This document is downloaded from DR‑NTU (https://dr.ntu.edu.sg) Nanyang Technological University, Singapore.

Inside‑out : pathogenesis of enterococcal‑associated wound infections

Tay, Wei Hong

2018

Tay, W. H. (2018). Inside‑out : pathogenesis of enterococcal‑associated wound infections. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/87455 https://doi.org/10.32657/10220/46716

Downloaded on 30 Sep 2021 23:08:00 SGT INSIDE-OUT: PATHOGENESIS OF ENTEROCOCCAL- ASSOCIATED WOUND INFECTIONS

TAY WEI HONG B.Sc. Biological Sciences (Hons) Nanyang Technological University

Interdisciplinary Graduate School Singapore Centre for Environmental Life Sciences Engineering

A thesis submitted to the Nanyang Technological University in partial fulfillment of the requirement for the degree of Doctor of Philosophy

2018 ACKNOWLEDGEMENTS

This is, without doubt, the most difficult section to write. Nevertheless, I will try my best. First and foremost, I would like to express my gratitude and appreciation to my supervisor,

Kimberly Kline. I must say, I will be hard-pressed to find another supervisor that is exceedingly kind, patient, supportive and understanding than Kim. Thank you, for never doubting me, giving me the freedom to develop my own research and for guiding me along in asking the right questions. Without your unwavering guidance and support, I would have not grown to be the scientist that I am today.

I am thankful to my co-supervisor Swaine Chen and my mentor John Common for their guidance, supervision and technical expertise. I would not have made it thus far, without the critical inputs and suggestions given.

Special thanks to all members of the Kline lab, both past and present. Everyone’s input during the Monday, pathogenesis and BUG meetings allowed us to drive this project forward.

I am grateful to my partner in crime, Kelvin Chong. Thank you, for the boisterous discussions that we had. We will always make the best tag-team. I am especially thankful to have Pei Yi around, thank you, for all the scientific and non-scientific help given and for being there, when

I needed a listening ear. Life in the Kline lab will never be the same, without your infectious laughter and your cheerful personality. I am also thankful to Sharon Goh, Rafi Rashid, Eliza

Soh, Artur Matysik, Irina Afonina and Sumitra Mitra for their scientific advises and friendship. I am also incredibly fortunate to have worked with Rayson Ho, Hui Yi, Joyce Pek,

Jun Jie, Eunice and Wan Xin. Thank you for being such wonderful students and I hope that all of you will be successful in all your future endeavors. I have learnt a lot from all of you!

I am indebted to Brenda Tien, Foo Kiong, Anu, Wooi Keong, Song Lin and Sarah

Kay for helping me with the maintenance and upkeep of the Tissue Culture Facility. I am also grateful to the following laboratory support personnel: Krager, Wey Yeeng, Kartini,

Deborah, Ley Byan, Wahyuna and Lee Chin Weng for assisting me with purchasing, facility management as well as consumables collections. All of you have always been under- appreciated by the rest of the researchers. Without all of you, how could we carry on with our research in SCELSE?

I would also like to extend my appreciation to my fellow PhD course mates Hafiz, Jia

Wang, Milton Kwek, Erica Teo, Lian Yilong, Wendy Phua, Penny Oh, Jeremy Lim and

Mark Tan for their friendship and help throughout this journey. Thank you for your support throughout these 4 long years.

I must also thank Evelyn Wong, Serene Tan, Melissa Tan and Lee Jing for supporting and helping me. Purchasing equipment and consumables would have been a lot more expensive and troublesome without your assistance.

Finally, I am immensely fortunate to have the full support of my wife, Yi Qi and my family members. This journey would not have been possible without your understanding and accommodation to my schedule for family events. Thank you for always believing in me and I dedicate this thesis to all of you. I do not usually say this out loud, or put it in writing, but I love you all!

This section took an unbelievable two weeks to craft. I hope that I did not miss anyone out. I would like to end off, with a quote from the 1997 science fiction film, Gatacca:

“There is no gene for the human spirit” TABLE OF CONTENTS

Sections Page

SUMMARY i

LIST OF TABLES iii

LIST OF FIGURES iv

LIST OF ABBREVIATIONS vii

PUBLICATIONS x

PRESENTATIONS xi

CHAPTER 1: BACKGROUND AND SIGNIFICANCE 1

1.1 Research motivation 1

1.2 Epidemiology of wound infections and factors involved 2

1.2.1 Microbiology of wound infections 3

1.2.2 Biofilms in wound infections and the host response 4

1.2.3 Polymicrobial wound infections 7

1.3 Enterococci 8

1.3.1 Epidemiology of Enterococcal infections 9

1.3.2 Enterococcal wound infections 10

1.3.3 Pathophysiology and virulence factors involved in 11 infection

1.3.4 Enterococcal biofilms and infection 13

1.3.5 Intracellular survival and dissemination of E. faecalis 14 1.4 Keratinocytes and the innate immune response 16

1.5 Thesis outlines and specific aims 18

1.6 References for Chapter 1 21

CHAPTER 2: ENTEROCOCCAL METABOLITE CUES 31 FACILITATE INTERSPECIES NICHE MODULATION AND POLYMICROBIAL INFECTION

2.1 Introduction and motivation 31

2.2 Results 38

2.2.1 Enhanced E. coli growth by E. faecalis is dependent on 38 enterobactin siderophore production 2.2.2 E. coli ferrous uptake systems are not important for E. 39 faecalis-driven E. coli growth promotion 2.2.3 Co-culture of E. faecalis and E. coli results in enhanced 41 biofilm/biomass under iron deplete conditions 2.2.4 E. faecalis Transposon library screen and validation 42

2.2.5 E. faecalis Transposon library single species secondary 46 validation 2.2.6 L-ornithine, but not branched-chain amino acids, 48 augments E. coli biofilm formation under iron limitation 2.3 Discussions 51

2.4 Materials and methods 56

2.5 Contributions 62

2.6 References for Chapter 2 64

CHAPTER 3: CHARACTERIZATION OF THE 69 INTRACELLULAR LIFESTYLE OF ENTEROCOCCUS FAECALIS

3.1 Introduction and motivation 69

3.2 Results 74

3.2.1 E. faecalis forms microcolonies on the wound surface and 74 under the wound bed 3.2.2 Intracellular E. faecalis can be recovered from CD45+ and 76 CD45- cells 3.2.3 E. faecalis adheres to and enters keratinocytes 77

3.2.4 E. faecalis entry is not cell line specific 79

3.2.5 E. faecalis survives up to 72 hpi within keratinocytes 80

3.2.6 E. faecalis is able to multiply within keratinocytes and 82 escape 3.2.7 E. faecalis escapes from infected keratinocytes and 83 subsequently infects other keratinocytes when conditions are favourable 3.2.8 E. faecalis exists within early and late endosomes 84

3.2.9 Lamp1+ vacuoles that contain E. faecalis lack Cathepsin D 87

3.2.10 E. faecalis is present in compartments lacking Lamp1 at 88 late timepoints in the cell periphery 3.3 Discussions 93

3.4 Materials and methods 98

3.5 Contributions 107

3.6 References for Chapter 3 108

CHAPTER 4: ENTRY OF ENTEROCOCCUS FAECALIS INTO 112 KERATINOCYTES AND IMMUNE-MODULATION

4.1 Introduction and motivation 112

4.2 Results 117

4.2.1 E. faecalis entry into keratinocytes is dependent on 117 microfilament polymerization but independent of microtubule polymerization 4.2.2 E. faecalis entry into keratinocytes is clathrin and 119 caveolae independent 4.2.3 E. faecalis entry into keratinocytes is dependent on PI3K 121 signalling 4.2.4 Classical E. faecalis virulence factors are dispensable for 122 attachment and uptake into keratinocytes 4.2.5 Intracellular E. faecalis can modulate the immune 124 response in keratinocytes 4.3 Discussions 127

4.4 Materials and methods 131

4.5 Contributions 134

4.6 References for Chapter 4 135

CHAPTER 5: SUMMARY AND PERSPECTIVES 140

5.1 Summary of findings 142

5.2 Future perspectives 145

5.3 References for Chapter 5 152

Inside-Out: Pathogenesis of Enterococcal-associated wound infections

by

Wei Hong TAY

Interdisciplinary Graduate School

Singapore Centre for Environmental Life Sciences Engineering

Nanyang Technological University

Summary

Wound infections represent a significant health problem, due to their difficulty in management and the cost associated with their treatment. Surgical wounds are often infected with Enterococci, which is one of the top 3 most frequently isolated bacterial species from this site. However, mechanisms that govern and contribute to Enterococcal pathogenesis in wounds have not been elucidated. The following thesis hypothesized that polymicrobial wounds with Enterococci will result in a poorer disease outcome, due to synergistic effects with other microbial species. We also proposed that persistent and recurring wound infections are due to Enterococci having a dual extracellular and intracellular lifestyle. We used an unbiased screening approach mimicking in vivo wound conditions to screen for Enterococcal factors that are responsible for enhanced E. coli biofilm growth in wounds. We demonstrated that the Enterococcal metabolite, L-ornithine can induce siderophore production in E. coli, resulting in enhanced virulence under iron limiting conditions. Using both in vitro and in vivo models, we also showed that Enterococci have a dual extracellular and intracellular infective lifestyle. Together, these studies demonstrate the contributions of Enterococci in wound infections, both outside of the host cells and inside of the host cells. Our work provides evidence of cooperative extracellular interactions between Enterococci and other co-infecting microbial species in wounds, and demonstrates that it has a dual infective lifestyle. Our findings will be useful to clinicians in managing persistent wound infections and could potentially result in the development of improved treatment strategies or therapeutics.

Page | i

Thesis supervisors:

Kimberly Ann Kline

Associate Professor, School of Biological Sciences, Nanyang Technological University

Principle Investigator, Singapore Centre for Environmental Life Sciences Engineering, Nanyang Technological University

Swaine Lin Chen

Assistant Professor, Yong Loo Lin School of Medicine, National University of Singapore

Senior Research Scientist, Genome Institute of Singapore, Agency for Science and Technology Research (A*STAR)

John Common

Principle Investigator, Institute of Medical Biology, Agency for Science and Technology Research (A*STAR)

Page | ii

LIST OF TABLES

Table 2.1 Mutants identified following primary validation of transposon library screen. Table 2.2 Final list of transposon mutants following library screening and validation. Table 3.1 Primers used for generation and verification of the EF1α Lamp1- mCherry construct

Page | iii

LIST OF FIGURES

Figure 1.1 Properties of bacterial biofilms

Figure 1.2 Chronic wounds are characterized as having prolonged inflammation, presence of

biofilms, hyperproliferative epidermis with defective migration as well a constant

influx of PMNLs

Figure 1.3 Cytokine receptors and secretion in keratinocytes

Figure 2.1 Various siderophores from the four major structural families

Figure 2.2 E. faecalis promotes E. coli biofilm growth under iron limitation

Figure 2.3 E. faecalis proximity causes growth changes in E. coli

Figure 2.4 Enterobactin deletion mutants fail to exhibit E. faecalis-driven growth promotion

Figure 2.5 Ferrous uptakes systems are dispensable for E. faecalis-driven E. coli growth

promotion despite local pH changes

Figure 2.6 Optical density (OD595) of single species E. faecalis (OG1RF), single species

UPEC (UTI89) and mixed species (OG1RF + UTI89) biofilms at the 120-hour post

incubation timepoint

Figure 2.7 Schematic representation of the E. faecalis transposon library screen process

Figure 2.8 Primary screen of E. faecalis transposon mutants with decreased biofilm/biomass

during co-culture with E. coli under iron limiting conditions

Figure 2.9 Extracted data of E. faecalis transposon mutants with impaired growth kinetics

following 14hrs incubation at 37 oC

Figure 2.10 Extracted data of E. faecalis transposon mutants with impaired biofilm formation

following 24 hrs incubation at 37 oC

Figure 2.11 L-Ornithine supplementation under iron limiting conditions enhances wild-type E.

coli biofilm but not E. coli UTI89ΔentB

Page | iv

Figure 2.12 Branched-chain amino acid supplementation does not enhance E. coli

biofilm/biomass accumulation at 120hrs in TSBG supplemented with 22D

Figure 2.13 E. faecalis co-infection increases the growth of E. coli in a mouse model of wound

infection

Figure 3.1 E. faecalis persists in wounds up to 5 dpi or 7 dpi depending on the initial

inoculum

Figure 3.2 E. faecalis elicits a robust immune response at 8 hpi

Figure 3.3 E. faecalis persists in infected wounds

Figure 3.4 Recovery of viable E. faecalis from within host cells

Figure 3.5 Time and dose-dependent increase of E. faecalis CFU numbers

Figure 3.6 E. faecalis entry into keratinocytes

Figure 3.7 E. faecalis entry into host cell is not cell line specific.

Figure 3.8 E. faecalis survives for prolonged periods within keratinocytes in large numbers

Figure 3.9 E. faecalis is present in large numbers in keratinocytes at late timepoints

Figure 3.10 Large clusters of E. faecalis is present the periphery of dead keratinocytes.

Figure 3.11 Intracellular E. faecalis is able to escape from keratinocytes and cause subsequent

infections

Figure 3.12 Internalized E. faecalis can be found within early and late endosomes at both 3hpi

and 24hpi

Figure 3.13 E. faecalis survives in Lamp1+ compartments

Figure 3.14 Expansion of Lamp1+ compartments at late timepoints

Figure 3.15 Lamp1+ E. faecalis compartments lack the lysosomal Cathepsin D

Figure 3.16 Lysosomes do not fuse with E. faecalis containing compartments

Figure 3.17 E. faecalis containing compartments do not contain more than 2 diplococci

Page | v

Figure 3.18 Membranes of E. faecalis containing compartments are single-walled and

discontinuous

Figure 4.1 “Trigger” and “zipper” mechanisms utilized by intracellular bacteria to invade non- phagocytic cells. Figure 4.2 E. faecalis entry is dependent on actin polymerization

Figure 4.3 E. faecalis internalization is microtubule independent

Figure 4.4 E. faecalis internalization is dynamin independent

Figure 4.5 E. faecalis internalization is dynamin independent

Figure 4.6 E. faecalis internalization is dependent on P13K signaling

Figure 4.7 E. faecalis internalization is independent of collagen, fibronectin and laminin

Figure 4.8 E. faecalis internalization is increased in a pilus null mutant.

Figure 4.9 Intracellular E. faecalis modulates the cytokine and chemokine secretion at 4hpi by

keratinocytes

Figure 4.10 Intracellular E. faecalis modulates the cytokine and chemokine secretion at 24 hpi

by keratinocytes.

Page | vi

LIST OF ABBREVIATIONS

2,2’ DP 2,2’Dipyridyl

ANOVA Analysis of Variance

BHI Brain-Heart Infusion

BSA Bovine Serum Albumin

CAUTI Catheter Associated Urinary Tract Infections

CCD Charge-Coupled Device

CLEM Correlative Light and Electron Microscopy

CLSM Confocal Laser Scanning Microscopy

CFU Colony Forming Unit

CNA Colistin and Nalidixic Acid

DMEM Dulbecco's Modified Eagle Medium

DPI Day Post Infection

EEA1 Early Endosome Antigen 1

EDTA Ethylenediaminetetraacetic Acid

EPS Extracellular Polymeric Substance

FACS Fluorescence-Activated Cell Sorting

FBS Fetal Bovine Serum

FISH Fluorescence in-situ Hybridization

FLM Fluorescence Light Microscopy

GFP Green Fluorescent Protein

GI Gastrointestinal

HPI Hour Post Infection

IACUC Institutional Animal Care and Use Committee

Page | vii

IFN-γ Interferon-Gamma

IL Interleukins

LAMP-1 Lysosomal-Associated Membrane Protein 1

LTA Lipoteichoic Acid

LB Luria-Bertani

LM Light Microscopy

M6PR Mannose 6-Phosphate Receptor

MMP Matrix Metalloprotease

MOI Multiplicity of Infection

NLR Nod-Like Receptor

OD Optical Density

PAMP Pathogen-Associated Molecular Pattern

PBS Phosphate Buffered Saline

PFA Paraformaldehyde

PMNL Polymorphonuclear Leukocyte

PRR Pattern Recognition Receptor

RLR RIG-I-Like Receptor

ROS Reactive Oxygen Species

SD Standard Deviation

SEM Scanning Electron Microscopy

SSI Surgical Site Infection

TEM Transmission Electron Microscopy

TLR Toll-Like Receptors

TNF-α Tumor Necrosis Factor α

TSBG Tryptic Soy Broth with Glucose

Page | viii

UTI Urinary Tract Infection

WT Wild Type

Page | ix

PUBLICATIONS

Tay WH, Chong KK, Kline KA. Polymicrobial-Host Interactions during Infection. Journal of Molecular Biology. 2016 Aug 28;428(17):3355-71. doi: 10.1016/j.jmb.2016.05.006. Epub 2016 May 8.

Keogh D, Tay WH, Ho YY, Dale JL, Chen S, Umashankar S, Williams RBH, Chen SL, Dunny GM, Kline KA. Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and Polymicrobial Infection. Cell Host & Microbe. 2016 Oct 12;20(4):493-503. doi: 10.1016/j.chom.2016.09.004.

Chong KKL*, Tay WH*, Janela B, Yong AMH, Liew TH, Madden L, Keogh D, Barkham TMS, Ginhoux F, Becker DL, Kline KA. Enterococcus faecalis Modulates Immune Activation and Slows Healing During Wound Infection. The Journal of Infectious Diseases. 2017 Dec 19;216(12):1644-1654. doi: 10.1093/infdis/jix541. *Co-first authors.

Vajjala A, Biswas D, Tay WH, Hanski E, Kline KA. Streptolysin-induced Endoplasmic Reticulum Stress Promotes Group A Streptococcal In Vivo Biofilm Formation and Necrotizing Fasciitis. Cellular Microbiology. Submitted.

Page | x

PRESENTATIONS

Tay WH, Chong KKL and Kline KA. (2017). Inside-Out: Insights into the intracellular lifestyle of Enterococcus faecalis. CSHL Conference on Microbial Pathogenesis & Host Response, New York, USA. 11-16 September 2017 (poster presentation).

Tay WH, Chong KKL and Kline KA. (2018). The Secret Life of Enterococcus faecalis. 5th International Conference on Enterococci, Chamonix, France. 15-20 April 2018 (poster presentation).

Tay WH, Chong KKL, Ludwig A and Kline KA. (2018). The Secret Life of Enterococcus faecalis. EMBO Conference on Innate Immunity in Host Pathogen Interactions 2018, Heidelberg, Germany. 24-27 June 2018 (poster presentation).

Page | xi

CHAPTER 1

Background and Significance

1.1 Research Motivation

The motivation for this thesis dissertation are twofold. First, wound infections impose a

significant financial burden on our healthcare system in addition to contributing to a lower

quality of life for patients. Wound infectious are often polymicrobial, where Enterococci

are associated with 5%-30% of all surgical wound infections and up to 35% of wounds in

diabetic individuals. Despite this, the basic microbiology of the influence of infecting

microbes are poorly understood. The lack of understanding therefore necessitates research

to understand the influence and impact of Enterococcal associated wound infections to

develop better treatment outcomes.

Second, despite antibiotic treatment of wounds infected with Enterococci, they remain

recalcitrant to treatment. Though some studies have attributed their persistence to biofilm

formation, recent studies have also shown that Enterococci can survive within host immune

cells. Moreover, studies from our group have demonstrated that Enterococci can also

modulate the host immune response (Tien et al. 2017). Hence, we hypothesized that the

ability to enter host cells might confer protective benefits to Enterococci, allowing for

immune evasion and survival during wound infection. My goal was to fully characterize

the intracellular lifecycle of Enterococci, which will enable us to understand about the

pathogenesis of Enterococcal associated wound infections and to develop better treatment

plans for recalcitrant infections.

Page | 1

1.2 Epidemiology of wound infections and factors involved

Wound infections affect close to 11 million people worldwide (Demidova-Rice et al. 2012)

and an estimated 20 billion USD is spent annually for the treatment (Sen et al. 2009).

Wound infections are thought to occur due to the expression of virulence factors by

contaminating microorganisms that allows them to circumvent or survive the host immune

system, allowing colonization, infection, invasion and dissemination to take place (Bowler

et al. 2001).

While infections can occur in individuals with superficial burns, abrasions and traumatic

wounds, the bulk of wound infections tend to be surgical in nature (Driscoll 2013), with an

estimated 100 million surgical wounds being created yearly (Driscoll 2013). Surgical site

infections (SSIs) are estimated to occur in 5% of all operations, with a mortality rate of

25% (Cheadle 2006, Edmiston et al. 2015). SSIs are defined as infections that occur within

30 days of surgery, occurring at either the site of incision, organ or interstitial space

following an invasive procedure (Berrios-Torres et al. 2017, Horan et al. 1992, Mangram

et al. 1999a).

SSIs represents the bulk of nosocomial infections in hospitals and are also the more

expensive post-surgical complications to treat (de Lissovoy et al. 2009, Magill et al. 2012,

Sen et al. 2009). Epidemiological data from the United States and Europe estimates the

incidence of SSIs at 38% and 20% respectively (Emori and Gaynes 1993, Mangram et al.

1999a). The incidence of SSI is expected to increase exponentially, due to both an aging

population and an increase in the number of surgical procedures performed (Byrne et al.

1994, DeFrances and Podgornik 2006, Moro et al. 2005). SSIs also impose a significant

Page | 2

burden on the healthcare resources by increasing the average hospital stay by 9.7 days,

which in turn increases healthcare costs (de Lissovoy et al. 2009, DiPiro et al. 1998).

Despite improvements in both infection control practices and surgical techniques, SSIs still

represent a major cause of morbidity and mortality in hospitals (Humphreys 2009, Leaper

et al. 2004, Mangram et al. 1999a).

1.2.1 Microbiology of wound infections

All surgical wounds are contaminated by microbes, and these microbes are usually

composed of the individual’s endogenous flora which can come from the skin, gut, oral or

vaginal cavity (Bowler et al. 2001). In some cases, contaminating microbes can also be

from an exogenous source, due to environmental contaminants in the air or traumatic injury.

The exposed subcutaneous tissue provides a favourable substratum for colonization,

especially when the tissue is devitalized and/or if there is a compromised host immune

response. However, in normal situations where the tissue is perfused, and the host immune

response is normal, opportunities for microorganisms to colonize and proliferate is limited.

Most acute and chronic surgical wounds tend to be polymicrobial in nature, and can

encompass both anaerobic and/or aerobic microorganisms (Bowler and Davies 1999,

Bowler et al. 2001, Rotstein et al. 1985). Commonly isolated aerobic and facultatively

anaerobic microorganisms include Staphylococcus aureus, Staphylococcus epidermidis

and coagulase-negative staphylococci, Enterococcus spp., Escherichia coli, and

Pseudomonas aeruginosa (Bowler et al. 2001, Emori and Gaynes 1993, Giacometti et al.

2000, Mangram et al. 1999b). Anaerobic microorganisms include: Bacteroides spp.,

Prevotella spp., Porphyromonas spp., and Peptostreptococcus spp (Bowler et al. 2001).

Page | 3

While the consequence of anaerobic microorganisms in wound infections are not clearly

understood, extensive studies have shown that infection with S. aureus, P. aeruginosa, and

beta-hemolytic streptococci will result in delayed wound healing (Bowler et al. 2001).

1.2.2 Biofilms in wound infections and the host response

Microbial biofilms represent a considerable problem and approximately 80% of all SSIs

involve biofilms (Hall-Stoodley et al. 2004, Romling and Balsalobre 2012). Biofilms are an

aggregation of microorganisms that attach onto a surface or air/liquid interface and are the

natural state of microorganisms in the environment or in clinical settings. These

microorganisms are encased in a matrix of self-produced extracellular polymeric substances

(EPSs), which are composed of proteins, lipids, and polysaccharides. Bacterial biofilms are

distinct from their planktonic counterparts, in that they often exhibit enhanced tolerance to

antimicrobials, as well as to immunological, chemical, and predatory attacks. Bacterial

biofilms also heterogenous in nature, where different metabolic capacities and physiological

gradients exists within the bacterial community (Flemming et al. 2016). Due to the genetic,

morphological, and spatial heterogeneity of microbial biofilms, therapeutic failures often

result, complicating treatment outcomes (Figure 1.1) (Flemming et al. 2016).

While there are no clear diagnostic guidelines for biofilm-associated infection, it is generally

accepted that for an infection to be classified as biofilm-associated, it should encompass the

following four parameters. 1) The infecting bacteria are surface-associated 2) The attached

infecting bacteria are encased within a matrix that is composed of bacteria or host

constituents 3) The infecting bacteria is localized and 4) The infection is resistant to

Page | 4 antibiotic concentrations that would otherwise successfully eradicate the infecting bacteria, in a planktonic state (Hall-Stoodley et al. 2004).

Figure 1.1: Properties of bacterial biofilms. Reprinted with permission from (Flemming et al. 2016).

Wounds are an ideal niche for biofilm formation due to the presence of necrotic tissue and cell debris which facilitate bacterial attachment, colonization, and subsequent infection.

Bacterial biofilms have been implicated in the transition of an acute wound to a chronic wound. A chronic wound is one that is fails to heal in a timely manner, resulting in an infected site which lacks functional and anatomical integrity (Sen et al. 2009). An estimated

6.5 million individuals in the United States suffer from chronic wounds, and this number is expected to rise, due to increased incidences of comorbidities such as diabetes and an aging population (Sen et al. 2009). Interestingly, all chronic wounds exhibit increased levels of cytokines and matrix metalloproteases (MMPs) together with a constant influx of polymorphonuclear leukocytes (PMNLs) upon transition to the chronic state (Bjarnsholt et al. 2008). A prolonged inflammatory phase thus results in a constant release of cytotoxic and oxygen radicals into the wound site together with inflammatory mediators

(Falanga 2000, Schultz et al. 2003). Although immune infiltration and inflammation are required for proper wound healing, inflammatory disequilibrium results in extensive

Page | 5 damage to the host tissue, ultimately resulting in a chronic and non-healing wound (Figure

1.2). Biofilms are thought to contribute to the impediment of the host response, by preventing PMNLs from penetrating and the bacteria microcolonies or aggregates

(Bjarnsholt et al. 2008). In one study, 60% of all chronic wounds were found to be biofilm associated, compared to only 6% in acute wounds (James et al. 2008).

Figure 1.2: Chronic wounds are characterized as having prolonged inflammation, presence of biofilms, hyperproliferative epidermis with defective migration as well a constant influx of PMNLs. Reprinted with permission from (Eming et al. 2014).

The major limiting factor or bottleneck in the healing process of chronic wounds is the failure of keratinocyte migration across the wound bed and subsequent differentiation, rather than defects in keratinocyte proliferation (Andriessen et al. 1995, Stojadinovic et al. 2005,

Usui et al. 2008). Keratinocyte migration is diminished in the presence of neutrophils (Fazli et al. 2011). Moreover, in-vivo studies have also shown that a sustained inflammatory response impairs epithelial migration (Gill and Parks 2008). Therefore, since keratinocytes

Page | 6

account for 95% of all cells in the epidermis (Sprenger et al. 2013), it is not surprising that

any defects or deficiencies in migration can adversely affect the re-epithelization process.

1.2.3 Polymicrobial wound infections

Epidemiological studies have shown a that polymicrobial presence is correlated with the

chronic non-healing state of wounds and such communities often involve biofilms (Davies

et al. 2004, James et al. 2008, Misic et al. 2014). These observations suggest that there is

some form of synergy between the different microbial species. While there is a growing

appreciation that infecting microbes can interact synergistically to induce virulence traits

and modulate the host immune response, very few mechanistic studies have been done, to

investigate other microbial species in wounds.

In the first polymicrobial experimental wound infection study, performed in guinea pigs,

co-infection with both E. coli and Bacteriodes fragilis resulted in a >100-fold increase in

bacterial CFU for each species after 7 dpi, compared to monomicrobial infection.

Moreover, compared to monomicrobial infection, co-infection with B. fragilis and E. coli

resulted in increased inflammation and purulent wounds with impaired healing developed,

demonstrating pathogenic synergy (Kelly 1978). Similarly, wounds infected with both P.

aeruginosa and S. aureus have been associated with a chronic non-healing state (Schierle

et al. 2009, Seth et al. 2012, Watters et al. 2013, Zhao et al. 2010). Interestingly, co-

infection studies of P. aeruginosa and S. aureus have shown that the P. aeruginosa limits

S. aureus growth (Kessler et al. 1993, Korgaonkar et al. 2013, Mashburn et al. 2005,

Mitchell et al. 2010, Park et al. 2012, Pastar et al. 2013). Using a murine wound excisional

model, peptidoglycan sensing by P. aeruginosa was shown to be important for its

competitive advantage in the presence of other Gram-positive species. P. aeruginosa is able

Page | 7 to respond to N-acetylglucosamine or peptidoglycan fragments via a putative two- component response regulator PA0601, resulting in the induction and secretion of elastase and pyocyanin (Korgaonkar et al. 2013). Upon co-infection with equal numbers of P. aeruginosa and S. aureus, P. aeruginosa outnumbered S. aureus by >100-fold at 4 dpi in the wound. However, a P. aeruginosa PA0601 deletion mutant was not able to outcompete

S. aureus in the same manner, suggesting that P. aeruginosa specifically senses S. aureus peptidoglycan leading to virulence factor induction that allow it to outcompete S. aureus in co-infected wounds (Korgaonkar et al. 2013). However, during co-infection in porcine wounds with P. aeruginosa, the S. aureus virulence factor protein A was significantly downregulated by >3-fold at both 2 and 4 dpi, whereas Panton-Valentine leukocidin (PVL) and α-hemolysin (hla) were significantly upregulated at 4 dpi compared to monomicrobial infection (Pastar et al. 2013). PVL and hla are toxic to leukocytes and result in necrosis within wounds (Inoshima et al. 2011, Kennedy et al. 2010, Lina et al. 1999, Lipinska et al.

2011). These findings therefore suggest that S. aureus mounts countermeasures that promote its virulence within polymicrobial wounds (Korgaonkar et al. 2013, Pastar et al.

2013). Moreover, co-infection with P. aeruginosa and S. aureus together resulted in decreased production of pro-inflammatory cytokines IL-1α, IL-1β, IL-6 and IL-8 compared to either monomicrobial infection (Pastar et al. 2013). Taken together, these studies show that competition between microbes can result in altered virulence factor expression, as well as host immune response modulation, thus providing a competitive edge to a subset of species within a polymicrobial infection.

Page | 8

1.3 Enterococci

1.3.1 Epidemiology of Enterococcal infections Enterococci are Gram-positive, ovococci that are facultative anaerobes and commensals of

the human gastrointestinal (GI) tract. However, Enterococci are also opportunistic

pathogens and have been implicated in numerous infections such as bacteraemia, infective

endocarditis, pelvic infections, urinary tract infections (UTIs), and wound infections (Arias

and Murray 2012). Alarmingly, large-scale epidemiological studies have shown that

Enterococci have become one of the most frequently isolated microbial species from

nosocomial infections (Arias and Murray 2012, Fisher and Phillips 2009). Enterococcal

infections in humans are usually caused by two predominant species: Enterococcus faecalis

and Enterococcus faecium where, together, these two-species account for 90% of all

clinical Enterococcal isolates obtained from various infection sites. Of these two species,

E. faecalis is a major cause of nosocomial infections in both Europe and North America

(McCracken et al. 2013).

Enterococci are remarkably adaptable to environmental pressures, and are intrinsically

resistant to certain classes of antibiotics such as beta-lactams, cephalosporins and

aminoglycosides (Chow 2000, Hollenbeck and Rice 2012). Moreover, their malleable

genomes allow them to easily acquire and/or disseminate determinants of antibiotic

resistance (Dunny et al. 1995). Numerous E. faecalis clinical isolates have been reported

to harbour resistance to last-resort antibiotics such as vancomycin and daptomycin over the

past few decades (Arias et al. 2011, Arias and Murray 2012, Munoz-Price et al. 2005).

Infection with vancomycin resistant strains are associated with increased risk of patient

mortality, of up to 75% as opposed to 45% for infections with vancomycin susceptible

Page | 9

strains (Bearman and Wenzel 2005). Enterococci are also extremely hardy (Bradley and

Fraise 1996), and reports have documented their survival in bile salts and growth under

high salt (Arias and Murray 2012, Flahaut et al. 1996). Taken together, these factors have

complicated treatment outcomes for individuals with Enterococcal infections and accounts

for their widespread dissemination in the healthcare setting. To complicate matters, many

Enterococcal nosocomial infections occur in immunocompromised and in elderly

individuals, which worsens an already unfavourable treatment outcome.

1.3.2 Enterococcal wound infections Many clinical studies have reported E. faecalis as one of the dominant bacteria species that

is commonly isolated from various types of wound infections (Citron et al. 2007, Dowd et

al. 2008, Giacometti et al. 2000, Gjodsbol et al. 2006). In the US, Enterococcus spp. were

the second most commonly isolated organism from SSIs during the period of 2006-2007

(Hidron et al. 2008, Sievert et al. 2013). In one study involving surgical wounds, E. faecalis

accounted for approximately 80% of all Enterococcus spp. and all these strains exhibited

resistance to cefazolin (Giacometti et al. 2000). Another study in New Delhi showed that

E. faecalis accounted for 40% of all trauma and soft tissue infections and exhibited high

level of resistance to gentamicin (Rajkumari et al. 2014). Vancomycin resistant E. faecalis

has also been co-isolated with E. coli, P. aeruginosa from a femoral wound following

surgery (Dombrádi et al. 2009).

