.05, Figure 3.1B). Colonies of Lactobacillus were not detectable in co-culture after the first time point due to the higher concentration of GBS; therefore, we were unable to calculate its growth in co-culture. Lactobacillus does not affect GBS biofilm formation As biofilms are thought to be important for colonization and persistence,31,32 we sought to determine if co-culturing GBS with live Lactobacillus would affect the ability of GBS to form biofilms. Biofilm production at 48 hours was compared between each GBS strain in mono- culture and to each GBS strain in co-culture with all four Lactobacillus strains (Figure 3.2). The colonizing GB112 strain formed a weak biofilm of 0.86 in mono-culture (Figure 3.2A). Although co-culture of GB112 with Lactobacillus increased biofilm production to absorbance values of 1.36, 1.37, 1.49, and 0.97 for L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390, respectively, the difference was not statistically significant (p > 0.05). The invasive GB411 strain formed a stronger biofilm than GB112 with an absorbance value of 1.22 (p > 0.05) but there were not statistical differences between GB112 and GB411 with a given Lactobacillus strain (Figure 3.2B). Co-culturing with L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. 70 crispatus 19390 increased biofilm production to 1.44, 1.50, 1.80 and 1.25, respectively, but this change was also not significant (p > 0.05). Lactobacillus variably affects GBS association with dT-HESCs As we have demonstrated previously, dT-HESCs serve as a model of GBS attachment to and invasion of the outermost layer of the placental membrane.28–30 We have also previously established that Lactobacillus is capable of associating with this cell line (Figure 2.3). Hence, the goal of this experiment was to determine whether live Lactobacillus affects the ability of GBS to interact with this important barrier. The colonizing (Figure 3.3A) and invasive (Figure 3.3B) strains of GBS did not significantly differ in the ability to associate with dT-HESCs, associating at 0.025% and 0.021%, respectively (p > 0.05). Both GBS strains also demonstrated similar responses to the addition of Lactobacillus, with no observed statistical differences between each co-culture condition. Notably, co-culture with L. gasseri 33323 significantly increased association of the colonizing GBS strain to 6.4% (Figure 3.3A; p < 0.005) and of the invasive strain to 4.8% (Figure 3.3B; p < 0.05). Association to dT-HESCs by L. gasseri 33323 plus both GBS strains was significantly greater when compared to all other Lactobacillus/GBS combinations. Additionally, no statistical differences were observed between each GBS strain in mono-culture compared to co-culture with Lr 6475, Lr 17938, or Lc 19390 (Figure 3.3A&B, p > 0.05). GBS invasion of the dT-HESCs was also examined. Although there was no significant difference in percent invasion between the colonizing (0.003%) and invasive (0.00076%) strains of GBS (p > 0.05), a trend of greater invasion in the colonizing strain relative to the invasive strain was observed (Figure 3.3C&D). The addition of each of the Lactobacillus strains did not 71 affect the ability of either GBS strain to invade the dT-HESCs (p > 0.05). It is interesting to note, however, that if calculating the percent of associated cells that invaded, there was a reduction for all cells when co-cultured with Lactobacillus. An average of 11.05% of the colonizing GBS cells invaded, while 5.75%, 0.10%, 0.94% and 1.04% of GBS invaded while in co-culture with Lr 6475, Lg 33323, Lr 17938 and Lc 19390, respectively. A similar reduction was observed with the invasive strain of GBS, with 3.68% of GBS invading in mono-culture compared to 1.00%, 0.03%, 0.10% and 0.13% when co-cultured with each Lactobacillus strain, respectively. Because these experiments were conducted separately, we cannot determine if these differences are significant as they cannot be paired in a way to allow for biological replicates. Co-culture with Lactobacillus variably affects host cell death GBS is known to lyse host cells using the β-hemolysin, CylE, and this hemolysis could influence the ability of GBS to cross barriers such as the placental membranes.5 To determine if Lactobacillus is capable of preventing from GBS-mediated cell lysis, we performed host cell permeability assays. We have previously established that Lactobacillus alone does not induce dT- HESC death (Figure 2.5). Both the colonizing and invading GBS strains significantly damaged the host cells, causing 71.16% (p < 0.00005) and 70.78% (p < 0.00005) cell death in the four-hour period compared to the mock infection (24.85%) (Figure 3.4). No significant difference in cell damage was observed between the colonizing and invasive strains (p > 0.05). Certain strains of Lactobacillus significantly reduced cell death when co-cultured with GBS. These effects, however, were dependent on the GBS strain. For the colonizing GBS strain, L. reuteri 17938 and L. crispatus 19390, reduced host cell death from 71.16% to 58.21% (p < 0.05) and 56.59% (p < 0.005), 72 respectively (Figure 3.4A). Conversely, in the invasive GBS strain, only L. reuteri 6475 significantly reduced host cell death from 70.78% to 53.25% (p < 0.05; Figure 3.4B) Lactobacillus supernatants inhibit GBS growth Because Lactobacillus is unable to grow in the infection media, we hypothesized that it may not be producing the secondary metabolites that could inhibit GBS. To assess secreted metabolites or other inhibitory compounds, we grew Lactobacillus overnight and collected the supernatant. Supernatants were added at 10% v/v of supernatant to GBS, and differences in growth were assessed by calculating the Area Under the Curve (AUC). Supernatants from Lr 6475, L. gasseri 33323 and Lr 17938 reduced the AUC of the colonizing strain from 1.44 to 0.46 (p < 0.00005), 0.25 (p < 0.00005) and 0.53 (p < 0.00005), respectively (Figure 3.5A). Similarly, the same strains reduced the AUC of the invasive strain from 1.21 to 0.39 (p < 0.00005), 0.37 (p < 0.00005), and 0.68 (p < 0.0005), respectively (Figure 3.5B). The supernatant of L. crispatus 19390 did not affect GBS growth (p > 0.05; 1.44 to 1.39 & 1.21 to 1.50) and no statistical difference was observed between the colonizing and invasive GBS strains. One characterized secreted compounds is As reuterin is created by altering glycerol,33 we also assessed the growth effects of supernatants created by Lactobacillus grown with glycerol. The addition of glycerol, however, did not significantly reduce the AUC for either GBS strain upon co-culture with L. reuteri 6475, L. gasseri 33323, or L. crispatis 19390 (p > 0.05; Figure 3.5). Conversely, addition of glycerol did increase the inhibitory effect of Lr 17938 (p < 0.05). For the colonizing strain of GBS, the addition of glycerol further reduced the AUC to 0.15 (p < 0.05), while it further reduced the AUC of the invasive strain to 0.29 (p < 0.005). 73 Lactobacillus supernatants prevent GBS biofilm formation Since we found that supernatants affected growth of GBS, we hypothesized they may also impact GBS biofilm formation. Given that only one of the two L. reuteri strains was affected by glycerol addition, we only examined the effect of supernatant without glycerol, which was added (10% v/v) at the beginning of the 48-hour incubation period. Statistical differences in GBS biofilm formation were observed for all the supernatants compared to the control except for the invasive strain with L. reuteri 6475 supernatant (Figure 3.6). Supernatants of L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390 decreased biofilm formation in the colonizing strain from 1.72 to 0.16 (p < 0.00005), 0.31 (p < 0.00005), 0.21 (p < 0.00005) and 0.70 (p < 0.0005), respectively. The same supernatants also decreased biofilm production in the invasive strain from 1.26 to 0.72 (p > 0.05), 0.29 (p < 0.005), 0. 15 (p < 0.0005) and 0.43 (p < 0.005). No significant differences were found between how a given supernatant affected the colonizing versus invasive strains (p > 0.05). Lactobacillus supernatants variably affect association of GBS, but do not affect invasion To further assess the effects of Lactobacillus supernatants, we characterized how the addition of this supernatant would affect association and invasion of dT-HESCs. As this experimental design accounts for growth within the time span of the experiment, differences would be independent of the growth changes observed in Figure 3.7. Because Lactobacillus is a lactic acid producing bacteria, we were concerned that the supernatant may damage the host cells, which are sensitive to pH changes. We confirmed there was no significant effect of supernatant on host cell permeability for all supernatants, except that of L. gasseri (p < 0.05; Figure 3.7). L. gasseri supernatant significantly increased the host cell permeability from 27.27% 74 to 59.84% in three hours (p < 0.05; Figure 3.7); therefore, the following results using this supernatant in combination with dT-HESCs must be interpreted carefully as results may be due to supernatant-induced host cell permeability. Generally, the addition of all Lactobacillus supernatants, with the exception of L. reuteri, reduced association of GBS with dT-HESCs. 10% v/v of supernatant was added at the beginning of the incubation period and percent association was examined as a percentage of each GBS strain alone to account for variation between trials. When Lactobacillus was grown without glycerol, the supernatant of L. reuteri 6475 increased association of the colonizing strain to 169.90% of the strain alone. Conversely, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390 reduced the association of the colonizing strain to 35.85% (p > 0.05), 29.74% (p < 0.05) and 4.41% (p < 0.005), respectively (Figure 3.8A). The addition of glycerol during Lactobacillus growth also resulted in reduced association for the colonizing strain, but this decrease was not statistically significant. Indeed, supernatants with glycerol from L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390 reduced association to 41.85% (p > 0.05), 14.16% (p < 0.005), 72.27% (p > 0.05) and 50.00% (p > 0.05) of the invasive strain without supernatant added. Association of the invasive strain was also reduced by Lactobacillus supernatants (Figure 3.8B). Supernatants without glycerol from L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390 reduced association of the colonizing strain to 44.50% (p < 0.00005), 21.36% (p < 0.00005), 38.17% (p < 0.00005) and 30.49% (p < 0.00005). Similarly, supernatants made with glycerol also significantly reduced association. Association of the invasive strain was reduced to 21.95% (p < 0.00005), 34.38% (p < 0.00005), 27.36% (p < 0.00005) and 23.99% (p < 75 0.00005) by L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390, respectively. There were no significant differences between how the colonizing and invasive strains interacted with a given supernatant, with the exception of L. reuteri grown without glycerol, which increased association of the colonizing strain but reduced association of the invasive strain (p < 0.00005). Invasion of the placental membrane is an important step in in utero infection for GBS; therefore, we also sought to assess the effect of these supernatants on intracellular invasion. As with the association experiments, supernatant was added at 10% v/v, and the percent invasion was normalized to the invasion of the GBS strain alone to account for variation between replicates. We found that even controlling for variation between the GBS strain alone between biological replicates did not account for all the variation observed in these experiments (Figure 3.8C&D). The effect of supernatants from L. crispatus 19390 were particularly variable (Figure 3.8C). Given this high variability, we did not observe any statistical differences with the addition of Lactobacilli supernatants (p > 0.05, Figure 3.8 C&D). Lactobacillus supernatants increase host cell death in the invading GBS strain As maintenance of the placental membranes is important for a successful pregnancy, we examined if the supernatants altered the amount of host cell damage GBS does. To this end, we examined host cell death in the presence of both GBS and Lactobacillus supernatants five hours post-infection. Indeed, we observed an increase in host cell death that was dependent on the GBS strain. The colonizing strain of GBS caused 60.5% cell death alone, but, with the addition of supernatants from L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390, cell death increased to 74.8%, 77.0%, 75.1% and 81.0%. These increases were not statistically 76 significant (p > 0.05). The invasive strain of GBS also increased host cell death in the presence of Lactobacillus supernatant. Percent host cell death increased from 53.6% alone to 77.8% (p < 0.05), 79.9% (p < 0.005), 83.6% (p < 0.005) and 77.1% (p < 0.05) with L. reuteri 6475, L. gasseri 33323, L.reuteri 17938, and L. crispatus 19390, respectively. No statistical differences were observed between host cell death with the colonizing strain versus the invasive strain with the same supernatant. Supernatant from Lactobacillus reuteri 6475 broadly inhibits GBS strains To assess if growth inhibition of GBS was specific to the two strains used in the experiments above, an extended list of strains was used that spanned ST, capsule and clinical type (Table 3.1). We found that GBS growth was broadly inhibited, averaging 46.6% of each GBS strain alone and ranging from 39.2% to 54.7% (Figure 3.10). We did not observe any statistical differences between the strains’ growth inhibition (unpaired ANOVA). Further, we observed no differences based on sequence, capsule or clinical type. 77 DISCUSSION Efforts to reduce the burden of GBS through antibiotic treatment have not been fully successful.3 While reducing early-onset disease, there has been no effect on late-onset disease or adverse pregnancy outcomes. Further, the use of antibiotics is not without their own negative effects including effects on maternal and neonatal microbiomes, increased risk of E. coli- associated disease and antibiotic resistance.10 For this reason, alternative therapies, like probiotics, have been suggested. Herein, we sought to characterize the effects of Lactobacillus and its secreted factors on GBS and key components of its interactions with dT-HESCs. To determine if GBS-Lactobacillus interactions could play a role in differences in disease outcome between different strains of GBS, we also compared interactions following exposure to two strains of GBS from different sources, a neonate with infection and a colonized pregnant woman. Surprisingly, we found that live Lactobacillus does not affect the growth or biofilm production of either GBS strain (Figure 3.1&2). This result is likely due to the inability of Lactobacillus to grow in both the infection media and the biofilm media (Figure 2.1). It would be interesting to determine the relative composition the Lactobacillus and GBS within the biofilm as previous work has found that Lactobacillus can inhibit biofilm formation in other species, including Pseudomonas fluorescens and Bacillus cereus in biofilms.34 Further, Lactobacillus did not greatly alter GBS interactions with dT-HESCs, which serve as a model of the outermost layer of the placental membranes. While being able to attach (Figure 2.3), live Lactobacillus cannot inhibit the association of GBS (Figure 3.3). The MOI of this experiment may be too low to be able to assess whether these two organisms would have to compete for binding sites on the cells. While most strains had no effect, L. gasseri significantly increased association of both the colonizing and invasive GBS, suggesting that the L. gasseri and GBS may interact in some way 78 to increase the attachment. It has been previously observed that this strain of Lactobacillus can aggregate with other pathogens, which may increase the colonization of both L. gasseri and GBS if the aggregate is secured to the host cells.35 This increased association may not be negative as it did not also lead to an increase in invasion (Figure 3.3) or host cell death (Figure 3.5). Because Lactobacillus is not capable of growing in the medias used for the experiments, we were not assessing the impact of any secreted factors; therefore, we isolated the supernatants of Lactobacillus grown over-night and challenged GBS with it. We first found effects on growth and biofilm production. Biofilms are a congregation of cells surrounded by a polysaccharide matrix.36 We hypothesize that the reduction in biofilm production may be due to the limited growth of GBS. Previous work has demonstrated the importance growth conditions for GBS biofilm production.31 It is interesting to note that though supernatant from L. crispatus did not significantly inhibit growth, it did significantly inhibit biofilm formation (p < 0.005). While this growth deficiency may explain the effects on biofilm production, it does not explain the effects on association. Because the experimental design accounts for growth, the reduction observed is due to another factor. Further characterization of the supernatants and the effects of association to determine how the association in being lowered. We also observed high variation in the impact of the supernatant on GBS invasion. While the host cell death assay time points are three hours later than those for the invasion assay, these results could suggest that the host cells are becoming more permeable within the time points of the invasion assay. If this cell death is increased or decreased during the first two hours of the experiment, it may explain the high variability between biological replicates, as the assay requires the host cell to be non-permeable to protect the invaded GBS from antibiotic treatment. 79 It is also interesting that host cell death is increased with the combination of GBS and supernatant (Figure 3.9) as the supernatant itself did not affect host cell death, with the exception of L. gasseri (Figure 3.7). This suggests that the GBS alters its interactions in some way in response to the supernatant, which was more pronounced in the invasive strain of GBS compared to the colonizing strain (Figure 3.9). Hence, these data suggest that there may also be differences in how these two strains respond. Previous research has found that stressors, including antibiotics and reactive oxygen species, alter GBS interaction with host cells including increasing macrophage uptake.37 Further experiments will be needed to access how GBS alters its transcription or protein profile to better understand these changes. This increase in virulence in response to Lactobacillus supernatants is of particular importance because the placental membranes are key to maintaining a health pregnancy. Lactobacillus is known to secrete a large number of factors that could be affecting GBS. We hypothesized that reuterin may be one. Reuterin is created by altering glycerol;33 therefore, we hypothesized that glycerol addition would result in increased GBS inhibition. However, we only observed increased inhibition by L. reuteri 17938, suggesting that it is the only strain capable of producing it. Since we observed effects on growth and association, it would be interesting to determine what components of the supernatant are responsible for each phenotype. Further work is needed to examine the different components of the supernatants. The supernatants could be fractionated and examined individually for the effects observed on association and growth. This fractionation could also allow for separation of positive effects of the supernatant, such as inhibition of GBS growth (Figure 3.5), from negative effects including increased GBS-induced host cell death (Figure 3.9). Identification of the growth inhibitory 80 compound would be of particular interest due to its broad impact on GBS growth across strains (Figure 3.10). Collectively, these data suggest that different strains of Lactobacillus have variable effects on GBS strains and that these effects may be further complicated by the GBS strain. While current literature examines the effect of Lactobacillus on vaginal-rectal colonization, we sought to determine if this organism could also impact GBS phenotypes that are known to be important for GBS ascending infections, including growth, biofilm production, association with and invasion of decidual cells and host cell death. We found that while live Lactobacillus has minimal impact on GBS, its supernatant could impact growth, biofilm production, association and induced host cell death. Some of these effects, namely increased association and induced host cell death, are concerning when considering the importance of the placental membranes in maintaining a healthy pregnancy, demonstrating the importance of thorough examination of the effects of any potential alternative therapy. Further examination of the complex supernatants may reveal an individual compound that could serve such a purpose. 81 ACKNOWLEDGEMENTS We would like to thank Dr. Cindy Arvidson and Dr. Poorna Visvanathan for donation of Lactobacillus strains. We would also like to thank Dr. Neal Hammer and Dr Arvidson for their careful editing of the chapter. 82 APPENDIX 83 Table 3.1. Extended GBS strain list. ST: Sequence Type; CPS: capsule type, VRC: vaginal- rectal swab, EOD: early onset disease, LOD: late onset disease GBS Strain Information Colonization Site Clinical Type Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive Colonizing Colonizing Invasive Colonizing Colonizing Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive ST ST-1 ST-1 ST-23 ST-19 ST-1 ST-17 ST-19 ST-19 ST-1 ST-17 ST-26 ST-23 ST-12 ST-12 ST-12 ST-1 ST-12 ST-19 ST-23 ST-17 ST-12 ST-12 ST-17 ST-19 ST-19 ST-19 ST-1 ST-19 ST-12 ST-19 ST-1 ST-12 ST-19 ST-1 ST-12 ST-23 CPS cpsV cpsV cpsIa cpsIII cpsV cpsIII cpsIII cpsIII cpsVIII cpsIII cpsV cpsII cpsIb cpsII cpsIa cpsV cpsIb cpsIII cpsIa cpsIII cpsIb cpsIb cpsIII cpsV cpsIII cpsIII cpsIa cpsIb cpsII cpsIII cpsV cpsII cpsIII cpsV cpsII cpsIII VRC VRC EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis Unknown VRC VRC VRC VRC VRC VRC EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis LOD/sepsis VRC VRC VRC VRC VRC VRC VRC VRC VRC Stillbirth Stillbirth Stillbirth Stillbirth EOD/sepsis EOD/sepsis EOD/sepsis/meningitis Strain GB00012 GB00020 GB00033 GB00036 GB00037 GB00045 GB00066 GB00079 GB00084 GB00097 GB00121 GB00279 GB00285 GB00291 GB00305 GB00310 GB00374 GB00377 GB00390 GB00418 GB00438 GB00555 GB00557 GB00561 GB00571 GB00590 GB00620 GB00651 GB00653 GB00663 GB00686 GB00910 GB01007 GB01454 GB01455 NEM316 84 Figure 3.1. Lactobacillus does not impact GBS growth. Individual GBS and Lactobacillus strains were added at a 1:1 ratio in co-culture. Growth was observed by colony-forming units for six hours. The Area Under the Curve (AUC) was calculated with PRISM 6. Statistical analysis was performed using an unpaired ANOVA to evaluate the AUC. Error bars represent standard deviation between three biological trials. 85 Figure 3.2. Lactobacillus does not affect GBS biofilm formation. GBS and Lactobacillus were added in a 1:1 ratio to a 96-well plate. After 48 hours, wells were washed to remove non- adherent cells and stained with crystal violet. Normalized absorbance was calculated by taking OD595, subtracting the media control and multiplying by four. Error bars represent standard deviation between biological trials. No statistical differences were found by unpaired ANOVA. 86 Figure 3.2 (cont’d) 87 Figure 3.3. Lactobacillus variably affects GBS association with dT-HESCs. dT-HESCs were infected with GBS at a MOI of 10 for two hours with or without an equivalent amount of Lactobacillus. The percent of associated and invaded bacteria were calculated relative to the total number of bacteria in the well. Experiments were completed in biological quadruplets of technical triplicates. Error bars represent standard deviation between biological trials. Significance was determined by an unpaired ANOVA. p < 0.005 ** 88 Figure 3.4. Lactobacillus variably affects host cell death. dT-HESCs were infected with GBS at a MOI of 10 with or without the equivalent amount of Lactobacillus and incubated for four hours. Cell permeability was detected using an ethidium homodimer, and percent permeability was calculated by lysing the remaining cell in each well. Graphed data represents three biological replicates, and the error bars represent the standard deviation of the data. Significance was determined using an unpaired ANOVA between mono-culture and co-culture. p < 0.05 *; p < 0.005 **; p < 0.00005 ####All other conditions were significantly higher than the media control. 89 Figure 3.4 (cont’d) 90 Figure 3.5. Lactobacillus supernatants inhibit GBS growth. GBS growth was monitored for eight hours by plate reader (OD595) with or without Lactobacillus supernatant (25% v/v). The Area Under the Curve (AUC) was calculate with PRISM 6. Error bars represent standard deviation between three biological trials. A Student’s T-test was used to evaluate differences between the addition of glycerol. p < 0.05 # An unpaired ANOVA was used to assess differences between each supernatant condition and GBS alone. p < 0.00005 **** 91 Figure 3.5 (cont’d) 92 Figure 3.6. Lactobacillus supernatants prevent GBS biofilm formation. GBS wase added to a 96-well plate with or without 10% v/v Lactobacillus supernatant without glycerol added. After 48 hours, wells were washed to remove non-adherent cells and stained with crystal violet. Normalized absorbance was calculated by taking OD595, subtracting the media control and multiplying by four. Error bars represent standard deviation between biological trials. Statistical differences between GBS alone and with supernatant added were calculated in PRISM 6 with unpaired ANOVA. p < 0.005 ** p < 0.0005 *** p < 0.00005 **** 93 Figure 3.7. Lactobacillus supernatants variably affect host cell permeability. dT-HESCs were incubated with Lactobacillus supernatants for four hours. Cell permeability was detected using an ethidium homodimer assay, and percent permeability was calculated by lysing the remaining cell in each well. Graphed data represents three biological replicates, and the error bars represent the standard deviation of the data. Significance was determined using an unpaired ANOVA to compare each condition to the media alone. p < 0.005 ** 94 Figure 3.8. dT-HESCs were infected with GBS at a MOI of 10 for two hours with or without 10% v/v Lactobacillus supernatant. The percent of associated and invaded bacteria were calculated relative to the total number of bacteria in the well. Experiments were completed in biological quadruplets of technical triplicates. Error bars represent standard deviation between biological trials. Significance between GBS alone and with supernatant was determined by an unpaired ANOVA. p < 0.05 * 95 Figure 3.9. Lactobacillus supernatants increase GBS-induced host cell death. dT-HESCs were infected with GBS at a MOI of 10 with or without 10% v/v Lactobacillus supernatant and incubated for four hours. Cell permeability was detected using an ethidium homodimer, and percent permeability was calculated by lysing the remaining cell in each well. Graphed data represents three biological replicates, and the error bars represent the standard deviation of the data. Significance was determined using an unpaired ANOVA between GBS alone and with supernatant. p < 0.05 *; p < 0.005 ** 96 Figure 3.10. Supernatant from Lr6475 broadly inhibits GBS growth. GBS growth was monitored for eight hours by plate reader (OD595) with or without Lr6475 supernatant (25% v/v). The Area Under the Curve (AUC) was calculate with PRISM 6. Data is presented as the percent of growth with the supernatant compared to without. Error bars represent standard deviation between three biological trials. No statistical difference was found between strains by unpaired ANOVA. n o i t i b i h n I h t w o r G t n e c r e P 1 0 0 8 0 6 0 4 0 2 0 0 G B01455 G B01454 G B01007 G B00910 G B00686 G B00663 G B00653 G B00651 G B00620 G B00590 G B00571 G B00561 G B00557 G B00555 G B00438 G B00418 G B00390 G B00377 G B00374 G B00310 G B00305 G B00291 G B00285 G B00279 G B00121 G B00097 G B00084 G B00079 G B00066 G B00045 G B00037 G B00036 G B00033 G B00020 G B00012 N E M 316 97 REFERENCES 98 REFERENCES 1. Nan, C. et al. Maternal group B Streptococcus-related stillbirth: A systematic review. BJOG: An International Journal of Obstetrics and Gynaecology vol. 122 1437–1445 (2015). 