In chronic wounds, several studies have shown that Enterococcus spp. were one of the most

consistently isolated microorganism, together with E. coli, Klebsiella spp, Proteus spp. and

Streptococcus spp (Bowler et al. 1999, Brook and Frazier 1998, Davies et al. 2001, Davies

Page | 10

et al. 2004, Gjodsbol et al. 2006). Moreover, E. faecalis was also frequently found in

chronic venous ulcers (71.2%), in association with other microbial species such as S.

aureus, P. aeruginosa, coagulase-negative staphylococci, Proteus spp. as well as other

anaerobes (Gjodsbol et al. 2006). In another study of diabetic foot ulcers, Enterococcus

spp. were also the fourth most frequently isolated microbial species (Dowd et al. 2008).

However, despite all these epidemiological data, mechanisms that contribute E. faecalis

pathogenesis in wounds have not been elucidated, nor have the clinical consequence of its

presence in surgical and chronic wounds.

1.3.3 Pathophysiology and virulence factors involved in infection

Enterococci encode an extensive array of virulence factors, allowing E. faecalis to

successfully colonize many niches and to cause infection. These virulence factors can be

secreted, or surface attached (Arias and Murray 2012). Secreted virulence factors include:

cytolysin, and serine protease (Arias and Murray 2012, Garsin et al. 2014).

Cytolysin (Cyl) is a type-A lantibiotic, belonging to the class of bacteriocins, and exerts

cytotoxic effects on macrophages and neutrophils (Day et al. 2003, Garsin et al. 2014).

Gelatinase (GelE, encoded by gelE) is a matrix metalloprotease that has hydrolytic

properties, allowing for the degradation of host matrix and is also able to inhibit the

chemotaxis of pro-inflammatory cells (Arias and Murray 2012). The serine protease (SprE,

encoded by sprE) lies downstream of GelE and is co-transcribed with gelE (Qin et al. 2000).

Both GelE and SprE regulate biofilm formation, and clinical isolates encoding these two

genes have been associated with increased mortality and complications of treatment

outcome (Garsin et al. 2014).

Page | 11

Surface attached virulence factors include: aggregation substance (AS), adhesin to collagen of Enterococcus faecalis (Ace), endocarditis and biofilm-associated pili (Ebp), and the

Enterococcal surface protein (Esp) (Arias and Murray 2012, Garsin et al. 2014). AS is important for the binding to host epithelium, survival within polymorphonuclear leukocytes and internalization into intestinal epithelial cells (Arias and Murray 2012). AS has also been shown to bind to lipoteichoic acid of neighbouring Enterococci which facilitates clumping (Waters et al. 2004). AS has also been shown to be important for endocarditis, and the deletion resulted in a decrease of recoverable bacterial titres in a mouse model

(Johnson et al. 2004). However, in a mouse model of ascending UTI, deletion of AS did not impair colonization, suggesting that it is dispensable for urinary tract colonization

(Johnson et al. 2004). Ace is important for the binding to collagen, laminin. Deletion of

Ace resulted in impaired colonization in murine models for endocarditis and UTI (Lebreton et al. 2009). Ebp plays an important role in pili biogenesis (Nallapareddy et al. 2006), biofilm formation (Nallapareddy et al. 2006) and subsequent attachment to platelets, fibrinogen and collagen (Nallapareddy et al. 2011a). Several studies have shown that Ebp is also important for early biofilm formation in murine models for endocarditis, UTI and

CAUTI, where the deletion of Ebp resulted in significant attenuation of infection

(Nallapareddy et al. 2006, Nallapareddy et al. 2011a, Nallapareddy et al. 2011b, Nielsen et al. 2013). Esp is also involved biofilm formation (Arias and Murray 2012), attachment to bladder epithelial cells and subsequent colonization in ascending UTI (Shankar et al. 2001).

While all the described virulence factors are important for initial attachment and subsequent establishment of a biofilm, these studies also demonstrate that these virulence factors are niche specific. Given the functional redundancy and the specificity of these virulence factors, it is unlikely that the abrogation of a single factor will affect all Enterococcal associated infections.

Page | 12

Though numerous studies have been done to investigate the virulence factors and

mechanisms responsible for the host pathogen interactions of E. faecalis in the context of

various infections such as endocarditis, UTIs and bacteraemia, very little studies have been

done to investigate the pathogenesis of E. faecalis in wounds. Recently, E. faecalis has

been shown to be the causative agent of intestinal anastomotic leaks following surgery,

where it degrades collagen I within intestinal epithelial cells and activates host

metalloproteinase 9 (MMP9), resulting in degrading of scar tissue and leakage of gut flora

(Shogan et al. 2015).

1.3.4 Enterococcal biofilms and infection

Enterococci can form robust biofilms on a variety of substrates in the natural environment

and on artificial substrates such as prosthetic implants and on catheters (Fernandez

Guerrero et al. 2007). Enterococcal biofilms represent a significant problem in the clinic

because biofilm producing clinical isolates have been reported to exhibit enhanced

tolerance to vancomycin and tigecycline (Hashem et al. 2017, Torelli et al. 2017).

Moreover, mature E. faecalis biofilms that were 3 and 5 days old were also reported to have

increased tolerance to vancomycin (Holmberg and Rasmussen 2016). Enterococcal

biofilms therefore represent a significant healthcare problem and can complicate treatment

outcomes.

Though the spatiotemporal dynamics of Enterococcal biofilm have not been extensively

characterized, there have been some recent developments in the field describing the

development. Several studies have shown that several virulence factors are important for

the initial attachment and subsequent biofilm initiation. These virulence factors are mostly

Page | 13

adhesins, and includes Ace, AS, Ebp, Esp as well as various glycolipids and

(Ch’ng et al. 2018). Following initial attachment, rhamnopolysaccharides are produced and

microcolonies form. Deletion of Epa, which is involved in the production of

rhamnopolysaccharides, resulted in reduced microcolony structural integrity and increased

detachment (Dale et al. 2017). This is then followed by the development of a mature biofilm

which is characterized by extensive production of extracellular DNA, extracellular

proteases and polysaccharides. Lipoteichoic acid (LTA) is important for the biofilm

development and maturation, as well as protection against antimicrobial peptides and

attachment to epithelial cells (Fabretti et al. 2006). Work done by other and our group have

also shown that the multiple peptide resistance factor (MprF), is important for biofilm

growth and virulence. Deletion of a paralogue of MprF, MprF2, resulted in significant

increase of biofilm production and increased resistance to opsonic killing (Bao et al. 2012).

Deletion of both paralogues, MprF1 and MprF2, resulted in reduced biofilm persistence

and the double mutant was out-competed during competitive infection in a mouse model

of wound infection (Chong et al. 2017). While little is known about biofilm dispersal in

Enterococci, gelE has been implicated in the dispersal process. Deletion of gelE resulted in

increased fibrinous matrix deposition on the aorta in a rabbit model of endocarditis,

suggesting that it is involved in digestion of the biofilm matrix and the subsequent dispersal

(Thurlow et al. 2010). Altogether, future studies into Enterococcal biofilm development

and dispersal will aid in the development of new therapeutics or strategies in the clinic.

1.3.5 Intracellular survival and dissemination of E. faecalis

Although E. faecalis has been traditionally considered as an extracellular pathogen, studies

in the past two decades have also shown that E. faecalis is capable of survival and

persistence within eukaryotic cells, including intestinal epithelial cells (Olmsted et al. 1994,

Page | 14

Wells et al. 1988, Wells et al. 1990), macrophages (Baldassarri et al. 2001, Gentry-Weeks et al. 1999), monocytes (Baldassarri et al. 2005), cervical epithelial cells (Bertuccini et al.

2002), endothelial cells (Millan et al. 2013), osteoblasts (Campoccia et al. 2015), and enterocytes (Olmsted et al. 1994).

The only cell type for which E. faecalis intracellular survival has been mechanistically studied are macrophages. Macrophages have received special interest because of their ability to infiltrate to the infection site, and transit the bloodstream to secondary lymphoid organs. Early studies showed that E. faecalis can survive for up to 72 hours post infection

(hpi) within macrophages (Baldassarri et al. 2005, Gentry-Weeks et al. 1999, Sabatino et al. 2015, Süßmuth et al. 2000). Moreover, at timepoints of 24 hpi, infected macrophages began to exhibit a degraded vacuolar compartment and the presence of bacteria within the cytoplasm (Gentry-Weeks et al. 1999). At 72 hpi, the infected macrophages were dead and both live and degraded bacteria can be observed within the cellular remnants of the infected macrophages (Gentry-Weeks et al. 1999). Subsequent studies showed that the glycosaminoglycans heparin, heparan sulfate and chondroitin sulfate A on macrophages mediated E. faecalis entry (Baldassarri et al. 2005). Recent studies also showed that E. faecalis containing vacuoles in macrophages were able to resist acidification and to resist autophagy by inhibiting LC3-II production following phagocytosis by macrophages, together resulting in intracellular survival (Zou and Shankar 2016).

By contrast, limited studies have been done to investigate the intracellular survival and dissemination of E. faecalis within non-professional phagocytes. Early characterization studies utilizing transmission electron microscopy (TEM), suggests that E. faecalis can attach to, be taken up into, and translocate across intestinal epithelial cells to cause

Page | 15

bacteraemia and disseminate to distant organs in mice (Olmsted et al. 1994, Wells et al.

1990). The presence of the virulence factor, aggregation substance, was also found to be

associated with increased E. faecalis uptake into cultured intestinal epithelial cells

(Olmsted et al. 1994). However, later studies showed that Asc10, a member of the

aggregation substance family of surface proteins, was not critical for entry into human

umbilical vein endothelial cells as null mutants were still taken up in similar numbers as

the parental wild type strain. Subsequent studies also demonstrated the importance of GelE

in the translocation of E. faecalis across enterocytes in vitro (Zeng et al. 2005).

Collectively, these studies strongly suggest that E. faecalis possesses a dual

extracellular/intracellular lifestyle, necessitating the need for mechanistic studies into the

otherwise unexplored intracellular lifestyle.

1.4 Keratinocytes and the innate immune response

Keratinocytes contribute to the protective barrier of the skin, while playing a major role in

both the innate defense and recognition of pathogens. Keratinocytes express numerous

pattern recognition receptors (PRRs), including Toll-like receptors (TLRs), Nod-like

receptors (NLRs) and RIG-I-like receptors (RLRs). These receptors function to detect

various evolutionary conserved pathogen-associated molecular patterns (PAMPs), such as

lipoteicoic acid (LTA) and peptidoglycan of Gram positive bacteria, liposaccharides of

Gram negative bacteria (Janeway 1989).

In addition to possessing PRRs, keratinocytes also secrete cytokines which are important

in the recruitment of leukocytes from the circulating blood and the signalling to other

Page | 16 neighbouring cutaneous cells. These cytokines include interleukins (ILs)-1, 6, 7, 8, 10, 12,

15, 18 and 20, tumor necrosis factor α (TNF-α) (Figure 1.3) and interferons α, β and γ

(Grone 2002). Keratinocytes also possesses receptors for IL-1, 2, 4, 6, 10, 13, 17, 18, 20,

24 and TNF- α (Figure 1.3), even though they do not secrete some of these cytokines, which allows them to respond to cytokines secreted by infiltrating immune cells or neighbouring epithelial cells (Grone 2002). Keratinocytes can also secrete chemokines in addition to expressing chemokine receptors. Chemokines include CC-chemokine ligand

(CCL)-20, CXC-chemokine ligand (CXCL)-9, 10, 11 (Olaru and Jensen 2010). Moreover, keratinocytes also secrete many cationic antimicrobial peptides, including β-defensins and cathelicidins. Keratinocytes therefore function not only as a protective barrier but are also well equipped to detect pathogens and initiate the immune response.

Figure 1.3: Cytokine receptors and secretion in keratinocytes. Reprinted with permission from (Grone 2002).

Page | 17

Interestingly, studies have demonstrated that other organisms, such as S. aureus and P.

aeruginosa can be internalized into keratinocytes (Garzoni and Kelley 2009, Kierbel et al.

2005, Kintarak et al. 2004, Soong et al. 2015), and are able to survive within them (Garzoni

and Kelley 2009, Soong et al. 2015). These internalized bacteria are protected from the

extracellular environment and can possibly account for or contribute to delayed wound

healing, recurrent and persistent infections.

1.5 Thesis outlines and specific aims

This thesis was motivated by the fact that Enterococcal wound infections are often ignored

in the clinic as they are often regarded as commensals by clinicians. However, studies by

our group and others have shown that Enterococci can become opportunistic pathogens

when the host is immunocompromised. Several studies have already shown that

Enterococci are frequently co-isolated together with other microorganisms in wounds and

that polymicrobial wounds often have a poorer disease outcome. Moreover, recent studies

by others and our group have also shown that Enterococci can persist in wounds and within

cultured cells, suggesting that they might be utilizing this intracellular modality to escape

from the host immune surveillance, emerging to cause re-infections and thereby resulting

in a persistent and chronic wound. The goal of my thesis is, therefore, to 1) elucidate the

contributions of Enterococci to polymicrobial wound infections and to 2) understand and

determine the mechanism(s) and consequences of intracellular Enterococci during wound

infections. I hypothesized that, in the case of polymicrobial wound infections, Enterococci

can interact with other microorganisms to enable them to better colonize this nutrient poor

niche. In addition, I also hypothesized that their ability to enter host cells can contribute to

Page | 18 persistent infections and immune evasion from host surveillance mechanisms. I have organized the work into three specific aims:

Aim 1: Determine the consequence of the presence of Enterococci in wounds, in association with another commonly co-isolated microorganism, E. coli. To achieve this, we performed crystal violet biofilm assays to quantify biofilm biomass and a large scale saturated E. faecalis transposon library to screen for mutants that displayed lower biofilm/biomass during co-culture as compared to their single species controls. This work is described in Chapter 2.

Aim 2: Define the intracellular lifestyle of Enterococci in wounds. We used an in-vivo mouse model of wound infection as well as an in-vitro 2D tissue culture model for initial characterization of intracellular persistence. We then used a variety of imaging modalities using both light and electron microscopy to determine the localization of intracellular

Enterococci as well as to characterize the structural compartments that contain them. This work is described in Chapter 3.

Aim 3: Elucidate how Enterococci enters the host cell. To understand this, we utilized an in-vitro 2D tissue culture model, applying specific pharmacological inhibitors that block specific host uptake pathways, to determine which uptake pathways are required for

Enterococci entry into host cells. This work is described in Chapter 4.

Page | 19

Chapter 5 of this thesis will summarize our findings and discuss the remaining knowledge gaps in addition to future experimental work. Altogether, these studies advance our understanding of Enterococcal wound pathogenesis, which has largely been ignored until now. Our work provides evidence that Enterococci can influence polymicrobial wound infections, leading to enhanced fitness of E. coli. We have also demonstrated that

Enterococci can exist intracellularly within the host, under in vitro and in vivo settings, where it can survive up to 72 hpi and replicates within keratinocytes. Collectively, our work demonstrates that Enterococcal wound infections poses a significant problem in the clinic, where the presence has been largely ignored. Our work will improve treatment outcomes in the clinic, where the provision of drugs to treat Enterococci infections could improve wound healing and disease outcome.

Page | 20

1.6 References for Chapter 1

Andriessen MP, van Bergen BH, Spruijt KI, Go IH, Schalkwijk J, van de Kerkhof PC (1995). Epidermal proliferation is not impaired in chronic venous ulcers. Acta dermato-venereologica 75: 459-462.

Arias CA, Panesso D, McGrath DM, Qin X, Mojica MF, Miller C et al (2011). Genetic Basis for In Vivo Daptomycin Resistance in Enterococci. New England Journal of Medicine 365: 892-900.

Arias CA, Murray BE (2012). The rise of the Enterococcus: beyond vancomycin resistance. Nat Rev Microbiol 10: 266-278.

Baldassarri L, Cecchini R, Bertuccini L, Ammendolia MG, Iosi F, Arciola CR et al (2001). Enterococcus spp. produces slime and survives in rat peritoneal macrophages. Med Microbiol Immunol 190: 113- 120.

Baldassarri L, Bertuccini L, Creti R, Filippini P, Ammendolia MG, Koch S et al (2005). Glycosaminoglycans mediate invasion and survival of Enterococcus faecalis into macrophages. J Infect Dis 191: 1253-1262.

Bao Y, Sakinc T, Laverde D, Wobser D, Benachour A, Theilacker C et al (2012). Role of mprF1 and mprF2 in the pathogenicity of Enterococcus faecalis. PLoS One 7: e38458.

Bearman GM, Wenzel RP (2005). Bacteremias: a leading cause of death. Archives of medical research 36: 646-659.

Berrios-Torres SI, Umscheid CA, Bratzler DW, Leas B, Stone EC, Kelz RR et al (2017). Centers for Disease Control and Prevention Guideline for the Prevention of Surgical Site Infection, 2017. JAMA surgery 152: 784-791.

Bertuccini L, Ammendolia MG, Superti F, Baldassarri L (2002). Invasion of HeLa cells by Enterococcus faecalis clinical isolates. Med Microbiol Immunol 191: 25-31.

Bjarnsholt T, Kirketerp-Moller K, Jensen PO, Madsen KG, Phipps R, Krogfelt K et al (2008). Why chronic wounds will not heal: a novel hypothesis. Wound Repair Regen 16: 2-10.

Bowler P, Davies BJ (1999). The Microbiology of Acute and Chronic Wounds, vol. 11.

Bowler PG, Davies BJ, Jones SA (1999). Microbial involvement in chronic wound malodour. J Wound Care 8: 216-218.

Bowler PG, Duerden BI, Armstrong DG (2001). Wound Microbiology and Associated Approaches to Wound Management. Clin Microbiol Rev 14: 244-269.

Page | 21

Bradley CR, Fraise AP (1996). Heat and chemical resistance of enterococci. The Journal of hospital infection 34: 191-196.

Brook I, Frazier EH (1998). Aerobic and anaerobic microbiology of chronic venous ulcers. Int J Dermatol 37: 426-428.

Byrne DJ, Lynch W, Napier A, Davey P, Malek M, Cuschieri A (1994). Wound infection rates: the importance of definition and post-discharge wound surveillance. J Hosp Infect 26: 37-43.

Campoccia D, Testoni F, Ravaioli S, Cangini I, Maso A, Speziale P et al (2015). Orthopedic implant- infections. Incompetence of Staphylococcus epidermidis, Staphylococcus lugdunensis and Enterococcus faecalis to invade osteoblasts. J Biomed Mater Res A.

Ch’ng J-H, Chong KKL, Lam LN, Wong JJ, Kline KA (2018). Biofilm-associated infection by enterococci. Nature Reviews Microbiology.

Cheadle WG (2006). Risk factors for surgical site infection. Surg Infect (Larchmt) 7 Suppl 1: S7-11.

Chong KKL, Tay WH, Janela B, Yong AMH, Liew TH, Madden L et al (2017). Enterococcus faecalis modulates immune activation and slows healing during wound infection. J Infect Dis.

Chow JW (2000). Aminoglycoside resistance in enterococci. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 31: 586-589.

Citron DM, Goldstein EJ, Merriam CV, Lipsky BA, Abramson MA (2007). Bacteriology of moderate-to- severe diabetic foot infections and in vitro activity of antimicrobial agents. J Clin Microbiol 45: 2819- 2828.

Dale JL, Nilson JL, Barnes AMT, Dunny GM (2017). Restructuring of Enterococcus faecalis biofilm architecture in response to antibiotic-induced stress. NPJ biofilms and microbiomes 3: 15-15.

Davies CE, Wilson MJ, Hill KE, Stephens P, Hill CM, Harding KG et al (2001). Use of molecular techniques to study microbial diversity in the skin: chronic wounds reevaluated. Wound Repair Regen 9: 332-340.

Davies CE, Hill KE, Wilson MJ, Stephens P, Hill CM, Harding KG et al (2004). Use of 16S ribosomal DNA PCR and denaturing gradient gel electrophoresis for analysis of the microfloras of healing and nonhealing chronic venous leg ulcers. J Clin Microbiol 42: 3549-3557.

Day AM, Cove JH, Phillips-Jones MK (2003). Cytolysin gene expression in Enterococcus faecalis is regulated in response to aerobiosis conditions. Mol Genet Genomics 269: 31-39. de Lissovoy G, Fraeman K, Hutchins V, Murphy D, Song D, Vaughn BB (2009). Surgical site infection: incidence and impact on hospital utilization and treatment costs. Am J Infect Control 37: 387-397.

Page | 22

DeFrances CJ, Podgornik MN (2006). 2004 National Hospital Discharge Survey. Adv Data: 1-19.

Demidova-Rice TN, Hamblin MR, Herman IM (2012). Acute and impaired wound healing: pathophysiology and current methods for drug delivery, part 1: normal and chronic wounds: biology, causes, and approaches to care. Advances in skin & wound care 25: 304-314.

DiPiro JT, Martindale RG, Bakst A, Vacani PF, Watson P, Miller MT (1998). Infection in surgical patients: effects on mortality, hospitalization, and postdischarge care. Am J Health Syst Pharm 55: 777-781.

Dombrádi Z, Bodnár F, Orosi P, Dombrádi V, Szabó J (2009). A case report of vancomycin-resistant Enterococcus faecalis colonization of a femoral wound in central europe. Central European Journal of Medicine 4: 259-261.

Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA et al (2008). Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiol 8: 43.

Driscoll P (2013). Wound prevalence and wound management, 2012-2020. MedMarket Diligence.

Dunny GM, Leonard BA, Hedberg PJ (1995). Pheromone-inducible conjugation in Enterococcus faecalis: interbacterial and host-parasite chemical communication. Journal of Bacteriology 177: 871- 876.

Edmiston CE, Jr., McBain AJ, Roberts C, Leaper D (2015). Clinical and microbiological aspects of biofilm-associated surgical site infections. Adv Exp Med Biol 830: 47-67.

Eming SA, Martin P, Tomic-Canic M (2014). Wound repair and regeneration: mechanisms, signaling, and translation. Sci Transl Med 6: 265sr266.

Emori TG, Gaynes RP (1993). An overview of nosocomial infections, including the role of the microbiology laboratory. Clin Microbiol Rev 6: 428-442.

Fabretti F, Theilacker C, Baldassarri L, Kaczynski Z, Kropec A, Holst O et al (2006). Alanine esters of enterococcal lipoteichoic acid play a role in biofilm formation and resistance to antimicrobial peptides. Infect Immun 74: 4164-4171.

Falanga V (2000). Classifications for wound bed preparation and stimulation of chronic wounds. Wound repair and regeneration : official publication of the Wound Healing Society [and] the European Tissue Repair Society 8: 347-352.

Fazli M, Bjarnsholt T, Kirketerp-Moller K, Jorgensen A, Andersen CB, Givskov M et al (2011). Quantitative analysis of the cellular inflammatory response against biofilm bacteria in chronic

Page | 23 wounds. Wound repair and regeneration : official publication of the Wound Healing Society [and] the European Tissue Repair Society 19: 387-391.

Fernandez Guerrero ML, Goyenechea A, Verdejo C, Roblas RF, de Gorgolas M (2007). Enterococcal endocarditis on native and prosthetic valves: a review of clinical and prognostic factors with emphasis on hospital-acquired infections as a major determinant of outcome. Medicine (Baltimore) 86: 363-377.

Fisher K, Phillips C (2009). The ecology, epidemiology and virulence of Enterococcus. Microbiology 155: 1749-1757.

Flahaut S, Hartke A, Giard JC, Benachour A, Boutibonnes P, Auffray Y (1996). Relationship between stress response toward bile salts, acid and heat treatment in Enterococcus faecalis. FEMS microbiology letters 138: 49-54.

Flemming HC, Wingender J, Szewzyk U, Steinberg P, Rice SA, Kjelleberg S (2016). Biofilms: an emergent form of bacterial life. Nat Rev Microbiol 14: 563-575.

Garsin DA, Frank KL, Silanpaa J, Ausubel FM, Hartke A, Shankar N et al (2014). Pathogenesis and Models of Enterococcal Infection. In: Gilmore MS, Clewell DB, Ike Y, Shankar N (eds). Enterococci: From Commensals to Leading Causes of Drug Resistant Infection. Massachusetts Eye and Ear Infirmary: Boston.

Garzoni C, Kelley WL (2009). Staphylococcus aureus: new evidence for intracellular persistence. Trends Microbiol 17: 59-65.

Gentry-Weeks CR, Karkhoff-Schweizer R, Pikis A, Estay M, Keith JM (1999). Survival of Enterococcus faecalis in mouse peritoneal macrophages. Infect Immun 67: 2160-2165.

Giacometti A, Cirioni O, Schimizzi AM, Del Prete MS, Barchiesi F, D'Errico MM et al (2000). Epidemiology and microbiology of surgical wound infections. J Clin Microbiol 38: 918-922.

Gill SE, Parks WC (2008). Metalloproteinases and their inhibitors: regulators of wound healing. The international journal of biochemistry & cell biology 40: 1334-1347.

Gjodsbol K, Christensen JJ, Karlsmark T, Jorgensen B, Klein BM, Krogfelt KA (2006). Multiple bacterial species reside in chronic wounds: a longitudinal study. International wound journal 3: 225-231.

Grone A (2002). Keratinocytes and cytokines. Vet Immunol Immunopathol 88: 1-12.

Hall-Stoodley L, Costerton JW, Stoodley P (2004). Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2: 95-108.

Page | 24

Hashem YA, Amin HM, Essam TM, Yassin AS, Aziz RK (2017). Biofilm formation in enterococci: genotype-phenotype correlations and inhibition by vancomycin. Sci Rep 7: 5733.

Hidron AI, Edwards JR, Patel J, Horan TC, Sievert DM, Pollock DA et al (2008). NHSN annual update: antimicrobial-resistant pathogens associated with healthcare-associated infections: annual summary of data reported to the National Healthcare Safety Network at the Centers for Disease Control and Prevention, 2006-2007. Infect Control Hosp Epidemiol 29: 996-1011.

Hollenbeck BL, Rice LB (2012). Intrinsic and acquired resistance mechanisms in enterococcus. Virulence 3: 421-433.

Holmberg A, Rasmussen M (2016). Mature biofilms of Enterococcus faecalis and Enterococcus faecium are highly resistant to antibiotics. Diagn Microbiol Infect Dis 84: 19-21.

Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG (1992). CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol 13: 606-608.

Humphreys H (2009). Preventing surgical site infection. Where now? J Hosp Infect 73: 316-322.

Inoshima I, Inoshima N, Wilke GA, Powers ME, Frank KM, Wang Y et al (2011). A Staphylococcus aureus pore-forming toxin subverts the activity of ADAM10 to cause lethal infection in mice. Nature medicine 17: 1310-1314.

James GA, Swogger E, Wolcott R, Pulcini E, Secor P, Sestrich J et al (2008). Biofilms in chronic wounds. Wound Repair Regen 16: 37-44.

Janeway CA, Jr. (1989). Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harb Symp Quant Biol 54 Pt 1: 1-13.

Johnson JR, Clabots C, Hirt H, Waters C, Dunny G (2004). Enterococcal aggregation substance and binding substance are not major contributors to urinary tract colonization by Enterococcus faecalis in a mouse model of ascending unobstructed urinary tract infection. Infect Immun 72: 2445-2448.

Kelly M (1978). The quantitative and histological demonstration of pathogenic synergy between Escherichia coli and Bacteroides fragilis in guinea-pig wounds. Journal of medical microbiology 11: 513-523.

Kennedy AD, Wardenburg JB, Gardner DJ, Long D, Whitney AR, Braughton KR et al (2010). Targeting of alpha-hemolysin by active or passive immunization decreases severity of USA300 skin infection in a mouse model. Journal of Infectious Diseases 202: 1050-1058.

Kessler E, Safrin M, Olson JC, Ohman DE (1993). Secreted LasA of Pseudomonas aeruginosa is a staphylolytic protease. Journal of Biological Chemistry 268: 7503-7508.

Page | 25

Kierbel A, Gassama-Diagne A, Mostov K, Engel JN (2005). The phosphoinositol-3-kinase-protein kinase B/Akt pathway is critical for Pseudomonas aeruginosa strain PAK internalization. Mol Biol Cell 16: 2577-2585.

Kintarak S, Whawell SA, Speight PM, Packer S, Nair SP (2004). Internalization of Staphylococcus aureus by human keratinocytes. Infect Immun 72: 5668-5675.

Korgaonkar A, Trivedi U, Rumbaugh KP, Whiteley M (2013). Community surveillance enhances Pseudomonas aeruginosa virulence during polymicrobial infection. Proceedings of the National Academy of Sciences 110: 1059-1064.

Leaper DJ, van Goor H, Reilly J, Petrosillo N, Geiss HK, Torres AJ et al (2004). Surgical site infection - a European perspective of incidence and economic burden. International wound journal 1: 247-273.

Lebreton F, Riboulet-Bisson E, Serror P, Sanguinetti M, Posteraro B, Torelli R et al (2009). ace, Which encodes an adhesin in Enterococcus faecalis, is regulated by Ers and is involved in virulence. Infect Immun 77: 2832-2839.

Lina G, Piémont Y, Godail-Gamot F, Bes M, Peter M-O, Gauduchon V et al (1999). Involvement of Panton-Valentine leukocidin—producing Staphylococcus aureus in primary skin infections and pneumonia. Clinical Infectious Diseases 29: 1128-1132.

Lipinska U, Hermans K, Meulemans L, Dumitrescu O, Badiou C, Duchateau L et al (2011). Panton- Valentine leukocidin does play a role in the early stage of Staphylococcus aureus skin infections: a rabbit model. Plos one 6: e22864.

Magill SS, Hellinger W, Cohen J, Kay R, Bailey C, Boland B et al (2012). Prevalence of healthcare- associated infections in acute care hospitals in Jacksonville, Florida. Infect Control Hosp Epidemiol 33: 283-291.

Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR (1999a). Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 20: 250-278; quiz 279-280.

Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR (1999b). Guideline for Prevention of Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control 27: 97-132; quiz 133-134; discussion 196.

Mashburn LM, Jett AM, Akins DR, Whiteley M (2005). Staphylococcus aureus serves as an iron source for Pseudomonas aeruginosa during in vivo coculture. Journal of bacteriology 187: 554-566.

McCracken M, Wong A, Mitchell R, Gravel D, Conly J, Embil J et al (2013). Molecular epidemiology of vancomycin-resistant enterococcal bacteraemia: results from the Canadian Nosocomial Infection Surveillance Program, 1999–2009. Journal of Antimicrobial Chemotherapy.

Page | 26

Millan D, Chiriboga C, Patarroyo MA, Fontanilla MR (2013). Enterococcus faecalis internalization in human umbilical vein endothelial cells (HUVEC). Microb Pathog 57: 62-69.

Misic AM, Gardner SE, Grice EA (2014). The Wound Microbiome: Modern Approaches to Examining the Role of Microorganisms in Impaired Chronic Wound Healing. Adv Wound Care 3: 502-510.

Mitchell G, Séguin DL, Asselin A-E, Déziel E, Cantin AM, Frost EH et al (2010). Staphylococcus aureus sigma B-dependent emergence of small-colony variants and biofilm production following exposure to Pseudomonas aeruginosa 4-hydroxy-2-heptylquinoline-N-oxide. BMC microbiology 10: 33.

Moro ML, Morsillo F, Tangenti M, Mongardi M, Pirazzini MC, Ragni P (2005). Rates of surgical-site infection: an international comparison. Infect Control Hosp Epidemiol 26: 442-448.

Munoz-Price LS, Lolans K, Quinn JP (2005). Emergence of resistance to daptomycin during treatment of vancomycin-resistant Enterococcus faecalis infection. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 41: 565-566.

Nallapareddy SR, Singh KV, Sillanpaa J, Garsin DA, Hook M, Erlandsen SL et al (2006). Endocarditis and biofilm-associated pili of Enterococcus faecalis. J Clin Invest 116: 2799-2807.

Nallapareddy SR, Sillanpaa J, Mitchell J, Singh KV, Chowdhury SA, Weinstock GM et al (2011a). Conservation of Ebp-type pilus genes among Enterococci and demonstration of their role in adherence of Enterococcus faecalis to human platelets. Infect Immun 79: 2911-2920.

Nallapareddy SR, Singh KV, Sillanpää J, Zhao M, Murray BE (2011b). Relative Contributions of Ebp Pili and the Collagen Adhesin Ace to Host Extracellular Matrix Protein Adherence and Experimental Urinary Tract Infection by Enterococcus faecalis OG1RF. Infect Immun 79: 2901-2910.

Nielsen HV, Flores-Mireles AL, Kau AL, Kline KA, Pinkner JS, Neiers F et al (2013). Pilin and sortase residues critical for endocarditis- and biofilm-associated pilus biogenesis in Enterococcus faecalis. J Bacteriol 195: 4484-4495.

Olaru F, Jensen LE (2010). Chemokine expression by human keratinocyte cell lines after activation of Toll-like receptors (TLRs). Exp Dermatol 19: e314-e316.

Olmsted SB, Dunny GM, Erlandsen SL, Wells CL (1994). A plasmid-encoded surface protein on Enterococcus faecalis augments its internalization by cultured intestinal epithelial cells. J Infect Dis 170: 1549-1556.

Park J-H, Lee J-H, Cho MH, Herzberg M, Lee J (2012). Acceleration of protease effect on Staphylococcus aureus biofilm dispersal. FEMS microbiology letters 335: 31-38.

Page | 27

Pastar I, Nusbaum AG, Gil J, Patel SB, Chen J, Valdes J et al (2013). Interactions of Methicillin Resistant Staphylococcus aureus USA300 and Pseudomonas aeruginosa in Polymicrobial Wound Infection. PLoS ONE 8: e56846.

Qin X, Singh KV, Weinstock GM, Murray BE (2000). Effects of Enterococcus faecalis fsr genes on production of gelatinase and a serine protease and virulence. Infect Immun 68: 2579-2586.

Rajkumari N, Mathur P, Misra MC (2014). Soft Tissue and Wound Infections Due to Enterococcus spp. Among Hospitalized Trauma Patients in a Developing Country. J Glob Infect Dis 6: 189-193.

Romling U, Balsalobre C (2012). Biofilm infections, their resilience to therapy and innovative treatment strategies. J Intern Med 272: 541-561.

Rotstein OD, Pruett TL, Simmons RL (1985). Mechanisms of microbial synergy in polymicrobial surgical infections. Rev Infect Dis 7: 151-170.

Sabatino R, Di Cesare A, Pasquaroli S, Vignaroli C, Citterio B, Amiri M et al (2015). Adherence and intracellular survival within human macrophages of Enterococcus faecalis isolates from coastal marine sediment. Microbes Infect 17: 660-664.

Schierle CF, De la Garza M, Mustoe TA, Galiano RD (2009). Staphylococcal biofilms impair wound healing by delaying reepithelialization in a murine cutaneous wound model. Wound repair and regeneration 17: 354-359.

Schultz GS, Sibbald RG, Falanga V, Ayello EA, Dowsett C, Harding K et al (2003). Wound bed preparation: a systematic approach to wound management. Wound repair and regeneration : official publication of the Wound Healing Society [and] the European Tissue Repair Society 11 Suppl 1: S1-28.

Sen CK, Gordillo GM, Roy S, Kirsner R, Lambert L, Hunt TK et al (2009). Human skin wounds: a major and snowballing threat to public health and the economy. Wound Repair Regen 17: 763-771.

Seth AK, Geringer MR, Hong SJ, Leung KP, Mustoe TA, Galiano RD (2012). In vivo modeling of biofilm- infected wounds: a review. journal of surgical research 178: 330-338.

Shankar N, Lockatell CV, Baghdayan AS, Drachenberg C, Gilmore MS, Johnson DE (2001). Role of Enterococcus faecalis surface protein Esp in the pathogenesis of ascending urinary tract infection. Infect Immun 69: 4366-4372.

Shogan BD, Belogortseva N, Luong PM, Zaborin A, Lax S, Bethel C et al (2015). Collagen degradation and MMP9 activation by Enterococcus faecalis contribute to intestinal anastomotic leak. Sci Transl Med 7: 286ra268.

Sievert DM, Ricks P, Edwards JR, Schneider A, Patel J, Srinivasan A et al (2013). Antimicrobial- resistant pathogens associated with healthcare-associated infections: summary of data reported to

Page | 28 the National Healthcare Safety Network at the Centers for Disease Control and Prevention, 2009- 2010. Infect Control Hosp Epidemiol 34: 1-14.

Soong G, Paulino F, Wachtel S, Parker D, Wickersham M, Zhang D et al (2015). Methicillin-resistant Staphylococcus aureus adaptation to human keratinocytes. mBio 6.

Sprenger A, Weber S, Zarai M, Engelke R, Nascimento JM, Gretzmeier C et al (2013). Consistency of the Proteome in Primary Human Keratinocytes With Respect to Gender, Age, and Skin Localization. Molecular & Cellular Proteomics : MCP 12: 2509-2521.

Stojadinovic O, Brem H, Vouthounis C, Lee B, Fallon J, Stallcup M et al (2005). Molecular pathogenesis of chronic wounds: the role of beta-catenin and c-myc in the inhibition of epithelialization and wound healing. The American journal of pathology 167: 59-69.

Süßmuth SD, Muscholl-Silberhorn A, Wirth R, Susa M, Marre R, Rozdzinski E (2000). Aggregation substance promotes adherence, phagocytosis, and intracellular survival of Enterococcus faecalis within human macrophages and suppresses respiratory burst. Infection and immunity 68: 4900- 4906.

Thurlow LR, Thomas VC, Narayanan S, Olson S, Fleming SD, Hancock LE (2010). Gelatinase contributes to the pathogenesis of endocarditis caused by Enterococcus faecalis. Infect Immun 78: 4936-4943.

Tien BYQ, Goh HMS, Chong KKL, Bhaduri-Tagore S, Holec S, Dress R et al (2017). Enterococcus faecalis promotes innate immune suppression and polymicrobial catheter-associated urinary tract infection. Infect Immun.

Torelli R, Cacaci M, Papi M, Paroni Sterbini F, Martini C, Posteraro B et al (2017). Different effects of matrix degrading enzymes towards biofilms formed by E. faecalis and E. faecium clinical isolates. Colloids and Surfaces B: Biointerfaces 158: 349-355.

Usui ML, Mansbridge JN, Carter WG, Fujita M, Olerud JE (2008). Keratinocyte migration, proliferation, and differentiation in chronic ulcers from patients with diabetes and normal wounds. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 56: 687-696.

Waters CM, Hirt H, McCormick JK, Schlievert PM, Wells CL, Dunny GM (2004). An amino-terminal domain of Enterococcus faecalis aggregation substance is required for aggregation, bacterial internalization by epithelial cells and binding to lipoteichoic acid. Mol Microbiol 52: 1159-1171.

Watters C, DeLeon K, Trivedi U, Griswold JA, Lyte M, Hampel KJ et al (2013). Pseudomonas aeruginosa biofilms perturb wound resolution and antibiotic tolerance in diabetic mice. Medical microbiology and immunology 202: 131-141.

Page | 29

Wells CL, Jechorek RP, Maddaus MA, Simmons RL (1988). Effects of clindamycin and metronidazole on the intestinal colonization and translocation of enterococci in mice. Antimicrob Agents Chemother 32: 1769-1775.

Wells CL, Jechorek RP, Erlandsen SL (1990). Evidence for the translocation of Enterococcus faecalis across the mouse intestinal tract. J Infect Dis 162: 82-90.

Zeng J, Teng F, Murray BE (2005). Gelatinase is important for translocation of Enterococcus faecalis across polarized human enterocyte-like T84 cells. Infect Immun 73: 1606-1612.

Zhao G, Hochwalt PC, Usui ML, Underwood RA, Singh PK, James GA et al (2010). Delayed wound healing in diabetic (db/db) mice with Pseudomonas aeruginosa biofilm challenge: a model for the study of chronic wounds. Wound repair and regeneration 18: 467-477.

Zou J, Shankar N (2016). The opportunistic pathogen Enterococcus faecalis resists phagosome acidification and autophagy to promote intracellular survival in macrophages. Cell Microbiol 18: 831- 843.

Page | 30

CHAPTER 2

Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and

Polymicrobial Infection

2.1 Introduction and motivation:

Many, if not most, infections are not monomicrobial, either in origin or in manifestation

(Bakaletz 2004, Brogden et al. 2005, Nelson et al. 2012). We now appreciate that many

infections either initiate from a polymicrobial inoculation into the infection site, as is the

case of the introduction of a bolus of gastrointestinal (GI) tract microbes into the urinary

tract prior to urinary tract infection (UTI), or they persist as a polymicrobial community as

in the case in catheter-associated urinary tract infections (CAUTI) and in many wound

infections. E. faecalis is frequently part of a polymicrobial community that usually includes

E. coli in catheter-associated urinary tract infections (CAUTIs) (Flores-Mireles et al. 2015).

It is also one of the most frequently isolated organisms in surgical site infections and wound

infections (Dowd et al. 2008, Giacometti et al. 2000).

Both microbes and the human host require iron. Iron is an essential nutrient that is required

for many cellular processes, including respiration and DNA replication, and is an ideal

redox catalyst due to its ability to cycle between the ferrous or ferric state. Iron levels within

the human host must be tightly regulated, due to the generation of reactive oxygen species

which catalyze the Fenton reaction to generate hydroxyl radicals which can damage DNA

(Inoue and Kawanishi 1987), lipids and proteins in the host (Bogdan et al. 2016). In

addition, this regulation serves as a form of nutritional immunity against invading or

opportunistic pathogens. The availability of iron in the human host is therefore very limited

and most it is sequestered intracellularly in erythrocytes where it is complexed with

Page | 31 haemoglobin. Any remaining free extracellular iron is bound to transferrin, albumin, citrate or amino acids in the blood plasma, further limiting bioavailability of iron (Miethke and

Marahiel 2007).

During infection, iron levels are further limited through several mechanisms. Hypoferremia can be induced by the release of hepcidin from the liver which reduces the transmembrane iron transporter ferroportin to limit iron egress. This release can be triggered by the presence of pro-inflammatory cytokines, TLR activation or the induction of the unfolded protein response (Drakesmith and Prentice 2012, Nairz et al. 2010, Weiss 2005). At the local infection site, immune infiltrates consisting of neutrophils and macrophages also synthesize hepcidin, siderocalin/lipocalin-2, further modulating extracellular iron availability at the infection site (Peyssonnaux et al. 2006). Neutrophils also carry lactoferrin in their secondary granules, which can bind iron, even at low pH (Masson et al. 1969). Pro- inflammatory cytokines such as interleukin-1 (IL-1), interleukin-6 (IL-6), interferon- gamma (IFN-γ) and tumor necrosis factor-alpha (TNF-α) can further reduce iron levels, via repression of the divalent metal transporter 1 (DMT1) mediated iron absorption and activation of ferritin synthesis (Cassat and Skaar 2013).

Given how tightly iron is regulated in the human host, infecting microbes must therefore possess mechanisms to overcome iron limitation in the human host for successful colonization and infection to take place. Commonly used strategies include the production of siderophores, which are small, high affinity iron chelators that can bind iron with an association constant of >1050, outcompeting transferrin which has only an association constant of 1036 (Holden and Bachman 2015). There are three major classes of siderophores, catecholates, hydroxamates, and carboxylates and they are classified

Page | 32 according to their chemical groups (Crumbliss and Harrington 2009, Sandy and Butler

2009). There is also another class of siderophores, the mixed family, and they have two or more elements from the three major siderophore classes (Crumbliss and Harrington 2009).

Figure 2.1: Various siderophores from the four major structural families. The binding moieties are depicted in colour. Reprinted with permission from (Crumbliss and Harrington 2009).

Siderophores are synthesized within the cytoplasm before being exported to the extracellular space. These export systems usually consist of cognate membrane-receptor and inner membrane uptake system (Garenaux et al. 2011). In the extracellular space, iron- bound siderophores are recognized by the TonB-ExbB-ExbD transduction system, which transports the ligand into the periplasmic space (Higgs et al. 2002). From the periplasm, specific ABC-transporters transport the iron-bound siderophores to the inner membrane.

During the transport from the outer membrane to the inner membrane, chaperones can also be involved. In the case of ferric enterobactin transport, it requires the FepD chaperone

(Chenault and Earhart 1991). Upon entry into the periplasm, the siderophore bound iron can be released, either though reduction of ferric to ferrous iron which destabilizes the

Page | 33 complex or through enzymatic degradation (Miethke and Marahiel 2007). Other strategies that infecting microbes utilize to fulfil their metal requirements include the use of alternative metals such as manganese, utilization of heme uptake systems and direct acquisition of iron from transferrin and lactoferrin (Cassat and Skaar 2013).

E. coli possesses several high affinity iron acquisition systems and uropathogenic strains of E. coli typically encode at least three siderophores. Iron acquisition systems are important for the virulence of uropathogenic Escherichia coli (UPEC) strains and studies have found over 14 gene clusters that encode uptake systems, receptors for heme, ferrous iron and ferric citrate as well as siderophore biosynthesis (Garcia et al. 2011, Hagan and

Mobley 2009, Johnson et al. 2005, Torres et al. 2001). These siderophores include: enterobactin, salmochelin, aerobactin and yersiniabactin (Searle et al. 2015,

Subashchandrabose et al. 2013). UPEC also possesses two ferrous transporters FeoB and

EfeUOB. The FeoB pathway is activated under anaerobic or microaerophilic conditions and transports ferrous iron into the cytoplasm from the inner membrane via a GTP dependent manner (Cao et al. 2007). EfeUOB, on the other hand, is induced under aerobic conditions together with low pH and low iron (Cao et al. 2007). While it is also involved in ferrous iron uptake, it also has the capability to oxidize ferrous iron, and is thought to be able to mediate uptake of ferric iron as well (Rajasekaran et al. 2010).

Interestingly, past studies have shown that lactic acid bacteria, which includes E. faecalis, is remarkably tolerant of low iron, where it can grow and thrive in conditions where iron is almost absent in vitro (Archibald 1986, Bruyneel et al. 1989, Weinberg 1997). In E. faecalis, iron is substituted by manganese as a in my cellular enzymes and this includes enzymes of the DNA biosynthetic pathway and energy generation (Colomer-

Page | 34

Winter et al. 2017, Kehres and Maguire 2003). Manganese is also important for the oxidative stress response in E. faecalis (Colomer-Winter et al. 2017, Kehres and Maguire

2003). Though studies have detected compounds with broad chelation capacity in E. faecalis, they have not been characterized in detail (Lisiecki and Mikucki 2004, Lisiecki and Mikucki 2006). Moreover, iron acquisition systems for siderophore transport or heme iron transport has not been functionally characterized in E. faecalis (Miethke and Marahiel

2007, Neilands 1995). To, date, no siderophores have been identified in E. faecalis. This is in stark contrast to E. coli, which possesses several high affinity iron acquisition systems and transporters.

Studies have shown that iron availability is restricted during infection, and this also applies to wound infections (Nairz et al. 2010, Peyssonnaux et al. 2006, Weiss 2005). Lipocalin-2 is secreted at the wound edge by neutrophils to aid wound healing and cell migration (Miao et al. 2014). Liopcalin-2 can also bind and sequester enterobactin, which is a siderophore that E. coli produces (Flo et al. 2004, Goetz et al. 2002). Despite this, E. coli is still frequently co-isolated together with E. faecalis in many iron limited niches, suggesting a degree of dependence on one another. We therefore hypothesized that E. faecalis is helping

E. coli to overcome iron limitation during infection, allowing E. coli to overcome the host nutritional immunity. To test this hypothesis, we developed an in-vitro macrocolony and biofilm assay where iron is chelated, mimicking iron limitation in the host.

We have previously shown that co-culture of E. coli and E. faecalis under iron limitation resulted in an enhanced microcolony phenotype, which was not seen in the single species control (Figure 2.2A, B, C) (Keogh et al. 2016). Further, we have also shown that this enhanced growth phenotype was also positively correlated with an increased in E. coli CFU

Page | 35 counts under co-culture conditions but not in mono-culture (Figure 2.2D, E) (Keogh et al.

2016).

Figure 2.2: E. faecalis promotes E. coli biofilm growth under iron limitation. (A and B) E. coli, E. faecalis, and Mix (1Ec:1Ef) biofilm (A), and Mix (1Ec:4Ef), and Mix (1Ec:19Ef) biofilm (B) were grown in 22D- supplemented TSBG. (C) Supplementation with 50 μM FeCl3 restored biomass to iron-replete levels. (A–C) Representative macrocolony images at 120 hr; scale bars represent 1 cm. (D and E) Time course enumeration of E. coli (D) and E. faecalis (E), respectively, from macrocolonies with single or mixed inoculums under iron limitation. (F) E. coli, E. faecalis, and Mix (1Ec:19Ef) biofilm biomass in TSBG supplemented with 22D. Statistical significance was determined by two-way ANOVA with Tukey’s test for multiple comparisons; ∗∗∗∗p < 0.0001. (G) Enumeration of planktonic growth at 120 hr for E. coli, E. faecalis, and Mix (1Ec:19Ef) in TSBG with 0.1, 0.2, and 0.7 mM 22D. (D–G) n = 3 with three technical replicates; error bars represent SD from the mean. Reprinted with permission from (Keogh et al. 2016).

Page | 36

We have also demonstrated that this growth enhancement was due to a secreted factor from

E. faecalis, using a proximity assay, where E. coli and E. faecalis macrocolonies were grown in increasing distances apart from one another (Figure 2.3A, B) (Keogh et al. 2016).

We then performed RNA-sequencing on the proximal and distal colonies, but this experiment did not yield any useful information, regarding the secreted factor.

Figure 2.3: E. faecalis proximity causes growth changes in E. coli. (A) Proximity assay of E. coli and E. faecalis (120 hr) macrocolonies in TSBG supplemented with 22D, inoculated from 1 cm (proximal) to 5 cm (distal). Scale bar represents 1 cm. Black brackets indicate proximal and distal sample sites. (B) Enumeration at 24 and 120 hr for E. coli and E. faecalis at both proximal and distal sites; n = 3 with three technical replicates; error bars represent SD from the mean. Reprinted with permission from (Keogh et al. 2016).

Since this enhanced E. coli growth was only observed under iron depleted conditions, we hypothesized that the secreted factor might be influencing metal uptake or siderophore production. To test this hypothesis, we performed a targeted screen of E. coli mutants that were deficient for metal uptake or siderophore synthesis. Following this, we screened a saturated E. faecalis transposon library to search for mutants that are important for E. coli growth under iron limiting conditions. Our results demonstrate the E. faecalis secretes a metabolite that E. coli can utilize to kickstart siderophore production, enabling it to proliferate even under iron limitation. Further, we demonstrate that this soluble, diffusible factor is important for siderophore synthesis in E. coli.

Page | 37

2.2 Results:

2.2.1 Enhanced E. coli growth by E. faecalis is dependent on enterobactin siderophore

production

E. coli UTI89 has for three siderophore biosynthetic operons, which are yersiniabactin,

enterobactin, and salmochelin. These three siderophores play an important role in

colonization and are important virulence factors. We thus hypothesized that E. faecalis

could possibly be influencing or inducing siderophore synthesis under iron deplete

conditions. To test this, we grew deletion mutants of each the E. coli UTI89 siderophores

in increasing distances using the agar plate biofilm assay.

Our results showed that the ΔentB mutant, which was unable to produce enterobactin, was

attenuated in growth and does not exhibit the enhanced overgrowth phenotype when grown

near E. faecalis (Figure 2.4). However, the ΔiroA and the ΔybtS mutants which are

deficient for salmochelin and yersiniabactin production respectively, were dispensable for

E. faecalis-mediated E. coli growth enhancement (Figure 2.4). The ΔentBΔybtS double

mutant also lacked the enhanced overgrowth phenotype (Figure 2.4). Collectively, our

results suggest that enterobactin, but not salmochelin or yersiniabactin, is important for E.

faecalis-mediated enhancement of E. coli growth under iron limitation.

Page | 38

Figure 2.4: Enterobactin deletion mutants fail to exhibit E. faecalis-driven growth promotion. Agar plate biofilm assays of (A) E. coli UTI89ΔentB, (B) E. coli UTI89ΔiroA, (C) E. coli UTI89ΔybtS, and (D) E. coli UTI89ΔentBΔybtS and E. faecalis (right) macrocolonies at 120 hrs in TSBG with 22D. Reprinted with permission from (Keogh et al. 2016).

2.2.2 E. coli ferrous uptake systems are not important for E. faecalis-driven E. coli

growth promotion

We previously observed differential expression of the two iron transporters, feoB and

efeUOB in RNA-Seq of E. coli macrocolonies. Comparison of the proximal and distal E.

coli macrocolonies in relation to E. faecalis macrocolonies indicated that feoB was

downregulated while efeUOB upregulated under iron restricted conditions. Given that

efeUOB is induced only under low iron and low pH conditions (Cao et al. 2007), and that

E. faecalis is a lactic acid-producing bacterium, we hypothesized that E. faecalis could

possibly be affecting the pH of the local environment, increasing iron solubility and

inducing the ferrous uptake systems in E. coli. To test this hypothesis, we overlaid

macrocolonies with an agar solution of bromothymol blue. Our results show that the local

pH does indeed change, with E. coli macrocolonies in the acidic range of pH 5.5 at 7 hrs

and 48 hrs. E. coli macrocolonies near to E. faecalis were in the alkaline range of pH 8 at

72 hrs (Figure 2.5A), due to the metabolic by-product accumulation as a result of increased

growth.

Page | 39

Since our results demonstrated that the local pH does indeed change, we hypothesized that the ferrous uptake systems might be important for the E. faecalis-driven E. coli growth promotion and constructed single and double in-frame deletion mutants of the permease components of each system, feoB and efeU. However, we did not observe any attenuation in E. coli feoB and efeU macrocolonies that were grown in close proximity to E. faecalis macrocolonies, suggesting that the E. coli response mechanism is independent of the ferrous iron uptake systems (Figure 2.5B).

Page | 40

Figure 2.5: Ferrous uptakes systems are dispensable for E. faecalis-driven E. coli growth promotion despite local pH changes. (A) Proximity assay of E. coli and E. faecalis macrocolonies in TSBG supplemented with 22D at 7, 48 and 72hrs. Bromothymol Blue pH indicator in 1.5% agar solution overlay prior to image capture. (B) Proximity assay of E. coli ferrous uptake mutants at 120hrs. Reprinted with permission from (Keogh et al. 2016).

2.2.3 Co-culture of E. faecalis and E. coli results in enhanced biofilm/biomass under

iron deplete conditions

While the macrocolony assay was useful in examining information about the mixed species

interactions between E. coli and E. faecalis, it was not amenable to a large-scale screen. To

circumvent this, we utilized a modified 96 well microtiter plate biofilm assay. This assay

was chosen as it is amenable to high throughput screens, as compared to the agar plate

biofilm assay. Given our previous findings using the macrocolony plate assay, we reasoned

that co-culture of both species using a microtiter plate biofilm assay would result in similar

results, where there would be enhanced biofilm/biomass for the co-culture instead of the

monoculture single species strains. Our results show that co-culture both E. coli and E.

faecalis together resulted in enhanced biofilm/biomass in a reproducible manner, similar to

that of the agar plate biofilm assay (Figure 2.6).

C ry s ta l V io le t A s s a y

1 :1 9 M ix

*

*

*

*

* *

E . fa e c a lis * *

E . c o li

M e d ia

.0 .5 .0 .5 .0 0 0 1 1 2 A b s o r b a n c e 5 9 5 n m

Figure 2.6: Optical density (OD595) of single species E. faecalis (OG1RF), single species UPEC (UTI89) and mixed species (OG1RF + UTI89) biofilms at the 120-hour post incubation timepoint. Mixed species cultures have a significant increase in biofilm production as compared to UPEC single species (p<0.0005, two-tailed

Page | 41

T-test). Data obtained represent 4 biological replicates with 4 technical replicates each. Error bars represent the standard deviation of means of each biological replicate. Reprinted with permission from (Keogh et al. 2016).

Interestingly, this enhanced biofilm/biomass phenotype was not observed when planktonic

cultures of E. coli and E. faecalis were used (data not shown), suggesting specificity for

biofilm cultures. However, the same experimental condition, without iron chelation,

resulted in significantly higher E. coli absorbance readings (data not shown).

Unsurprisingly, E. faecalis absorbance readings were similar even without iron chelation

(data not shown). This assay demonstrates the restriction of E. coli biofilm growth while

showing the uninhibited E. faecalis biofilm growth. E. coli remained sensitive to iron

chelation in both biofilm and planktonic conditions where they display inhibited growth.

The significant differences in biomass between the two-single species and the 1Ec:19Ef

mixed species can therefore be exploited in the transposon library screen.

2.2.4 E. faecalis Transposon library screen and validation

An E. faecalis mariner transposon library (Kristich et al. 2008) which contains a total of

14,978 mutants was subjected to a high-throughput biofilm primary screen for reduced

interspecies biofilm biomass in iron-depleted media of 0.1mM 2,2D. An overview of the

entire screening process is shown in Figure 2.13, in the Materials and Methods section.

We expected a significant reduction in the biofilm/biomass readout, if the mutated gene

was important for 1) biofilm formation 2) normal growth and metabolism 3) synthesizing

the soluble, diffusible factor. We defined significant reduction as ≥50% decrease in the

biofilm/biomass readout as compared to the wildtype 1Ec:19Ef mixed species controls.

Page | 42

Following primary screening, 335 mutants are found to have significant reduction in biomass as compared to wild-type controls. An example of the data, where 11 of the 27 transposon mutants displayed a significant reduction in biofilm formation following 5 days of incubation under iron-limiting conditions and crystal violet staining is shown in Figure

2.7. As controls, we inoculated single species E. coli and E. faecalis as well as mixed species E. coli and E. faecalis, since we already know from previous results that co-culture of both will result in a drastic increase in biofilm/biomass over the single species controls.

M e d ia E . c o li E . fa e c a lis M ix e d S p e c ie s (1 EC :1 9 EF ) 1 9 .2 I1 B 2 3 4 .2 I2 C 0 9 3 4 .2 I2 B 0 5 2 1 .2 I2 A2 2 4 .2 I2 C 1 2 8 .2 I1 D 2 2 5 .2 A2 E 1 0 1 9 .2 I1 B 8 1 9 .2 I1 B 1 2 9 .2 I2 A1 2 2 1 .2 I2 A1 1 6 .2 A2 G 0 9 1 3 .2 I1 G 0 4 6 .2 A2 A1 1 1 3 .2 I1 B 1 0 1 .2 A2 A8 8 .2 A1 A9 1 1 .2 A1 C 3 1 4 .2 A1 A7 1 3 .2 A2 H 3 1 .2 A1 D 7 1 6 .2 I1 C 1 0 1 4 .2 A1 G 2 3 8 .2 I2 A1 2 .2 A1 C 2 2 1 .2 A2 E 3 2 1 .2 A2 H 5 0 2 0 4 0 6 0 8 0 1 0 0 B io film F o rm a tio n (% )

Figure 2.7: Primary screen of E. faecalis transposon mutants with decreased biofilm/biomass during co- culture with E. coli under iron limiting conditions. Relative expression (%) of interspecies biofilm formation among E. faecalis transposon mutants and E. coli were compared to WT E. faecalis and E. coli mixed species

(1EC:19EF ) biofilm formation, set at 100%. Error bars represent the standard deviation of mean.

This screen was then repeated for all 335 mutants that exhibited significant decrease in biofilm/biomass readout. The validation assay was performed using 2 independent biological replicates with 6 technical replicates to rule out any false-positives. 161 mutants with poor inter-species biofilm growth remained after primary validation, of which 66 were

Page | 43

selected. These mutants were grouped according to 3 classes: 1) Class 1, which are mutants

known to be defective in biofilm formation in normal media from previous work 2) Class

2, which are biofilm-defective mutants with transposon insertions in genes unique to this

study and 3) Class 3, which are mutants with transposon insertions in intergenic regions

(Table 2.1).

Category 1) Genes previously characterized to contribute to E. faecalis single species biofilm formation in iron-replete media Locus Tag Gene Product OG1RF_10472 carbamate kinase OG1RF_10716 hemolysin A OG1RF_10716 hemolysin A OG1RF_10717 arginine repressor OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10868 M protein trans-acting positive regulator OG1RF_10869 von Willebrand factor type A domain-containing protein OG1RF_10869 von Willebrand factor type A domain-containing protein OG1RF_10870 cell wall surface anchor family protein OG1RF_10870 cell wall surface anchor family protein OG1RF_10870 cell wall surface anchor family protein OG1RF_10871 cell wall surface anchor family protein OG1RF_10956 dihydrodipicolinate synthase OG1RF_11076 heat-inducible transcription repressor HrcA OG1RF_11080 chaperone DnaJ OG1RF_11526 gelatinase OG1RF_11526 gelatinase OG1RF_11527 sensor histidine kinase FsrC OG1RF_11527 sensor histidine kinase FsrC OG1RF_11528 FsrB protein OG1RF_11528 FsrB protein OG1RF_11529 FsrA response regulator

Category 2) Transposon insertions in genes unique to this study Locus Tag Gene Product OG1RF_10103 UIT3 family protein OG1RF_10187 LIVCS family branched chain amino acid:cation symporter

Page | 44

OG1RF_10301 efflux transporter OG1RF_10308 Kline Team Sequencing: MerR family transcriptional regulator OG1RF_10350 transcriptional regulator>>>>>>> upstream of glucosamine-6-phosphate deaminase, nagB OG1RF_10506 hypothetical protein OG1RF_10620 ABC superfamily ATP binding cassette transporter, ABC protein OG1RF_10621 amino acid ABC superfamily ATP binding cassette transporter, membrane protein OG1RF_11138 oligo-1,6-glucosidase OG1RF_11674 hypothetical protein OG1RF_11802 Kline Team Sequencing: Hypothetical protein OG1RF_12034 phosphoglycerate mutase OG1RF_12417 hypothetical protein Unsequenced: Plasmid lost Unsequenced: Plasmid lost Unsequenced: Plasmid lost Unsequenced: Plasmid lost Unsequenced: Plasmid lost Kline Team Sequencing: Plasmid still remains Kline Team Sequencing: Plasmid still remains Kline Team Sequencing: Insertion site not found Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains Unsequenced: Plasmid still remains

Category 3) Transposon insertions in intergenic regions Locus Tag Gene Product M protein trans-acting positive regulator > TA> von Willebrand factor type A domain- NA containing protein NA LuxS: s-ribosylhomocysteine asd:aspartate-semialdehyde dehydrogenase NA dnaK > TA >Hypothetical protein NA GelE:coccolysin < TA TA> Hypothetical protein Table 2.1: Mutants identified following primary validation of transposon library screen.

Page | 45

2.2.5 E. faecalis Transposon library single species secondary validation

Since any single mutant in this list of 66 mutants could be defective for the 1) normal

growth, 2) biofilm formation, or 3) synthesizing the soluble, diffusible factor, we next

performed single species experiments. To differentiate between these possibilities and to

eliminate any mutants that had general growth defects, we performed single-species

planktonic and biofilm growth assays of E. faecalis transposon mutants in iron-deplete and

iron-replete media.

We first subjected each of the mutants to a growth kinetics assay under iron deplete and

iron replete conditions. Here, poor planktonic growth was defined by the following: 1)

delayed exponential phase or 2) lower stationary phase. We reasoned that if growth was

attenuated, it would in turn impact the overall biofilm/biomass during co-culture with E.

coli. Indeed, some mutants had reduced absorbance readings as compared to wildtype E.

faecalis under iron deplete or replete conditions (Figure 2.8). These mutants were thus

eliminated from follow-up since any defects in their growth kinetics would affect their

biofilm/biomass readout during co-culture with E. coli.

A B 0 .8 0 .8

O G 1 R F O G 1 R F

0 .6 0 .6 )

2 .2 A 1 C 9 ) 8 .2 I2 F 4

m

m n

M e d ia B la n k n M e d ia B la n k

(

(

0

0 .4 0 0 .4

0

0

6

6

D

D

O O 0 .2 0 .2

0 .0 0 .0

0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 1 1 1 1 1 1 1 1 1 1

T im e ( H r s ) T im e ( H r s )

Figure 2.8: Extracted data of E. faecalis transposon mutants with impaired growth kinetics following 14hrs incubation at 37 oC. Mutants are deemed to have growth defects if they have (A) delayed exponential phase and/or (B) lower stationary phase relative to the growth profile of the parental E. faecalis OG1RF strain in iron replete or deplete media.

Page | 46

We next subjected the remaining mutants to single species biofilm assays. Similarly, defective biofilm formation in mutants was defined by a reduced absorbance reading at

OD595 nm as compared to wildtype E. faecalis under iron deplete and replete conditions

(Figure 2.9). Mutants exhibiting pronounced defects in either or both planktonic growth and biofilm profile were indicative of false positives and were removed from follow-up.

A B Iro n R e p le te M e d ia Iro n D e p le te M e d ia

O G 1 R F O G 1 R F

8 .2 I1 D 0 2 **** 8 .2 I1 D 0 2 ***

3 4 .2 I2 C 0 9 ** 3 4 .2 I2 C 0 9 *

3 8 .2 I2 A 0 1 **** 3 8 .2 I2 A 0 1 **

.0 .5 .0 .0 .5 .0 0 0 1 0 0 1 A b s o rb a n c e 5 9 5 n m A b s o rb a n c e 5 9 5 n m

Figure 2.9: Extracted data of E. faecalis transposon mutants with impaired biofilm formation following 24 hrs incubation at 37 oC. Mutants are deemed to have growth defects if they exhibit significant reduction in biofilm/biomass under (A) iron replete or (B) iron deplete conditions.

The remaining E. faecalis transposon mutants all exhibited similar levels of fitness in both the growth kinetics and biofilm/biomass as compared to the wild type controls, demonstrated that the reduced biofilm/biomass during co-culture was due to defects in interspecies interactions and not the result of single species growth or biofilm defects. A total of 9 E. faecalis mutants remained, out of which 5 were already sequenced and carried insertions genes of known functions. The insertional sites of the remaining 4 mutants were not mapped and we next performed whole-genome sequencing. Following sequencing, we eliminated 3 additional mutants, because the insertional sites could not be mapped. The final 6 transposon mutants are listed in Table 2.2.

Page | 47

Gene Gene Locus Description Genome Position

arcD OG1RF_10103 UIT3 family protein, antiporter 116023 LIVCS family branched-chain amino brnQ OG1RF_10187 184362 acid cation symporter – OG1RF_10350 transcriptional regulator 362782 ABC superfamily ATP binding – OG1RF_10620 657774 cassette transporter, ABC protein malL OG1RF_11138 Oligo-1,6-glucosidase 1185608 MerR family transcriptional merR OG1RF_10308 323110 regulator Table 2.2: Final list of transposon mutants following library screening and validation. Reprinted with permission from (Keogh et al. 2016).