2. Doran, K. S. & Nizet, V. Molecular pathogenesis of neonatal group B streptococcal infection: No longer in its infancy. Molecular Microbiology vol. 54 23–31 (2004). 3. Center for Disease Control. Active Bacterial Core Surveillance (ABCs) Report Emerging Infections Program Network Group B Streptococcus, 2014. (2015). 4. Whidbey, C. et al. A streptococcal lipid toxin induces membrane permeabilization and pyroptosis leading to fetal injury. EMBO molecular medicine 7, 488–505 (2015). 5. Whidbey, C. et al. A hemolytic pigment of Group B Streptococcus allows bacterial penetration of human placenta. The Journal of experimental medicine 210, 1265–81 (2013). 6. Vornhagen, J. et al. Bacterial hyaluronidase promotes ascending GBS infection and preterm birth. mBio 7, 1–10 (2016). 7. Yoon, B. H. et al. Clinical significance of intra-amniotic inflammation in patients with preterm labor and intact membranes. in American Journal of Obstetrics and Gynecology vol. 185 1130–1136 (Mosby Inc., 2001). 8. Schrag SJ, Hadler JL, Arnold KE, Martell-Cleary P, Reingold A, S. A. Risk factors for invasive, early-onset Escherichia coli infections in the era of widespread intrapartum antibiotic use. Pediatrics 118, 570–6 (2006). 9. Roesch, L. F. W. et al. Diversity and composition of vaginal microbiota of pregnant women at risk for transmitting Group B Streptococcus treated with intrapartum penicillin. PLoS ONE 12, (2017). 10. Azad, M. B. et al. Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: a prospective cohort study. doi:10.1111/14710528.13624. 11. Stearns, J. C. et al. Intrapartum antibiotics for GBS prophylaxis alter colonization patterns in the early infant gut microbiome of low risk infants. Scientific Reports 7, (2017). 12. Aloisio, I. et al. Influence of intrapartum antibiotic prophylaxis against group B Streptococcus on the early newborn gut composition and evaluation of the anti- Streptococcus activity of Bifidobacterium strains. Applied Microbiology and Biotechnology 98, 6051–6060 (2014). 13. Aloisio, I. et al. Evaluation of the effects of intrapartum antibiotic prophylaxis on newborn 99 intestinal microbiota using a sequencing approach targeted to multi hypervariable 16S rDNA regions. Applied Microbiology and Biotechnology 100, 5537–5546 (2016). 14. Olsen, P., Williamson, M., Traynor, V. & Georgiou, C. The impact of oral probiotics on vaginal Group B Streptococcal colonisation rates in pregnant women: A pilot randomised control study. Women and Birth 31, 31–37 (2018). 15. Ho, M. et al. Oral Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 to reduce Group B Streptococcus colonization in pregnant women: A randomized controlled trial. Taiwanese Journal of Obstetrics and Gynecology 55, 515–518 (2016). 16. Hanson, L., Vandevusse, L., Duster, M., Warrack, S. & Safdar, N. Feasibility of oral prenatal probiotics against maternal group B streptococcus vaginal and rectal colonization. JOGNN - Journal of Obstetric, Gynecologic, and Neonatal Nursing 43, 294–304 (2014). 17. Li, W. et al. Lipopolysaccharide-induced profiles of cytokine, chemokine, and growth factors produced by human decidual cells are altered by lactobacillus rhamnosus gr-1 supernatant. Reproductive Sciences 21, 939–947 (2014). 18. Yeganegi, M. et al. Effect of Lactobacillus rhamnosus GR-1 supernatant and fetal sex on lipopolysaccharide-induced cytokine and prostaglandin-regulating enzymes in human placental trophoblast cells: implications for treatment of bacterial vaginosis and prevention of preterm labor. American Journal of Obstetrics and Gynecology 200, 532.e1- 532.e8 (2009). 19. Yang, S. et al. Probiotic Lactobacillus rhamnosus GR-1 supernatant prevents lipopolysaccharide-induced preterm birth and reduces inflammation in pregnant CD-1 mice. American Journal of Obstetrics and Gynecology 211, 44.e1-44.e12 (2014). 20. Prince, A. L. et al. The placental membrane microbiome is altered among subjects with spontaneous preterm birth with and without chorioamnionitis. American Journal of Obstetrics and Gynecology 214 627e1-627e16 (2016). 21. Jones N. et al. Multilocus Sequence Typing System for Group B Streptococcus. J Clin Microbiol. 41(6): 2530–2536, (2003). 22. Manning S. D. et al. Mulitlocus swquence types associated with neonatal Group B streptococcal sepsis and meningitis in Canada. Journal of Clinical Microbiology 47, 1143- 1148 (2009). 23. Spaetgens, R. et al. Perinatal antibiotic usage and changes in colonization and resistance rates of group B streptococcus and other pathogens. Obstetrics and Gynecology 100, 525– 533 (2002). 24. Davies, H. D., Raj, S., Adair, C., Robinson, J. & McGeer, A. Population-based active surveillance for neonatal group B streptococcal infections in Alberta, Canada: implications for vaccine formulation. The Pediatric infectious disease journal 20, 879–884 (2001). 25. Urbańska, M. & Szajewska, H. The efficacy of Lactobacillus reuteri DSM 17938 in 100 infants and children: a review of the current evidence. European Journal of Pediatrics 173, 1327–1337 (2014). 26. Krikun, G. et al. A novel immortalized human endometrial stromal cell line with normal progestational response. Endocrinology 145, 2291–2296 (2004). 27. Brosens, J. J. et al. Human Endometrial Fibroblasts Immortalized by Simian Virus 40 Large T Antigen Differentiate in Response to a Decidualization Stimulus. Endocrinology. 137(6): 2225-31 (1996). 28. Parker, R. E. et al. Association between genotypic diversity and biofilm production in group B Streptococcus. BMC Microbiology 16, (2016). 29. Korir, M. L. et al. Association and virulence gene expression vary among serotype III group B Streptococcus isolates following exposure to decidual and lung epithelial cells. Infection and Immunity 82, 4587–4595 (2014). 30. Flaherty, R. A. et al. Modulation of death and inflammatory signaling in decidual stromal cells following exposure to Group B Streptococcus. Infection and Immunity (2019) doi:10.1128/iai.00729-19. 31. Costerton. Introduction to biofilm. Int J Antimicrob Agents 11, (1999). 32. Lewis K. Biochemistry (Mosc). Persister cells and the riddle of biofilm survival. 33. Biochemistry. 70(2):267-74, (2005). Spinler J. K. et al. Human-derived probiotic Lactobacillus reuteri demonstrate antimicrobial activities targeting diverse enteric bacterial pathogens. Anaerobe 14(3):166- 71, (2008) 34. Jalilsood, T. et al. Inhibition of pathogenic and spoilage bacteria by a novel biofilm- forming Lactobacillus isolate: A potential host for the expression of heterologous proteins. Microbial Cell Factories 14, (2015). 35. Spurbeck, R. R. & Arvidson, C. G. Inhibition of Neisseria gonorrhoeae epithelial cell interactions by vaginal Lactobacillus species. Infection and Immunity 76, 3124–3130 (2008). 36. Botto, M. et al. Bacterial Biofilms: A Common Cause of Persistent Infections. Annu. Rev. Plant Physiol. Plant Mol. Biol vol. 64 http://science.sciencemag.org/ (1998). 37. Korir, M. L. et al. Differing mechanisms of surviving phagosomal stress among group b streptococcus strains of varying genotypes. Virulence 8, 924–937 (2017). 101 THE EFFECT OF PHAGE COMMNUNITIES ISOLATED FROM HUMAN FECAL SAMPLES ON GROUP B STREPTOCOCCUS CHAPTER 4 102 ABSTRACT Group B Streptococcus colonizes the rectal-vaginal tract of approximately 30% of women but is simultaneously a leading cause of neonatal disease and a major contributor to adverse pregnancy outcomes. Maternal colonization is the greatest risk factor in neonatal infections; however, determinants of this colonization are not fully understood. The bacteriophage component of the microbiome has been shown to preferentially lyse invading pathogens in the gut and is thought to be important in maintaining a healthy bacterial microbiota. Because GBS can be frequently isolated by rectal swab, we hypothesized that phage communities isolated from fecal samples could contain phage capable of inhibiting GBS as GBS is commonly isolated using fecal swabs. Indeed, we found that 6% of phage communities isolated from fecal samples were capable of inhibiting GBS. We further characterized the interactions between complex phage communities and GBS hosts and found that capsule, sequence and clinical types of the strains did not affect lysis patterns. However, the number of spacers in Clustered regularly interspaced palindromic repeats (CRISPR) did correlate to sensitivity to a phage community. As we had observed the ability of these communities to inhibit GBS in vitro, we hypothesized that we would observe a difference in GBS genetic presence in the metagenomic reads of the associated phage community. While this correlation did not exist across all GBS strains tested, strains that were generally sensitive to phage communities were significantly more likely to be inhibited by phage communities with a lower abundance of GBS. Collectively, these data indicate the importance of both the GBS strain and the phage community in this interaction and suggest that the phage component of the intestinal microbiome could impact GBS colonization. To further examine this interaction, we sought to isolate and 103 characterize an individual phage from these communities; however, we were unsuccessful in doing so. 104 INTRODUCTION Group B Streptococcus (Streptococcus agalactiae, GBS) is a leading cause of neonatal infections and deaths in the United States and contributes globally to adverse pregnancy outcomes, including premature birth and stillbirth. While GBS is less recognized for its ability to negatively affect maternal health, studies have found that GBS is responsible for 20% of bacteremia in Ireland1 and 25% of puerperal bacteremia in the United States.2 A recent meta- analysis also estimated that the global incidence of GBS maternal effects is 0.38 per 1,000 pregnancies.3 Combined with the increased risk of preterm birth (Risk Ratio = 1.21), maternal GBS colonization continues to be a major health concern worldwide.4 GBS colonization is assessed in the United States by the combination of a vaginal and rectal swab at 37 weeks gestation. Swabs from both areas are taken because GBS is thought to colonize the vaginal tract in a rectal-vaginal route and rectal colonization has been shown to be a good predictor of vaginal colonization.5 In spite of the known risk of GBS colonization for both the mother and the fetus, the only preventative measure currently employed is Intrapartum Antibiotic Prophylaxis (IAP), which is only effective against early onset neonatal disease. This antibiotic treatment is known to negatively impact both the maternal vaginal microbiome and the neonatal gut microbiome, which are important for neonatal development.6–10 The shortcomings of IAP combined with growing concern of antibiotic over-use have resulted in increased interest in alternative therapies for the prevention of GBS-associated neonatal disease and pregnancy outcomes. As maternal colonization is the most important risk factor,11 therapies should ideally reduce maternal colonization; however, they should be more specific than the antibiotic course so that multiple 105 doses could be used to prevent adverse pregnancy outcomes and neonatal disease while keeping the microbiomes of the mother and neonate intact. Phage therapy is the use of a virulent bacteriophage to preferentially kill an undesirable bacterium and is of particular interest for GBS because bacteriophage are more specific than antibiotics.12 The earliest successful publication of phage therapy was in 1917, when Félix d’Hérelle published on the use of a virus to treat bacilli dysentery.13 While largely overlooked in the United States, phage therapy has been studied and implemented in France, Russia, Poland and Georgia.13 In spite of this history, few well-structured, accessible studies have been conducted. However, a recent insurgence of interest has produced several key studies that support the efficacy of phage therapy as a therapeutic and alternative to antibiotics14–18 Most recently, a large, well-controlled study examined the use of bacteriophage in burn wounds.19 While the trial was canceled because the bacteriophage treatment was not as effective as the standard care of antibiotics, it was able to reduce infections in the burns, suggesting adjustments in the dosing may lead to better results. Though phage therapy has only recently been applied to GBS,20 the use of phage-derived lysins has been researched. Success with such lysins has been demonstrated against GBS in two different mouse models that represent vaginal colonization in humans and bovine mastitis, which is also caused by GBS.21,22 While lysins are more specific than antibiotics, allowing for minimal effect on the microbiota, an added advantage of a phage would be self-propagation. Self-propagation allows for phage amplification only in the presence of host bacteria, resulting in limited amplification.12 Research on lysins combined with the