2.2.6 L-Ornithine, but Not Branched-Chain Amino Acids, Augments E. coli Biofilm

Formation under Iron Limitation

We hypothesized that the E. faecalis-driven E. coli growth promotion was a result of

systems involved in signal production or mechanisms of signal exchange to E. coli. Of the

6 mutants identified in the transposon screen, we considered ArcD (OG1RF_10103) as the

primary candidates because ArcD was the sole system that facilitates export of L-ornithine,

making it a compelling candidate for signal exchange between E. coli. Previous work has

shown that ArcD is a transporter, where it forms part of the arginine deiminase pathway in

E. faecalis, and is responsible for importing L-arginine and exporting L-ornithine

(Barcelona-Andres et al. 2002). Since ArcD was the sole system the facilitates export of a

substrate, we hypothesized that L-ornithine might be involved in this signal exchange. To

test this hypothesis, we performed single-species E. coli biofilm assays under iron-limited

conditions and supplemented with L-ornithine in a concentration range up to 3.12 mM. We

observed significantly increased E. coli biofilm formation at 120 hrs at concentrations less

than 1.5 mM L-ornithine (Figure 2.10A). In contrast, L-ornithine did not augment biofilm

formation by the E. coli enterobactin mutant, suggesting that utilization of L-ornithine was

linked to the biosynthesis of the enterobactin siderophore (Figure 2.10B).

Page | 48

Figure 2.10: L-Ornithine supplementation under iron limiting conditions enhances wild-type E. coli biofilm but not E. coli UTI89ΔentB. (A) Wild-type E. coli UTI89 (B) UTI89ΔentB biofilm/biomass at 120 hrs in TSBG supplemented with 22D. L-ornithine supplementation is indicated by the black bar with concentration indicated above. Representative data from three independent experiments are shown, where the trend is consistent among all. Error bars represent SD from the mean. Statistical significance of the L-ornithine supplemented samples compared to the non-supplemented E. coli control was determined by two-way ANOVA with Tukey’s test for multiple comparisons, ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001. Reprinted with permission from (Keogh et al. 2016).

Sequence homology and alignment analysis of the second transporter system

(OG1RF_10187) from the transposon library screen predicted that putative substrates for this MFS symporter were branched-chain amino acids. Symporters generally function by an inward-directed polarity facilitated by membrane potential; however, in the presence of the substrate on the trans side of the membrane, this polarity can be reversed (Pao et al.

1998). To determine whether branched-chain amino acids were involved in signal exchange, we supplemented E. coli biofilm assays with the branched-chain amino acids valine, leucine, and isoleucine and analyzed biofilm/biomass accumulation at 120 hrs. We did not observe increased biomass across a concentration range for any of these substrates

(Figure 2.11).

Page | 49

Figure 2.11: Branched-chain amino acid supplementation does not enhance E. coli biofilm/biomass accumulation at 120hrs in TSBG supplemented with 22D. Amino acid supplementation is indicated by the black bar with concentration indicated above. Representative data from three independent experiments is shown, where the trend is consistent among all, error bars represent standard deviation from the mean. Statistical significance was determined by Two-Way ANOVA with Tukey’s test for multiple comparisons, *P≤0.05, **P≤0.01. Reprinted with permission from (Keogh et al. 2016).

Page | 50

2.3 Discussions

Determining the etiologic agents of infection have long relied on isolation of culturable

bacteria from the site of infection, often focusing on the predominant isolated organism

during the height of infection. The widespread application of next-generation sequencing

technologies to the study of infectious disease has resulted in a greater appreciation that

many infections are polymicrobial in nature, both in time and space, rendering inadequate

the application of Koch’s postulates to some polymicrobial infections, especially those

involving non-culturable organisms. Moreover, our growing understanding of the delicate

interplay between the native microbiome, invading pathogens, and the host response to

each, has given rise to many new questions about the fundamental basis of bacterial

pathogenesis. In our case, E. coli and E. faecalis are frequently co-isolated together in

CAUTIs and in wound infections, but no studies have characterized the consequence and/or

the significance of this clinical presentation.

In this study, we established that co-culture of E. faecalis and E. coli resulted in growth

augmentation, one that is dependent on proximity to E. faecalis. We also determined that a

soluble, diffusible, secreted factor from E. faecalis was important for the growth

augmentation, and appears to dependent on enterobactin but not on the ferrous uptake

systems. Our findings demonstrate that E. faecalis is a major influence on E. coli growth

in biofilms, when E. coli numbers are initially low and experiencing iron limitation. Using

a murine wound model later in this study, performed by a postdoc in the lab DK, we also

demonstrated that co-infection of E. faecalis augments virulence of E. coli. However, when

the E. coli enterobactin mutant (ΔentB) was co-infected with E. faecalis in the mouse

wound excisional model, the E. coli ΔentB CFU from mixed-species infections at 24 hpi

did not increase to the levels of wild-type E. coli co-infected with E. faecalis (Figure 2.12),

Page | 51 demonstrating that E. faecalis mediated augmentation of E. coli involves enterobactin.

Altogether, our data nicely recapitulates our in vitro biofilm assays, mimicking in vivo infection settings.

Figure 2.12: E. faecalis Co-infection Increases the Growth of E. coli in a Mouse Model of Wound Infection. Mouse wound infection with bacterial burdens determined at 24 hpi. The E. coli inoculum in single-species controls and/or mixed-species infections was 2–4 × 102 CFU/wound for E. coli and 2–4 × 106 CFU/wound for E. faecalis. (A) E. coli CFU (24 hpi) for mono-infected and co-infected wounds. (B) E. coli wild-type CFU (24 hpi) for co-infected compared to E. coli UTI89ΔentB co-infected wounds. (C) E. faecalis CFU at 24 hpi from wounds co-infected with E. coli wild-type or E. coli UTI89ΔentB. Recovered titers of zero were set to the limit of detection of the assay for statistical analyses and graphical representation in all figures. Horizontal bars represent the median value for each group of mice. N = 3 biological replicates. Statistical significance was determined by the Mann-Whitney test with Dunn’s post-test for multiple comparisons, ∗∗∗p = 0.0009. Percentages indicate the proportion of data points falling above 1 × 105 CFU/wound for each group. Reprinted with permission from (Keogh et al. 2016).

Because we observed changes in the local pH in our agar plate macrocolony assays, we postulate that this local change might give E. coli an advantage in iron-limited niches. Since the pH influences chelation efficiency of siderophores, E. coli growth under iron limiting conditions in the presence of the lactic acid producing E. faecalis might influence iron-

Page | 52 siderophore dynamics. The mixed-phenolate type siderophore, yersiniabactin, functions more efficiently at neutral to alkaline conditions (Valdebenito et al. 2006). However, the catecholate siderophore, enterobactin, functions more efficiently at neutral pH and is more efficient than carboxylate siderophores at pH 5 or more (Miethke and Marahiel 2007).

Enterobactin is also the strongest siderophore known, with a binding constant of 1052

(Holden and Bachman 2015).

To establish the identity of the secreted, diffusible, soluble E. faecalis factor, we performed a large scale, unbiased transposon screen. We capitalized on the enhanced biofilm/biomass production that occurred during co-culture of E. faecalis and E. coli under iron limitation, which was not observed in single species growth. We expected a loss of this phenotype, when there is an insertion in the in E. faecalis that produces the factor. In our primary validation, we isolated mutants with insertions in genes that were known to be important in biofilm formation. One of such mutants had insertions in gelatinase (gelE), which lies downstream of the fsrABC systems and has been shown to be important for biofilm formation and virulence in E. faecalis (Hancock and Perego 2002). Another mutant had insertions in the cell wall surface anchor family protein, which is involved in the production of the Ebp pilin subunit. Ebp has been previous shown to be a biofilm associated virulence factors and is important in the pathogenesis of UTIs and in infective endocarditis

(Nallapareddy et al. 2006). Altogether, isolation of these known biofilm mutants, in multiple different positions within the same gene, validates our screening approach.

The arginine deiminase pathway is important for the generation of ATP and ammonia in other bacteria species and is also important for biofilm formation in Staphylococci

Page | 53

(Lindgren et al. 2014). Other polymicrobial studies have also shown that the metabolic exchange of ornithine with Fusobacterium nucleatum from Streptococcus gordonii through

ArcD is important in F. nucleatum biofilm development (Sakanaka et al. 2015). E. faecalis has a homologous arginine deiminase pathway and is encded by the arcABC genes in addition to the arcD antiporter (Barcelona-Andres et al. 2002). Our results showed that ornithine supplementation of single-species E. coli biofilms enhanced biofilm/biomass accumulation in an enterobactin-dependent manner. While ornithine is a precursor for the biosynthesis of hydroxamate siderophores, E. coli UTI89 only produces the catecholates enterobactin and salmochelin and the mixed-phenolate yersiniabactin (Henderson et al.

2009), all of which do not necessitate the use of ornithine as a precursor (Yuan et al. 2001).

However, ornithine has also been shown to be important in the arginine biosynthesis pathway, and is interconnected with other biosynthesis and degradative pathways in pyrimidine and polyamine biosynthesis (Caldara et al. 2006). The arginine biosynthesis pathway is transcriptionally repressed by argR, and the accumulation of arginine will result in feedback inhibition by argA. This ultimately leads to an increase in amino acid precursors that has been linked to enterobactin synthesis (Charlier and Glansdorff 2004), thus accounting for E. faecalis ornithine-driven induction of E. coli siderophore biosynthesis during early stages of infection, where E. coli numbers are low and under iron limitation. Metabolomic studies performed by our collaborators later confirmed that the arginine biosynthesis pathway was indeed active, where N-acetyl-L-glutamate (C00624),

N-acetylornithine (C00437), and L-aspartate (C00049) in addition to L-ornithine (C00077) are detected in all samples.

Page | 54

Collectively, our results suggest that L-ornithine production from E. faecalis can induce enterobactin siderophore biosynthesis in E. coli during polymicrobial infections, allowing it to overcome iron limitation and growth restriction. During iron limitation in E. coli, the ferric uptake regulator (Fur) lifts the repression of the biosynthesis of siderophore genes

(Seo et al. 2014). However, RNA-seq also showed that during monocultures of E. coli, siderophore biosynthesis genes were still upregulated. Our results therefore suggest that L- ornithine functions independently of Fur to cue the enhanced synthesis of enterobactin in

E. coli. Our work highlights that the presence of other microbial species within the infection niche can have important ramifications to disease outcome. Given that most wound infections are polymicrobial, biofilm-associated and are difficult to eradicate, targeting of

L-ornithine or the presence of other microbial species might improve treatment outcomes.

Our work also raises some interesting questions. While E. faecalis has been shown to aid

E. coli colonization and virulence in this study, we found no evidence that E. faecalis benefits from this interaction. What is E. faecalis doing in the wound? Does the mere presence of it warrant any clinical intervention? Why is E. faecalis so commonly isolated from surgical and chronic wounds? We aim to answer these questions, in Chapter 3, which focuses on understanding and unravelling the pathogenesis of Enterococcal associated wound infections.

Page | 55

2.4 Materials and methods:

2.4.1 Bacterial strains and cultures: Escherichia coli strain UTI89 and its derivative strains were grown using Difco Luria-

Bertani (LB) broth (Becton, Dickinson and Company, Franklin Lakes, NJ). Enterococcus

faecalis strain OG1RF (ATCC47077) was grown using Bacto Brain Heart Infusion (BHI)

(Becton, Dickinson and Company, Franklin Lakes, NJ). Unless otherwise stated, bacteria

strains were streaked from glycerol stocks stored at -80 °C, inoculated and grown overnight

statically for 16-20 hours in 20 ml of liquid broth as indicated above. This was followed by

another overnight subculture at a 1:1000 dilution for 16-20 hours. For biofilm assays, both

E. coli and E. faecalis were grown using Tryptic Soy Broth/Agar (Oxoid, Singapore)

supplemented with 0.175% of glucose. Iron was depleted from the broth/agar media by

addition of 2,2’dipyridyl (2,2’DP) (Sigma-Aldrich, St. Louis, MO).

2.4.2 Preparation of standard bacterial suspensions:

Cells were harvested by centrifugation at 5000 RPM (4 °C) for 4 minutes. The supernatant

was discarded and the pellet washed with 1 ml of 1X sterile PBS before being harvested.

The pellet was then resuspended in 3 ml of 1X sterile PBS. The concentration of the cell

suspension was then adjusted to an optical density (OD) of 0.4 for E. coli and 0.7 for E.

faecalis at 600 nm. Following adjustment, the cell population was equivalent to 2 – 4×108

cells per ml. E. coli and E. faecalis mixtures at ratios of 5:95 were prepared by mixing 50

µl of normalized E. coli cultures and 950 µl of normalized E. faecalis cultures together in

a 1.5 ml Eppendorf tube.

Page | 56

2.4.3 Agar plate biofilm assay for proximity studies:

Five microlitre of normalized E. coli, E. faecalis or mixture cultures was pipetted onto TSB

agar plates supplemented with glucose and the iron chelator 2,2’dipyridyl as indicated

above. E. coli UTI89 was spotted on one side and E. faecalis OG1RF on the other side,

separated by a horizontal distance of 1 cm in large 50 ml square Petri dishes. A total of 10

spots were pipetted for each side with the last spot having a horizontal distance of 4.5 cm

apart from the other. A slanted vertical interval of 1 cm was maintained in between each

spot for the same bacterial species.

2.4.4 Microtitre plate biofilm assay:

Normalized bacteria cultures were inoculated at a 1:25 ratio into 200 µl of TSBG with the

necessary supplementations indicated above. The microtitre plate was then incubated for

120 hours at 37 °C statically. Supernatant was discarded from the 96 well plate and the

plate blotted dry. 250 µl of 1X PBS was then added into each well, discarded and the plates

blotted dry for a total of 2 times. A 200 µl aliquot of 0.1% w/v of crystal violet solution

(Sigma-Aldrich, St. Louis, MO) was then added into each well. Plates were then incubated

at 4 °C for 30 minutes and the crystal violet solution discarded thereafter. A 250 µl aliquot

of 1X PBS was then added into each well, discarded and the plates blotted dry for a total

of 2 times. A 200 µl aliquot of 80:20 ethanol-acetone solution was then added into each

well and the plates incubated for 45 minutes at room temperature. OD595 was then measured

with the Tecan Infinite® 200 PRO spectrophotometer (Tecan Group Ltd., Männedorf,

Switzerland).

Page | 57

2.4.5 Transposon library screen:

An overview of the entire screen process is shown in Figure 2.12. The E. faecalis OG1RF

mariner transposon library containing 14978 mutants was cryogenically stocked in

individual 96-well format microtitre plates and was kindly provided by Gary M. Dunny

from the University of Minnesota (Kristich et al. 2008). These were cultured with a cryo-

replicator (Adolf Kühner AG) to inoculate 2ml DeepWell blocks (Greiner Bio-One

#780270) containing 1ml BHI medium with overnight incubation at 37oC 220rpm. The 96

8 well DeepWell blocks were normalized to an OD600 of 0.1 (2-4x10 CFU/ml) in a 96 well

8 microtiter plate. E. coli culture stocks were normalized to an OD600 0.4 to yield 2-4x10

CFU/ml. A primary screen of the library was performed by inoculating the cultures at a

1:25 volume into 200µl of iron depleted TSBG medium. This was achieved by seeding the

stock iron depleted TSBG medium with 8.0-16.0x104 CFU/ml E. coli, dispensing 200µl

per well into the microtitre plate and finally inoculating each individual well with 1.5-

3.0x106 CFU/ml of the normalized transposon mutant cultures. This achieved an inoculum

ratio of 1Ec:19Ef. The microtitre plates were then incubated for 120 hours at 37 °C statically

inside a moistened bag to prevent evaporation of media. The microtiter plate biofilm assay

was then performed, as indicated above.

Page | 58

Figure 2.13: Schematic representation of the E. faecalis transposon library screen process. Red arrows refer to mutants that were eliminated for not fulfilling the indicated criteria.

2.4.6 Growth kinetics assay:

Transposon mutants identified in the primary validation were grown as described above.

Cultures where then normalized to obtain a final concentration of 103 cells/ ml and

inoculated at a 1:25 ratio into 200 µl of iron replete or deplete TSBG with the necessary

supplementations indicated above. Growth rates were then measured by recording the

OD600 over a period of 14 hours at 37oC with the Tecan Infinite® 200 PRO

spectrophotometer (Tecan Group Ltd., Männedorf, Switzerland). A total of 6 technical

replicates and 2 biological replicates were prepared for each transposon mutant.

2.4.7 Colony forming unit (CFU) enumeration:

Macrocolonies were excised by use of a scalpel and resuspended in 1X sterile PBS in 10

ml Falcon tubes (Becton, Dickinson and Company, Franklin Lakes, NJ) by vortexing.

Page | 59

Serial dilution was then performed for each macrocolony sample by pipetting 10 µl of

bacterial suspension into 90 µl of 1X PBS for a total of 4 technical replicates up to 10-8 in

a 96 well microtiter plate (Thermo Fisher Scientific, Waltham, MA). Five microliters of

bacterial suspension from each dilution with a total of 4 technical replicates was then

pipetted onto LB agar for single species E. coli, BHI agar for single species E. faecalis.

BHI agar supplemented with 10 µg/ml of colistin and 10 µg/ml of nalidixic acid (BHI-

CNA) was used to select for E. faecalis and McConkey agar was used to select for E. coli

within the mixed species macrocolonies. All media were purchased from BD (Becton,

Dickinson and Company, Franklin Lakes, NJ).

2.4.8 Whole genome sequencing of transposon mutants:

Genomic DNA was extraction with the Wizard Genomic DNA Purification Kit (Promega,

Madison, USA). 1 mL of overnight culture was spun down at 16,000 x g for 2 minutes and

the supernatant discard. Cells were then resuspending in 480 µl of 50mM EDTA, mixed

with 120 µl of 10 mg/mL lysozyme (Sigma-Aldrich, St. Louis, MO) and incubated at 37

°C for 30 minutes. Samples were then centrifuged at 16,000 x g for 2 minutes and the

supernatant discarded. 600 µl of nucleic lysis solution was then added and the samples were

incubated at 80 °C for 5 minutes before cooling to room temperature. 3 µl of RNase solution

was then added, mixed well by inversion and then incubated at 37 °C for 30 minutes. 200

µl of protein precipitation was then added, vortexed at maximum speed for 20 seconds

before cooling on ice for 5 minutes. Samples were then centrifuged at 16,000 x g for 5

minutes and the supernatant transferred to 600 µl of ice-cold isopropanol (Fisher Scientific,

Hampton, USA) to precipitate DNA. Precipitated DNA was then washed with 70% ethanol

(Fisher Scientific, Hampton, USA) and air-dried. 100 µl of DNA rehydration solution was

then added and the samples were incubated overnight at room temperature. Purified DNA

Page | 60 samples were then quantified with the Qubit High Sensitive dsDNA assay (Invitrogen,

Carlsbad, USA). Samples were then processed for Illumina MiSeq as per manufacturer’s instructions. CLC Genomics Workbench version 8.0 (Qiagen, Hilden, Germany) was then used to assemble the de novo reads. The resulting contigs were then aligned against the mariner transposon that was used to generate the E. faecalis OG1RF transposon library to identify the insertional site (Kristich et al. 2008).

Page | 61

2.5 Contributions:

The results here, were published as a peer-reviewed article, in the journal Cell Host &

Microbe (Keogh et al. 2016). I performed all the experiments and data analyses presented

in this section. I was also assisted by an undergraduate student, Rayson Ho Yao Yong,

whom I co-supervised with the lead author. E. coli deletion mutants of the ferrous

transporters were generated by Siyi Chen. I was also involved in the development and

optimization of the wound excisional model that was used in this study, and I was the co-

1st author on a manuscript reporting that work (Chong et al. 2017). In the work presented

in this chapter, I performed the animal work, together with Damien Keogh and Adeline

Yong Hui Mei. A part of the introduction was included in a peer-reviewed article, in the

Journal of Molecular Biology, of which I was the lead co-author (Tay et al. 2016).

A Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and

Polymicrobial Infection

Damien Keogh1 Wei Hong Tay1,2 Yao Yong Ho1,2 Jennifer L. Dale6 Siyi Chen4 Shivshankar

Umashankar3 Rohan B. H. Williams3 Swaine L. Chen4,5 Gary M. Dunny6 Kimberly A.

Kline1,2

1Singapore Centre for Environmental Life Science Engineering, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore; 2School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore; 3Singapore Centre for Environmental Life Sciences Engineering, National University of Singapore, 28 Medical Drive, Singapore 114756, Singapore; 4Division of Infectious Diseases, Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, 1E Kent Ridge Road, NUHS Tower Block, Level 10, Singapore 119074, Singapore; 5GERMS and Infectious Disease Group, Genome Institute of Singapore, 60 Biopolis Street, Genome, #02- 01, Singapore 138672, Singapore; 6Department of Microbiology, University of Minnesota Medical School, Minneapolis, MN 55455, USA

Page | 62

B Polymicrobial-Host Interactions During Infection

Wei Hong Taya,b*, Kelvin Kian Long Chonga,b*, Kimberly A. Klinea aSingapore Centre for Environmental Life Sciences Engineering, School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551; bSingapore Centre for Environmental Life Sciences Engineering, Interdisciplinary Graduate School, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551 *Authors contributed equally.

Page | 63

2.6 References for Chapter 2:

Archibald F (1986). Manganese: its acquisition by and function in the lactic acid bacteria. Crit Rev Microbiol 13: 63-109.

Bakaletz LO (2004). Developing animal models for polymicrobial diseases. Nat Rev Microbiol 2: 552- 568.

Barcelona-Andres B, Marina A, Rubio V (2002). Gene structure, organization, expression, and potential regulatory mechanisms of arginine catabolism in Enterococcus faecalis. J Bacteriol 184: 6289-6300.

Bogdan AR, Miyazawa M, Hashimoto K, Tsuji Y (2016). Regulators of Iron Homeostasis: New Players in Metabolism, Cell Death, and Disease. Trends Biochem Sci 41: 274-286.

Brogden KA, Guthmiller JM, Taylor CE (2005). Human polymicrobial infections. Lancet 365: 253-255.

Bruyneel B, vande Woestyne M, Verstraete W (1989). Lactic acid bacteria: Micro-organisms able to grow in the absence of available iron and copper. Biotechnol Lett 11: 401-406.

Caldara M, Charlier D, Cunin R (2006). The arginine regulon of Escherichia coli: whole-system transcriptome analysis discovers new genes and provides an integrated view of arginine regulation. Microbiology 152: 3343-3354.

Cao J, Woodhall MR, Alvarez J, Cartron ML, Andrews SC (2007). EfeUOB (YcdNOB) is a tripartite, acid- induced and CpxAR-regulated, low-pH Fe2+ transporter that is cryptic in Escherichia coli K-12 but functional in E. coli O157:H7. Mol Microbiol 65: 857-875.

Cassat JE, Skaar EP (2013). Iron in Infection and Immunity. Cell host & microbe 13: 509-519.

Charlier D, Glansdorff N (2004). Biosynthesis of Arginine and Polyamines. EcoSal Plus 1.

Chenault SS, Earhart CF (1991). Organization of genes encoding membrane proteins of the Escherichia coli ferrienterobactin permease. Mol Microbiol 5: 1405-1413.

Chong KKL, Tay WH, Janela B, Yong AMH, Liew TH, Madden L et al (2017). Enterococcus faecalis modulates immune activation and slows healing during wound infection. J Infect Dis.

Colomer-Winter C, Gaca AO, Lemos JA (2017). Association of Metal Homeostasis and (p)ppGpp Regulation in the Pathophysiology of Enterococcus faecalis. Infect Immun 85.

Page | 64

Crumbliss AL, Harrington JM (2009). Iron sequestration by small molecules: Thermodynamic and kinetic studies of natural siderophores and synthetic model compounds. In: van Eldik R, Hubbard CD (eds). Advances in Inorganic Chemistry. Academic Press. pp 179-250.

Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA et al (2008). Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiol 8: 43.

Drakesmith H, Prentice AM (2012). Hepcidin and the iron-infection axis. Science 338: 768-772.

Flo TH, Smith KD, Sato S, Rodriguez DJ, Holmes MA, Strong RK et al (2004). Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature 432: 917-921.

Flores-Mireles AL, Walker JN, Caparon M, Hultgren SJ (2015). Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat Rev Microbiol 13: 269-284.

Garcia EC, Brumbaugh AR, Mobley HL (2011). Redundancy and specificity of Escherichia coli iron acquisition systems during urinary tract infection. Infect Immun 79: 1225-1235.

Garenaux A, Caza M, Dozois CM (2011). The Ins and Outs of siderophore mediated iron uptake by extra-intestinal pathogenic Escherichia coli. Vet Microbiol 153: 89-98.

Giacometti A, Cirioni O, Schimizzi AM, Del Prete MS, Barchiesi F, D'Errico MM et al (2000). Epidemiology and microbiology of surgical wound infections. J Clin Microbiol 38: 918-922.

Goetz DH, Holmes MA, Borregaard N, Bluhm ME, Raymond KN, Strong RK (2002). The neutrophil lipocalin NGAL is a bacteriostatic agent that interferes with siderophore-mediated iron acquisition. Mol Cell 10: 1033-1043.

Hagan EC, Mobley HL (2009). Haem acquisition is facilitated by a novel receptor Hma and required by uropathogenic Escherichia coli for kidney infection. Mol Microbiol 71: 79-91.

Hancock L, Perego M (2002). Two-component signal transduction in Enterococcus faecalis. J Bacteriol 184: 5819-5825.

Henderson JP, Crowley JR, Pinkner JS, Walker JN, Tsukayama P, Stamm WE et al (2009). Quantitative Metabolomics Reveals an Epigenetic Blueprint for Iron Acquisition in Uropathogenic Escherichia coli. PLoS Pathog 5: e1000305.

Higgs PI, Larsen RA, Postle K (2002). Quantification of known components of the Escherichia coli TonB energy transduction system: TonB, ExbB, ExbD and FepA. Mol Microbiol 44: 271-281.

Holden VI, Bachman MA (2015). Diverging roles of bacterial siderophores during infection. Metallomics : integrated biometal science 7: 986-995.

Page | 65

Inoue S, Kawanishi S (1987). Hydroxyl radical production and human DNA damage induced by ferric nitrilotriacetate and hydrogen peroxide. Cancer Res 47: 6522-6527.

Johnson JR, Jelacic S, Schoening LM, Clabots C, Shaikh N, Mobley HL et al (2005). The IrgA homologue adhesin Iha is an Escherichia coli virulence factor in murine urinary tract infection. Infect Immun 73: 965-971.

Kehres DG, Maguire ME (2003). Emerging themes in manganese transport, biochemistry and pathogenesis in bacteria. FEMS Microbiol Rev 27: 263-290.

Keogh D, Tay WH, Ho YY, Dale JL, Chen S, Umashankar S et al (2016). Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and Polymicrobial Infection. Cell Host Microbe 20: 493-503.

Kristich CJ, Nguyen VT, Le T, Barnes AMT, Grindle S, Dunny GM (2008). Development and Use of an Efficient System for Random mariner Transposon Mutagenesis To Identify Novel Genetic Determinants of Biofilm Formation in the Core Enterococcus faecalis Genome. Appl Environ Microbiol 74: 3377-3386.

Lindgren JK, Thomas VC, Olson ME, Chaudhari SS, Nuxoll AS, Schaeffer CR et al (2014). Arginine deiminase in Staphylococcus epidermidis functions to augment biofilm maturation through pH homeostasis. J Bacteriol 196: 2277-2289.

Lisiecki P, Mikucki J (2004). [Citric acid as a siderophore of enterococci?]. Med Dosw Mikrobiol 56: 29-40.

Lisiecki P, Mikucki J (2006). Iron supply of enterococci by 2-oxoacids and hydroxyacids. Pol J Microbiol 55: 195-202.

Masson PL, Heremans JF, Schonne E (1969). Lactoferrin, an iron-binding protein in neutrophilic leukocytes. J Exp Med 130: 643-658.

Miao Q, Ku AT, Nishino Y, Howard JM, Rao AS, Shaver TM et al (2014). Tcf3 promotes cell migration and wound repair through regulation of lipocalin 2. Nature communications 5: 4088.

Miethke M, Marahiel MA (2007). Siderophore-based iron acquisition and pathogen control. Microbiol Mol Biol Rev 71: 413-451.

Nairz M, Schroll A, Sonnweber T, Weiss G (2010). The struggle for iron - a metal at the host-pathogen interface. Cell Microbiol 12: 1691-1702.

Nallapareddy SR, Singh KV, Sillanpaa J, Garsin DA, Hook M, Erlandsen SL et al (2006). Endocarditis and biofilm-associated pili of Enterococcus faecalis. J Clin Invest 116: 2799-2807.

Page | 66

Neilands JB (1995). Siderophores: structure and function of microbial iron transport compounds. J Biol Chem 270: 26723-26726.

Nelson A, De Soyza A, Perry JD, Sutcliffe IC, Cummings SP (2012). Polymicrobial challenges to Koch's postulates: ecological lessons from the bacterial vaginosis and cystic fibrosis microbiomes. Innate Immun 18: 774-783.

Pao SS, Paulsen IT, Saier MH (1998). Major Facilitator Superfamily. Microbiol Mol Biol Rev 62: 1-34.

Peyssonnaux C, Zinkernagel AS, Datta V, Lauth X, Johnson RS, Nizet V (2006). TLR4-dependent hepcidin expression by myeloid cells in response to bacterial pathogens. Blood 107: 3727-3732.

Rajasekaran MB, Nilapwar S, Andrews SC, Watson KA (2010). EfeO-cupredoxins: major new members of the cupredoxin superfamily with roles in bacterial iron transport. Biometals 23: 1-17.

Sakanaka A, Kuboniwa M, Takeuchi H, Hashino E, Amano A (2015). Arginine-Ornithine Antiporter ArcD Controls Arginine Metabolism and Interspecies Biofilm Development of Streptococcus gordonii. J Biol Chem 290: 21185-21198.

Sandy M, Butler A (2009). Microbial Iron Acquisition: Marine and Terrestrial Siderophores. Chem Rev 109: 4580-4595.

Searle LJ, Meric G, Porcelli I, Sheppard SK, Lucchini S (2015). Variation in siderophore biosynthetic gene distribution and production across environmental and faecal populations of Escherichia coli. PLoS One 10: e0117906.

Seo SW, Kim D, Latif H, O'Brien EJ, Szubin R, Palsson BO (2014). Deciphering Fur transcriptional regulatory network highlights its complex role beyond iron metabolism in Escherichia coli. Nature communications 5: 4910.

Subashchandrabose S, Hazen TH, Rasko DA, Mobley HL (2013). Draft genome sequences of five recent human uropathogenic Escherichia coli isolates. Pathogens and disease 69: 66-70.

Tay WH, Chong KK, Kline KA (2016). Polymicrobial-Host Interactions during Infection. J Mol Biol 428: 3355-3371.

Torres AG, Redford P, Welch RA, Payne SM (2001). TonB-dependent systems of uropathogenic Escherichia coli: aerobactin and heme transport and TonB are required for virulence in the mouse. Infect Immun 69: 6179-6185.

Valdebenito M, Crumbliss AL, Winkelmann G, Hantke K (2006). Environmental factors influence the production of enterobactin, salmochelin, aerobactin, and yersiniabactin in Escherichia coli strain Nissle 1917. Int J Med Microbiol 296: 513-520.

Page | 67

Weinberg ED (1997). The Lactobacillus anomaly: total iron abstinence. Perspect Biol Med 40: 578- 583.

Weiss G (2005). Modification of iron regulation by the inflammatory response. Best Pract Res Clin Haematol 18: 183-201.

Yuan WM, Gentil GD, Budde AD, Leong SA (2001). Characterization of the Ustilago maydis sid2 gene, encoding a multidomain peptide synthetase in the ferrichrome biosynthetic gene cluster. J Bacteriol 183: 4040-4051.

Page | 68

CHAPTER 3

Characterization of the intracellular lifestyle of Enterococcus faecalis

3.1 Introduction and motivation:

In Chapter 2, we demonstrated that co-infection with E. faecalis and E. coli resulted in

increased virulence of E. coli during iron limitation and in wound infections. The increased

virulence of E. coli was due to the export of L-ornithine by E. faecalis, which can be used

by E. coli to promote the induction and production of the siderophore, enterobactin.

However, this study raised the question of whether E. faecalis was truly altruistic during

wound infections, and how it persisted within wound infections in the first place. To answer

these questions, we set out to understand the significance and consequence of the presence

of E. faecalis in wound infections. The work was in part motivated by the lack of

mechanistic insights into the pathogenesis of E. faecalis in wounds, despite several

epidemiological studies characterizing E. faecalis as one of the dominant species in wound

infections. Using a mouse model of wound infection, we demonstrated that E. faecalis can

persist in wounds, up to 5 dpi at a colonization dose of 102 or 7 dpi at an infection dose of

106 (Figure 3.1) (Chong et al 2017). Moreover, we also showed that wound healing was

impaired in the infected mice as compared to the mock infected mice (Chong et al 2017).

Page | 69

Figure 3.1: E. faecalis persists in wounds up to 5 dpi or 7 dpi depending on the initial inoculum. Wounds were harvested at the indicated time points post-inoculation and the colony forming units (CFU) were enumerated. Mice were inoculated with (A) 102 CFU of OG1RF, (B) 106 CFU of OG1RF. Each dot represents one mouse, and the solid horizontal lines indicate the median, N=2, n=≥5. Statistical analysis was performed using Kruskall-Wallis test with Dunn’s post-test to correct for multiple comparisons. * = p < 0.05, ** = p < 0.01. Reprinted with permission from (Chong et al 2017).

Studies have also shown that E. faecalis is capable of persisting within a variety of host cells, including intestinal epithelial cells (Olmsted et al 1994, Wells et al 1988, Wells et al

1990), macrophages (Baldassarri et al 2001, Gentry-Weeks et al 1999b), monocytes, cervical epithelial cells (Bertuccini et al 2002), endothelial cells (Millan et al 2013), osteoblasts (Campoccia et al 2015) and enterocytes (Olmsted et al 1994). Building upon these studies and our current findings that E. faecalis can persist in wounds up to 5 dpi or

7 dpi depending on the initial inoculum, we hypothesized that Enterococcal wound infections may give rise to a subpopulation of intracellular bacteria, which may contribute to persistence which we observe in our mouse model.

Pathogenic bacteria have been traditionally classified into the following categories, based on their infective lifestyles: extracellular, facultative intracellular, and obligate intracellular

(Silva 2012). These classifications are important in the context of microbial pathogenesis,

Page | 70 as it can dictate antibiotic choice for treatment. However, due to advances in our understanding of host-pathogen interactions, we now know that this classification is inadequate and ambiguous at best. For example, some extracellular microbes will spend some of their microbial life inside host cells. The converse is also true as well for obligate intracellular pathogens, as at some point in time, they may be found in extracellular spaces upon release/escape from the host cell. This secondary extracellular/intracellular infection phase can therefore influence disease outcomes and affect treatment outcomes as well.

Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, Streptococcus pyogenes and Enterococcus faecalis have been previously classified as extracellular pathogens and are commonly found in wound infections, where they can persist (Dowd et al 2008, Giacometti et al 2000, Gjodsbol et al 2006). However, recent work has shown that this paradigm may no longer hold true, as all 5 microorganisms also have an intracellular infection phase. S. aureus can survive within immune (Gresham et al 2000, Kubica et al

2008, Voyich et al 2005) and non-immune cells (Garzoni et al 2007, Jarry and Cheung

2006) within the eukaryotic host and it has been suggested that the ability to persist intracellularly is important for subsequent dissemination to distant sites (Garzoni and

Kelley 2009, Qazi et al 2004). The ability to persist within an intracellular niche is also thought to contribute to chronic carriage or relapsing Staphylococcal infections. Similar studies have also shown that S. pyogenes can also be taken up by both immune (Bastiat-

Sempe et al 2014) and non-immune cells (Osterlund and Engstrand 1997), where it can replicate, survive host defenses and disseminate to distant sites. This is also true for E. coli strain K1, where it has been shown to survive and multiply within the phagosome of murine macrophages, before subsequent killing and lysis of the host cell (Sukumaran et al 2003).

Collectively, these studies highlight the need to investigate the intracellular infection phase

Page | 71 of classically extracellular pathogens, to develop better treatment plans and to understand the pathogenic mechanisms supporting intracellular infection and persistence.

The ability to adapt to an intracellular lifestyle is advantageous to bacterial pathogens as it allows them to escape the host humoral immune response, avoiding antibodies and complement. However, upon entry into host cells, the bacteria must now content with the host intracellular defense mechanisms. Under normal circumstances, upon entry into host cells, internalized bacteria are encapsulated within a vacuole that is derived from the host plasma membrane. The membrane bound vacuole progressively matures into a phagolysosome where it is subsequently destroyed (Haas 2007). The phagolysosome is an extremely harsh environment where there is low pH, high concentrations of free radicals

(Klebanoff 2005), lack of nutrients and antimicrobial proteins (Bulet et al 2004). To avoid this, these classical intracellular bacteria employ strategies like preventing lysosomal fusion, environmental modification within the phagolysosome or preventing vacuole maturation (Cornejo et al 2017, Pizarro-Cerdá and Cossart 2006, Ray et al 2009, Thi et al

2012). Some other classical intracellular bacteria such as Listeria monocytogenes and

Shigella flexneri can also escape from this vesicular compartment, where they continue their lifecycle within the cytosol (Ray et al 2009). Classical intracellular bacteria can reside in two main areas of the host cell: vesicular and non-vesicular compartments. However, these strategies are also employed by non-classical intracellular pathogens such as S. aureus, where extensive studies have shown that they can escape the phagolysosome as well as evade both oxidative and non-oxidative killing by phagocytes (Horn et al 2017).

These studies further reinforce the point that having an intracellular pathogenic strategy is not a special property, and that it may be a common attribute among extracellular

Page | 72 pathogens, necessitating the need to relook at the infective strategies of these otherwise extracellular pathogens.

In this Chapter, I have shown that E. faecalis possesses a dual extracellular and intracellular lifecycle. Using both in vitro and in vivo models, I have demonstrated that a sub-population of bacterial cells gets internalized into the host cell, where they can survive, replicate and subsequently escape to cause reinfection. I have also shown that intracellular E. faecalis remains in the late endosome up to 72 hpi, where there is a lack of co-localization with the lysosomal protease, Cathepsin D suggesting a failure to fuse with the lysosome. Finally, using correlative-light and electron microscopy, I have shown that E. faecalis containing vesicles are single membrane-bound compartments and do not display lysosomal fusion, suggesting that there could be a defect in the fusion process because of bacteria manipulation or a modification of the E. faecalis containing compartments, resulting in a lack of lysosomal fusion. These results provide the foundation for investigating the route of bacterial internalization and the bacterial virulence factors involved as well as the possibility of immunomodulation in Chapter 4.

Page | 73

3.2 Results:

3.2.1 E. faecalis forms microcolonies on the wound surface and under the wound bed

Using a mouse model of wound infection, we previously showed that wounds infected with

106 CFU of E. faecalis resulted in an increase to 108 CFU by 8 hpi and later decreases to

105 at 3 dpi (Figure 3.1) (Chong et al 2017). The number of recoverable CFU remains

constant at 105, up to 7 dpi, even though the wound has already begun closing. We

hypothesized that E. faecalis could be persisting in the wound for long periods of time by

actively evading the immune system. To begin to test this, we performed scanning electron

microscopy (SEM) on 8 hpi and 3 dpi wounds, since these two timepoints represent the

height of the infection and the initiation of steady state persistence respectively. At 8 hpi,

we observed the presence of E. faecalis microcolonies on the wound bed, in addition to the

presence of immune infiltrates, suggesting that there is an immune response in response to

high titer acute infection (Figure 3.2). Some of these E. faecalis microcolonies were also

embedded within a matrix and is suggestive of biofilm development (Figure 3.2). At 3 dpi,

we were unable to observe any E. faecalis, or any immune infiltrates at the wound surface,

due to remodeling of the infected tissue and scab formation (data not shown).

Page | 74

Figure 3.2: E. faecalis elicits an immune response at 8 hpi. E. faecalis microcolonies are indicated by the red arrows. Images are representative of 3 independent experiments and the scale bar represents 5 µm. Reprinted with permission from (Chong et al 2017).

Since we were still able to obtain 105 CFU from the infected wound homogenates from 3 dpi, we hypothesized that E. faecalis might be persisting under the wound bed. To determine whether this was true, we performed fluorescence in-situ hybridization (FISH) with specific probes for E. faecalis on sectioned 3 dpi wound tissues. We could detect E. faecalis microcolonies at both the wound edge and the wound bed but not in the dermis or the deeper tissue layers (Figure 3.3). These results show that E. faecalis can persist within infected wounds and possibly suggests tropism at the epidermal layer, since we did not observe any microcolonies at the deeper layers at 3 dpi. However, we were still uncertain whether E. faecalis was persisting extracellularly or intracellularly within host cells in the infected area.

Page | 75

EF – Cy3 T-PMT Eub1 - 488 Merge

WoundEdge

WoundBed

Figure 3.3: E. faecalis persists in infected wounds. Wounds were harvested at 3 days post-infection, cryosectioned, and subjected to FISH staining with E. faecalis specific probes and probes specific for the domain bacteria (Eub). Bar represents 2 μm. Images shown are representative from three independent experiments. Reprinted with permission from (Chong et al 2017).

3.2.2 Intracellular E. faecalis can be recovered from CD45+ and CD45- cells

To test whether E. faecalis was persisting extracellularly or intracellularly within infected

wounds, we infected mice with E. faecalis with 106 CFU and harvested the infected wounds

at 1, 3 and 5 dpi. We then dissociated the infected wounds to a single cell suspension,

treated them with antibiotics to kill off any extracellular bacteria, before staining them with

a CD45 antibody. CD45 is a marker that selectively labels most immune cells. We then

performed fluorescence activated cell sorting, resulting in two populations of cells, CD45+

immune cells and CD45- non-immune cells. These cells were then lysed, to release any

intracellular bacteria, before plating for bacterial CFU counts.

Page | 76

Interestingly, while we were able to recover viable E. faecalis from the immune cell

population, we were also able to recover E. faecalis from the non-immune cell population,

up to 5 dpi as well (Figure 3.4). From our results, it appears that approximately 10% of the

bacterial population are intracellular at 3 and 5 dpi. Collectively, our studies demonstrate

that E. faecalis is able to exist intracellularly during infection and is not an exclusive

extracellular pathogen, unlike previously thought.

Im m u n e c e lls (C D 4 5 + ) N o n -im m u n e c e lls (C D 4 5 - )

1 0 5 1 0 4 1 d p i

1 0 4 3 d p i

3 s

s 1 0

l

l l

l 5 d p i

e e

3 C

C 1 0

6

6 2 0

0 1 0

1

1 /

/ 1 0 2

U

U

F F

1 C C 1 0 1 0 1

1 0 0 1 0 0 1 d p i 3 d p i 5 d p i 1 d p i 3 d p i 5 d p i

Figure 3.4: Recovery of viable E. faecalis from within host cells. Male C57BL/6 mice were wounded and infected with 106 CFU of E. faecalis OG1RF. Wounds were harvested at 1/3/5 days post-infection, dissociated to single cell suspension, treated with antibiotics to kill off any extracellular bacteria, before staining and cell sorting into (A) CD45+ or (B) CD45- populations. CD45+ and CD45- cell populations were then lysed and plated for bacterial CFU counts. CD45 selectively labels immune cells. Data shown are representative of at least 3 independent experiments. N≥3, n≥4.

3.2.3 E. faecalis adheres to and enters keratinocytes

To determine the mechanisms by which E. faecalis can infect keratinocytes, we used the

gentamicin protection assay. The gentamicin protection assay is well described in the

literature and has been used to quantify the number of extracellular, adhered and

internalized E. faecalis within other cell lines (Bertuccini et al 2002, Elhadidy and Zahran

2014, Millan et al 2013). We infected the spontaneously immortalized human keratinocyte

cell line, HaCaT, with E. faecalis at a multiplicity of infection (MOI) of 1, 10 and 100 for

Page | 77 a period of up to 3 hours. We demonstrated that E. faecalis can attach to and enter keratinocytes, with detectable counts of bacterial CFU as early as 1-hour post-infection

(hpi) at an MOI of 10 and 100 (Figure 3.5). We have also performed cell cytotoxicity experiments and have shown that E. faecalis infection does not negatively affect the keratinocytes at early timepoints <5 hpi, without the presence of gentamicin/penicillin

(data not shown).

A d h e re d B a c te ria In te rn a liz e d B a c te ria

1 0 8 1 0 7 3 H o u r 7 H a C a T 1 0 1 0 6 2 H o u r 1 0 6 1 0 5

H a C a T 1 H o u r l l 1 0 5

4 m

m 1 0

/ /

4 U U 1 0

3 F F 1 0

3 C C 1 0 1 0 2 1 0 2

1 1 0 1 1 0

1 0 0 1 0 0

I1 0 0 I1 0 0 1 0 1 0 O I 1 O I 1 M O I M O I M O M O M M

Figure 3.5: Time and dose-dependent increase of E. faecalis CFU numbers. Dashed lines indicate the number of keratinocytes used for infection while solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

To confirm E. faecalis entry into the keratinocytes, we used confocal laser scanning microscopy (CLSM) on keratinocytes infected with a green fluorescent protein (GFP) expressing derivative of E. faecalis. When we compared images taken at early infection timepoints of 3 hpi, followed by 1 hr gentamicin protection to ensure that all extracellular bacteria were, we observed that there were only 1 to 10 bacteria within each infected keratinocyte (Figure 3.6).

Page | 78

Figure 3.6: E. faecalis entry into keratinocytes. CLSM orthogonal view of internalized E. faecalis within keratinocytes at 3 hpi. Blue, dsDNA stained with Hoechst 33342; green, E-GFP E. faecalis; red, F-actin. Data and images shown are representative of 3 independent experiments, N=3, n=3.

Unfortunately, we were unable to maintain the infection any longer than 4 hours in the

absence of gentamicin, as by 4 hpi, keratinocytes were already beginning to detach from

the dishes, eventually sloughing off entirely by 6 hpi (data not shown). We speculate that

since E. faecalis is a lactic-acid bacterium, the rapid acidification of the media over time

results in the detachment of the keratinocytes. Given that attachment precedes

internalization or uptake (Pizarro-Cerda and Cossart 2006, Ribet and Cossart 2015), we

chose a MOI of 100:1 for subsequent infections as by 1 hpi, the number of adhered bacteria

has already matched the number of keratinocytes.

3.2.4 E. faecalis entry is not cell line specific

To ensure that E. faecalis internalization is not cell type specific, we infected two other cell

lines, Raw264.7 murine macrophages and NIH/3T3 murine fibroblasts with E. faecalis. We

hypothesized that since we were able to recover intracellular E. faecalis from infected

wounds in mice, we would also recover intracellular bacteria from mouse fibroblasts and

macrophages. Our results show that E. faecalis entry is not cell line specific, with similar

intracellular CFU obtained from keratinocytes and fibroblasts (Figure 3.7). We speculate

that the higher bacterial CFU burden in macrophages is because this cell line is naturally

Page | 79

phagocytic, accounting for the higher intracellular counts. Consistent with previous reports,

we also show that E. faecalis can survive in macrophages for extended periods (Gentry-

Weeks et al 1999b), further validating our assay conditions. Unfortunately, we were unable

to test on human primary epithelial cell lines, as those require defined media formulations,

which cannot support E. faecalis growth (data not shown).

In te rn a liz e d B a c te ria (M O I1 0 0 :1 )

1 0 9

1 0 8 M a c ro p h a g e s ) l 1 0 7

m F ib ro b la s ts /

U 6

F 1 0 K e ra tin o c y te s

C (

5

s 1 0

t n

u 1 0 4

o C

3

U 1 0 F

C 1 0 2

1 0 1

1 0 0 1 2 3 2 4 T im e (H o u rs )

Figure 3.7: E. faecalis entry into host cell is not cell line specific. Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate (N=3).

3.2.5 E. faecalis survives up to 72 hpi within keratinocytes

Previous studies have demonstrated that E. faecalis is able to survive within professional

phagocytes for up to 48 hpi, before the host macrophages undergo lysis (Gentry-Weeks et

al 1999a). However, no studies have been done to investigate this in human keratinocytes.

We hypothesized that E. faecalis is similarly able to survive within keratinocytes or even

replicate within them, as they are unable to kill internalized bacteria effectively due to their

non-professional phagocytic nature (Mempel et al 2002). Consistent with previous studies

using professional phagocytes, and we could also recover viable intracellular bacteria up to

72 hpi (Figure 3.8).

Page | 80

In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 4

1 0 3

l

m / 2

U 1 0

F C

1 0 1

1 0 0 1 h r A b x 2 2 h r A b x 4 6 h r A b x 7 0 h r A b x

T im e (2 h o u r p o s t in fe c tio n + A b x ) Figure 3.8. E. faecalis survives for prolonged periods within keratinocytes in large numbers. HaCaTs were infected with E. faecalis at a MOI of 100 for 2 hrs. Following which, infected cells were vigorously washed and treated with gentamicin & penicillin for 1/22/46/70 hours before lysis to obtain the intracellular population.

However, we observed a decrease in intracellular bacterial CFU with increasing time which we speculate can be attributed to the following three reasons: host cell death, penetration of gentamicin into the host cell over time, or the destruction of intracellular bacteria within the host cells. The death of host cells as a result of infection results in the release of intracellular bacteria into the extracellular media, where the presence of gentamicin then kills E. faecalis, reducing the number of viable CFU recovered over time in the absence of intracellular replication to offset this loss. Moreover, studies have also shown that gentamicin is able to penetrate into host cells to affect cytosolic bacteria, albeit very slowly

(Drevets et al 1994). Due to the continuous exposure of infected cells to gentamicin up to

72 hours in our experiments, it is possible that a decrease in bacterial CFU numbers over time is due to this phenomenon. Lastly, the decrease in intracellular numbers could also be due to host cell defence mechanisms, where some intracellular bacteria are being killed or destroyed.

Page | 81

3.2.6 E. faecalis is able to multiply within keratinocytes and escape

Since at least some internalized E. faecalis can persist within the host cells, and since the

bacterial CFUs recovered decreased very slowly over 72 hrs, we hypothesized that E.

faecalis may be able to multiply within host cells. The ability to grow and multiply within

the infected keratinocytes could account for the persistence of Enterococci in chronic non-

healing wounds, where there could be repeated cycles of infection. To test this hypothesis,

we infected keratinocytes for 3 hours, then added gentamicin for the next 21 hrs. We

reasoned that if intracellular replication does indeed occur, we would observe increased

numbers of E. faecalis within the infected keratinocytes. Consistent with our hypothesis,

we observed large numbers of E. faecalis within the infected keratinocytes distributed

throughout the cytoplasm and often in a perinuclear manner within the infected host cell

(Figure 3.9). This phenotype was absent in the 3 hpi keratinocytes and suggests that E.

faecalis is indeed replicating within the cells over the subsequent 21 hrs.

Figure 3.9: E. faecalis is present in large numbers in keratinocytes at late timepoints. CLSM orthogonal view of infected keratinocytes following 3 hours of infection and 21 hours of incubation in antibiotic laced media, suggestive of trafficking through the host endo-lysosomal pathway. Blue, dsDNA stained with Hoechst 33342; green, E-GFP E. faecalis; red, F-actin. Data shown are representative of at least 3 independent experiments, N=3, n=3.

Moreover, we also show that large clusters of fluorescent bacteria can be detected in the

peripheral of apoptotic keratinocytes (Figure 3.10). The presence of large clusters of E.

Page | 82

faecalis suggests that the intracellular bacteria do escape from the keratinocytes. This was

not seen in early timepoints of 3 hpi and is only observed after 24 hpi.

Figure 3.10: Large clusters of E. faecalis is present the periphery of dead keratinocytes. CLSM orthogonal view of infected keratinocytes following 3 hours of infection and 21 hours of incubation in antibiotic laced media. Blue, dsDNA stained with Hoechst 33342; green, E-GFP E. faecalis; red, F-actin. Data shown are representative of at least 3 independent experiments, N=3, n=3.

3.2.7 E. faecalis escapes from infected keratinocytes and subsequently infects other

keratinocytes when conditions are favourable

Since we could detect large numbers of E. faecalis in what appeared to be the periphery of

dead host cells, we hypothesized that upon escape from the infected host cells, E. faecalis

can subsequently be taken up by and infect other keratinocytes. To test this hypothesis, we

infected keratinocytes with E. faecalis for 3 hrs, then incubated E. faecalis with gentamicin

for 3 hrs to kill all extracellular bacteria, before finally switching back to antibiotic-free

media for 17 hrs. A final gentamicin protection step was then performed for 1 hr, to kill

off all extracellular bacteria. Following CFU enumeration of bacteria after this

“reinfection” experiment, we found a marked 150-fold increase in internalized E. faecalis

as compared to 3 hpi plus 21 hrs gentamicin protection, control. (Figure 3.11). These

findings suggest that E. faecalis has an intracellular life cycle, where it is able to enter,

replicate and subsequently escape from host cells under favourable conditions.

Page | 83

In te r n a liz e d B a c te r ia (M O I 1 0 0 :1 ) 1 0 6 * * * *

5

l 1 0

m

/

U

F C 1 0 4

1 0 3 2 1 h r A b x 3 h r A b x + 1 7 h r w /o A b x + 1 h r A b x

Figure 3.11: Intracellular E. faecalis is able to escape from keratinocytes and cause subsequent infections (p≤0.0001, two-tailed T-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=2) of each biological replicate.

3.2.8 E. faecalis exists within early and late endosomes

Given that large numbers of E. faecalis were present within keratinocytes in the perinuclear

region at late timepoints of 24 hpi, we hypothesized that E. faecalis may be trafficked

through the host endo-lysosome pathway. Given that possibility, we infected cells for 3 hpi

or 24 hpi and subsequently labeled the infected cells with fluorescently labelled antibody

conjugates for the following markers: EEA1 for the early endosome, M6PR for late

endosome, Lamp1 for the late endosome/early lysosome and Cathepsin D for the early

lysosome. At the early timepoint of 3 hpi, we observed E. faecalis in the early endosomes,

where it is surrounded by the EEA1 marker (Figure 3.12A). We were also able to observe

E. faecalis confined within the late endosomes, which was indicated by E. faecalis in close

proximity to the the M6PR and Lamp1 markers, but not the EEA1 marker (Figure 3.12B).

Page | 84

A EEA1 Lamp1 E. faecalis Merge

3hpi

B Lamp1 M6PR E. faecalis Merge

3hpi

24hpi

Figure 3.12: Internalized E. faecalis can be found within early and late endosomes at both 3hpi and 24hpi. (A) CLSM of infected HaCaTs with fluorescent labelling of EEA-1 (early endosome) and Lamp1 (late endosome/lysosome). (B) CLSM of infected HaCaTs with fluorescent labelling of M6PR (late endosome) and Lamp1 (late endosome/lysosome). Images shown are representative of 3 independent experiments, N=3, n=3.

Moreover, when we performed time course immunofluorescence on the infected cells, we were still able to observe live EF within the Lamp1+ compartments, up to 72 hpi (Figure

3.13). While E. faecalis has been shown to survive in murine macrophages by resisting acidification which in turn prevents fusion with lysosomes (Zou and Shankar 2016), this has not been documented to occur in epithelial cells.

Page | 85

EEA1 Lamp1 E. faecalis Merge

24hpi

48hpi

72hpi

Figure 3.13: E. faecalis survives in Lamp1+ compartments. CLSM of infected HaCaTs with fluorescent labelling of EEA1 (red), Lamp1 (green), showing presence of the E. faecalis in Lamp1+/EEA1- vacuoles. Images shown are representative of 3 independent experiments, N=3, n=3.

Importantly, we also found that intracellular E. faecalis accumulates in Lamp1+ vacuoles which are severely distended by 24 hpi but not at early timepoints of 3 hpi (Figure 3.14).

It is therefore tempting to speculate that internalized E. faecalis can survive and replicate within this compartment. Based on these observations, we propose that replication occurs until a certain point whereby the vacuole is unable to accommodate any additional bacteria, and eventually burst, releasing free bacteria out into the host cell cytoplasm.

Page | 86

Lamp1 E. faecalis dsDNA Merge

24hpi

Figure 3.14: Expansion of Lamp1+ compartments at late timepoints. CLSM of infected HaCaTs with fluorescent labelling of Lamp1, showing expansion of the E. faecalis containing vacuoles (white arrows). Images shown are representative of 3 independent experiments, N=3, n=3.

3.2.9 Lamp1+ vacuoles that contain E. faecalis lack Cathepsin D

Our results suggest that internalized E. faecalis is indeed trafficked through the host endo-

lysosomal pathway, which would normally lead to lysosome fusion, targeting the bacteria

for destruction. However, our imaging data also showed that this was not the case, because

E. faecalis is replicating in Lamp1+ compartments. We thus hypothesized that there might

be defects in the fusion of the late endosomes with the lysosomes, resulting in the

intracellular E. faecalis survival and proliferation within the late endosome. The lysosome

has a pH of 4.5, containing over 60 and they can fuse with late endosomes and

autophagosomes to degrade extracellular and intracellular substrates respectively

(Settembre et al 2013). To test this, when we stained for the lysosomal protease, Cathepsin

D, we found that there was no colocalization of the lysosomal protease with the E. faecalis

containing vacuoles even at late timepoints of 24hpi (Figure 3.15).

Page | 87

Lamp1 Cathepsin D E. faecalis Merge

3hpi

24hpi

Figure 3.15: Lamp1+ E. faecalis compartments lack the lysosomal protease Cathepsin D. CLSM of infected HaCaTs with fluorescent labelling of Cathepsin D (lysosomal protease) at both early and late timepoints. Arrows indicate lack of the lysosomal protease Cathepsin D where E. faecalis is present. Images shown are representative of 2 independent experiments, N = 3, n=3.

3.2.10 E. faecalis is present in compartments lacking Lamp1 at late timepoints in the

cell periphery

While fluorescence light microscopy (FLM) can be useful for the visualization of

internalized E. faecalis within keratinocytes, as demonstrated by our findings above,

several caveats exist for that approach. Firstly, FLM is limited by diffraction and there are

limits to the resolution achievable which may limit our interpretation of our findings like

whether the large clusters of E. faecalis are within a single compartment or within

individual compartments (Figure 3.14). Secondly, FLM only allows for the imaging of

fluorescently labelled structures or proteins. The remaining unlabelled structures or

proteins cannot be imaged, nor can the reference space, preventing us from learning about

the surrounding cellular structures that are around the E. faecalis containing compartments.

Do these compartments preferentially localize next to cellular organelles such as the

endoplasmic reticulum? Or are they near the mitochondria? Thirdly, the internalization

Page | 88 process is unlikely to be a synchronized event, and if the endocytic pathway is involved as our data suggest, a large number of structures must be fluorescently labelled or tagged in order to fully visualize the entire process among snapshots of many cells. To overcome some of these limitations, previous studies have used transmission electron microscopy

(TEM). Given that TEM is able to reveal subcellular details of both labelled and unlabelled structures, its use provided critical insights into the internalization process and the spatial localization of E. faecalis within the vacuolar compartments of host cells (Baldassarri et al

2005, Millan et al 2013, Zou and Shankar 2016). However, the limitations of TEM lies in that it has a very small field of view and cannot be used for live-cell imaging. Despite the limitations of both techniques, the combination of the two imaging modalities overcomes these problems, where light microscopy (LM) would allow us to spot tagged cellular structures as well as infected cells, before moving in for high resolution imaging with TEM.

This technique dramatically reduces time spent and is especially useful in our case, since only approximately 5-10 percent of all cells are infected, making it difficult to locate infected cells with conventional TEM alone (de Boer et al 2015). It also provides us details on the localization of the fluorescently labelled protein within an ultrastructural context (de

Boer et al 2015).

To perform CLEM on infected keratinocytes, we first created a stable cell line by transfecting keratinocytes with a linearized Lamp1-mCherry construct which will label the lysosomal compartments. Following this, we infected these cells with GFP expressing E. faecalis and utilized LM to locate, image and mark the infected cells before processing them for TEM. Our results show that E. faecalis can be found in single walled vacuolar membranes that is Lamp1+ at 24 hpi (Figure 3.16B). While these single walled vacuolar compartments are Lamp1+, there appears to be a lack of lysosomal fusion with these E.

Page | 89 faecalis containing compartments as both compartments were spatially apart in all the images obtained (Figure 3.16A). Internalized bacteria also appear to be morphologically intact, with high cytoplasmic density and envelop integrity (Figure 3.16A).

Figure 3.16: Lysosomes do not fuse with E. faecalis containing compartments. (A) TEM of infected keratinocytes at 24 hpi. Large arrows indicate E. faecalis containing compartments, small arrows indicate lysosomes. (B) LM of infected keratinocytes with fluorescent labelling of E. faecalis (GFP, green) and LAMP1 positive compartments (mCherry, red). Large arrows indicate E. faecalis containing compartments, small arrows indicate lysosomes. Images shown are representative of 3 independent experiments, N = 3, n=3.

Surprisingly, unlike what was previously seen in Figure 3.14, we did not observe clusters of Enterococci within a single compartment. Rather, we only observed two diplococci at most, enclosed within a single vacuole that appears to be budding off (Figure 3.17).

Page | 90

Figure 3.17: E. faecalis containing compartments do not contain more than 2 diplococci. (A) TEM of infected keratinocytes at 24 hpi. Large arrows indicate E. faecalis containing compartments, small arrows indicate lysosomes. (B) LM of infected keratinocytes with fluorescent labelling of E. faecalis (GFP, green) and LAMP1 positive compartments (mCherry, red). Large arrows indicate E. faecalis containing compartments, small arrows indicate lysosomes. Images shown are representative of 3 independent experiments, N = 3, n=3.

We also observed that the E. faecalis compartments were defective, where the single- walled membranes appeared to be discontinuous and broken (Figure 3.18). Our results raise the possibility that E. faecalis can degrade the compartments that they are sequestered in, allowing for the eventual escape. Moreover, given that the cytosol is rich in nutrients, such a compartmental breech might allow E. faecalis to acquire nutrients, allowing it to replicate or survive.

Page | 91

Figure 3.18: Membranes of E. faecalis containing compartments are single-walled and discontinuous. TEM of infected keratinocytes at 24 hpi. Arrows indicate breaks in the E. faecalis containing compartments. Images shown are representative of 3 independent experiments, N = 3, n=3.

Together, data from our CLEM experiments suggests that lysosomes are not fusing with E. faecalis containing compartments. The integrity of these single-walled compartments is also compromised. Our results raise the possibility that E. faecalis could be hijacking the pathway, altering membrane markers to prevent lysosomal recognition, disrupting the compartment that they are in, allowing for the survival, replication and eventual escape.

Page | 92

3.3 Discussions:

Globally, E. faecalis is among the top three most commonly isolated microbial species

cultured from chronic wound infections. Despite this, little is known about the pathogenesis

of E. faecalis in the context of wound infections. The ability of E. faecalis to adapt to an

intracellular lifecycle has only been minimally investigated, in one study demonstrating

that E. faecalis can enter into epithelial cells of a cervical and mammary origin (Baldassarri

et al 2005). However, how long/where/etc was not previously investigated. Here, we report

for the first time that E. faecalis can be taken up, replicate and survive within non-immune

cells; raising the possibility that this intracellular lifecycle is linked to the prevalence and

chronic, non-healing nature in wounds. We also demonstrate that this was physiologically

relevant in a mouse model of wound infection, where we could detect E. faecalis

intracellularly within infected cells.

Using in-vitro infection assays, we demonstrate that E. faecalis is able to successfully enter

into keratinocytes, which are non-professional phagocytes and are also the most abundant

cell type in the skin, and is similar to previous reports using HeLa (Bertuccini et al 2002)

and HUVEC cell lines (Millan et al 2013). Although the number of internalized bacteria

from the HUVEC were similar to our studies across time points and MOIs (Millan et al

2013), the number of internalized E. faecalis recovered from the infected HeLa cells were

as much as 3-fold higher compared to our results (Baldassarri et al 2005). This difference

can be attributed to methodological differences in determining the number of internalized

bacteria, where the authors used light microscopy to count the infected cells instead of

performing CFU enumeration. We also show that E. faecalis entry into host cells was not

cell type specific, because intracellular bacteria can be recovered from cultured immune

(murine macrophages) and non-immune cells (murine fibroblasts and human epithelial

Page | 93 cells). Consistent with previous studies, we observed 10-fold fewer internalized bacteria in infected keratinocytes as compared to murine macrophages in our in-vitro infection assays.

Due to the phagocytic nature of macrophages, it is unsurprising that the number of internalized bacteria recovered from macrophages is different. Macrophages and monocytes are professional phagocytes and they differ from non-professional phagocytes in that they have receptors that can efficiently recognize foreign substances (Rabinovitch

1995). Nonetheless, non-professional phagocytes such as epithelial cells and fibroblasts can efficiently take up particles in-vivo and in cell culture (Rabinovitch 1995). We also observed that the uptake of E. faecalis into cells was not confined only to cultured cell lines, and we were also able to recover intracellular bacteria, in-vivo, from infected mice up to 5 dpi from both immune and non-immune cells.

Previous studies using either professional or non-professional phagocytic cell lines have reported the internalization, but not the replication of intracellular E. faecalis (Baldassarri et al 2005, Bertuccini et al 2002, Gentry-Weeks et al 1999a) as they used only the gentamicin protection assay and transmission electron microscopy at single timepoints.

Here, we performed the gentamicin protection assay across multiple timepoints, and our results showed that E. faecalis can enter and survive up to 72 hpi, validating work in previous cell lines. Importantly, through the use of a modified gentamicin protection assay, where antibiotic pressure was lifted for a period, we were able to show that intracellular E. faecalis does escape from the host cells to cause re-infection. Furthermore, using confocal laser scanning microscopy, we were able to visualize and quantify the number of intracellular E. faecalis within infected keratinocytes across different timepoints. Imaging data obtained at the end of an experiment consisting of a 3 hr infection followed by a 21 hr gentamicin exposure, showed that infected cells contained large numbers of internalized

Page | 94 bacteria, ranging from 4-38 enterococci in a single cell, validating results from our modified gentamicin protection assay that E. faecalis does multiply and escape. By contrast, our imaging data obtained at the end of an experiment consisting of a 3 hr infection followed by a 1 hr gentamicin protection, revealed that infected cells contained very few internalized bacteria, ranging from 1 to 6 diplococci. Together these data suggest that once E. faecalis has entered the host keratinocytes, at least some of the bacteria are able to replicate within those cells. While it was previously thought that only certain obligate or facultative intracellular bacteria species were capable of intracellular replication, it is becoming increasingly clear that other “classical” extracellular bacteria including S. aureus and P. aeruginosa are also able to replicate intracellularly (Flannagan et al 2016, Jolly et al 2015).