recent study on tilapia validates that phage may be used to specifically attack GBS, but more work is needed to isolate phage and validate them for use as a therapeutic during pregnancy. Phage 106 capable of infecting GBS have not been extensively studied, though first described by Russell et al. in 1969.23 Available research has focused on phage isolated from cow milk samples or induction of lysogenic phage residing in GBS strains.24–26 These phage are typically Siphoviridae, which has a double-stranded DNA genome and a non-contractile tail.27 Additionally, studies that investigated host range found a limited host range that is restricted to Streptococci, for example, Group D Streptococcus or Group B Streptococcus.24,28 Another remaining gap in our understanding of GBS-specific phage is their role in GBS colonization. GBS is estimated to colonize one of every five women worldwide,29 but we do not fully understand why some women are colonized while others are not. Additionally, this colonization is often transient, but the reason for this transience is also unknown. While a generally overlooked component of the human microbiome, phage outnumber human cells and rival bacteria in the human body and are thought to greatly contribute to the stability and diversity of the human microbiota.30,31 To date, no studies have examined the prevalence of GBS-specific phage in the human microbiome; however, a recent study has demonstrated that there is a relevant relationship between phage and GBS in the vaginal tract by examining the Clustered regularly interspaced short palindromic repeats (CRIPSR) regions of GBS isolates. CRISPR is thought to act as an adaptive immune system against invading phage and foreign DNA by storing small complementary segments of phage or invading DNA so that it can be used to guide targeted degradation of the invading DNA.32 Across two studies, Beauruelle et al. have demonstrated that GBS strains are actively incorporating CRISPR spacers during vaginal carriage, supporting that GBS is interacting with bacteriophage in the vaginal tract.33,34 107 As GBS is thought to colonize the vaginal tract from a rectal route,5 we utilized our large collection of phage communities that were previously isolated from fecal samples (unpublished, Nohomovich) to identify GBS-specific phage in these communities. In all, we examined 130 phage communities for their ability to reduce the growth of a GBS strain isolated from a neonate with sepsis. A subset of these communities was then examined using a larger GBS strain subset to examine if any GBS characteristics were determinant of the phage community’s ability to reduce GBS growth. Further, the abundance of GBS-assigned reads in certain fecal metagenomic samples was examined to determine if the infectivity of a phage community is associated with GBS abundance in the gastrointestinal tract. Finally, we attempted to isolate lytic phage from three communities of interest, but were unsuccessful to in isolating an individual phage at high enough titer to characterize. 108 MATERIALS AND METHODS Bacterial strains and growth conditions GBS strains were selected based on multilocus sequence type (ST) designation, capsular serotype (CPS), and source. A total of 38 GBS strains were used and are detailed in Table 1; all strains were characterized in prior studies.35,36 Unless otherwise noted, GBS strains were grown up overnight in Todd Hewitt Broth (THB) with 5% CO2 at 37°C. The area under the curve (AUC) was determined from individual growth curves using PRISM 6. Growth inhibition was calculated as the percent of AUC with the phage community (PC) over the AUC of the GBS strain alone. [(AUCPC / AUCAlone) * 100]. GBS genome sequencing and extraction of GBS CRISPR spacers Genomes were available for 11 GBS strains and were sequenced in prior studies (Table 1) as described previously.37 Sixteen additional strains were sequenced as part of this study, resulting in a total of 27 strains. Briefly, Isolates were grown in THB overnight DNA was isolated using the Wizard® Genomic DNA purification kit and subsequently prepped for sequencing using the Nextera XT kit (Illumina, San Diego, CA, USA) following manufacturer’s instructions. Libraries were sequenced. Spades, 3.10.1 was used to de novo genome assemblies following trimming and quality checking with Trimmomatic and FastQC, respectively. Multiple k-mers (21, 33, 55, 77, 99, 127) were used and k-mers that passed quality control were cross- assembled to generate the assembly used for downstream analyses. Error correction was performed during the assembly process to minimize the number of mismatches present in the assembled contigs. (www.bioinformatics.babraham. ac.uk/projects/fastqc/). Contigs that contained both genes that flank the CRIPSR repeats (csn2 and ndk) were extracted and analyzed in Geneious.38 to identify the CRISPR spacers. Any strain that did not have complete loci 109 containing both genes flanking the CRISPR spacers/repeats was not included in the analysis. Spacers were matched to annotated sequences available at the National Center for Biotechnology Information (NCBI) using the following criteria: <15% difference in alignment length and 85% identity. Isolation of phage communities from human fecal samples Phage communities were isolated from stool samples collected from otherwise healthy individuals. These samples were collected as part of a larger study to examine the intestinal microbiome in patients with enteric infections and was described previously.39 Following collection, all samples were transported in Cary Blair medium, homogenized and centrifuged. The supernatant was collected and filtered with a .22µm filter (Nohomovich, unpublished). The resulting supernatant is referred to as a phage community (PC) herein. Metagenomics of fecal samples and estimation of GBS abundance in the microbiome Fecal samples were previously sequenced using the Illumina HighSeq 2500 Rapid Run platform in a 1x150 or 2x250 paired-end format as described (Nohomovich, unpublished) In brief, the output was demultiplexed and converted to FastQ format using Bcl2Fastq v.1.8.4 (Illumina), and reads were quality checked using Trimmomatic.40 All reads passing quality control were compared to human RefSeq genome, GRCh38_1118, downloaded November 2018 from NCBI with Bowtie 2.41 SAMtools was used to remove the humans reads.42 Reads that could be annotated as Streptococcus agalactiae at the species level using Kaiju43 with default parameters were extracted. The percent GBS in the metagenomic sample was calculated as the number of GBS reads divided by the total number of reads per sample and multiplied by 100 to 110 calculate percent GBS in a sample. Only samples with greater than 80x coverage were used, resulting in 26 samples for analysis. GBS growth inhibition by Phage Communities (PCs) In all, 130 PCs were initially screened for the ability to negatively affect the growth of two ST-17, serotype III GBS strains, GB00411 and/or GB00112. These two strains were recovered from a neonate with sepsis and an asymptomatically colonized mother, respectively.35,36 (Manning 2009, Manning 2008). Early log phase cultures of GBS were mixed with different PCs and added to a 96-well plate. The OD595 was monitored over 8 hours in a plate reader (BioTek Cytation 3 Imager). PCs that caused a decrease in growth of these two strains were of interest and were targeted for downstream experiments aimed at isolating individual phage and for examining PC host range using 36 additional GBS strains for a total of 38 GBS strains. PC sensitivity was determined by calculating the AUC of a GBS strain grown alone and with a PC. GBS strains that had at least a 10% reduction in growth compared to the strain alone [(AUCPC / AUCAlone) * 100 < 90] were considered sensitive to inhibition by the PC. Combinations of PCs and GBS strains were coded into 1’s (inhibition) and 0’s (no inhibition) based on GBS sensitivity to each PC. Uncorrected Chi square tests were used to examine associations between GBS sensitivity to a given PC and the number of spacer sequences present within the GBS genome. GBS strain source was also taken into consideration. We also examined associations between GBS sensitivity and the percent of GBS within 14 of the 26 fecal metagenomes from which each PC originated. 111 Assessment of infectivity by spot plating To determine if a given PC and bacterial host were compatible, a spot plating method was employed. Briefly, soft agar was overlaid on a plate of the same media. After allowing the soft agar to solidify for 10 minutes, and 10µl of phage sample was pipetted on top without piercing the soft agar. The plates were left to solidify for at least 30 minutes and were incubated overnight at 37°C with 5% CO2. This version of spot plating was attempted with a range of soft agar concentrations (.3% to .7%) and media types including Brain Heart Infusion (BHI), Tryptic Soy Agar (TSA), Todd Hewitt Agar (THA), TSA + 5% Sheep blood, Mueller-Hinton Agar and modified THA (mTHA). A liter of mTHA consists of 30 g of THB, 2 g of yeast extract, 12 mg of CaCl2, and 10 mg of l-tryptophan (Domelier, 2009). After troubleshooting, BHI with 0.7% agar was selected for remaining experiments because it resulted in the most confluent bacterial lawns. Plaque assays To isolate individual phage from the PCs showing the greatest level of GBS inhibition, we attempted to perform plaque assays. The same media types and agar concentrations were attempted. Briefly, soft agar was prepared and cooled to 50°C. 300-500µl of log-phase host bacteria were added to 10-50µl of phage community or enriched phage sample that has been previously diluted in phage buffer containing Tris (10 mM), pH 7.5, MgSO4 (10 mM), NaCl (68 mM), and CaCl2 (1 mM). After an incubation period of 10-15 minutes, soft agar (3-5mL) was added to the mixture and poured over agar plates that matched the soft agar; multiple media types were evaluated. Enrichment of GBS-specific phage To enrich a GBS specific phage, PCs were added to actively growing GBS strain GB411 when it reached 0.2 OD595. During the trouble-shooting phase, a range of temperature (25-37°C), 112 shaking speeds (0 – 200), atmospheric or 5% CO2 and lengths of incubation (two to sixteen hours) were tried. Final conditions were 28°C, atmospheric CO2, 50rpm shaking and a four-hour incubation. Afterwards, samples were centrifuged to pellet bacteria, and the resulting supernatant was collected and filter sterilized with a 0.45µm filter. The resulting enriched phage sample was given the original number of the phage community followed by the name of the GBS strain used. For example, if PC 561 was grown with GB112, it was termed PC561GB112. Three different concentration methods were also tested. First, an additional Polyetheleneglycol (PEG) precipitation step was included after filter sterilization for portions of the trouble shooting. Briefly, PEG was added to precipitate the phage, and the sample was incubated overnight at 4°C. The sample was then centrifuged and resuspended in phage buffer. Second, a 100kd centricon was used to remove any particles less than 100kd. Finally, a cesium chloride gradient was used to separate samples based on size as previously described.44 Visible bands were isolated via needle extraction, and 500µl fractions of the remaining gradient were collected. Individual fractions or bands were then tested for infectivity on GB112 or GB411. The success of each of these steps was examined by spot plating, a growth screen in broth or both. Microscopy of enriched phage samples Phage samples were added to continuous carbon TEM grids, which had been plasma cleaned for 20 seconds prior to sample application. They were be negatively stained using 1% uranyl acetate and imaged using a JEOL 2200FS TEM operating at 200 keV. 113 RESULTS Screening phage communities for GBS-specific phage isolation We hypothesized that phage communities that greatly impacted GBS growth would contain GBS-specific phage that could be isolated for further examination. A total of 130 communities were screened using two GBS strains of interest, GB112 and GB411. Approximately 6% (n=8) were considered sensitive to the phage communities (PCs) based on at least a 10% reduction in Area Under the Curve (AUC). The average reduction across these samples was 52%, ranging from an 11 to 86% reduction in AUC. Of the eight PCs of interest, three were selected for further analysis based on substantial growth reduction in the growth curves and the amount of sample available. PC561 reduced the maximum OD600 of GB411 from 0.233 alone to 0.155 with PC561 (Figure 1). PC801 and PC895 reduced the growth of GB411 by approximately 90%. Notably, some communities increased the growth of GBS, suggesting other components of the phage community may affect GBS growth. The infectivity of PC561, 801 and 895 was also confirmed via spot plating. All three communities resulted in large spots, or zones of clearing, in both GB112 and GB411. Host range of select communities To assess if certain strain characteristics of GBS made them more likely to be lysed by the communities, the host range of PC801 and PC895 communities was assessed using 38 GBS strains of various clinical types, CPS types and STs as well as three commensal Streptococcus salvarius strains. PC561 was not examined for inhibition in any of the GBS strains due to low sample volume. Similarly, the number of GBS strains tested with PC801 was reduced (n=17 GBS, 20 total) due to a lower sample volume compared to PC895. The impact of PCs on all GBS 114 strains was examined using both spot plating and growth inhibition in broth. PC801 lysed 15 of the 17 GBS strains (Table 2), while PC895 lysed 36 of the 38 GBS strains tested. It is notable that the two PCs had different lysis