Given that non-professional phagocytes often lack the ability to produce reactive oxygen species (ROS) or cytokines in response to infection (Rabinovitch 1995), it is therefore possible that keratinocytes are less able to eradicate internalized bacteria and may provide a safe haven for E. faecalis.

In addition, our microscopy data also showed that at early time-points of 3 hpi, E. faecalis is sometimes localized to the perinuclear region. However, at 24 hpi, the majority of the infected cells contained bacteria at the perinuclear region, suggesting that they are being transported via the host endolysosomal pathway. We later show that internalized E. faecalis is indeed trafficked through the host cell endolysosomal pathway, where it first starts off in the early endosome which is EEA1+, before transiting to the late endosome, which is

Lamp1+ and M6PR+ (Humphries et al 2011). Moreover, we also observed that these compartments are distended at late timepoints of 24 hpi or more, and that not all E. faecalis are tightly encapsulated by the Lamp1 marker. Furthermore, even at late timepoints of 24 hpi or later, there was no Cathepsin D in these E. faecalis containing compartments,

Page | 95 suggesting that the late endosome could be missing markers or that the endosomal compartment has been modified, making lysosomal fusion impossible (Flannagan et al

2009). The lack of colocalization of intracellular E. faecalis with the lysosomal protease,

Cathepsin D, was not likely due to delayed acquisition, because we observed no Cathepsin

D in E. faecalis compartments even at late timepoints of 72 hpi (data not shown). Other intracellular pathogens such as Coxiella burnetii (Ghigo et al 2002), Francisella tularensis

(Clemens et al 2004) do not have Cathepsin D in bacteria containing vacuoles, or have very low amounts of Cathepsin D. The authors suggest that this was achieved through either permeabilization of the compartmental membranes containing the bacteria, allowing

Cathepsin D to diffuse out or that the compartments harbouring the bacteria simply do not fuse with lysosomes (Clemens et al 2004, Ghigo et al 2002). Transmission electron microscopy on infected keratinocytes at late timepoints of 24 hpi revealed that all membrane bound E. faecalis did not have lysosomal fusion, as the lysosomes were spatially apart from the E. faecalis containing compartments, and there was no indication of membrane fusion between the two. We also observed that these E. faecalis containing compartments are damaged. Collectively our results strongly suggest that the late endosome containing E. faecalis vacuoles have either been modified or damaged, preventing the expected destruction of intracellular E. faecalis by lysosomal fusion and allowing them to replicate from within. We have also shown that infected host cells can eventually die, releasing large numbers of E. faecalis into the periphery of dead host cells.

Based on our reinfection studies, we propose that those bacteria released from dead cells can infect other cells, resulting in a cycle of reinfection.

Altogether, our work has demonstrated that E. faecalis can enter, survive, replicate and escape from keratinocytes in vitro. If this intracellular lifecycle also exists in vivo, these

Page | 96 findings may allow for a protective niche for bacterial persistence that could contribute to the chronic nature of non-healing wounds. In Chapter 4 in this thesis, we focus on understanding the intracellular lifecycle in greater mechanistic detail, examining the mode of bacterial entry, the bacterial virulence factors involved in adhesion and entry into host cells, as well as the potential of intracellular immune modulation.

Page | 97

3.4 Materials and methods:

3.4.1 Bacterial strains and growth conditions:

Enterococcus faecalis strain OG1RF (ATCC47077) and its derivative strains were grown

using Bacto Brain Heart Infusion (BHI) (Becton, Dickinson and Company, Franklin Lakes,

NJ). Unless otherwise stated, bacteria strains were streaked from glycerol stocks stored at

-80 °C, inoculated and grown overnight statically for 16-20 hours in 20 ml of liquid broth

as indicated above. Cells were harvested by centrifugation at 5000 RPM (4 °C) for 4

minutes. The supernatant was discarded, and the pellet washed with 1 ml of 1X sterile

phosphate buffered saline (PBS) before being harvested. The pellet was then resuspended

in 3 ml of 1X sterile PBS. The concentration of the cell suspension was then adjusted to an

optical density (OD) of 0.7 for E. faecalis at 600 nm. Following adjustment, the cell

population was equivalent to 2–3×108 cells per ml.

3.4.2 Mouse wound excisional model

Mouse wound infections were modified from a previous study (Keogh et al 2016). Male

wild-type C57BL/6 mice (7-8 weeks old, 22 to 25g; InVivos, Singapore) were anesthetized

with 3% isoflurane and the dorsal hair trimmed. Following trimming, Nair™ cream

(Church and Dwight Co, Charles Ewing Boulevard, USA) was applied and the fine hair

removed via shaving with a scalpel. The skin was then disinfected with 70% ethanol. A 6-

mm biopsy punch (Integra Miltex, New York, USA) was used to create a full-thickness

wound and 10 μl of the respective bacteria inoculum applied. The wound site was then

sealed with a transparent dressing (Tegaderm™ 3M, St Paul Minnesota, USA). At the

indicated time points, mice were euthanized and a 1 cm by 1 cm squared piece of skin

surrounding the wound site was excised and collected in sterile 1X PBS. Skin samples were

homogenized, and the viable bacteria enumerated by plating onto both BHI plates and

Page | 98

antibiotic selection plates to ensure all recovered colony forming units (CFU) correspond

to the inoculating strain.

3.4.3 Gene probe and Fluorescence in-situ Hybridization (FISH)

Detection of E. faecalis was achieved with the oligonucleotide probe 5’-GGT GTT GTT

AGC ATT TCG/Cy3/-3’ (IDT Technologies, Iowa, United States). The general

oligonucleotide probe 5’- GCT GCC TCC CGT AGG AGT/Alexa Fluor® 488/-3’ (IDT

Technologies, Iowa, United States) was used as a counterstain and targets the 16S rRNA

of organisms in the domain of Bacteria (Kempf et al 2000). Cryo-sectioned tissue sections

were dehydrated in a graded ethanol series (70% and 80%) for 3 minutes each. Tissue

sections were then immersed in a 0.2% Sudan Black solution (prepared in 96% ethanol) for

20 minutes and washed thrice with a 0.02% Tween solution (prepared in 1X PBS). A total

of 25 μl of 25% formamide hybridization buffer (20 mM Tris-HCl [pH 8.0], 5M NaCl,

0.1% sodium dodecyl sulfate, and 25% formamide) containing 100 pmol of the labelled

probe (50 μg/ml stock) was added to the sections and incubated overnight at 48°C. Slides

were then immersed in 50 ml of wash buffer (0.5M EDTA and 5M NaCl, 20mM Tris-HCl

[pH 8.0]) for 30 minutes in a 46°C water bath. After washing, slides were plunged into ice

cold water for 5 seconds and left to dry.

3.4.4 Scanning Electron Microscopy

Excised skin samples were fixed using 2.5% glutaraldehyde (prepared in 0.1M PBS pH

7.4) for 48 hours at 4°C and then washed three times with 0.1M PBS. Fixed samples were

then dehydrated with a graded ethanol series (once with 30%, 50%, 70%, 80%, 90% and

twice with 100% for 15 minutes at each step) together with gentle agitation. Samples were

then subjected to amyl acetate immersion for 30 minutes. Samples were next critical point

Page | 99

dried with the Bal-Tec CPD-030 Critical Point Dryer (Bal-Tec AG, Balzers, Liechtenstein)

overnight and deposited onto SEM specimen stubs using NEM Tape (Nisshin Em. Co. Ltd,

Tokyo, Japan). Samples were then sputter coated with gold using a Bal-Tec SCD 005

sputter coater (Bal-Tec AG, Balzers, Liechtenstein). Samples were viewed using a JSM-

6360LV (JEOL, Tokyo, Japan) scanning electron microscope.

3.4.5 Flow cytometry

Excised skin samples were harvested, placed in 1.5 mL Eppendorf tubes containing

2.5U/ml liberase prepared in complete DMEM with 500 μg/mL of gentamicin–penicillin

G (Sigma-Aldrich, St. Louis, MO) and minced into small sections using a pair of surgical

scissors. The mixture was then transferred into 6-well plates and incubated for 1 hour at

37°C in a 5% CO2 humidified atmosphere with constant agitation. Dissociated cells were

then passed through a 70 μm cell strainer to remove undigested tissues and spun down at

1350 RPM for 5 minutes at 4°C. The enzymatic solution was then aspirated, and cells were

blocked in 500 μl of 1X FACs buffer (10ml heat-inactivated fetal bovine serum (FBS)

(Gibco, Thermo Fisher Scientific, Singapore), 1ml of 100mM ethylenediaminetetraacetic

acid (EDTA) and 500 ml PBS (Gibco, Thermo Fisher Scientific, Singapore). Cells were

then incubated with 10μl of CD16/CD32 antibody for 20 minutes to block the Fc receptors.

Cells were incubated with an anti-mouse CD45-cy7 conjugated antibody (BD

Pharmingen™, Singapore) (1:400 dilution) for 20 minutes at room temperature. Cells were

then centrifuged at 1350 RPM for 5 minutes at 4°C and washed in 1X FACs buffer before

a final resuspension in 1X FACs buffer. Following which, cells were then sorted using a

BD FACSAria™ 3 sorter, equipped with 4 air-cooled lasers (355nm UV, 488 nm Blue, 561

nm Yellow/Green and 633 nm Red) (Becton Dickinson, Franklin Lakes, NJ). Following

that, cells were lysed, by the addition of 0.1% Triton–X100 (Sigma-Aldrich, St. Louis, MO)

Page | 100

and intracellular bacteria plated onto both BHI plates and antibiotic selection plates to

ensure all recovered colony forming units (CFU) correspond to the inoculating strain.

3.4.6 Cell culture:

The spontaneously immortalized human keratinocyte cell line, HaCaT (AddexBio, San

Diego, CA) was routinely cultured at 37 °C in a 5% CO2 humidified atmosphere. Cells

were grown and maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco;

Thermo Fisher Scientific, Singapore) with 10% heat-inactivated fetal bovine serum (FBS)

(PAA, GE Healthcare, Singapore), 100 U of penicillin–streptomycin (Gibco, Thermo

Fisher Scientific, Singapore). The culture medium was replaced once every three days, and

upon reaching 80% confluency, cultures were passaged. Passaging was achieved by

treatment with 0.25% trypsin-EDTA (Gibco; Thermo Fisher Scientific, Singapore) for 6

minutes and seeding cells at a density of 2×106 cells/T75 flask (Nunc; Thermo Fisher

Scientific, Singapore).

3.4.7 Invasion Assay:

Keratinocytes were seeded at a density of 5×105 cells/well in a 6-well tissue culture plate

(Nunc; Thermo Fisher Scientific, Singapore) and grown for 3 days at 37 °C in a 5% CO2

humidified atmosphere. After 3 days, each well had approximately 1–1.5×106

keratinocytes. Keratinocytes were infected with bacteria at a multiplicity of infection

(MOI) of 100, 10 or 1 for 1–3 hours. Following infection, the media was aspirated, and the

cells washed thrice in 1X PBS before performing the following to determine either the

adhered or intracellular bacteria population. To determine the adhered bacteria population,

0.1% Triton–X100 (Sigma-Aldrich, St. Louis, MO) solution was then added and the cells

mechanically disrupted with a cell scraper. Appropriate dilutions were then plated on BHI

Page | 101

agar plates and bacteria CFUs were counted after 18 hours incubation at 37 °C. To

determine the intracellular bacteria population, the infected cells were incubated with 500

μg/mL of gentamicin–penicillin G (Sigma-Aldrich, St. Louis, MO) in complete DMEM for

1 hour to selectively kill extracellular bacteria. The antibiotic containing medium was then

removed and the cells washed for 3 times in 1X PBS. Intracellular bacteria were then

released by the addition of 0.1% Triton–X100 (Sigma-Aldrich, St. Louis, MO) solution and

mechanically disrupted with a cell scraper. Appropriate dilutions were then plated on BHI

agar plates and bacteria CFUs were counted after 18 hours incubation at 37 °C. To ensure

that all the extracellular bacteria were killed, an aliquot of the medium was taken after 1

hour of antibiotic treatment, diluted and then plated on BHI agar plates to ensure that there

was no bacterial growth.

3.4.8 Immunofluorescence:

Following infection, coverslips with cells were washed 3 times in 1X PBS and fixed with

4% paraformaldehyde at 4 °C for 15 minutes. Cells were then permeabilized with 0.1%

Triton X–100 (Sigma-Aldrich, St. Louis, MO) (actin) or 0.1% saponin (endosomal

compartments) 15 minutes at room temperature and washed 3 times in 1X PBS or 1X PBS

with 0.1% saponin respectively. Cells were then blocked using a solution of 2% BSA in

1X PBS with 0.1% saponin. For actin labelling, the phalloidin–Alexa Fluor 568 conjugate

(Thermo Fisher Scientific, Singapore) was diluted 1:40 in 1X PBS. For antibody labelling

of endosomal compartments, antibody solutions were diluted in 1X PBS with 0.1% saponin

at a 1:10 dilution for mouse α-LAMP-1 (ab25630, Abcam, Cambridge, UK), 1:50 for rabbit

α-EEA1-Alexa Fluor 647 (ab196186, Abcam, Cambridge, UK), 1:100 for rabbit α-M6PR-

568 (ab202535, Abcam, Cambridge, UK) or a 1:100 for rabbit α-Cathepsin D (ab75852,

Abcam, Cambridge, UK) and subjected to overnight incubation at 4 °C. Coverslips were

Page | 102

then washed 3 times in 1XPBS with 0.1% saponin the following day. Coverslips were then

incubated with a 1:500 dilution of the following secondary antibodies (Thermo Fisher

Scientific, Singapore): goat α-Mouse IgG (H+L) Alexa Fluor Plus 647, goat α-Rabbit IgG

(H+L) Alexa Fluor Plus 647, goat anti-Rabbit IgG (H+L) Alexa Fluor 568, goat anti-Mouse

IgG (H+L) Alexa Fluor 568 for 1 hour at room temperature. Coverslips were then washed

3 times in 1X PBS with 0.1% sapnonin, and incubated with a 1:500 dilution of Hoechst

33342 (Thermo Fisher Scientific, Singapore) for 20 minutes at room temperature.

Coverslips were then subjected to a final wash, 3 times with 1XPBS with 0.1% saponin and

2 times with 1X PBS. Coverslips were then mounted with SlowFade™ Diamond Antifade

(Thermo Fisher Scientific, Singapore) and sealed with clear nail polish.

3.4.9 Microscopy:

Confocal Laser Scanning Microscopy (CLMS): Confocal images were then acquired with

a Zeiss Elyra PS.1 inverted laser scanning confocal microscope (Carl Zeiss, Göttingen,

Germany) equipped with three Helium/Neon-lasers (633 nm, 594 nm and 561 nm), one

Argon-ion laser (458-514 nm) and one Diode laser (405 nm) using the Zeiss Zen Black

2012 SP2 software suite. Laser power and gain were kept constant between experiments.

Control labeling experiments were performed in parallel infected cells. Z-stacked images

were processed using Zen 2.1 (Carl Zeiss, Göttingen, Germany).

3.4.10 Stable cell line generation:

pLamp1-mCherry (Addgene plasmid #45147) (Addgene, Cambridge) and EF1α-mCherry-

N1 plasmid (Thermo Fisher Scientific, Singapore) were isolated using the Monarch®

Plasmid Miniprep Kit (New England BioLabs Inc., USA), according to manufacturer’s

instructions. The Lamp1 gene was then sub cloned into the pEF1α-mCherry-N1 vector

Page | 103 using the In-Fusion® HD Cloning Kit (Clontech, Takara, Japan), according to manufacturer’s instructions. The plasmid construct was then transformed into DH5α

Stellar™ competent cells by incubating at 42°C for 1 minute. Colony PCR was performed to select for successfully transformed colonies with the desired construct and primers used for the cloning and subsequent verification are shown in Table 3.1. EF1α Lamp1-mCherry plasmid was extracted from successful transformants with the Monarch® Plasmid

Miniprep Kit (New England BioLabs Inc., USA), according to manufacturer’s instructions.

Keratinocytes were grown in 6-well tissue culture plates as previously described, where each well was seeded with 2 ×105 cells. 2.5 μg of plasmid DNA was transfected into keratinocytes using Lipofectamine 3000 (Invitrogen; Thermo Fisher Scientific, Singapore), according to manufacturer’s instructions. The culture media was replaced after 6 hours of incubation, followed by a subsequent replacement 18 hours later. Keratinocytes were then subjected to Geneticin selection (1 mg/ml) (Invitrogen; Thermo Fisher Scientific,

Singapore) to select for transfected clones. Following which, clones that were stably over- expressing Lamp1-mCherry were subjected to validation by western blotting and flow cytometry. Clonal populations were selected and subjected to fluorescence activated cell sorting (FACS) to ensure that the entire population were expressing the fluorescent construct.

Primer Annealing Purpose Primer Sequences (5’-3’) Type Temperature (°C)

Lamp1 sub- F 72°C ATGGCGGCCCCCGGCA

cloning R 72°C GATAGTCTGGTAGCCTGCGTGACTCCTCT

Lamp1 F 72.7°C GACAGCGCTACCATGGCGGCCCC linker R 70.5°C GCTCACCATGGTGGCGATAGTCTGGT addition

Page | 104

pEF1-α

mCherry- F 63.9°C GCCACCATGGTCAGCAAGGGC

N1 R 61.5°C GGTAGCGCTAGCGTCACGACA

linearization

Colony F 60.8°C TCCATTTCAGGTGTCGTGACGCT PCR R 61.5°C CTACTTGTACAGCTCCATGCCG verification

Table 3.1: Primers used for generation and verification of the EF1α Lamp1-mCherry construct.

3.4.11 Correlative light and electron microscopy

Keratinocytes stably expressing Lamp1-mCherry were grown on 35mm glass-bottom

dishes (MatTec Corp., Ashland, MA) and infected with fluorescent E. faecalis at an MOI

of 100:1. Infected cells were washed to remove extracellular bacteria and imaged live in

colourless DMEM (Thermo Fisher Scientific, Singapore) supplemented with 10% FBS

(Thermo Fisher Scientific, Singapore) using a CorrSight confocal spinning disk microscope

(FEI Company, Oregon, US). Cells were fixed using 2.5% glutaraldehyde (prepared in

0.1M cacodylate buffer pH 7.4 with 2.5mM CaCl2) for 1 hour at room temperature and then

washed five times with 0.1M cacodylate buffer. Cells were then post-fixed with 2%

osmium tetraoxide (prepared in 0.1M cacodylate buffer pH 7.4 with 2.5mM CaCl2) for 1

hour at 4 °C and then washed five times with 0.1M cacodylate buffer. Fixed cells were then

incubated with 1% low molecular weight tannic acid (prepared in 0.1M cacodylate buffer

pH 7.4) for 45 minutes at room temperature and then washed five times with 0.1M

cacodylate buffer followed by an additional wash with distilled water for five times. Fixed

cells were then incubated in a 2% uranyl acetate solution (prepared in water) overnight at

4 °C. Fixed cells were then dehydrated in an ice-cold graded ethanol series (once with 20%,

50%, 70%, 90%, 100%) together with gentle agitation. Fixed cells were subjected to a final

Page | 105

dehydration step by washing five times with 100% ethanol at room temperature. The

coverslip was then detached from the MatTek dishes and infiltrated with a graded series of

Durcupan and ethanol (once with 1:2, 1:1, 2:1 for 30 minutes each and 100% overnight).

Cells were then subjected to a final exchange of Durcupan for four times at 50 °C before

curing for 72 hours. Samples were then trimmed and cut before viewing with a JSM-

6360LV (JEOL, Tokyo, Japan) transmission electron microscope. Electron micrographs

were recorded on a Tecnai T12 (FEI Company, Oregon, US) at 120 kV equipped with a

4k×4k Eagle CCD camera (FEI Company, Oregon, US).

3.4.12 Statistical analysis:

Statistical analysis was done using Prism 6.05 (Graphpad, San Diego, CA). Data from

gentamicin protection assays and pharmacological inhibitors of internalization were

reported as the mean ±SD, consisting of at least three independent experiments. The two-

way analysis of variance (ANOVA) Sidak multiple comparison test was used to statistically

determine whether there were any significant differences between the bacterial CFU

between keratinocytes treated with pharmacological inhibitors and untreated keratinocytes.

In all cases, a p value of ≤0.001 was considered statistically significant.

3.4.13 Ethics statement

All procedures were approved and performed in accordance with the Institutional

Animal Care and Use Committee (IACUC) in Nanyang Technological University,

School of Biological Sciences (ARF SBS/NIEA0198Z).

Page | 106

3.5 Contributions:

I performed all the experiments and data analyses presented in this section. We would like

to thank Alexander Ludwig for his assistance with the correlative-light and electron

microscopy. Part of the results here, were published as a peer-reviewed article, in the

Journal of Infectious Diseases. The remaining results are unpublished work.

A Enterococcus faecalis modulates immune activation and slows healing during

wound infection

Kelvin Kian Long Chong1,2, *, Wei Hong Tay1,3, *, Baptiste Janela4, Mei Hui Adeline

Yong1,5, Tze Horng Liew1, Leigh Madden6, Damien Keogh1, Timothy Mark Sebastian

Barkham7, Florent Ginhoux4, David Laurence Becker6, Kimberly A. Kline1,4

1Singapore Centre for Environmental Life Sciences Engineering, Nanyang Technological University, SBS- B1n-27, 60 Nanyang Drive, Singapore 637551; 2Nanyang Institute of Technology in Health and Medicine (NITHM), Research Techno Plaza, Nanyang Technological University, XFrontiers Block #02-07, 50 Nanyang Drive, Singapore 637553; 3Singapore Centre for Environmental Life Sciences Engineering, Interdisciplinary Graduate School, Nanyang Technological University, SBS-B2n-27, 60 Nanyang Drive, Singapore 637551; 4Singapore Immunology Network (SIgN), Agency for Science, Technology and Research (A*STAR), 8A Biomedical Grove, Singapore 138648; 5School of Biological Sciences, Nanyang Technological University, SBS-B2n-27, 60 Nanyang Drive, Singapore 637551; 6Lee Kong Chian School of Medicine, Nanyang Technological University, 11 Mandalay Road, Singapore 308232; 7Department of Laboratory Medicine, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, Singapore 308433 *These authors contributed equally to this work.

Page | 107

3.6 References for Chapter 3:

Baldassarri L, Cecchini R, Bertuccini L, Ammendolia MG, Iosi F, Arciola CR et al (2001). Enterococcus spp. produces slime and survives in rat peritoneal macrophages. Med Microbiol Immunol 190: 113- 120.

Baldassarri L, Bertuccini L, Creti R, Filippini P, Ammendolia MG, Koch S et al (2005). Glycosaminoglycans mediate invasion and survival of Enterococcus faecalis into macrophages. Journal of Infectious Diseases 191: 1253-1262.

Bastiat-Sempe B, Love JF, Lomayesva N, Wessels MR (2014). Streptolysin O and NAD-glycohydrolase prevent phagolysosome acidification and promote group A Streptococcus survival in macrophages. mBio 5: e01690-01614.

Bertuccini L, Ammendolia MG, Superti F, Baldassarri L (2002). Invasion of HeLa cells by Enterococcus faecalis clinical isolates. Med Microbiol Immunol 191: 25-31.

Bulet P, Stocklin R, Menin L (2004). Anti-microbial peptides: from invertebrates to vertebrates. Immunol Rev 198: 169-184.

Campoccia D, Testoni F, Ravaioli S, Cangini I, Maso A, Speziale P et al (2015). Orthopedic implant- infections. Incompetence of Staphylococcus epidermidis, Staphylococcus lugdunensis and Enterococcus faecalis to invade osteoblasts. J Biomed Mater Res A.

Chong KKL, Tay WH, Janela B, Yong AMH, Liew TH, Madden L et al (2017). Enterococcus faecalis modulates immune activation and slows healing during wound infection. J Infect Dis.

Clemens DL, Lee BY, Horwitz MA (2004). Virulent and avirulent strains of Francisella tularensis prevent acidification and maturation of their phagosomes and escape into the cytoplasm in human macrophages. Infect Immun 72: 3204-3217.

Cornejo E, Schlaermann P, Mukherjee S (2017). How to rewire the host cell: A home improvement guide for intracellular bacteria. The Journal of Cell Biology 216: 3931. de Boer P, Hoogenboom JP, Giepmans BN (2015). Correlated light and electron microscopy: ultrastructure lights up! Nat Methods 12: 503-513.

Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA et al (2008). Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiol 8: 43.

Drevets DA, Canono BP, Leenen PJ, Campbell PA (1994). Gentamicin kills intracellular Listeria monocytogenes. Infection and Immunity 62: 2222-2228.

Page | 108

Elhadidy M, Zahran E (2014). Biofilm mediates Enterococcus faecalis adhesion, invasion and survival into bovine mammary epithelial cells. Letters in applied microbiology 58: 248-254.

Flannagan RS, Cosio G, Grinstein S (2009). Antimicrobial mechanisms of phagocytes and bacterial evasion strategies. Nat Rev Microbiol 7: 355-366.

Flannagan RS, Heit B, Heinrichs DE (2016). Intracellular replication of Staphylococcus aureus in mature phagolysosomes in macrophages precedes host cell death, and bacterial escape and dissemination. Cellular microbiology 18: 514-535.

Garzoni C, Francois P, Huyghe A, Couzinet S, Tapparel C, Charbonnier Y et al (2007). A global view of Staphylococcus aureus whole genome expression upon internalization in human epithelial cells. BMC Genomics 8: 171.

Garzoni C, Kelley WL (2009). Staphylococcus aureus: new evidence for intracellular persistence. Trends Microbiol 17: 59-65.

Gentry-Weeks CR, Karkhoff-Schweizer R, Pikis A, Estay M, Keith JM (1999a). Survival of Enterococcus faecalis in mouse peritoneal macrophages. Infection and immunity 67: 2160-2165.

Gentry-Weeks CR, Karkhoff-Schweizer R, Pikis A, Estay M, Keith JM (1999b). Survival of Enterococcus faecalis in mouse peritoneal macrophages. Infect Immun 67: 2160-2165.

Ghigo E, Capo C, Tung CH, Raoult D, Gorvel JP, Mege JL (2002). Coxiella burnetii survival in THP-1 monocytes involves the impairment of phagosome maturation: IFN-gamma mediates its restoration and bacterial killing. J Immunol 169: 4488-4495.

Giacometti A, Cirioni O, Schimizzi AM, Del Prete MS, Barchiesi F, D'Errico MM et al (2000). Epidemiology and microbiology of surgical wound infections. J Clin Microbiol 38: 918-922.

Gjodsbol K, Christensen JJ, Karlsmark T, Jorgensen B, Klein BM, Krogfelt KA (2006). Multiple bacterial species reside in chronic wounds: a longitudinal study. International wound journal 3: 225-231.

Gresham HD, Lowrance JH, Caver TE, Wilson BS, Cheung AL, Lindberg FP (2000). Survival of Staphylococcus aureus inside neutrophils contributes to infection. J Immunol 164: 3713-3722.

Haas A (2007). The phagosome: compartment with a license to kill. Traffic (Copenhagen, Denmark) 8: 311-330.

Horn J, Stelzner K, Rudel T, Fraunholz M (2017). Inside job: Staphylococcus aureus host-pathogen interactions. Int J Med Microbiol.

Humphries WH, Szymanski CJ, Payne CK (2011). Endo-Lysosomal Vesicles Positive for Rab7 and LAMP1 Are Terminal Vesicles for the Transport of Dextran. PLoS One 6: e26626.

Page | 109

Jarry TM, Cheung AL (2006). Staphylococcus aureus escapes more efficiently from the phagosome of a cystic fibrosis bronchial epithelial cell line than from its normal counterpart. Infect Immun 74: 2568-2577.

Jolly AL, Takawira D, Oke OO, Whiteside SA, Chang SW, Wen ER et al (2015). Pseudomonas aeruginosa-Induced Bleb-Niche Formation in Epithelial Cells Is Independent of Actinomyosin Contraction and Enhanced by Loss of Cystic Fibrosis Transmembrane-Conductance Regulator Osmoregulatory Function. MBio 6: e02533-02514.

Kempf VA, Trebesius K, Autenrieth IB (2000). Fluorescent In situ hybridization allows rapid identification of microorganisms in blood cultures. J Clin Microbiol 38: 830-838.

Keogh D, Tay WH, Ho YY, Dale JL, Chen S, Umashankar S et al (2016). Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and Polymicrobial Infection. Cell Host Microbe 20: 493-503.

Klebanoff SJ (2005). Myeloperoxidase: friend and foe. J Leukoc Biol 77: 598-625.

Kubica M, Guzik K, Koziel J, Zarebski M, Richter W, Gajkowska B et al (2008). A potential new pathway for Staphylococcus aureus dissemination: the silent survival of S. aureus phagocytosed by human monocyte-derived macrophages. PLoS One 3: e1409.

Mempel M, Schnopp C, Hojka M, Fesq H, Weidinger S, Schaller M et al (2002). Invasion of human keratinocytes by Staphylococcus aureus and intracellular bacterial persistence represent haemolysin-independent virulence mechanisms that are followed by features of necrotic and apoptotic keratinocyte cell death. The British journal of dermatology 146: 943-951.

Millan D, Chiriboga C, Patarroyo MA, Fontanilla MR (2013). Enterococcus faecalis internalization in human umbilical vein endothelial cells (HUVEC). Microb Pathog 57: 62-69.

Olmsted SB, Dunny GM, Erlandsen SL, Wells CL (1994). A plasmid-encoded surface protein on Enterococcus faecalis augments its internalization by cultured intestinal epithelial cells. J Infect Dis 170: 1549-1556.

Osterlund A, Engstrand L (1997). An intracellular sanctuary for Streptococcus pyogenes in human tonsillar epithelium--studies of asymptomatic carriers and in vitro cultured biopsies. Acta Otolaryngol 117: 883-888.

Pizarro-Cerda J, Cossart P (2006). Bacterial adhesion and entry into host cells. Cell 124: 715-727.

Pizarro-Cerdá J, Cossart P (2006). Bacterial Adhesion and Entry into Host Cells. Cell 124: 715-727.

Qazi SN, Harrison SE, Self T, Williams P, Hill PJ (2004). Real-time monitoring of intracellular Staphylococcus aureus replication. J Bacteriol 186: 1065-1077.

Page | 110

Rabinovitch M (1995). Professional and non-professional phagocytes: an introduction. Trends in cell biology 5: 85-87.

Ray K, Marteyn B, Sansonetti PJ, Tang CM (2009). Life on the inside: the intracellular lifestyle of cytosolic bacteria. Nat Rev Microbiol 7: 333-340.

Ribet D, Cossart P (2015). How bacterial pathogens colonize their hosts and invade deeper tissues. Microbes and infection / Institut Pasteur 17: 173-183.

Settembre C, Fraldi A, Medina DL, Ballabio A (2013). Signals for the lysosome: a control center for cellular clearance and energy metabolism. Nature reviews Molecular cell biology 14: 283-296.

Silva MT (2012). Classical Labeling of Bacterial Pathogens According to Their Lifestyle in the Host: Inconsistencies and Alternatives. Front Microbiol 3: 71.

Sukumaran SK, Shimada H, Prasadarao NV (2003). Entry and Intracellular Replication of Escherichia coli K1 in Macrophages Require Expression of Outer Membrane Protein A. Infect Immun 71: 5951- 5961.

Thi EP, Lambertz U, Reiner NE (2012). Sleeping with the Enemy: How Intracellular Pathogens Cope with a Macrophage Lifestyle. PLoS Pathog 8: e1002551.

Voyich JM, Braughton KR, Sturdevant DE, Whitney AR, Said-Salim B, Porcella SF et al (2005). Insights into mechanisms used by Staphylococcus aureus to avoid destruction by human neutrophils. J Immunol 175: 3907-3919.

Wells CL, Jechorek RP, Maddaus MA, Simmons RL (1988). Effects of clindamycin and metronidazole on the intestinal colonization and translocation of enterococci in mice. Antimicrob Agents Chemother 32: 1769-1775.

Wells CL, Jechorek RP, Erlandsen SL (1990). Evidence for the translocation of Enterococcus faecalis across the mouse intestinal tract. J Infect Dis 162: 82-90.

Zou J, Shankar N (2016). The opportunistic pathogen Enterococcus faecalis resists phagosome acidification and autophagy to promote intracellular survival in macrophages. Cell Microbiol 18: 831- 843.

Page | 111

CHAPTER 4

Entry of Enterococcus faecalis into keratinocytes and immune-modulation

4.1 Introduction and motivation:

In Chapter 3, we demonstrated that E. faecalis can exist intracellularly, under both in vivo and in vitro conditions. We also showed that intracellular E. faecalis can survive, replicate and subsequently escape from host cells to cause re-infection. We also demonstrated that there was a lack of the lysosomal protease in the E. faecalis containing intracellular compartments. While these results provided new insights into the intracellular lifestyle, it does not address the following questions: 1) How is E. faecalis getting into the host cell? 2) Are there any virulence factors in E. faecalis that are important for the entry and/or intracellular survival? 3) Can intracellular E. faecalis be modulating the host immune response, helping them to survive/persist intracellularly for extended periods?

Separation of the intracellular compartment of the mammalian cell from the extracellular milieu is governed by the highly dynamic plasma membrane. However, the plasma membrane is hydrophobic and effectively prevents passive diffusion of molecules across. For cells to interact with extracellular signals, express or degrade receptors on the plasma membrane, and respond to environmental cues to trigger intracellular signaling cascades, the mammalian cell has to actively traffic molecules across the plasma membrane. Endocytosis is a process where internal membranes are produced de novo from the integral plasma membrane lipid bilayer, allowing lipids, integral proteins and extracellular fluids to enter into the cell (Conner and

Schmid 2003, Doherty and McMahon 2009). This process is important for the cells to regulate expression of proteins and molecules, as well as for nutrient acquisition from the extracellular milieu (Conner and Schmid 2003, Doherty and McMahon 2009). Several modalities of

Page | 112 endocytosis have been described and include caveolin-dependent, clathrin-dependent, caveolin and clathrin-independent, macropinocytosis and phagocytosis (Conner and Schmid 2003).

Both macropinocytosis and phagocytosis are actin-dependent processes and the other remaining modalities are actin-independent processes (Veiga and Cossart 2006).

It is not surprising that bacterial pathogens have evolved to exploit endocytosis to get into host cells, where they are protected from the extracellular milieu. Active triggering of macropinocytosis in non-phagocytic cells has been widely documented in numerous obligate intracellular bacterial pathogens such as Haemophilus influenza (Ketterer et al 1999), Listeria monocytogenes (Garcia-del Portillo and Finlay 1994), Salmonella enterica (Francis et al 1993),

Shigella flexneri (Clerc and Sansonetti 1987) and Yersinia spp. (Andor et al 2001, Eitel and

Dersch 2002) to induce entry. Previous studies on obligate intracellular bacteria such as

Listeria, Salmonella and have shown that they can mediate their entry into non-professional phagocytes through two main mechanisms: zipper or trigger (Figure 4.1) (Cossart and

Sansonetti 2004). Both mechanisms involve hijacking the host signaling mechanism resulting in a reorganization of the actin cytoskeleton of the plasma membrane, allowing bacterial entry into host cells (Cossart and Roy 2010, Pizarro-Cerda and Cossart 2009). L. monocytogenes and

Yersinia spp. utilize the zipper mechanism, using bacterial effectors that bind and engage host effectors such as cadherins and integrins, both of which are involved in cellular adhesion. The induction of signaling cascades consequently leads to strengthening of cell adhesion through actin reorganization and uptake of bacteria. L. monocytogenes achieves this with two surface proteins, which are known as internalins (InIA and InIB), to bind to E-cadherin and c-Met, respectively, on the host cells (Hamon et al 2006). Shigella flexneri and Salmonella typhimurium both utilize the trigger mechanism, where effectors are injected into the host cell through the type III secretion system (T3SS) (Ham et al 2011). These effectors induce signaling

Page | 113 cascades by activating Rho GTPases such as Rac1 and Cdc42 (Schlumberger and Hardt 2006).

This in turn causes rearrangement of the actin cytoskeleton, resulting in the formation of membrane ruffles that protrude outwards and eventually folds back, bringing bacteria into the host cell (Cossart and Roy 2010, Pizarro-Cerdá and Cossart 2006).

Figure 4.1: “Trigger” and “zipper” mechanisms utilized by intracellular bacteria to invade non- phagocytic cells. Upon entry into host cells, these microorganisms then employ various strategies to persist within an intracellular vacuole and/or escape from the vacuole later. Reproduced with permission from (Ham et al 2011).

Unlike bacterial invasion which was described above, extracellular bacteria can also enter into host cells independent of the injection of bacterial effectors. In S. aureus, a classical extracellular opportunistic pathogen, uptake into endothelial cells is mediated by fibronectin- binding protein A (FnBPA) interaction with integrins (Schröder et al 2006). This is similar to

S. pyogenes, where the fibronectin-binding protein (SfbI) can mediate the adhesion to epithelial cells and mediate uptake (Molinari et al 1997). In uropathogenic Escherichia coli, bacterial internalization into bladder epithelial cells is dependent on the mannose-binding adhesin FimH, which is expressed on the tip of Type I pili (Martinez et al 2000). Pseudomonas aeruginosa encodes Type IV pili, which recognize and bind the disaccharide β-GalNAc(1-4)βGal (Suh et

Page | 114 al 2001), resulting in subsequent internalization into epithelial cells (Comolli et al 1999).

Collectively, these studies suggest that bacterial virulence factors that are important for attachment can be important for uptake into host cells as well. These studies also suggest that attachment precedes internalization, at least in the case of traditionally extracellular bacteria.

Though there are very limited studies investigating virulence factors that are important for E. faecalis uptake into host cells, an early study has shown that the surface protein, aggregation substance, resulted in increased uptake into epithelial cells (Olmsted et al 1994). A subsequent study also showed that aggregation substance was important for adhesion and survival within human macrophages (Sussmuth et al 2000). However, a subsequent study suggested that this virulence factor was dispensable for uptake into endothelial cells (Millan et al 2013).

Altogether, these studies warrant an in-depth investigation into Enterococcal virulence factors that might be important for adhesion and/or uptake into epithelial cells.

While entry into the host cell can help bacteria to escape from the extracellular milieu, they will still encounter the intracellular host defense systems (Akira et al 2006). One of the many ways that intracellular bacteria subvert the host defense system is by targeting mediators of the inflammatory response. These mediators are known as cytokines and they are small, soluble proteins or peptides that form the communication network of the host immune system

(Dinarello 2007). Several studies have shown that intracellular bacteria can down-regulate pro- inflammatory cytokines while up-regulating anti-inflammatory cytokines (Cornejo et al 2017,

Stewart and Cookson 2016, Thi et al 2012). Studies from our lab and others have also demonstrated that E. faecalis can modulate the host immune response (Chong et al 2017, Park et al 2008, Tien et al 2017). Work in a mouse model of wound infection has shown that the levels of pro-inflammatory cytokines IL-2, IL-5, IL-10, IL12-p70, CCL11, IFN-γ and CSF2 was similar to mock-infected controls at 3 days post infection, despite the presence of relatively

Page | 115 high E. faecalis titers, suggesting active immune suppression (Chong et al 2017). Building on these studies, we hypothesized that E. faecalis can similarly modify the host-immune response intracellularly, allowing the bacteria to persist for extended periods.

In this Chapter, I have shown that E. faecalis entry into host cells is dependent on microfilament polymerization, P13K signaling, as well as Rac GTPase activity. To date, our work has found no evidence of receptor-mediated endocytosis, as a targeted screen of mutants with surface expressing virulence factors did not result in any decrease in recoverable intracellular CFU.

Instead, our studies suggest that E. faecalis may enter into host cells in a passive manner, independent of the tested virulence factors, likely through macropinocytosis as a consequence of solute acquisition. Furthermore, our results indicate that E. faecalis can modulate the host immune response, where pro-inflammatory cytokines are down-regulated at high MOIs. These results form the basis of future work to investigate potential bacterial/host protein interactions within the cytokine signaling pathways.

Page | 116

4.2 Results:

4.2.1 E. faecalis entry into keratinocytes is dependent on microfilament

polymerization but independent of microtubule polymerization

Bacterial-mediated actin mediated rearrangement of the host cytoskeleton is well described

for both Gram-negative and -positive bacteria species (Haglund and Welch 2011), resulting

in the uptake of bacteria into the host cells. Previous work on E. faecalis in non-phagocytic

cell lines (endothelial and epithelial) suggest that uptake is dependent on microfilament and

microtubule polymerization (Bertuccini et al 2002, Millan et al 2013). However, these

studies utilized different cell lines and did not conclusively show how E. faecalis is entering

host cells. The authors also speculated that entry could be through macropinocytosis or

receptor mediated endocytosis.

To determine whether microfilament and microtubule polymerization is important for E.

faecalis entry into keratinocytes, we performed a modified gentamicin protection assay

where keratinocytes were pre-treated with the inhibitors of microfilament polymerization,

cytochalasin-D and lantruculin-A. The latter inhibits actin polymerization by binding to

monomeric G-actin in a 1:1 ratio (Coue et al 1987) while the former inhibits actin

polymerization by binding to F-actin, preventing the addition of monomeric G-actin

(Flanagan and Lin 1980). While inhibition of microfilament polymerization did not alter

bacteria adhesion to keratinocytes (Figure 4.2A), we observed a 100-fold decrease in

recoverable intracellular bacteria, demonstrating that microfilament polymerization is

important for the entry process (Figure 4.2B). Studies have shown that bacteria can enter

and/or invade non-phagocytic cells through stimulation of the endogenous uptake processes

like macropinocytosis or phagocytosis. Actin polymerization is critical for both processes

(Haglund and Welch 2011).

Page | 117

A ) B )

A d h e r e d B a c te r ia (M O I 1 0 0 :1 ) In te r n a liz e d B a c te r ia (M O I 1 0 0 :1 )

1 0 7 n .s 1 0 0 0 0 0 0 n .s L a n A (0 .2 5  g /m L ) * 1 0 0 0 0 0 C y to D (1  g /m L )

* C o n tro l l l 1 0 0 0 0

n .s

m

m

/ /

6 U

U 1 0 1 0 0 0 F

F *

C C 1 0 0

1 0

1 0 5 1 r r r r r r u u u u u u o o o o o o H H H H H H 1 2 3 1 2 3

T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 h r )

Figure 4.2: E. faecalis entry is dependent on actin polymerization. Actin inhibitors do not affect (A) bacterial adhesion but affects the (B) entry (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=4) of each biological replicate.

Previous studies have demonstrated that the entry of some bacteria species is dependent on microtubules (Dhakal and Mulvey 2009, McGee et al 2003, Oelschlaeger et al 1993) and we sought to test whether this is necessary for E. faecalis using the microtubule inhibitor, colchicine. Colchicine inhibits microtubule polymerization by binding to both α and β- tubulin (Andreu and Timasheff 1982). The use of this inhibitor functions as a good control to test for the specificity of the actin dependent entry, as previous studies have demonstrated that microtubule depolymerisation by colchicine results in the increase of filamentous actin

(F-actin) and activates stress fiber formation (Danowski 1989, Jung et al 1997). We showed that E. faecalis entry is independent on microtubules, where the selective inhibition of microtubule polymerization using colchicine did not decrease or limit the number of internalized bacteria (Figure 4.3). Rather, we observed a slight 3-5-fold increase in the number of internalized bacteria. Since we demonstrated that internalization of E. faecalis is an actin-dependent process, the increase in intracellular bacteria recovered can be attributed to an increase of F-actin, brought about by the colchicine treatment, that was previously described and characterized (Jung et al 1997).

Page | 118

A ) B ) A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

8 n .s 1 0 n .s 1 0 0 0 0 0 * C o lc h ic in e (1 0  g /m L ) 1 0 0 0 0 C o n tro l n .s *

1 0 7

l l

1 0 0 0

m

m

/

/

U

U F

F 1 0 0

C C 1 0 6

1 0

1 0 5 1

r r r r r r u u u u u u o o o o o o H H H H H H 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 h r)

Figure 4.3: E. faecalis internalization is microtubule independent. Colchicine does not affect (A) bacterial adhesion or decrease (B) entry (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

4.2.2 E. faecalis entry into keratinocytes is clathrin and caveolae independent

Clathrin mediated endocytosis has been reported to have an upper limit of only 200 nm in

cargo size (Ehrlich et al 2004). This size limit effectively eliminates the possibility of

bacterial internalization by this pathway, since the majority of bacteria exceed the limit of

the cargo size. However, recent findings by several groups have shown that clathrin plays

an important role in bacteria endocytosis where bacterial bound host receptors were taken

up by clathrin mediated endocytosis (Veiga and Cossart 2005, Veiga and Cossart 2006).

The large GTPase dynamin is important for the formation of clathrin-coated vesicles (CCV)

and can be inhibited by dynasore, which inhibits the GTPase activity of dynamin

(Kirchhausen et al 2008). We showed that dynasore does not inhibit E. faecalis adhesion

or affect internalization in a time-dependent manner (Figure 4.4). Though there was a slight

decrease in the number of internalized bacteria across all timepoints, we found that

dynasore kills >20% of all cells by 3 hpi (data not shown). These results suggest that

dynamin is not essential for the uptake of E. faecalis into keratinocytes.

Page | 119

A ) B ) A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 8 1 0 0 0 0 0 0 n . s * C o n tro l 1 0 0 0 0 0 D y n a s o re n . s (2 5  g /m l)

1 0 7 1 0 0 0 0

l l

n . s

m

m

/

/ U

U 1 0 0 0

F

F

C C 1 0 6 1 0 0

1 0

1 0 5 1

r r r r r r u u u u u u o o o o o o h h h h h h 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 h r)

Figure 4.4: E. faecalis internalization is dynamin independent. Dynamin does not affect (A) bacterial adhesion or decrease (B) entry (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

Next, to test whether E. faecalis entry involves caveolae mediated endocytosis, we treated keratinocytes with nystatin. Nystatin selectively affects caveolae mediated endocytosis by binding sterols, causing caveolae and cholesterol disassembly in the plasma membrane

(Ivanov 2008, Rothberg et al 1992). We showed that there was no significant reduction in

E. faecalis attachment and internalization (Figure 4.5).

A ) B ) A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 8 n . s 1 0 0 0 0 0 0

n . s C o n tro l 1 0 0 0 0 0 N y s ta tin (2 5  g /m l) n . s n . s n . s

1 0 7 1 0 0 0 0

l l

n . s

m

m

/

/ U

U 1 0 0 0

F

F

C C 1 0 6 1 0 0

1 0

1 0 5 1

r r r r r r u u u u u u o o o o o o h h h h h h 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 h r)

Figure 4.5: E. faecalis internalization is dynamin independent. Dynamin does not affect (A) bacterial adhesion or decrease (B) entry (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

Page | 120

Altogether, our results suggest to us that E. faecalis is likely entering into cells via

macropinocytosis, since chemical inhibitors of the clathrin and caveolae mediated

pathways failed to significantly decrease intracellular bacterial counts.

4.2.3 E. faecalis entry into keratinocytes is dependent on PI3K signalling

Since macropinocytosis involves other proteins such as phosphoinositide 3-kinase (PI3K),

Ras and PAK1 (Doherty and McMahon 2009), we tested the effect of a well-known

covalent inhibitor of the P13K, wortmannin (Wymann et al 1996), on E. faecalis uptake.

Again, we showed that wortmannin does not impact bacteria adhesion to the treated cells

(Figure 4.6A). We also demonstrate that there was approximately 90% reduction, or a 10-

fold decrease, in recoverable intracellular CFU numbers, in a time dependent manner, as

compared to the untreated controls, suggesting that P13K signalling is important for the

uptake process (Figure 4.6B).

A ) B )

A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 9 1 0 0 0 0 0 * C o n tro l n . s 1 0 0 0 0 * W o rtm a n n in (0 .1  g /m l)

1 0 8

l l

n . s 1 0 0 0

m

m / / *

7 n . s U

U 1 0

F F

1 0 0

C C

1 0 6 1 0

1 0 5 1

r r r r r r u u u u u u o o o o o o h h h h h h 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 h r)

Figure 4.6: E. faecalis internalization is dependent on P13K signalling. Wortmannin does not affect (A) bacterial adhesion or (B) but does decrease entry compared to untreated controls (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

Page | 121

4.2.4 Classical E. faecalis virulence factors are dispensable for attachment and uptake

into keratinocytes

Given that previous studies in E. coli, P. aeruginosa, S. aureus and S. pyogenes have

found that bacterial adhesins are important for uptake, we performed a targeted screen of

known adhesins or adhesins with homology to other bacteria species. We showed that the.

faecalis mutants defective for collagen-binding (Ace) (Figure 4.7A) and fibronectin-

binding protein A (EfbA) (Figure 4.7B) were not deficient for either attachment and uptake

into keratinocytes. Ace has been shown to bind to collagen Type I and IV and laminin

(Nallapareddy and Murray 2006, Nallapareddy et al 2011a). Likewise, EfbA has been

shown to bind to fibronectin, collagen Type I and IV with high affinities (Singh et al 2015).

The lack of a reduction suggests that entry into keratinocytes were not dependent on

collagen, fibronectin and laminin binding. Interestingly, there was a significant increase in

intracellular bacteria CFU counts at both 2 hpi and 3 hpi for both the collagen-binding Ace

mutant and the fibronectin-binding protein A mutant. This increase might be due to a

compensatory effect, resulting in the upregulation of other virulence genes.

Page | 122

A ) A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 8 1 0 0 0 0 0 0 * 1 0 7 n . s 1 0 0 0 0 0 * n . s

1 0 6 O G 1 R F T n ::9 1 7 7 1 2 l l *

5 m 1 0 0 0 0 O G 1 R F

m /

/ 1 0

U

U F

F 1 0 4 C C 1 0 0 0

1 0 3 1 0 0 1 0 2

1 0 1 1 0 r r r r r r h h h h h h 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 )

B ) A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 8 1 0 0 0 0 0 0

1 0 7 1 0 0 0 0 0 *

1 0 6 O G 1 R F T n ::1 0 6 3 3 7 6

l l 1 0 0 0 0 5 m O G 1 R F

m * /

/ 1 0

U

U F

F 1 0 4 C C 1 0 0 0

1 0 3 1 0 0 1 0 2

1 0 1 1 0 r r r r r r h h h h h h 1 2 3 1 2 3 T im e (p o s t in fe c tio n ) T im e (p o s t in fe c tio n + 1 )

Figure 4.7: E. faecalis internalization is independent of collagen, fibronectin and laminin. The collagen- binding ace::Tn does not affect (A) bacterial adhesion or decrease entry. Similarly, the fibronectin-binding protein efbA::Tn does not affect (B) bacterial adhesion or decrease entry (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

Furthermore, we demonstrate that the endocarditis and biofilm associated pili (Ebp) null mutant were similarly not defective for attachment or uptake. Rather, we observed an increased in intracellular bacteria CFU counts for the ebp null mutant (Figure 4.8). A previous study showed that ebpABC was important in mediating binding to fibrinogen and collagen, but not fibronectin (Nallapareddy et al 2011a, Nallapareddy et al 2011b).

However, our results contrast that of previous studies where a pilus null mutant in

Streptococcus gallolyticus resulted in decreased opsonophagocytosis by macrophages

(Danne et al 2014). Our results suggest that the importance of pili could be dependent on the host cell type, whether it is a non-phagocytic or phagocytic cell. It also suggests to us that E. faecalis pili could possibly be attaching to other host proteins.

Page | 123

A d h e re d B a c te ria (M O I 1 0 0 :1 ) In te rn a liz e d B a c te ria (M O I 1 0 0 :1 )

1 0 8 1 0 0 0 0 0 0 * * * O G 1 R F  e b p A B C 7 1 0 * 1 0 0 0 0 0 O G 1 R F *

1 0 6

l l

5 1 0 0 0 0 m

m 1 0

/

/

U U

4 F F *

1 0 1 0 0 0

C C 1 0 3 1 0 0 1 0 2

1 0 1 1 0

r r r r r r h h h h h h 1 2 3 1 2 3 T im e T im e (p o s t in fe c tio n + 1 )

Figure 4.8: E. faecalis internalization is increased in a pilus null mutant. The ΔebpABC pilus null mutant exhibited increased attachment to keratinocytes and had increased bacterial CFU (p≤0.05, Student’s t-test). Solid lines indicate the median for each data set and error bars represent the standard deviation of means (n=3) of each biological replicate.

4.2.5 Intracellular E. faecalis can modulate the immune response in keratinocytes

Bacterial infection of host cells will invariably trigger the release of one or more cytokines

and chemokines. As mentioned in the Chapter introduction, we hypothesized that

intracellular E. faecalis infection could result in the modulation of the host

cytokine/chemokine production, thereby allowing for the survival and/or replication of the

internalized bacteria. To test our hypothesis, we infected keratinocytes with E. faecalis for

3 hours and then added antibiotics for 1 hour or 21 hours, killing off all extracellular

bacteria, before obtaining the supernatant to perform a multiplex analysis of proteins

(MAP). The cell-free supernatant would therefore contain soluble factors released over the

span of 1 hour or 21 hours from intracellularly infected keratinocytes.

At both early timepoints of 4 hpi (3 hr infection + 1 hr of antibiotic protection) and late

timepoints of 24 hpi (3 hr infection + 21 hr of antibiotic protection), we observed that

cytokine and chemokine production was altered in the infected cells (Figure 4.9, 4.10). At

early timepoints of 4 hpi, we did not observe any significant increase in pro-inflammatory

cytokines IL-6 and TNF-α at a MOI of 100:1. We also observed a significant down-

Page | 124 regulation of IL-8 and VEGF by 8-fold and 2-fold respectively. There was also a significant upregulation of FGF and GM-CSF by 2-fold.

IL-1b VGEF 3.000 IL-1ra TNF-a 2.500 IL-6 RANTES 2.000 IL-7 1.500 PDGF-bb 1.000 IL-8 Media 0.500 MOI 100:1 MIP-1a 0.000 IL-10 MOI 10:1 MCP-1 IL-12(p70) MOI 1:1

IP-10 IL-15 IFN-g Eotaxin GM-CSF FGF G-CSF

Figure 4.9: Intracellular E. faecalis modulates the cytokine and chemokine secretion by keratinocytes. Radar plot of data obtained from multiplex analysis of cytokines and chemokines from supernatant of 3 hpi cells after 1 hr of gentamicin protection at various multiplicity of infection (MOI, MOI 100:1, MOI 10:1 and MOI 1:1. Data presented as fold-change and are normalized against the media controls.

Page | 125

IL-1b VGEF 4.500 IL-1ra TNF-a 4.000 IL-6 3.500 RANTES 3.000 IL-7 2.500 2.000 PDGF-bb 1.500 IL-8 Media 1.000 0.500 MOI 100:1 MIP-1a 0.000 IL-10 MOI 10:1 MCP-1 IL-12(p70) MOI 1:1

IP-10 IL-15 IFN-g Eotaxin GM-CSF FGF G-CSF

Figure 4.10: Multiplex analysis of cytokines and chemokines from supernatant of 3 hpi cells after 21 hrs of gentamicin protection at various multiplicity of infection (MOI, MOI 100:1, MOI 10:1 and MOI 1:1. Data presented as fold-change and are normalized against the media controls.

However, by 24 hpi, we observed a significant increase in IL-1RA and IL-6, by approximately 2-fold, with no concurrent increase in IL-8, at MOI 100:1, when compared to the media controls. We also observed a significant decrease of MCP-1, at MOI 100:1, by 2-fold, which was not observed at early timepoints of 4 hpi. Levels of IL-1b, TNF-α,

VEGF and GM-CSF, at MOI 100:1, were also similar to the media controls. Our data suggests that intracellular E. faecalis can modulate the host immune response, where there is a lack of immune activation despite intracellular infection.

Page | 126

4.3 Discussions:

In the recent years, the fields of cellular biology and microbiology have converged due to increasing evidence that bacteria were manipulating the host to mediate uptake for the 1) escape from immuno-surveillance, 2) evasion of antibiotics and complement, and/or 3) replication and survival intracellularly. Gaining access to an intracellular niche within the host and surviving the host immune response afford bacteria a window of opportunity to persist and cause re- infection or disease later on. Understanding the mammalian host processes to bacterial internalization will undoubtedly aid in the targeting and treatment of infection.

In our results, we showed that E. faecalis is entering into keratinocytes in a manner that is dependent on actin polymerization and P13K signaling. Our results suggest that it is entering via a macropinocytosis-dependent manner and not via clathrin- or caveolae-mediated endocytosis because we did not see a reduction in intracellular bacteria CFU when we treated the cells with specific inhibitors for these two pathways. Our findings contradict a previous study involving HeLa cells and E. faecalis, where the authors argued that E. faecalis entry into host cells was dependent on clathrin-mediated endocytosis, in addition to macropinocytosis

(Bertuccini et al 2002). The authors of that study saw a decrease in intracellular bacterial CFU when host cells were treated with inhibitors of microtubule polymerization and cytosolic acidification (Bertuccini et al 2002). However, the concentrations of colchicine, an inhibitor of microtubule polymerization, was much higher than what was used in our study, by as much as

4-fold. We showed that 10 µg/ml of colchicine was sufficient to cause massive depolymerization of the microtubule network in infected cells, and concentrations higher than this might negatively affect the host cells, thereby accounting for the reduction in recoverable

CFU. Moreover, the authors only used ammonium chloride and amiloride, both of which causes cytosolic acidification, as inhibitors for clathrin-mediated endocytosis. While cytosolic

Page | 127 acidification is important for the budding of clathrin-coated pits from the membrane (Hansen et al 1993), cytosolic acidification has also been shown to impact macropinocytosis, blocking the uptake of fluid-phase markers (Cosson et al 1989, Ivanov 2008, Suzuki and Namiki 2007,

Swanson and Watts 1995). Altogether, these caveats could have resulted in misinterpretation of the results and we suggest that entry of E. faecalis is solely mediated by macropinocytosis.

Macropinocytosis is defined as the non-selective intake of external fluid from the environment and is achieved by the formation of wavy sheet-like extensions of the plasma membrane known as macropinosomes (Lim and Gleeson 2011, Swanson and Watts 1995). Chemical inhibition of actin polymerization by cytochalasin D and PI3K signaling by wortmannin has been shown to specifically affect macropinocytosis but not clathrin mediated endocytosis (Araki et al 1996,

Gaidarov et al 1999, Swanson and Watts 1995). Our results are consistent with earlier work in non-phagocytic epithelial cell lines (Bertuccini et al 2002) and phagocytic cell lines

(Baldassarri et al 2004, Baldassarri et al 2005, Sussmuth et al 2000) that cytoskeletal arrangements in the form of actin polymerization is important for entry of E. faecalis into host cells.

Our targeted screen of known Enterococcal virulence factors that are known to be important for adhesion and disease outcome in other niches, such as the urinary tract and the heart, showed that these virulence factors are dispensable for uptake and entry into keratinocytes. Our results suggest and reinforce earlier findings that classical virulence factors in E. faecalis that are important for urinary tract infections and endocarditis are less important in wound infections, because we do not see an attenuation or out-competition (Chong et al 2017). Other studies have suggested that other virulence factors such as antiphagocytic factors (IgA, IgM,

IgG proteases, short-chain fatty acids and capsule) as well as tissue damaging enzymes

Page | 128

(, elastase, gelatinase and protease) might also be important for infection in wounds

(Duerden 1994).

Keratinocytes are the major cell type of the integumentary system and can secrete chemotactic cytokines to recruit macrophages, neutrophils and other immune cells (Grone 2002, Pasparakis et al 2014, Pivarcsi et al 2005). At 24 hpi, where infected cells have large numbers of intracellular bacteria, we fail to see a robust immune response, and most of the inflammatory cytokines/chemokines are similar to the uninfected control, with the exception of IL-6. We also see a large spike in the IL-1ra at 24 hpi, with a concurrent decrease in MCP-1. IL-1ra is a competitive inhibitor of IL-1 and is important for the regulation of the pro-inflammatory effects of IL-1 (Banda et al 2005). IL-1ra can inhibit the production of IL-6 and IL-8 by IL-1α through inhibition of the phosphorylation of p38 (Banda et al 2005). While studies have reported that

IL-1ra concentrations increases with differentiation of keratinocytes (Bigler et al 1992,

Hammerberg et al 1998), we did not observe any elevated levels of IL-1ra in the uninfected media controls at 4 hpi and 24 hpi, suggesting that the increase in IL-1ra was due to the infection. Both IL-6 and IL-8 are potent mediators of the pro-inflammatory response and their main function is to amplify the local inflammatory response. IL-6 has been shown to play an important role in keratinocyte and lymphocyte proliferation/activation (Oppenheim et al 1991,

Sehgal 1990). IL-6 is released upon tissue damage and wounding and is not spontaneously produced in normal skin (Sugawara et al 2001). Elevated levels of IL-6 at high MOs at late timepoints could be due to death of infected cells, resulting in the release. IL-8 is involved in the recruitment of phagocytic cells and is a powerful chemoattractant. Previous studies have shown that maximal expression of IL-8 requires three different mitogen-activated protein kinases (MAPKs), Erk, p38 and JNK (Hoffmann et al 2002), and the lack of expression at 4hpi and 24 hpi suggests that E. faecalis could possibly be modulating these kinases. p38 is involved

Page | 129 in the phosphorylation of AP-1 and NF-κB, resulting in transcriptional activation (Olson et al

2007, Tanos et al 2005). Previous work done in our lab has shown that E. faecalis suppresses

NF-κB signalling in murine macrophages (Tien et al 2017) and suggests that the intracellular

E. faecalis in keratinocytes might be suppressing it through the p38 pathway.

VEGF is an important for the regulation of keratinocyte proliferation, migration and survival

(Johnson and Wilgus 2014). VEGF is also important for the recruitment of macrophages

(Murakami et al 2006, Murakami et al 2008, Muramatsu et al 2010). The downregulation of

VEGF at early timepoints of infection might result in impaired recruitment of macrophages and affect keratinocyte development. MCP-1 is chemotactic for basophils, monocytes and T- cells and is associated with inflammatory skin diseases such as psoriasis (Alam et al 1992, Carr et al 1994). A study has also shown that it is involved in the recruitment of dendritic and

Langerhans cells (Nakamura et al 1995). Downregulation of MCP-1 is associated with an anti- inflammatory state and was observed at late timepoints. Collectively, the lack of an inflammatory response at 24 hpi even at high MOIs suggest that E. faecalis is possibly modulating the host immune response.

Altogether, our work suggests that E. faecalis entry into keratinocytes occurs in an actin- dependent and P13K signaling dependent manner. Our findings also suggest that adhesins are dispensable for entry into host cells, as we do not see a decrease in intracellular CFU when we tested known mutants. Furthermore, we also show that infected cells display differential cytokine and chemokine secretion profiles, suggesting the possibility that intracellular E. faecalis can module the host immune response. In Chapter 5 in this thesis, we will attempt to summarize our findings as well as discuss about the future perspectives in this field of work.

Page | 130

4.4 Materials and methods:

4.4.1 Bacterial strains and growth conditions:

Enterococcus faecalis strain OG1RF (ATCC47077) and its derivative strains were

grown using Bacto Brain Heart Infusion (BHI) (Becton, Dickinson and Company,

Franklin Lakes, NJ). Unless otherwise stated, bacteria strains were streaked from

glycerol stocks stored at -80 °C, inoculated and grown overnight statically for 16-20

hours in 20 ml of liquid broth as indicated above. Cells were harvested by centrifugation

at 5000 RPM (4 °C) for 4 minutes. The supernatant was discarded, and the pellet

washed with 1 ml of 1X sterile phosphate buffered saline (PBS) before being harvested.

The pellet was then resuspended in 3 ml of 1X sterile PBS. The concentration of the

cell suspension was then adjusted to an optical density (OD) of 0.7 for E. faecalis at

600 nm. Following adjustment, the cell population was equivalent to 2–3×108 cells per

ml. Transposon insertional mutants were obtained from Gary M. Dunny, University of

Minnesota (Kristich et al 2008).

4.4.2 Cell culture:

The spontaneously immortalized human keratinocyte cell line, HaCaT (AddexBio, San

Diego, CA) was routinely cultured at 37 °C in a 5% CO2 humidified atmosphere. Cells

were grown and maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco;

Thermo Fisher Scientific, Singapore) with 10% heat-inactivated fetal bovine serum

(FBS) (PAA, GE Healthcare, Singapore), 100 U of penicillin–streptomycin (Gibco,

Thermo Fisher Scientific, Singapore). The culture medium was replaced once every

three days, and upon reaching 80% confluency, cultures were passaged. Passaging was

achieved by treatment with 0.25% trypsin-EDTA (Gibco; Thermo Fisher Scientific,

Page | 131

Singapore) for 6 minutes and seeding cells at a density of 2×106 cells/T75 flask (Nunc;

Thermo Fisher Scientific, Singapore).

4.4.3 Invasion Assay:

Keratinocytes were seeded at a density of 5×105 cells/well in a 6-well tissue culture

plate (Nunc; Thermo Fisher Scientific, Singapore) and grown for 3 days at 37 °C in a

5% CO2 humidified atmosphere. After 3 days, each well had approximately 1–1.5×106

keratinocytes. Keratinocytes were infected with bacteria at a multiplicity of infection

(MOI) of 100, 10 or 1 for 1–3 hours. Following infection, the media was aspirated, and

the cells washed thrice in 1X PBS and thereafter incubated with 500 μg/mL of

gentamicin–penicillin G (Sigma-Aldrich, St. Louis, MO) in complete DMEM for 1

hour to selectively kill extracellular bacteria. The antibiotic containing medium was

then removed and the cells washed for 3 times in 1X PBS. Intracellular bacteria were

then released by the addition of 0.1% Triton–X100 (Sigma-Aldrich, St. Louis, MO)

solution and mechanically disrupted with a cell scraper. Appropriate dilutions were then

plated on BHI agar plates and bacteria CFUs were counted after 18 hours incubation at

37 °C.