profiles. For example, both strains that PC895 could not lyse, GB00121 and GB00310, are CPS type V, while PC801 did not lyse GB00079 and GB910 which are CPS III and II, respectively. Due to the low number of strains that were resistant to phage, we could not determine if strains with specific capsules or sequence types were more or less likely to be lysed. The commensal strains of Streptococcus salvarius (n=3) were less frequently lysed than GBS. PC801 lysed two of the three strains, while PC895 lysed two. We did not assess differences because sample size of the commensal strains is drastically lower. CRISPR spacer regions differ across GBS strains Since we did not observe an effect of capsule or sequence type, we sought to examine CRISPR spacers regions as they have been shown to act as a adaptive immune system in bacteria. The CRISPR regions of 27 of the 38 GBS strains with genome sequencing data available were examined for CRIPSR spacers (Figure 2). In all, 192 unique spacers were observed across the 27 genomes. Most of the spacer sequences were unique, but 39 (20.3%) of the spacers were found in at least two strains. Spacer 13 predominated and was found in 48% (13/27) of the GBS strains. Notably, 65 (33.9%) of these spacers were homologous to different phage and plasmid sequences available in the NCBI database (Table 3). Many of the annotated spacers (57/62; 92%) match to known Streptococcus phage, including JX01 (n = 15), LF2 (n = 14) and LSSO9 (n = 16). Additionally, six spacers matched plasmids. These include: pCBU1 from Clostridium butyricum, pSRC1 from Selenomonas runinantium, pNCT2 from Bacillus megaterium, and pSAL813 from Streptococcus salvarius as well as a cloning vector from 115 Streptococcus mutans and a plasmid from Legionella adelaidensis strain NCTC12735. Interestingly, two of the four strains that could not be lysed by either PC801 and PC895, each strain had spacers that are homologous to Streptococcocus phage LYGO9. We do not have associated metagenomic data for these two PCs; therefore, we cannot examine these communities for the presence or abundance of this LYGO9 phage. GBS strains also varied in the number of CRISPR spacer repeats, ranging from 3 to 24. The number of spacer sequences was previously linked to clonal complexes, which represent closely related strains defined by multilocus sequence typing.45 In our dataset, we also found differences in the number of CRISPR spacers across sequence types (ST), with the lowest average number of spacers in ST-17 strains (7.8 spacers) and ST-1 (12.8 spacers). Other common STs including STs 12, 19 and 23, had similar numbers of spacers with 16.7, 15.1 and 14.5, respectively (Figure 3A). When stratifying strains by clinical type (colonizing versus invasive), colonizing strains averaged 15.4 spacers, while invasive strains averaged 11.5 (Figure 3B); this difference was not statistically significant. We also examined differences between capsule type (Figure 3C). While capsule type III, which is commonly associated with ST-17, had a significantly lower average number of spacers (10.6) than the other capsule types (16.6 (CPS I), 17.5 (CPS II) and 14.0 (CPS V)), these differences were not significant by ANOVA. Because this result could be due to the unbalanced distribution of samples, we grouped CPS types I, II and V together for comparison to CPS type III. Importantly, CPS III strains had significantly less spacers on average (10.6) compared to other capsule types (17.0) (Figure 3D; Mann-Whitney T-test, p = 0.0101). To further examine this association, strains were defined as having a low or high number of spacers; seven strains had fewer than nine spacers (first quartile), while seven strains had greater than 16 116 spacers (fourth quartile). Indeed, six of the 12 CPS III strains had a low number of spacers as calculated by the 25% quartile (≤ 9 spacers) compared to only one of the 14 strains representing the other CPS types (Fishers exact p = 0.047). The number of CRISPR spacers significantly impacts the ability of phage communities to inhibit growth Since CRISPR acts as an adaptive immune system in bacteria, we hypothesized that a reduced number of spacers would result in a higher frequency of growth inhibition by PCs. Fourteen of the 27 GBS strains were selected to test this hypothesis based on number of spacers identified in the genome analysis using the previously defined quartiles. Strains were grown in the presence of fourteen different phage communities and reduction in AUC was calculated as a percent of growth without the PC (Figure 4). We found that 67 of the 196 combinations resulted in growth inhibition (34.2%). We observed that certain PCs were also more likely to affect GBS strains, including ER0788. Likewise, there were certain GBS strains were more likely to be sensitive, including GB000390 and GB000910, which were inhibited by all the PCs. Combinations that resulted in at least a 10% reduction in growth were considered sensitive (0) while those above the cutoff were considered resistant (1) to the PC. GBS strains in the lower quartile that possess the lowest number of spacers, were 2.0 times less likely to be sensitive to inhibition by the PCs (95% CI: 1.09, 3.63). Within this smaller data set, however, there was not a high enough or even enough distribution to account for CPS or ST-type. GBS presence in the microbiome is significantly correlated with GBS inhibition A selection of the fecal samples used in this study were previously used for metagenomics (Nohomovich, unpublished). Using these data, we calculated the percent of GBS 117 within the total bacterial microbiome. The abundance of GBS ranged from 0.9% to 3.1%, with an average of 2.1% in reads with detectable GBS. Samples with undetectable (0%) GBS were removed from the analysis because the metagenomic data did not reach a high enough read redundancy to confidently say that a zero result is in fact negative. Eight of the sequenced samples had ≤1.8% (25% quartile), while another eight had ≥2.4% (75% quartile). These quartiles were selected for further analysis (Figure 4). As previous research determined that the phage community in the gut determines the bacterial component, we hypothesized that presence of GBS in the microbiome may be dictated by the phage community (PC). We found no correlation between the percent of GBS in the microbiome and the sensitivity of GBS strains to the PCs. We further hypothesized that this observation may be due to the resistance of the GBS strains to a given community, therefore, we examined only strains that were sensitive to PCs as defined in the methods. Within this subset of 14 strains, we found that if GBS abundance was high (≥2.4%), then the associated PC was 2.7 times less likely to be able to reduce the growth of GBS (95% CI: 1.00, 6.46). Indeed, the average growth inhibition across the phage communities with lower percent GBS-associated reads was 79.9%, while the average for those with high percent GBS associated reads was 85.1%. Further, the percent of GBS found in the microbiome significantly correlated to the percent growth reduction in sensitive strains (Pearson correlation = 0.88). Recovery of phage from PCs capable of inhibiting GBS A common technique used to isolate individual phage is picking an isolated plaque from a titer plate; however, this technique is known to be difficult for many phages, similar to difficulties with unculturable bacteria.46 The ability to form plaques is reliant on several 118 components, which we systematically attempted to troubleshoot in attempt to isolate individual phage. In spite of troubleshooting, we were unable to observe plaques. Titer assays resulted in either a clear plate or a confluent lawn, even with 1:2 dilutions (Figure 5). 1. Media type and agar concentration: Media can greatly affect the ability of a phage to successfully infect a host.47 As there was little available research on phage in GBS, we started in the most commonly used media for GBS, Todd Hewitt Broth (THB). This is a general-purpose broth used for Streptococcus that contains peptones, dextrose, and salts (VMR). With this broth, we observed minimal growth of the bacterial host. While lawns grew, they were translucent and easy to see through. We observed similar problems with Tryptic Soy Agar (TSA) and Mueller- Hinton Agar. The concentration of agar in the soft agar overlay can affect migration of the phage and affect the growth of the bacteria48 We used a range of 0.2 to 1% agar but did not observe improvement in the lawn or the appearance of plaques. We hypothesized that the poor lawn may be due to GBS being unable to grow in a soft agar overlay; therefore, we tried using the standard technique used to assess antibiotic resistance in GBS, which results in a confluent lawn on the surface of an agar plate. A bacterial culture is spiral streaked on TSA + 5% Sheep blood and allowed to grow. While this did result in confluent growth over the top of the plate, it was difficult to adapt this method to titer as such a small amount of the bacteria and phage mixture was being added to a plate. A recent publication on GBS phage in tilapia used Brain Heart Infusion (BHI) with 0.7% agar to pick isolated phage plaques (Luo, 2018). Use of this media resulted in more confluent lawns; therefore, this media was used for all other assays. The use of this media also led to better spot tests revealing the hazy ring around the spot from PC801 and 895 that is indicative of a lysogenic phage, which was not visible on the other media types (Figure 6). 119 2. Incubation conditions: As the physiology of the host bacteria affects the ability of a phage to form plaques, we tried different temperatures, periods of exponential phase, shaking conditions and CO2 concentrations. GBS is typically grown in 5% CO2 at 37°C; therefore, we began by incubating plates in these conditions; however, we observed more success with atmospheric CO2. Incubation at atmospheric CO2 resulted in more confluent lawns and an increased enrichment; therefore, future experiments were preformed at atmospheric CO2. To successfully infect, phage must have an actively replicating host; however, the period of exponential phase (early OD595 0.2, middle 0.3 or late 0.4) may have difference results. Early exponential phase cultures seemed to result in better lawns and a more concentrated enrichment; therefore, experiments were preformed with cultures at an OD595 of 0.2-0.3. Different temperatures also affect how the host is growing and the ability of the phage to attach to the host.47 We assessed varying temperatures including 28, 30, 32 and 37°C, and found the most success with 28°C as phage recovery was highest at this temperature. Finally, for liquid enrichment, shaking the culture allows for increased contact between the phage and bacterial host, but high amounts of aeration affects growth of GBS and could shear phage, rendering them incapable of infecting. Gentle shaking at 50 rotations per minute (rpm) resulted in the greatest reduction of growth of GBS during enrichments in comparison to each control. 3. Concentration steps: Polyetheleneglycol (PEG) precipitation can be used to concentrate phage by clumping the phage together so that they can be pelleted and concentrated. We did not observe a corresponding increase in infectivity, suggesting this procedure did not help concentrate out phage. Next, we tried using 100 kilodalton (kd) centricons to reduce the volume. This was performed before Cesium Chloride gradients and microscopy. The results from these experiments suggest that the samples were still too dilute. 120 Enriching GBS-specific phage Due to the difficulty getting an isolated plaque from the PCs, we attempted to enrich for the phage of interest multiple times, thereby reducing unwanted components from the fecal sample, such as nutrients or inhibitory compounds, while maintaining the desired phage. The addition of a suitable host allows a bacteriophage to infect and propagate, thereby preferentially increasing concentrations of the phage that are specific to the bacteria of interest. We attempted to enrich PC561, 801 and 895 using GB112 and GB411. As a control for these experiments, we also added the community to an equivalent amount of broth without a bacterial host to confirm that the agent responsible for GBS growth inhibition needed to propagate in order to have an effect on GBS (Figure 6). We found that these controls did not spot, confirming that any inhibition observed with the enrichment samples is due to a phage. Samples from the attempted enrichments had a smaller zone of clearance in spot plates (Figure 6). We confirmed a reduced infectivity of the resulting sample by measuring growth reduction in broth (data not shown). We hypothesized this reduction could be the result of a lysogenic phage. Indeed, the hazy perimeter observed on the spot plates is indicative of lysogenic phage (Figure 6). For PC561, the enriched sample maintained its killing ability of both GBS strains (Figure 7A), but also demonstrated the ability to inhibit a variety of other bacteria including Lactobacillus gasseri and Escherichia coli (Figure 7B&C). Because an individual phage is unlikely to infect Gram-positive and negative bacteria, it is likely that the enrichment step was not sufficient and other methods need to be applied in order to isolate the GBS-specific phage in these samples. 121 Cesium Chloride Gradient Cesium chloride gradients allow for separation of particles based on density. When phage samples are added an observable band can be often observed and removed by needle puncturing the band and withdrawing the band. We hypothesized this would allow for isolation of phages of interest from other contaminants in the PC resulting from the fecal matter. No bands were observed for PC801 and PC895, but two bands were observed for PC561. The bands and individual fractions of the gradient were taken and tested for infectivity. Fractions C through G maintained infectivity, while fractions A, B, H as well as the pellet had no effect on GBS growth (Figure 8, representative PC895); however, these fractions were not capable of forming spots on soft agar. Additional attempts to enrich the phage from the gradients were unsuccessful. 122 DISCUSSION GBS remains a significant health burden in the United States and globally; however, preventive measures are limited to antibiotic treatments with known flaws. As GBS can be isolated from rectal swabs, we hypothesized that fecal phage communities would contain GBS- specific phage. Indeed, we found that many samples inhibited the growth of GBS strains and further examined three communities of interest. We found broad host range for two of these communities, PC801 and PC895, as they inhibited 88% and 94% of the strains tested, respectively (Table 2). While there were no GBS traits associated with the ability of PC801 or PC895 to lyse GBS, a larger data set that includes resistant strains of GBS may reveal more about the host range of these specific communities. This broad host specificity could be due to the presence of multiple phage types or inhibitory compounds present in the phage communities. The phage characterization portion of this study is severely limited by the inability to isolate an individual phage by plaque isolation; however, this is also known to be a problem outside of this work. It was estimated that less than 1% of phage will develop a plaque, or zone of clearing in a bacterial host that allows for recovery of individual viruses.46 Due to this limitation, we are unable to determine if the host range of a given PC is due to a single phage or a combination of phages in the sample. The complex nature of a phage community, as it is isolated directly from fecal samples, further complicates data interpretation since other proteins could be present in the phage communities. Indeed, we found that some communities increased growth, which we hypothesized may be due to extra carbon sources present in the fecal matter. Hence, it was important to confirm that the inhibitory compound is a bacteriophage. As the inclusion of a bacterial host was required for continued lysis during enrichment, it is unlikely that the inhibitory agent is a protein or antibiotic. In spite of this, we do not know how the fecal 123 sample is contributing to bacterial-phage interactions. For example, a component of the complex community may play a role in increasing the virulence of the bacteriophage in PC 801 or 895. To reduce the complexity of the samples, we enriched PCs with two GBS hosts, GB112 and GB411, but still observed a broad host range including other species such as Lactobacillus and E. coli (Figure 8). Since this was an enriched sample, it is unlikely that this broad specificity is due to a challenge that is generally inhibitory to bacteria such as pH as the sample had been significantly diluted. It is more likely that another phage was still present, which is not surprising given that the feces contain a broad range of phage. Indeed, these communities have been used to lyse a variety of enteric pathogens including E. coli strains, and Lactobacillus phage have been previously observed in fecal samples via metagenomics. (Nohomovich, unpublished). Phage characterization is also limited by the low titer of this phage. While we were unable to properly titer these samples by plaque-forming units, we estimate that the titer is low based on the clearance of 10-2 plate. This low titer limited our work by making microscopy and sequencing difficult. Typically, a sample needs to be at least 104 PFU in order to continue with downstream methods, which we were never able to accomplish. Both of these methods would have contributed greatly to our understanding of the bacteriophage residing within these communities. Despite these limitations, troubleshooting for optimal conditions for these GBS phage helped to develop spot plating and growth inhibition assays to examine how traits of both metagenomes of the fecal samples and the GBS host strains affect interactions. This study is unique in its examination of the impact of complex phage communities on GBS isolates. There are still fundamental gaps in our understanding of GBS colonization and to date, there have not been studies conducted to examine the role of the phage community in the gut in GBS carriage. 124 Herein, we examined whether different GBS traits affected the ability of a complex phage community to inhibit the growth of GBS. While CPS, ST and clinical type have been previously shown to affect a number of traits in GBS,33,49–52 but did not affect the ability of the phage community to lyse, we demonstrated that the number of CRISPR spacers present within the genome may be linked to infectivity. Specifically, we found that a lower number of spacers correlated to a lower rate of infectivity suggesting that strains with higher numbers of CRISPR spacers could be more resistant to removal by the phage community. Unfortunately, two of the four strains that did not lyse with PC801 and 895 have yet to be sequenced so we cannot assess this idea well in the current data set. The idea that strains with a lower number of spacers would be more likely to be cleared from the gut is interesting given that strains belonging to ST-17, which had the lowest average number of spacers, are more frequently associated with sepsis and meningitis. It would also be interesting to evaluate the CRISPR regions of strains isolated in longitudinal studies to examine if persistence correlates with the number of CRISPR repeats. Although Beauruelle et al. followed women with GBS longitudinally and examined CRISPR spacers, no details were provided showing that continued carriage was correlated to the number of CRISPR spacers. 33 We also hypothesized that if the phage community was having an in vivo effect on GBS, then a reduced percent of GBS in the microbiome would correlate to a higher infectivity of GBS in vitro. We found that this was not true when all the strains were examined together; however, examination of just the sensitive strains showed such a correlation. This finding suggests that the phage community might impact GBS carriage, but only if the strain itself is less resistant to bacteriophage. As phage sensitivity is also correlated to the number of spacers, it may be 125 important to continue a more comprehensive examination of CRIPSR spacers when considering GBS disease globally. 126 ACKNOWLEDGEMENTS I would like to thank Dr. Kristin Parent for her careful editing of this chapter and for her guidance. I also like thank Dr. Jason Schrad for his assistance with microscopy and for all our phage discussions. 127 APPENDIX 128 Table 4.1. GBS Strain List. ST: sequence type; CPS: capsule type; VRC: vaginal-rectal colonization; EOD: early onset disease; LOD: late onset disease. CPS Isolation Site VRC VRC EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis ST-1 V ST-1 V ST-23 Ia ST-19 III ST-1 V ST-17 III ST-19 III ST-19 III ST-1 VIII VRC ST-17 III VRC ST-17 III VRC Unknown ST-26 V VRC ST-23 II VRC ST-12 Ib VRC ST-12 II VRC ST-12 Ia EOD/sepsis ST-1 V EOD/sepsis ST-12 Ib EOD/sepsis ST-19 III EOD/sepsis/meningitis ST-23 Ia EOD/sepsis ST-17 III EOD/sepsis ST-17 III LOD/sepsis ST-12 Ib ST-12 Ib VRC ST-17 III VRC ST-19 V VRC ST-19 III VRC ST-19 III VRC VRC ST-1 Ia ST-19 Ib VRC ST-12 II VRC ST-19 III VRC ST-1 V ST-12 II ST-19 III ST-1 V ST-12 II Stillbirth EOD/sepsis Stillbirth Stillbirth Stillbirth Streptococcus Strain Information Strain Number Clinical Type ST GB00012 GB00020 GB00033 GB00036 GB00037 GB00045 GB00066 GB00079 GB00084 GB00097 GB00112 GB00121 GB00279 GB00285 GB00291 GB00305 GB00310 GB00374 GB00377 GB00390 GB00411 GB00418 GB00438 GB00555 GB00557 GB00561 GB00571 GB00590 GB00620 GB00651 GB00653 GB00663 GB00686 GB00910 GB01007 GB01454 GB01455 Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive Colonizing Colonizing Colonizing Invasive Colonizing Colonizing Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive Invasive Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive 129 Table 4.1 (cont’d) NEM316 19232 19233 19234 Invasive ST-23 III Commensal Streptococcus salvarius EOD/sepsis Saliva Unknown Oral Cavity 130 Table 4.2. PC 801 and 895 have broad host range within Streptococcus. 10 microliters of PC was added to a lawn of bacteria (0.7% BHI agar overlay). A total of 41 strains were tested for PC895, while 20 were tested for PC801. Grayed boxes represent strain combinations that were not performed. ST: sequence type; CPS: capsule type; VRC: vaginal-rectal colonization; EOD: early onset disease; LOD: late onset disease ST: sequence type; CPS: capsule type; VRC: vaginal-rectal colonization; EOD: early onset disease; LOD: late onset disease. Streptococcus Strain Information Spot Plate Strain Number Clinical Type ST CPS Isolation Site 801 895 GB00012 GB00020 GB00033 GB00036 GB00037 GB00045 GB00066 GB00079 GB00084 GB00097 GB00112 GB00121 GB00279 GB00285 GB00291 GB00305 Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive ST-1 ST-1 ST-23 ST-19 ST-1 ST-17 ST-19 ST-19 V V Ia III V III III III Colonizing ST-1 VIII Colonizing Colonizing Invasive Colonizing Colonizing Colonizing Colonizing ST-17 ST-17 ST-26 ST-23 ST-12 ST-12 ST-12 III III V II Ib II Ia 131 EOD/sepsis Yes Yes Yes Yes Yes VRC VRC EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis EOD/sepsis VRC VRC VRC Unknown VRC VRC VRC VRC Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Table 4.2 (cont’d) GB00310 GB00374 GB00377 GB00390 GB00411 GB00418 GB00438 GB00555 GB00557 GB00561 GB00571 GB00590 GB00620 GB00651 GB00653 GB00663 GB00686 GB00910 GB01007 GB01454 GB01455 NEM316 19232 Invasive Invasive Invasive Invasive Invasive Invasive Invasive Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Colonizing Invasive Invasive Invasive Invasive Invasive Invasive ST-1 ST-12 ST-19 ST-23 ST-17 ST-17 ST-12 ST-12 ST-17 ST-19 ST-19 ST-19 ST-1 ST-19 ST-12 ST-19 ST-1 ST-12 ST-19 ST-1 ST-12 ST-23 V Ib III III III Ib Ib III V III III Ia Ib II III V II III V II III Commensal Streptococcus salvarius Saliva 132 EOD/sepsis Yes No EOD/sepsis EOD/sepsis Ia EOD/sepsis/meningitis EOD/sepsis Yes Yes EOD/sepsis Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes LOD/sepsis VRC VRC VRC VRC VRC VRC VRC VRC VRC Stillbirth EOD/sepsis Stillbirth Stillbirth Stillbirth EOD/sepsis Table 4.2 (cont’d) 19233 19234 Unknown No No Oral Cavity Yes Yes 133 Table 4.3. CRISPR spacers can be annotated using NCBI. Spacer Number Spacer Sequence NCBI Annotation CCGTCAAACAAGAGCGACAGCGAAACAAGC TTTTATTGGTTTTCTAAGTGCTCGACCATC AGTTACTTCTGCTTGGGTTTGATAAGGGTC ACCCTTGTATGTTAAATCCGCAAGATTTTA ACTAAAATCTAGATTTGAATAATAGTATAG AATACCATTTTCCACCCAATCAAATCCAAC ACCTCTTAAAATTTAAGTAAATCCTCTTGA AACATTAGCCTTTTCTAACTCTTCAGCTGT TATGTCTTCTAACAGTTGCTTCTTGTGCTT 1 2 3 4 5 6 7 8 9 10 GAAATGTGGAGTCATTCAGGTTGATGATGG Streptococcus phage LF2 11 AAAAAATAAATGACTTTAAAGCACTTGGAG 12 CTGTAAGCGGTAGACTTGACAAAATTAAAA 13 AAGCTAATTCTCATCTCACCGAGATGGATA 14 GGATGATTTCGATTATGCGGCGGTGGTTGA 15 AAATGTTAATTTCATATCTACATCTTGTTC 16 ACATAACGTTCAAAAGTTTCCACTAATAGC 17 18 19 TCAAACAACATTGGTGATCTTATTGCGGTA TTATTTAAGACGTGATTTAATGTTAAAACT CAAACTATCTTGATTATAAGTGGACTGATA Streptococcus phage JX01 Streptococcus phage LYGO9 20 ATTTTTACACGAGTGCTAGAAAACGGGGCA Streptococcus phage JX01 Streptococcus phage LYGO9 21 22 23 24 ACATTTGTTAAATTTGAACTTACTGGAAAT Bacillus virus PBS1 Bacillus phage AR9 TCCAAAACAAATACGAATGCTTGAGCGATA GATTACCTTAGATGATGTTCTAATCGGTAA TTAACAGTTTCAAGTCTGTCTTGTTACTTA 134 Table 4.3 (cont’d) 25 26 27 ACTCTAAATGATAGTTATGAGTTAAATGTT CAAATTACAGTTTCGACTGATTATGGAAAT ATATGTTCCACTCTATGAATTTAGGCTCAT Enterococcus phage vB_EfaP_Zip Enterococcus phage vB_EfaP_IME199 TTTTTACCAATGCTTCCATATCGCTTATAT 28 29 TACTTGACGAATTGAAGATGACGGAATTTA 30 ACATATCAGGAATATAAAAATCGTTCAAAT TGCCTTTTAATTCAGCTCCCACTTTTTATC 31 CGGCGACTGATTAGGTAACCGTTAGCTCGT Streptococcus phage PhiNIH1.1 32 33 AGATGGTTATAACGCTTGGGCTATATCAAT AAGGTACTTTTAGCTTGTTCTTGTGGTGTT 34 ACAGCTACTGTAAATTCTGCTTTTACGGTT 35 TAGTGCAGTTGTCAAGGAGATTGTGAGCGA Streptococcus phage IPP35/34 36 37 TTAAAAGATTTAAACTATCAAGCGTCAATT TTCTAAATGCTGGTGACTGCTTTGCATAAA Streptococcus phage LF2 38 AACATATTAGACCTACTTCACTTCAATATC 39 40 TGGTTATACATTTACTAATCCATCAGCATT CGTCACCTATTGTTTTATTATTTTTACTA 41 GTATTTGCCCATTTTCTAAATAAGTATATT 42 43 CCGTTCAATCTGTTCTTGCTTTTGGTCATC Streptococcus phage LYGO9 44 GCGATGATGGTAAGTCATCATGGACAGCGT Streptococcus phage JX01 Streptococcus phage LYGO9 TCCTTTTTTAGATAATGTGCGATCACGGAC 45 TTAATAACCTTATGTGATAGATAATCATTT 46 47 ACTGCTGGCGCTGTTTGTTCAACTGCTGGT 48 ACGACCCCAGGTTCTAATACAAATCAACCA 49 TTAATCAATAACAAACGCCAACGACCTAAA 50 AGACAAGATAAAAAGACCTTGTATTGTGAT Streptococcus phage SOCP Streptococcus phage Cp-1 135 Table 4.3 (cont’d) TTATCAATTTGTTTCACAAGAATTTGATAA Streptococcus phage Cp-7 Streptococcus phage SOCP Streptococcus phage Cp-1 Bacillus phage vB_BthM-Goe5 CGTACTTTGTTATAAACCCTGTTTCCAGCA GAAGCTATTGTCTTAAACTCAAAAACTTT Bacillus phage vB_BsuM-Goe3 CGATTTGCTCGTACATTATAAATACCACCA TTGTATATGACCCGTTACCGCTATCAAATA TACCTGGAATGTTAAATTTACCTTTTAGCA Streptococcus phage LF2 Streptococcus phage JX01 Streptococcus phage LYGO9 ATGTTGCCACTAGTTGCTCTTTATGAACAG AGTTTCTTTTAAGTTCAATGTTTGTCAAAT AGTCCCTCCTTAACTTTTTGAGCGTAAATT Streptococcus phage LF2 TTTATTTTTTCTCAGTTCCTTGATTTTAGA 57 58 59 60 61 GATGATGGCCGAACAAGTTACACCAGATGA 62 TTAAGCAACATATTTTGGATGATTAAAACA Streptococcus phage T12 Streptococcus phage PhiNIH1.1 ATAATTGATGTTAAATCATCAATATCGATA Streptococcus phage LF1/4 63 TCTGCATCATGCGGAACTCTGATTGTTTTG Streptococcus phage LF1/4 64 TAGTATAATGCTCGAGGACTATCCGTCTGA 65 66 TCAAAATATCACACATAAATCATCCTCCTT 67 AATGGAATTATCACAAGCGGGGCATATGGT Streptococcus phage LF1/4 AGACTGCTTGCGCTTTGGTGGATTGTATGT Streptococcus phage T12 68 ATCGGGTCTATATGCCCCCTAAAATCAATT Streptococcus phage LF1/4 69 70 AGTCTAATCGCCTATTATTATCCAGATAAA 71 TGAAAGCGAATTGACGATAAGCGAAAATGA 72 TCATTACTTGTATTTATAAATGATTTAGCA Streptococcus phage LF2 Streptococcus phage LYGO9 136 51 52 53 54 55 56 Table 4.3 (cont’d) 73 ATAGACCCATCACGAATCGAACGTGATTAA Streptococcus phage LF1/2/4 Streptococcus phage LYGO9 Streptococcus phage JX01 Streptococcus salivarious strain LAB813 plasmid pSAL813 ATCGAGCTTTCACTTTGTTTAACATGGCTT TCAAATCATCGACGGTTCAAAAGTAAAAGA TAAGTGAGTGACTTCTTCGGTTGAATAGTC AGAGCGGTCAGCATGTTGTGCAACGCTGG Streptococcus phage JX01 74 75 76 77 Streptococcus phage LYGO9 ACGTCTTCAGGTGTTATTCTTGGGTCTTCT AAGACTTAAAATCGATTAGAATTGATTTTA Streptococcus phage LF1/4 78 79 80 CCAGTTTAGAAACAGAACAAAACTAACGAA Streptococcus phage PhiNIH1.1 81 AGACAAAGAAGATGGCAAGTCTATCAACAA 82 AGTAGAAACATAACGATAATTCCATGAATA 83 ATGTTGTAGAATCATATCGACCATATAACC 84 GCTTGGTCTCGAAGAATTACAGAAGAACAT 85 CAATTGATTGCCGTTAAAACCGATAGAGGA Streptococcus phage LF1/2/4 Streptococcus phage LYGO9 Streptococcus phage JX01 86 AAATTGCTCTATCAGTCAATAAAGCAAGAT 87 TCATATTGAGTGGTTGTTTAATTAAATTGA TGAATTCCACGCCACCAAGTAAACCTGTGA Streptococcus phage LYGO9 88 Streptococcus phage LF2 89 ACAAGAAACAAACAGCCTTGATGACTTAAT Streptococcus phage LYGO9 90 AAAAAGCAATGGAAGCTAACGAGATTGCTA Streptococcus phage LYGO9 91 CAACCCTATGTTTGATAATATTTTAGACGT Streptococcus phage LF2 Streptococcus phage LF2 Streptococcus phage JX01 137 Table 4.3 (cont’d) 92 93 94 95 96 97 98 99 AAGTGAAGTTGAATTTTATTTGAGATACTA TAATACTTTTACAATATGTGTTTTCACTAC Streptococcus phage LF1/2/4 TTCTATCTTCTGAAGATATTTCACAAGTGA TCAGCGAGATGCTCTAAGTAAGCATGTTGA TCTTCTTTTTAATTCTTCTAACACTCCATC ATCTTCTTTTGACCTAACAAAAGGATATGT AAGTCCAAAGCTTTTTTCACTGTTGCTGGA TAATTCCTTAAGCTTGATTATAGTATAACA Streptococcus phage JX01 Selenomonas ruminantium subsp. Lactilytica TAM6421 plasmid pSRC1 Streptococcus phage DCC1738 CAACTTTTCTATGAAATTAAAATGGCTTCT 100 101 GAAGTATACTACGGTTAGAGATTGGCTCAA Streptoccocus phage LF2 102 GAAACTTCGATTAGTTTGCGTACTCGCTCA Streptococcus phage LF1/4 103 ATTAGTCCTGTTGTTATGATGGATGATTAT 104 GCAAACCTCTAATGGATAATATAGAACAAA 105 ACACAATTAAACATAAAGAAGTGCTTATTT Legionella adelaidensis strain NCTC12735 plasmid 106 107 AAATTGTTTTTGTTGTAATATAAAGTCGAC Streptococcus phage phi3396 108 AACCTTGATAGGCGTACTGTATGGATTCCA Erysipelothrix phage phi1605 109 CTACTGTATTATCTGAAATAGTATCGTTTT 110 TCTCGGGAAATGCAACTGATGCTGCAATCA 111 GCTAGCATGGCACAAAAATAGCGTTGGAAT Streptococcus phage JX01 TCAGGAGATTGTGTGTACTCACGAATTTTT Streptococcus phage LYGO9 112 ACCTTGCTCCGATGACACCATCGCGAACCT Streptococcus phage LF1/2/4 113 TTAATCGGATTGGTCAAAATATCAATCAGC 114 GAAGTCGAAAAAGATGACTTCTATTACTGG 115 AATGAAGATATTCTAGCTCGTGTAAAAATA 116 TTGGAAAAAACACGAAAGTGATATTACTTT TAAGAATTTTAGATACTCTACTTGAATGCT 117 138 118 AAGTGCCACAGTTTGTGGCTGATTGGATTG Streptococcus phage 20617 119 CATTCAAGGACTACCCTCAACAGTAACTCT Streptococcus phage 20617 Table 4.3 (cont’d) Streptococcus phage IPP42/41/5/54 Streptococcus phage 23782 Streptococcus phage phi ARI0131-2 Streptococcus phage phiBHN167 Streptococcus phage SW27/19/18/14/4/24/7 Streptococcus phage CHPC1033/1148/1073/ 1027/927/877/572/1151/1005/676/640 Streptococcus phage P9854/9851/8921/7601/ 7574/7151/7134/7133/5651/3681/0094/0093/0092 Streptococcus phage D4276 Streptococcus phage SWK1/2 TCAATTTAAATATTATTTAGCCTTCTCTAA TCATTTAAATCAGACTTGTAAGTCTCGACT Streptococcus phage T12 120 121 122 CACTATCAGCCTGACTTCTACGTTTAAGGT 123 ATTGCTTCGTGAGCTTCAGGACTATCCAAT 124 ATGATTATATCAAAAAAGCCCGACGGGAAT Streptococcus phage P9 Streptococcus phage PhiNIH1.1 TTAATTAAGGTATTTATACCACCTTTTTGT Streptococcus phage LF2 125 GCCTGCCCCACAAAAAAGTATATTATTATT Streptococcus phage LF1/4 126 127 AGAGTGTGTCCAAGACCAGAGTTACTGTTT Streptococcus phage LF1/4 128 CTATTGGTTAGAATTTTTTTACAGGAAGAA 129 CCACCTCTAGGTCCACGTAGAGTCTTATGT 130 CCACCCCCTTTCTGTGGTATAATTGAAATA 131 AACCCTCATAGCCTCATTTTTATTAGTCGT Streptococcus phage LF3 Streptococcus phage Str03 Bacillus megaterium NCT-2 plasmid pNCT2_1 132 ATTTATTTTTTTATGCATAGCAATTTGACT 139 Table 4.3 (cont’d) 133 CGCTCGATTGATGCTATCAACTATATTGAA 134 TTCTTCAAGAGAACTTGTAGAACAGCTTCA Clostridium butyricum NBRC 13949 plasmid pCBU1 135 TTTAACCTTTGAAAATGTGAAAGGCTCGTA TTTTACACACGATGTCAGATATAATGTCAA Streptococcus phage LF1/4 136 137 AGTACTGCACTAGGAATTGTAGAGATCAAA 138 CTAAAAATAAACTGTTTGGGTCCAGCAGCAA 139 ACGGTGTTGCACACTCTATCACTTATAAAA Streptococcus phage LYGO9 Streptococcus phage LF2 Streptococcus phage JX01 140 ACCACTAGCAGGATTTTCTATGATGAAATA 141 GCATATAGTCATAGACATCTTGAAAGTAAT 142 TAGAAATGTACATTCTAGGAAAAGACATTA 143 AAAATTACCGTCAAACGTTACAAGTTCGCC Streptococcus phage LF2 144 CACAAGTATTCCCACAATCACAATGACATA 145 GAGTAAACATGATATTATTCAAAATTAAAC 146 AAGATGAGAAACCATATAGCATTGATAACT Streptococcus phage LF1/4 147 ACATTGGCAATTGTTTTCGTCTCGTAGATA 148 CTACTAGGGATAAAACAAAATACTTATAGT 149 AGTAAAGAACCAGATGCGCCTAAGCCTATT 150 AAAACCAAAGGAAGATATGATAACACACTT 151 AAATTGTCCTGGATTATTGTGCAAATCGTT 152 TGAAATGGCTGGTTATGTCGACGGCGAGGA 153 AAAACCATCTGCACAAACTATTTCAATATT TAGATATATCCCTTTCGAGGAAGCTGTGTT 154 155 AAAGGTTCGAAAGTCATGAAAGCTAGTATG Streptococcus phage LYGO9 TCAGAATGATCATCTTGTAGAAATTATTGA 156 157 CGCCGTTTGTAATGGTTTGCCAGTAAGAGT Streptococcus phage LF2 Streptococcus phage JX01 140 Table 4.3 (cont’d) 158 ACAACCTCACCAATAATTCTAAAGTCGCTA Streptococcus phage T1 2Streptococcus phage JX01 Streptococcus phage T12 repressor (excisionase) 159 AGATGCAACTAAAAACGGTGCAGACTTCAT Streptococcus phage Str01 Streptococcus phage A25 Streptococcus phage JX01 Bacteriophage PSA 160 GACAAACTTTCCATCTTAACATCTTTACAA Streptococcus phage P9 161 TATGGTCATCTTCTTTGATAACTTTGGGG Bacillus clarkii bacteriophage BCJA1c Streptococcus phage IPP66/65/55/54/53/52/48/39/14 Streptococcus phage SpGS-1 Streptococcus phagephiARI0639b/0468- 4/0285- 2/0378 Streptococcus phage phiBHN167 Streptococcus phage MM1 162 TATTTCTATATTTATTATATAATATATTAT 163 ATTTTTAGCATAGATTGCTGATTTGGACCC 164 AAGATGGGACTGATGGAAAAGACGGGTTAC 165 ACGACTTCGTTTTCTTCGATTTCTGACCAT 166 AAGAAGCGGAGAGACGGTTAAAAGTGCCGT 167 CATACTGGGCTTTCTTGACCGCTTCCAGAT 168 AAATACACTCTATAAGTTGAAAACTCAAAA Streptococcus phage phiBHN167 169 ACAACTAATATTGCAAGAACTCCCATAAG 170 CTGCGTTAACCCCTCTGCCATCTTTCCAA 171 ACAAAATAAGGGACGTCTTCCCAAAGGCAA 172 TGAAGGCTTGTGTGATTATGCTGAAAGCAG 173 CAATGAAACCAAGTCTCAACATCATGGAGT 174 CACGATGGAGCGAACAGTGGTTTTTACCTT 175 TAATGTGTTCTAGCCTATGAAAAGAGCATA 141 Table 4.3 (cont’d) 176 ACTGCCTTGTCCTGTTGATTCAAGTCAGTT 177 ATCCGCATTCGTTACCGCCCCAATAGTCTC 178 AATCAGAGTTTTTAGCCGACAAACCAGATG 179 ACCACGAGCGAACGACTAACGTTAGCTTTA 180 CAAACGTATAGAAGATGAAGATTTTAAATT Streptococcus phage JX01 Streptococcus phage LYGO9 181 AAAAATCGAAAAATAGATGTGCGTCCAGCA 182 AAGGGTGTTAGATGATAATACCTTTTTTAA 183 ACACCGTTGCGGTTGTTGTCGGTCACTCAA Streptococcus phage LF2 TCTATTAACAATAGTTTTATCCAATTGTTT 184 185 TGAAAACAAGCGCAAAGCTGTCAGAAAACA 186 CGTACCATCTATCAATTTACCGCAAGCTGT Streptococcus phage LF1/2/4 187 CCTCAACATAGTAATAGCTCTTTCCCATAG TTTCATTTTTAATAACCTGCTGGCTCATAT 188 189 CTATCAACGGCTTTTTCAATTAGTGAGATA 190 GTCATGTTATAATTTTCTTGCAAAAAAAAT shuttle/cloning vectors Streptococcus mutans strain MD 191 TTAGTTTTGCTGATTTAAGAAAAGGGGGGT 192 TAGTCGACATAAAACCATTCTTACCACCTC Streptococcus phage IPP62 plasmid 142 Figure 4.1. Phage communities (PCs) can affect the growth of GBS. 10%v/v of three PCs was added to actively growing GBS in a 96-well plate. Growth was observed by OD595 for eight hours. 0 .2 5 0 .2 0 5 9 5 D O 0 .1 5 0 .1 0 0 .0 5 0 .0 0 G B 4 1 1 w ith P C 5 6 1 w ith P C 8 0 1 w ith P C 8 9 5 0 1 2 0 2 4 0 3 6 0 4 8 0 M in u te s 143 Figure 4.2. CRISPR spacer regions differ across GBS strains. GBS CRIPSR Spacers. CRISPR spacers were extracted from whole sequenced genomes and given a unique number. Red shading in the “Number of Spacers” column represents the 25% quartile of the number of spacers, while green shading represents the 75% quartile. Strains are organized by GB strain number. Spacer colors represent a unique spacer. 144 Figure 4.3. The number of CRISPR spacers varies by capsule type. The number of spacers were examined across sequence (A) clinical (B) and (C-D) capsule type. Statistical differences were calculated using chi square tests. 145 Figure 4.4. The number of CRISPR spacers and the percent GBS affect likelihood of lysis. The number of CRISPR spacers and the percent of GBS in the sample significantly impacts the ability of phage communities to inhibit growth. The 25 (Low) and 75% (High) quartile of phage communities were selected based on the percent of GBS reads in the metagenomic sample. The 25 (Low) and 75% (High) quartile of host samples were selected based on the number of spacers. Combinations that resulted in at least a 10% reduction of growth were coded as 0’s. Those above the cutoff were considered resistant (1). 146 Figure 4.5. Phage titer experiments result in clearing or a confluent lawn between dilutions. 1:1 dilutions of PC801GB112 was added THB with 0.7% soft agar and overlaid on THA plates. Dilutions (A.) 1:2 and (B.) 1:3 are shown. Similar patterns were observed in BHI. 147 Figure 4.6. Use of BHI media resulted in clearer spot plates with clear lysogenic rings. A lawn of GB411 was created in BHI with 0.7% BHI agar overlay. Ten microliters of (A) PC801, (B) PC801GB411 and (C) a negative control of PC801 enriched without a bacterial host added. 148 Figure 4.7. Enrichment is not sufficient to isolate phage. PC561 was added to actively growing GB112 and allowed to grow for five hours. Remaining bacteria were pelleted and the supernatant was filter sterilized. The resulting enrichment (PC561.112) was added to actively growing (A) GB411, (B) Lactobacillus and (C) E. coli. Growth was monitored for eight hours on a plate reader (OD595). 149 Figure 4.7 (cont’d) 150 Figure 4.8. Fractions from a cesium chloride gradient maintained infectivity. An enriched sample of PC895 was fractionated by a cesium chloride gradient, resulting in 0.5ml of sample. Lowest density samples are labeled A, while higher density components would be found in the pellet and fraction G. 10% v/v of individual fractions were added to actively growing GB112. Growth was monitored for eight hours on a plate reader (OD595). G B 1 1 2 P C 8 9 5 w h o le 8 9 5 < 1 0 0 k d 8 9 5 > 1 0 0 k d 8 9 5 A 8 9 5 B 8 9 5 C 8 9 5 D 8 9 5 E 8 9 5 F 8 9 5 G 8 9 5 p e lle t 5 9 5 D O 0 .4 0 .3 0 .2 0 .1 0 .0 -0 .1 2 4 6 8 1 0 T im e (H o u rs ) 151 REFERENCES 152 1. REFERENCES Meehan, M. et al. Real-time polymerase chain reaction and culture in the diagnosis of invasive group B streptococcal disease in infants: a retrospective study. European Journal of Clinical Microbiology and Infectious Diseases 34, 2413–2420 (2015). 2. Gibbs RS, B. JD. Streptococcal infections in pregnancy. A study of 48 bacteremias. Am J Obstet Gynecol 140, 405–11 (1981). 3. Hall, J. et al. Maternal Disease with Group B Streptococcus and Serotype Distribution Worldwide: Systematic Review and Meta-analyses. Clinical Infectious Diseases vol. 65 S112–S124 (2017). 4. Bianchi-Jassir, F. et al. Preterm Birth Associated with Group B Streptococcus Maternal Colonization Worldwide: Systematic Review and Meta-analyses. Clinical Infectious Diseases vol. 65 S133–S142 (2017). 5. Meyn, L. A., Krohn, M. A. & Hillier, S. L. Rectal colonization by group B Streptococcus as a predictor of vaginal colonization. American Journal of Obstetrics and Gynecology 201, (2009). 6. Roesch, L. F. W. et al. Diversity and composition of vaginal microbiota of pregnant women at risk for transmitting Group B Streptococcus treated with intrapartum penicillin. PLoS ONE 12, (2017). 7. Azad, M. B. et al. Impact of maternal intrapartum antibiotics, method of birth and breastfeeding on gut microbiota during the first year of life: a prospective cohort study. doi:10.1111/14710528.13624. 8. Stearns, J. C. et al. Intrapartum antibiotics for GBS prophylaxis alter colonization patterns in the early infant gut microbiome of low risk infants. Scientific Reports 7, (2017). 9. Aloisio, I. et al. Influence of intrapartum antibiotic prophylaxis against group B Streptococcus on the early newborn gut composition and evaluation of the anti- Streptococcus activity of Bifidobacterium strains. Applied Microbiology and Biotechnology 98, 6051–6060 (2014). 10. Aloisio, I. et al. Evaluation of the effects of intrapartum antibiotic prophylaxis on newborn intestinal microbiota using a sequencing approach targeted to multi hypervariable 16S rDNA regions. Applied Microbiology and Biotechnology 100, 5537– 5546 (2016). 11. Center for Disease Control. Active Bacterial Core Surveillance (ABCs) Report Emerging Infections Program Network Group B Streptococcus, 2014. (2015). 12. Loc-Carrillo, C. & Abedon, S. T. Pros and cons of phage therapy. Bacteriophage 1, 111– 114 (2011). 153 13. Abedon, S. T., Kuhl, S. J., Blasdel, B. G. & Kutter, E. M. Phage treatment of human infections. Bacteriophage 1, 66–85 (2011). 14. Maura, D., Galtier, M., le Bouguénec, C. & Debarbieux, L. Virulent bacteriophages can target O104:H4 enteroaggregative Escherichia coli in the mouse intestine. Antimicrobial Agents and Chemotherapy 56, 6235–6242 (2012). 15. Galtier, M. et al. Bacteriophages to reduce gut carriage of antibiotic resistant uropathogens with low impact on microbiota composition. Environmental microbiology 18, 2237–2245 (2016). 16. Alemayehu, D., Casey, P. G. & Mcauliffe, O. aeruginosa in the Murine Lung and on Cystic Fibrosis Lung Airway. mBio 3, 1–9 (2012). 17. Chhibber, S., Kaur, T. & Kaur, S. Co-Therapy Using Lytic Bacteriophage and Linezolid: Effective Treatment in Eliminating Methicillin Resistant Staphylococcus aureus (MRSA) from Diabetic Foot Infections. PLoS ONE 8, 1–11 (2013). 18. Mendes, J. J. et al. Wound healing potential of topical bacteriophage therapy on diabetic cutaneous wounds. Wound Repair and Regeneration 21, 595–603 (2013). 19. Jault, P. et al. Efficacy and tolerability of a cocktail of bacteriophages to treat burn wounds infected by Pseudomonas aeruginosa (PhagoBurn): a randomised, controlled, double-blind phase 1/2 trial. The Lancet Infectious Diseases 19, 35–45 (2019). 20. Luo, X. et al. Characterization of bacteriophage HN48 and its protective effects in nile tilapia oreochromis niloticus against streptococcus agalactiae infections. Journal of Fish Diseases 41, 1477–1484 (2018). 21. Cheng, Q., Nelson, D., Zhu, S. & Fischetti, V. A. Removal of group B streptococci colonizing the vagina and oropharynx of mice with a bacteriophage lytic enzyme. Antimicrobial Agents and Chemotherapy 49, 111–117 (2005). 22. Schmelcher, M., Powell, A. M., Camp, M. J., Pohl, C. S. & Donovan, D. M. Synergistic streptococcal phage λSA2 and B30 endolysins kill streptococci in cow milk and in a mouse model of mastitis. Applied Microbiology and Biotechnology 99, 8475–8486 (2015). 23. Russell H., Norcross N.L., K. D. E. Isolation and Characterization of Steptococcus agalactiae Bacteriophage. 1, 75–92 (1969). 24. Bai, Q. et al. Characterization and genome sequencing of a novel bacteriophage infecting Streptococcus agalactiae with high similarity to a phage from Streptococcus pyogenes. Archives of Virology 158, 1733–1741 (2013). 25. Domelier, A. S. et al. Molecular characterization and lytic activities of Streptococcus agalactiae bacteriophages and determination of lysogenic-strain features. Journal of Bacteriology 191, 4776–4785 (2009). 154 26. Pritchard, D. G., Dong, S., Baker, J. R. & Engler, J. A. The bifunctional peptidoglycan lysin of Streptococcus agalactiae bacteriophage B30. Microbiology 150, 2079–2087 (2004). 27. Furfaro, L. L., Chang, B. J. & Payne, M. S. Applications for bacteriophage therapy during pregnancy and the perinatal period. Frontiers in Microbiology vol. 8 (2018). 28. Brock, T. D. Host Range of Certain Virulent and Temperate Bacteriophages Attacking. Journal of bacteriology 88, 165–171 (1964). 29. Russell, N. J. et al. Maternal Colonization with Group B Streptococcus and Serotype Distribution Worldwide: Systematic Review and Meta-analyses. Clinical Infectious Diseases vol. 65 S100–S111 (2017). 30. Dalmasso, M., Hill, C. & Ross, R. P. Exploiting gut bacteriophages for human health. Trends in Microbiology 22, 399–405 (2014). 31. Rodriguez-Valera, F. et al. Explaining microbial population genomics through phage predation. Nature reviews. Microbiology 7, 828–36 (2009). 32. Barrangou, R. CRISPR-Cas systems and RNA-guided interference. Wiley Interdisciplinary Reviews: RNA 4, 267–278 (2013). 33. Beauruelle, C. et al. CRISPR: A useful genetic feature to follow vaginal carriage of group B Streptococcus. Frontiers in Microbiology 8, (2017). 34. Beauruelle, C., Pastuszka, A., Mereghetti, L. & Lanotte, P. Group B Streptococcus Vaginal Carriage in Pregnant Women as Deciphered by Clustered Regularly Interspaced Short Palindromic Repeat Analysis. http://www.mbio.ncsu.edu/bioedit/bioedit.html (2018). 35. Manning, S. D. et al. Genotypic Diversity and Serotype Distribution of Group B Streptococcus Isolated from Women Before and After Delivery. Clinical Infectious Diseases 46, 1829–1837 (2008). 36. Manning, S. D. et al. Multilocus sequence types associated with neonatal Group B streptococcal sepsis and meningitis in Canada. Journal of Clinical Microbiology 47, 1143–1148 (2009). 37. Medugu, N. et al. Group B streptococcal colonization and transmission dynamics in pregnant women and their newborns in Nigeria: implications for prevention strategies. Clinical Microbiology and Infection 23, 673.e9-673.e16 (2017). 38. Kearse, M. et al. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–1649 (2012). 39. Singh, P. et al. Intestinal microbial communities associated with acute enteric infections and disease recovery. Microbiome 3, 45 (2015). 155 40. Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014). 41. Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nature Methods 9, 357–359 (2012). 42. Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009). 43. Menzel, P., Ng, K. L. & Krogh, A. Fast and sensitive taxonomic classification for metagenomics with Kaiju. Nature Communications 7, (2016). 44. Summers, W. C. Biology and Applications. Bacteriophages : biology and applications (2005). 45. Lier, C. et al. Analysis of the type II-A CRISPR-Cas system of Streptococcus agalactiae reveals distinctive features according to genetic lineages. Frontiers in Genetics 6, (2015). 46. Serwer, P., Hayes, S. J., Thomas, J. A. & Hardies, S. C. Propagating the missing bacteriophages: A large bacteriophage in a new class. Virology Journal 4, (2007). 47. Mullan, W. M. A. Factors Affecting Plaque Formation. [On-line]. https://www.dairyscience.info/index.php/enumeration-of-lactococcal- bacteriophages/factors-affecting-plaque-formation.htm (2002). 48. Walker, J. M. Series Editor Methods in Molecular Biology. www.springer.com/series/7651. 49. Korir, M. L. et al. Differing mechanisms of surviving phagosomal stress among group b streptococcus strains of varying genotypes. Virulence 8, 924–937 (2017). 50. Korir, M. L. et al. Investigation of the role that NADH peroxidase plays in oxidative stress survival in group B Streptococcus. Frontiers in Microbiology 9, (2018). 51. Flaherty, R. A. et al. Genetically distinct Group B Streptococcus strains induce varying macrophage cytokine responses. PLOS ONE 14, e0222910 (2019). 52. Parker, R. E. et al. Association between genotypic diversity and biofilm production in group B Streptococcus. BMC Microbiology 16, (2016). 156 CHAPTER 5 CONCLUSIONS AND FUTURE DIRECTIONS 157 The microbiota, or the microbes that inhabit the human body, have been extensively studied for their role in human health but can also serve as a reservoir of opportunistic pathogens. The advent and application of metagenomic sequencing has increased our understanding of which microbes are present in the microbiome and has also begun to challenge the sterility of a variety of sites in the human body (e.g., upper reproductive tract, inner ear). Though controversial, such studies have found low loads of bacteria in the placental membranes, including Lactobacillus, a commonly used probiotic. While there are studies that assess the functionality of the microbiota in other body sites, no such study has been performed with a model of the placental membranes. Hence, this thesis sought to gain a better understanding of how a typically commensal bacteria, Lactobacillus, affects this important barrier during pregnancy. Further, we sought to examine if probiotic properties of Lactobacillus observed in other body sites could be extended to the placental membranes to impact an opportunistic pathogen, Group B Streptococcus (GBS), which invades these membranes from the vaginal tract and is a leading cause of preterm birth, stillbirth, and neonatal infections. Finally, this thesis sought to examine whether an understudied component of the microbiome, bacteriophage, could impact GBS. Herein, we characterized four different strains of Lactobacillus for their ability to interact with a cell line model representing the outermost layer of the placental membranes, decidual cells. We found that while each strain could associate with the cells, L. crispatus associated significantly better. As this is one of the Lactobacillus strains most commonly found in the placental membranes, it would be interesting to further examine properties of this strain that allow it to better attach to this cell type in comparison to the other Lactobacillus strains. Further, we found that none of Lactobacillus strains appear to cause an inflammatory response or host 158 cell death in the decidual cell model. These finding suggests that if Lactobacillus was present in the placental membranes, it would not likely induce inflammation through the decidual cells. However, this model does not evaluate interactions with other cell types in the placental membranes including macrophages, which as professional phagocytes, likely play a larger role in inducing inflammation in vivo. For this reason, future directions require investigation of these other cell types individually and in combination. Ex vivo models using isolated placental membranes attached to a transwell would be an interesting model to further assess the effects of Lactobacillus on intact placental membranes containing multiple cell types. Because we determined that all four Lactobacillus strains associate with decidual stromal cells without inducing damage, we next asked if these strains inhibit pathogens such as GBS, that are capable of invading the placental membranes. As a member of the vaginal microbiome of approximately 30% of the population, GBS has the opportunity to ascend from the vaginal tract to the uterine cavity, where it can associate with the placental membranes. Association with these membranes could result in inflammation leading to premature birth, while invasion of these membranes could result in an in utero infection of the fetus. We found that live Lactobacillus does not directly inhibit growth or impact biofilm production in two distinct GBS strains. While a certain strain of Lactobacillus, L. gasseri 33323, increased association of GBS with decidual stromal cells, this interaction did not result in increased invasion or cell death in the host cells. Hence, this finding suggests that L. gasseri may interact with the GBS on the placental membranes. It would be interesting to examine this relationship further to determine if L. gasseri and GBS are co-aggregating. The biofilms could also be examined to determine if the two species are forming a multi-species biofilm and how these strains are affecting each other in this environment. 159 As Lactobacillus spp. are known to secrete many inhibitory compounds, we also examined the effect of Lactobacillus supernatants on GBS. We found that these supernatants were able to inhibit growth and biofilm formation in GBS, though this inhibition was dependent on the strain of Lactobacillus. Nonetheless, these supernatants also negatively affected host cells through GBS-induced host cell death as Lactobacillus supernatants alone had no effect. Finally, to assess the potential use of the supernatant on other strains of GBS, we confirmed that supernatant from L. reuteri 6475 broadly inhibited 35 GBS strains of different sequence, capsule and clinical types. Collectively, these data suggest that both live Lactobacillus and its supernatant could impact GBS interactions with the placental membranes. The variability observed between strain combinations underlines the importance of studying multiple strains when examining these interactions. Future work could focus on further characterization of the supernatants. Fractionation of the supernatants will help determine the causative compound in the supernatants that is contributing the observed phenotypes. Reducing the complexity of the supernatants may also separate the different phenotypes and remove the negative effects we found on host cell death. Lastly, though bacteriophage are abundant in the microbiome, their role in the colonization of opportunistic pathogens like GBS, which are commonly found in the microbiome, remains unknown. We hypothesized that phage communities isolated from fecal samples would contain phage capable of lysing GBS because GBS is commonly isolated using fecal swabs. We found that 6% of the 130 tested phage communities were capable of lysing a representative GBS strain. We further characterized the interactions between two complex phage communities and multiple GBS hosts (n = 38) and found that capsule, sequence and clinical types of the strains did not affect lysis patterns. However, examination of Clustered regularly 160 interspaced palindromic repeats (CRISPR) in 27 GBS genomes did reveal differences. These repeats serve as an adaptive immune system against invading foreign DNA including phage; therefore, we hypothesized that higher numbers of sequences would result in strains that are more resistant to lysis by phage communities. Comparing GBS strains with high and low numbers of spacers revealed that strains with fewer spacers were more likely to be lysed by a phage community. As we had observed the ability to these communities to inhibit GBS in vitro, we hypothesized that they would also be able to affect GBS in vivo, resulting in reduced levels of GBS in the corresponding bacterial metagenomic reads of each sample. While this correlation did not exist across all GBS strains tested, sensitive strains of GBS were significantly more likely to be inhibited by phage communities with a lower abundance of GBS. This finding highlights the importance of both the GBS strain and the phage community in this interaction. Collectively, these data suggest that the phage component of the intestinal microbiome could impact GBS colonization. Outside of the potential role of phage communities in GBS colonization in the rectal tract, it is also attractive to consider the use of bacteriophage to preferentially remove GBS from the microbiota. To examine this, an individual lytic phage would need to be isolated; however, we could not successfully isolate an individual phage. While examination of growth inhibition allowed us to examine a large number of interactions, it may not be optimal for isolating a lytic phage. Instead, it may be more useful to begin with identifying a strain that can form plaques. This work contributes to several current gaps in our understanding of the microbiota’s impact on health. First, the placental membranes were previously considered sterile, but we have characterized potential interactions between a cell line model and Lactobacillus and GBS. We have also expanded our understanding of GBS rectal colonization in humans by examining the 161 potential role of the bacteriophage component of the microbiome. As Lactobacillus and bacteriophage have been proposed as alternative therapies for GBS, this work may have downstream impacts on GBS disease prevention. 162