4.4.4 Drug treatments:

Pharmacological inhibitors were added to cells 30 minutes prior to any infection and

maintained throughout the course of the infection. Actin polymerization was inhibited

by 1 μg/mL of cytochalasin D or 250 ng/mL of latrunculin A. Microtubule

polymerization was inhibited by 10 μg/ml of colchicine. Stock solutions were dissolved

in DMSO unless otherwise indicated and stored at −20 °C. Stock solutions of

cytochalasin D, latrunculin A, colchicine, dynasore, nystatin and wortmannin were at 1

Page | 132

mg/mL, 100 μg/mL, 10 mg/mL, 25 mg/mL, 25 mg/mL and 10 mg/mL respectively.

Addition of pharmacological inhibitors at the concentrations indicated had no effect on

keratinocyte viability, where cytotoxicity assays using the alamarBlue™ cell viability

reagent (Invitrogen; Thermo Fisher Scientific, Singapore) was similar to the untreated

control. Addition of pharmacological inhibitors at the concentrations indicated also had

no effect on bacteria viability, where growth kinetics and colony forming units (CFUs)

count were similar to the untreated bacteria control. All chemical inhibitors were

purchased from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO), unless otherwise

stated.

4.4.5 Cytokine Luminex MAP analysis

Luminex MAP analysis was performed using the Bio-Plex Pro™ Human Cytokine 27-

plex Assay (Cat #: M500KCAF0Y, Bio-Rad, California, USA) as previously described.

4.4.6 Statistical analysis:

Statistical analysis was done using Prism 6.05 (Graphpad, San Diego, CA). Data from

gentamicin protection assays and pharmacological inhibitors of internalization were

reported as the mean ±SD, consisting of at least three independent experiments. The

Student’s T-test was used to statistically determine whether there were any significant

differences between the bacterial CFU between keratinocytes treated with

pharmacological inhibitors and untreated keratinocytes. In all cases, a p value of ≤0.05

was considered statistically significant.

Page | 133

4.5 Contributions:

I performed all the experiments and data analyses presented in this section. We would like

to thank Joyce Pek, Sng Wan Xin and Eunice Heng for their assistance with the chemical

inhibitor assays.

Page | 134

4.6 References for Chapter 4:

Akira S, Uematsu S, Takeuchi O (2006). Pathogen Recognition and Innate Immunity. Cell 124: 783- 801.

Alam R, Lett-Brown MA, Forsythe PA, Anderson-Walters DJ, Kenamore C, Kormos C et al (1992). Monocyte chemotactic and activating factor is a potent histamine-releasing factor for basophils. J Clin Invest 89: 723-728.

Andor A, Trulzsch K, Essler M, Roggenkamp A, Wiedemann A, Heesemann J et al (2001). YopE of Yersinia, a GAP for Rho GTPases, selectively modulates Rac-dependent actin structures in endothelial cells. Cell Microbiol 3: 301-310.

Andreu JM, Timasheff SN (1982). Tubulin bound to colchicine forms polymers different from microtubules. Proc Natl Acad Sci U S A 79: 6753-6756.

Araki N, Johnson MT, Swanson JA (1996). A role for phosphoinositide 3-kinase in the completion of macropinocytosis and phagocytosis by macrophages. J Cell Biol 135: 1249-1260.

Baldassarri L, Bertuccini L, Ammendolia MG, Cocconcelli P, Arciola CR, Montanaro L et al (2004). Receptor-mediated endocytosis of biofilm-forming Enterococcus faecalis by rat peritoneal macrophages. Indian J Med Res 119 Suppl: 131-135.

Baldassarri L, Bertuccini L, Creti R, Filippini P, Ammendolia MG, Koch S et al (2005). Glycosaminoglycans mediate invasion and survival of Enterococcus faecalis into macrophages. J Infect Dis 191: 1253-1262.

Banda NK, Guthridge C, Sheppard D, Cairns KS, Muggli M, Bech-Otschir D et al (2005). Intracellular IL- 1 receptor antagonist type 1 inhibits IL-1-induced cytokine production in keratinocytes through binding to the third component of the COP9 signalosome. J Immunol 174: 3608-3616.

Bertuccini L, Ammendolia MG, Superti F, Baldassarri L (2002). Invasion of HeLa cells by Enterococcus faecalis clinical isolates. Med Microbiol Immunol 191: 25-31.

Bigler CF, Norris DA, Weston WL, Arend WP (1992). Interleukin-1 receptor antagonist production by human keratinocytes. J Invest Dermatol 98: 38-44.

Carr MW, Roth SJ, Luther E, Rose SS, Springer TA (1994). Monocyte chemoattractant protein 1 acts as a T-lymphocyte chemoattractant. Proc Natl Acad Sci U S A 91: 3652-3656.

Chong KKL, Tay WH, Janela B, Yong AMH, Liew TH, Madden L et al (2017). Enterococcus faecalis modulates immune activation and slows healing during wound infection. J Infect Dis.

Page | 135

Clerc P, Sansonetti PJ (1987). Entry of Shigella flexneri into HeLa cells: evidence for directed phagocytosis involving actin polymerization and myosin accumulation. Infect Immun 55: 2681-2688.

Comolli JC, Waite LL, Mostov KE, Engel JN (1999). Pili Binding to Asialo-GM1 on Epithelial Cells Can Mediate Cytotoxicity or Bacterial Internalization by Pseudomonas aeruginosa. Infect Immun 67: 3207-3214.

Conner SD, Schmid SL (2003). Regulated portals of entry into the cell. Nature 422: 37-44.

Cornejo E, Schlaermann P, Mukherjee S (2017). How to rewire the host cell: A home improvement guide for intracellular bacteria. The Journal of Cell Biology 216: 3931.

Cossart P, Sansonetti PJ (2004). Bacterial invasion: the paradigms of enteroinvasive pathogens. Science 304: 242-248.

Cossart P, Roy CR (2010). Manipulation of host membrane machinery by bacterial pathogens. Curr Opin Cell Biol 22: 547-554.

Cosson P, de Curtis I, Pouyssegur J, Griffiths G, Davoust J (1989). Low cytoplasmic pH inhibits endocytosis and transport from the trans-Golgi network to the cell surface. J Cell Biol 108: 377-387.

Coue M, Brenner SL, Spector I, Korn ED (1987). Inhibition of actin polymerization by latrunculin A. FEBS Lett 213: 316-318.

Danne C, Dubrac S, Trieu-Cuot P, Dramsi S (2014). Single Cell Stochastic Regulation of Pilus Phase Variation by an Attenuation-like Mechanism. PLoS Pathog 10: e1003860.

Danowski BA (1989). Fibroblast contractility and actin organization are stimulated by microtubule inhibitors. Journal of cell science 93 ( Pt 2): 255-266.

Dhakal BK, Mulvey MA (2009). Uropathogenic Escherichia coli invades host cells via an HDAC6- modulated microtubule-dependent pathway. The Journal of biological chemistry 284: 446-454.

Dinarello CA (2007). Historical Review of Cytokines. Eur J Immunol 37: S34-S45.

Doherty GJ, McMahon HT (2009). Mechanisms of endocytosis. Annu Rev Biochem 78: 857-902.

Duerden BI (1994). Virulence factors in anaerobes. Clin Infect Dis 18 Suppl 4: S253-259.

Ehrlich M, Boll W, van Oijen A, Hariharan R, Chandran K, Nibert ML et al (2004). Endocytosis by Random Initiation and Stabilization of Clathrin-Coated Pits. Cell 118: 591-605.

Page | 136

Eitel J, Dersch P (2002). The YadA protein of Yersinia pseudotuberculosis mediates high-efficiency uptake into human cells under environmental conditions in which invasin is repressed. Infect Immun 70: 4880-4891.

Flanagan MD, Lin S (1980). Cytochalasins block actin filament elongation by binding to high affinity sites associated with F-actin. J Biol Chem 255: 835-838.

Francis CL, Ryan TA, Jones BD, Smith SJ, Falkow S (1993). Ruffles induced by Salmonella and other stimuli direct macropinocytosis of bacteria. Nature 364: 639-642.

Gaidarov I, Santini F, Warren RA, Keen JH (1999). Spatial control of coated-pit dynamics in living cells. Nat Cell Biol 1: 1-7.

Garcia-del Portillo F, Finlay BB (1994). Salmonella invasion of nonphagocytic cells induces formation of macropinosomes in the host cell. Infect Immun 62: 4641-4645.

Grone A (2002). Keratinocytes and cytokines. Vet Immunol Immunopathol 88: 1-12.

Haglund CM, Welch MD (2011). Pathogens and polymers: microbe-host interactions illuminate the cytoskeleton. The Journal of cell biology 195: 7-17.

Ham H, Sreelatha A, Orth K (2011). Manipulation of host membranes by bacterial effectors. Nat Rev Microbiol 9: 635-646.

Hammerberg C, Bata-Csorgo Z, Voorhees JJ, Cooper KD (1998). IL-1 and IL-1 receptor antagonist regulation during keratinocyte cell cycle and differentiation in normal and psoriatic epidermis. Arch Dermatol Res 290: 367-374.

Hamon M, Bierne H, Cossart P (2006). Listeria monocytogenes: a multifaceted model. Nat Rev Microbiol 4: 423-434.

Hansen SH, Sandvig K, van Deurs B (1993). Clathrin and HA2 adaptors: effects of potassium depletion, hypertonic medium, and cytosol acidification. J Cell Biol 121: 61-72.

Hoffmann E, Dittrich-Breiholz O, Holtmann H, Kracht M (2002). Multiple control of interleukin-8 gene expression. J Leukoc Biol 72: 847-855.

Ivanov AI (2008). Pharmacological inhibition of endocytic pathways: is it specific enough to be useful? Methods Mol Biol 440: 15-33.

Johnson KE, Wilgus TA (2014). Vascular Endothelial Growth Factor and Angiogenesis in the Regulation of Cutaneous Wound Repair. Adv Wound Care 3: 647-661.

Page | 137

Jung HI, Shin I, Park YM, Kang KW, Ha KS (1997). Colchicine activates actin polymerization by microtubule depolymerization. Mol Cells 7: 431-437.

Ketterer MR, Shao JQ, Hornick DB, Buscher B, Bandi VK, Apicella MA (1999). Infection of primary human bronchial epithelial cells by Haemophilus influenzae: macropinocytosis as a mechanism of airway epithelial cell entry. Infect Immun 67: 4161-4170.

Kirchhausen T, Macia E, Pelish HE (2008). Use of dynasore, the small molecule inhibitor of dynamin, in the regulation of endocytosis. Methods Enzymol 438: 77-93.

Kristich CJ, Nguyen VT, Le T, Barnes AMT, Grindle S, Dunny GM (2008). Development and Use of an Efficient System for Random mariner Transposon Mutagenesis To Identify Novel Genetic Determinants of Biofilm Formation in the Core Enterococcus faecalis Genome. Appl Environ Microbiol 74: 3377-3386.

Lim JP, Gleeson PA (2011). Macropinocytosis: an endocytic pathway for internalising large gulps. Immunol Cell Biol 89: 836-843.

Martinez JJ, Mulvey MA, Schilling JD, Pinkner JS, Hultgren SJ (2000). Type 1 pilus-mediated bacterial invasion of bladder epithelial cells. The EMBO Journal 19: 2803-2812.

McGee K, Holmfeldt P, Fallman M (2003). Microtubule-dependent regulation of Rho GTPases during internalisation of Yersinia pseudotuberculosis. FEBS letters 533: 35-41.

Millan D, Chiriboga C, Patarroyo MA, Fontanilla MR (2013). Enterococcus faecalis internalization in human umbilical vein endothelial cells (HUVEC). Microb Pathog 57: 62-69.

Molinari G, Talay SR, Valentin-Weigand P, Rohde M, Chhatwal GS (1997). The fibronectin-binding protein of Streptococcus pyogenes, SfbI, is involved in the internalization of group A streptococci by epithelial cells. Infect Immun 65: 1357-1363.

Murakami M, Iwai S, Hiratsuka S, Yamauchi M, Nakamura K, Iwakura Y et al (2006). Signaling of vascular endothelial growth factor receptor-1 tyrosine kinase promotes rheumatoid arthritis through activation of monocytes/macrophages. Blood 108: 1849-1856.

Murakami M, Zheng Y, Hirashima M, Suda T, Morita Y, Ooehara J et al (2008). VEGFR1 tyrosine kinase signaling promotes lymphangiogenesis as well as angiogenesis indirectly via macrophage recruitment. Arterioscler Thromb Vasc Biol 28: 658-664.

Muramatsu M, Yamamoto S, Osawa T, Shibuya M (2010). Vascular endothelial growth factor receptor-1 signaling promotes mobilization of macrophage lineage cells from bone marrow and stimulates solid tumor growth. Cancer Res 70: 8211-8221.

Page | 138

Nakamura K, Williams IR, Kupper TS (1995). Keratinocyte-derived monocyte chemoattractant protein 1 (MCP-1): analysis in a transgenic model demonstrates MCP-1 can recruit dendritic and Langerhans cells to skin. J Invest Dermatol 105: 635-643.

Nallapareddy SR, Murray BE (2006). Ligand-signaled upregulation of Enterococcus faecalis ace transcription, a mechanism for modulating host-E. faecalis interaction. Infect Immun 74: 4982-4989.

Nallapareddy SR, Sillanpaa J, Mitchell J, Singh KV, Chowdhury SA, Weinstock GM et al (2011a). Conservation of Ebp-type pilus genes among Enterococci and demonstration of their role in adherence of Enterococcus faecalis to human platelets. Infect Immun 79: 2911-2920.

Nallapareddy SR, Singh KV, Sillanpää J, Zhao M, Murray BE (2011b). Relative Contributions of Ebp Pili and the Collagen Adhesin Ace to Host Extracellular Matrix Protein Adherence and Experimental Urinary Tract Infection by Enterococcus faecalis OG1RF. Infect Immun 79: 2901-2910.

Oelschlaeger TA, Guerry P, Kopecko DJ (1993). Unusual microtubule-dependent endocytosis mechanisms triggered by Campylobacter jejuni and Citrobacter freundii. Proceedings of the National Academy of Sciences of the United States of America 90: 6884-6888.

Olmsted SB, Dunny GM, Erlandsen SL, Wells CL (1994). A plasmid-encoded surface protein on Enterococcus faecalis augments its internalization by cultured intestinal epithelial cells. J Infect Dis 170: 1549-1556.

Olson CM, Hedrick MN, Izadi H, Bates TC, Olivera ER, Anguita J (2007). p38 mitogen-activated protein kinase controls NF-kappaB transcriptional activation and tumor necrosis factor alpha production through RelA phosphorylation mediated by mitogen- and stress-activated protein kinase 1 in response to Borrelia burgdorferi antigens. Infect Immun 75: 270-277.

Oppenheim JJ, Zachariae CO, Mukaida N, Matsushima K (1991). Properties of the novel proinflammatory supergene "intercrine" cytokine family. Annu Rev Immunol 9: 617-648.

Park SY, Shin YP, Kim CH, Park HJ, Seong YS, Kim BS et al (2008). Immune evasion of Enterococcus faecalis by an extracellular gelatinase that cleaves C3 and iC3b. J Immunol 181: 6328-6336.

Pasparakis M, Haase I, Nestle FO (2014). Mechanisms regulating skin immunity and inflammation. Nature Reviews Immunology 14: 289.

Pivarcsi A, Nagy I, Kemeny L (2005). Innate immunity in the skin: how keratinocytes fight against pathogens. Curr Immunol Rev 1: 29-42.

Pizarro-Cerda J, Cossart P (2009). Listeria monocytogenes membrane trafficking and lifestyle: the exception or the rule? Annu Rev Cell Dev Biol 25: 649-670.

Pizarro-Cerdá J, Cossart P (2006). Bacterial Adhesion and Entry into Host Cells. Cell 124: 715-727.

Page | 139

Rothberg KG, Heuser JE, Donzell WC, Ying Y-S, Glenney JR, Anderson RGW (1992). Caveolin, a protein component of caveolae membrane coats. Cell 68: 673-682.

Schlumberger MC, Hardt W-D (2006). Salmonella type III secretion effectors: pulling the host cell's strings. Curr Opin Microbiol 9: 46-54.

Schröder A, Schröder B, Roppenser B, Linder S, Sinha B, Fässler R et al (2006). Staphylococcus aureus Fibronectin Binding Protein-A Induces Motile Attachment Sites and Complex Actin Remodeling in Living Endothelial Cells. Mol Biol Cell 17: 5198-5210.

Sehgal PB (1990). Interleukin-6: molecular pathophysiology. J Invest Dermatol 94.

Singh KV, La Rosa SL, Somarajan SR, Roh JH, Murray BE (2015). The Fibronectin-Binding Protein EfbA Contributes to Pathogenesis and Protects against Infective Endocarditis Caused by Enterococcus faecalis. Infect Immun 83: 4487-4494.

Stewart MK, Cookson BT (2016). Evasion and interference: intracellular pathogens modulate caspase-dependent inflammatory responses. Nature Reviews Microbiology 14: 346.

Sugawara T, Gallucci RM, Simeonova PP, Luster MI (2001). Regulation and role of interleukin 6 in wounded human epithelial keratinocytes. Cytokine 15: 328-336.

Suh J-Y, Spyracopoulos L, Keizer DW, Irvin RT, Sykes BD (2001). Backbone Dynamics of Receptor Binding and Antigenic Regions of a Pseudomonas aeruginosa Pilin Monomer. Biochemistry 40: 3985- 3995.

Sussmuth SD, Muscholl-Silberhorn A, Wirth R, Susa M, Marre R, Rozdzinski E (2000). Aggregation substance promotes adherence, phagocytosis, and intracellular survival of Enterococcus faecalis within human macrophages and suppresses respiratory burst. Infect Immun 68: 4900-4906.

Suzuki K, Namiki H (2007). Cytoplasmic pH-dependent spreading of polymorphonuclear leukocytes: regulation by pH of PKC subcellular distribution and F-actin assembly. Cell Biol Int 31: 279-288.

Swanson JA, Watts C (1995). Macropinocytosis. Trends Cell Biol 5: 424-428.

Tanos T, Marinissen MJ, Leskow FC, Hochbaum D, Martinetto H, Gutkind JS et al (2005). Phosphorylation of c-Fos by members of the p38 MAPK family. Role in the AP-1 response to UV light. J Biol Chem 280: 18842-18852.

Thi EP, Lambertz U, Reiner NE (2012). Sleeping with the Enemy: How Intracellular Pathogens Cope with a Macrophage Lifestyle. PLoS Pathog 8: e1002551.

Page | 140

Tien BYQ, Goh HMS, Chong KKL, Bhaduri-Tagore S, Holec S, Dress R et al (2017). Enterococcus faecalis promotes innate immune suppression and polymicrobial catheter-associated urinary tract infection. Infect Immun.

Veiga E, Cossart P (2005). Listeria hijacks the clathrin-dependent endocytic machinery to invade mammalian cells. Nat Cell Biol 7: 894-900.

Veiga E, Cossart P (2006). The role of clathrin-dependent endocytosis in bacterial internalization. Trends Cell Biol 16: 499-504.

Wymann MP, Bulgarelli-Leva G, Zvelebil MJ, Pirola L, Vanhaesebroeck B, Waterfield MD et al (1996). Wortmannin inactivates phosphoinositide 3-kinase by covalent modification of Lys-802, a residue involved in the phosphate transfer reaction. Mol Cell Biol 16: 1722-1733.

Page | 141

CHAPTER 5

Summary and perspectives

5.1 Summary of findings:

Enterococci are important opportunistic pathogens that are responsible for a large number of infections including urinary tract infections, catheter-associated infections, bacteraemia, endocarditis and wound infections. While E. faecalis infections in some of these niches have been extensively characterized and studied, little is known about the pathogenesis of

Enterococci in wound infections. The aim of this thesis was to understand the pathogenesis of

Enterococcal associated wound infections. To this end, I undertook a series of experiments to investigate the consequence of Enterococcus-associated polymicrobial wound infections and uncover the reason of persistent Enterococci associated wound infections. My results demonstrate that the presence of Enterococci in wounds can strongly influence disease outcomes, necessitating the need to rethink and develop better treatment plans. In this concluding Chapter, I will briefly summarize my findings and contributions, culminating with future perspectives and implications to help further our understanding of Enterococcal pathogenesis in wounds.

Because E. faecalis is commonly co-isolated together with E. coli in wound infections (Dowd et al 2008, Giacometti et al 2000) and in catheter-associated urinary tract infections (CAUTI)

(Flores-Mireles et al 2015), we used an unbiased approach, mimicking physiological conditions of iron restriction during wound infection, to screen for Enterococcal factors that are responsible for augmented E. coli growth. We demonstrated that E. faecalis produces a metabolite, L-ornithine, that can be used by E. coli to kickstart their siderophore production process, allowing for enhanced virulence even under iron limited conditions (Keogh et al

Page | 142

2016). Our work is consistent with other studies, where polymicrobial infections were generally shown to result in higher bacterial titers and a poorer disease prognosis (Peters et al

2012, Tay et al 2016). This finding is also similar to other studies in the gut, where differential metabolite production has been shown to influence bacterial communities to cause dysbiosis

(Lee and Hase 2014, Nicholson et al 2012, Rooks and Garrett 2016), showing that primary/secondary metabolites secreted can play an important role in pathogenesis. However, polymicrobial interactions can be antagonistic in nature too. Studies have reported that E. coli produces indole, which inhibits AHL-regulated virulence factors in P. aeruginosa (Chu et al

2012) and attenuate virulence in S. aureus (Lee et al 2013), allowing E. coli to grow much better. Collectively, our result and others have highlighted the importance of metabolites in polymicrobial infections. While it is generally thought that enterococci display low level of virulence, as they are natural colonizers of the human gastrointestinal (GI) tract and have been widely used as probiotics (Arias and Murray 2012), our results suggest that their presence can affect the virulence of co-infecting organisms and therefore disease outcomes. [Chapter 2]

Although our work in Chapter 2 highlights the contribution of E. faecalis in polymicrobial wound infections, it does not explain why Enterococcal wound infections are persistent or the mechanisms behind this persistence. To address this knowledge gap, we used a mouse model of surgical wound infection to study the spatial and temporal dynamics of E. faecalis in this infection niche. We uncovered evidence of bacterial internalization into both immune and non- immune cells using this model, potentially accounting for the persistence. To further investigate this, we utilized a tissue culture model to study E. faecalis infection in keratinocytes, which is the dominant cell type in the epidermal layer which the bacteria would encounter in wounds. Our tissue culture model corroborated our animal model, showing that

E. faecalis can enter and survive in non-immune cells. Importantly, we also showed and

Page | 143 demonstrated that intracellular E. faecalis can replicate and avoid destruction despite being trafficked through the host endolysosomal pathway. Our findings demonstrate that E. faecalis possesses a dual intracellular/extracellular lifestyle, breaking the longstanding paradigm that it is an exclusive extracellular pathogen. [Chapter 3]

We subsequently sought to investigate the mechanisms of E. faecalis entry into keratinocytes and to determine the virulence factors involved. We demonstrated the importance of actin polymerization and P13K signalling in the uptake of E. faecalis into keratinocytes.

Furthermore, we also demonstrated that this uptake was independent of clathrin-mediated and caveolae-mediated endocytosis. We also demonstrated that classical E. faecalis virulence factors, in particular adhesins, are dispensable for the uptake process, reinforcing our previous work that these classical virulence factors are less important in wound infections (Chong et al

2017). Finally, based on previous findings from our group that E. faecalis is capable of immune-modulation, we embarked on cytokine profiling of infected keratinocytes. Our preliminary results suggest that internalized bacteria can modulate the host response, resulting in an anti-inflammatory state. These results may explain why E. faecalis can persist intracellularly, and account for chronic, non-healing wounds. [Chapter 4]

Altogether, results presented in Chapter 2-4 set the stage for several other avenues of investigation and will be discussed in the following perspectives section. Specific avenues of future research will be first discussed, followed by global outlooks in the field.

Page | 144

5.2 Future perspectives

The unbiased screen of Enterococcal factors responsible for augmented E. coli growth in

Chapter 2 represents only a partially characterized bed of novel information. While we only focused on Enterococcal mutants that exhibited a decreased biofilm/biomass phenotype during co-culture with E. coli in our screen, we also observed many mutants with significant increase in biofilm/biomass phenotype under iron limitation. Are there any potential benefits to either bacterial species in response to increase biofilm production under iron limitation? Some studies in single species settings have shown that low iron is correlated with increased biofilm production in Pseudomonas aeruginosa and Staphylococcus aureus (Johnson et al 2005,

Johnson et al 2008, Yang et al 2007). Why do bacteria upregulate biofilm production under low iron conditions? What about mixed species infections when there are more than two infecting organisms? High biofilm production will make it more difficult for substrates to diffuse across the biofilm and into the biofilm, making iron acquisition even more difficult under iron-limiting conditions. Solving this conundrum will allow us to better determine factors that govern biofilm production during mixed species infection. The presence of biofilms is a cause for concern because the resulting bacterial community often becomes tolerant to antibiotics (Costerton et al 1995, Flemming et al 2016, Hall-Stoodley et al 2004, Hall-Stoodley and Stoodley 2009, James et al 2008, Romling and Balsalobre 2012). Investigating the mechanism behind this increase in biofilm/biomass production during co-culture will allow us to develop better treatment strategies and to prevent antibiotic tolerance.

Next, while we showed that augmented E. coli biofilm growth in the presence of E. faecalis depends on the production of the E. coli siderophore enterobactin in response to L-ornithine, little is known about the remaining E. coli siderophores (yersiniabactin and salmochelin) expressed in the same strain. Could other siderophores be similarly upregulated in response to

Page | 145 other simple metabolites secreted by the host or other co-infecting bacteria? To explore this avenue of research, one could create transcriptional fusions of different fluorescent reporters to all the siderophore biosynthetic operons to study induction and switching. Such single-copy fusions could allow one to study the differential expression of siderophore biosynthetic operons over the course of infection in animals or during co-infection with other organisms in a quantitative manner. Another avenue of research could involve exploring how the bioavailability of nutrients and metabolites could influence siderophore production. A recent study has shown that in Vibrionaceae, growth in mannose and chitin preferentially induced the siderophore fluvibactin but not under growth of glucose (Giubergia et al 2016).

Given the broadening landscape of how microbial consortia can work together to cause infection, careful attention must be paid to the entire community present during the pathogenic cascade: from the time of inoculation to symptomology to resolution. Despite having shown that E. faecalis can influence E. coli virulence and exacerbate wound infections, little is known about the consequence of Enterococcal association with other microbial species. E. faecalis is commonly co-isolated with many other bacteria species in wounds, and we would expect that it could play similar roles in influencing disease outcome in association with other bacterial species as described in Chapter 2. Yet, due to the complexity of bacteria-host interactions and the difficulty in extricating the influences of one bacteria species on another in vivo, most polymicrobial infection studies to date have been limited to co-infection with two species.

Future work should move beyond the current two species co-infection models to integrate investigation of the molecular, spatial, and temporal dynamics after polymicrobial inoculation of the infection niche. The fields of oral microbiology and microbial ecology lead the way in mechanistic studies of these polymicrobial interactions, where lessons of how interspecies diversity promotes stress tolerance, how individual members of the community contribute to

Page | 146 biofilm development and resilience, and how diversity promotes human health have been demonstrated (Abusleme et al 2013, Belda-Ferre et al 2012, Griffen et al 2012, Kelvin Lee et al 2015, Lee et al 2014, Periasamy et al 2015, Wade 2013). Furthermore, the contribution of the host microbiome to infection dynamics is poorly understood. While it is increasingly clear that a diverse GI tract microbiome can limit infection, our understanding of the molecular interactions between the host and microbiota or between the microbiota and invading pathogens is poorly described. The native microbiomes of the skin, urinary tract, and lungs are even less well understood, and our appreciation of the role of these microbial communities during infection is its infancy (Belkaid and Segre 2014, Ingersoll and Albert 2013, Marsland and Gollwitzer 2014, SanMiguel and Grice 2015, Scales and Huffnagle 2013, van Rensburg et al 2015, Whiteside et al 2015). A more complete understanding of all of the microbial contributors to infection may ultimately impact diagnosis or treatment, wherein the presence of a subdominant or non-traditional organism at the site of infection, and/or in hosts with predisposing risk factors, might indicate a different course of treatment. This in turn paves way for personalized medicine, as there may be cases where specific targeting of microbial subsets, such as those that modulate the host response or augment the virulence of other pathogens, may drastically improve the disease outcome.

Our findings that E. faecalis has an intracellular lifestyle can potentially account for recurring wound infections and persistence in the wound, as discussed in Chapter 3. Our findings raise some new questions for future research. First, where is E. faecalis getting the nutrients and necessary amino acids from to replicate intracellularly? Studies in Salmonella have shown that the Salmonella containing vacuole interacts extensively with endocytic vacuoles, facilitating the acquisition of glucose from the extracellular environment (Abu Kwaik and Bumann 2015).

Salmonella can also hijack the host endosomal system, turning it into an extensive network of

Page | 147 interconnecting tubular vesicles to gain access to nutrients from endocytic materials (Liss et al

2017). Given that we observe large numbers of E. faecalis within Lamp1+ compartments, could

E. faecalis be doing a similar thing? Or, could E. faecalis be degrading the membranes of the compartments that they are bound in, allowing them access to nutrients in the cytosol? We postulate that this might also be the case, since we observe in TEM that the E. faecalis- containing compartments have a non-continuous membrane that appears to be disrupted and broken in some cases. Studies in other bacteria species such as Shigella flexneri and Listeria monocytogenes have shown that the intracellular bacteria breaks out of the compartments that they are enclosed in and escape to the cytosol of the host cell, where rapid proliferation occurs

(Hamon et al 2012, Mellouk and Enninga 2016).

Secondly, to fully characterize the spatial and temporal dynamics of E. faecalis intracellular infections, additional live cell imaging experiments should be performed. We propose transient transfections of keratinocytes with the various endocytic markers (EEA-1, LAMP-1, M6PR) and/or the lysosomal protease (Cathepsin D) to follow through the transition and progression of E. faecalis throughout the endolysosomal pathway. A previous study has also shown that intracellular E. faecalis can subvert the autophagy pathway in infected macrophages, by inhibition of LC3-II infection (Zou and Shankar 2016). Could this be a similar case for intracellular E. faecalis in keratinocytes? We also observed internalized E. faecalis in single walled compartments, and found no evidence of a double walled membrane that is representative of autophagosome, similar to the previous study in macrophages (Zou and

Shankar 2016). One possible avenue of exploration is the transfection of cells with RFP-LC3II to check whether the autophagy pathway is being triggered. Another possible avenue would be to perform live cell imaging on keratinocytes transfected with a Galectin-8 fluorescent reporter.

Galectin-8 is a cytosolic lectin that activates the autophagy pathway, upon binding to host

Page | 148 glycans on damaged vacuoles and compartments (Thurston et al 2012). Galectin-8 has been shown to detect vacuole damage to endosomes and lysosomes in Listeria, Salmonella and

Shigella (Thurston et al 2012). Since we observed damaged membranes in the E. faecalis containing compartments, and yet, do not observe any autophagosome, could E. faecalis be blocking Galectin-8 binding? Or does intracellular E. faecalis degrade the glycans within the compartment, resulting in a failure of activation even when the compartment is compromised at a later stage? Fully characterizing the host response to intracellular E. faecalis might provide critical insights or ideas as to how E. faecalis is possibly mediating the replication, survival and escape from the host cells.

Thirdly, despite a targeted screen of known adhesion mutants in Chapter 4, we did not uncover any bacterial genes that are important for uptake and/or intracellular survival. Our findings suggest that these virulence factors may be less important during this phase of infection. Future work should involve the use of transcriptomics to profile the expression of Enterococcal genes during the intracellular phase. Understanding which genes are necessary or important for entry, survival, replication and escape will provide critical information on Enterococcal pathogenesis in wounds. Are there genes with previously uncharacterized functions important for this intracellular phase? Are there any known genes that are important for both extracellular and intracellular infection? We also propose that transcriptomics should be done for both the bacteria and the host (dual-RNA sequencing). This is extremely beneficial and informative, as it will allow us to correlate the functions of bacterial effectors with outcomes in the host. In the first study that used this technique, they showed that the Salmonella PhoP-small RNA, PinT, was able to suppress activation of the JAK-STAT pathway in host cells (Westermann et al

2016). A recent study that capitalized on this method showed that the expression of virulence factors in Staphylococcus aureus was induced and dependent on the level of stress of the

Page | 149 infected host (Thänert et al 2017). This method might therefore uncover information about the susceptibility of the host cells to infection, as not all host cells contain intracellular bacteria.

The traditional classification of pathogens according to their infective lifestyles is quixotic and ambiguous, since classical extracellular pathogens might have an intracellular phase and that classical intracellular pathogens may have an extracellular phase when they escape or kill the host cells. Although this classification was useful in dictating antibiotic choice for treatments in majority of the cases, it is insufficient should the infecting organism have dual extracellular/intracellular phases. Antibiotic treatment might only be effective for the treatment of one phase, resulting in recurring, chronic infections. Our work has shown that the extracellular pathogen, E. faecalis, can enter, survive, replicate and escape from host cells in

Chapter 3. Further, we also show that their entry is mediated by the host, instead of manipulation of the host by the infecting organism in Chapter 4. Our work also raises possibilities that other extracellular pathogens in wounds could similarly have dual infective lifestyles, resulting in a non-healing chronic state of infection. Other commonly isolated bacteria species in wounds, such as Pseudomonas aeruginosa and Staphylococcus aureus, have been shown to have an intracellular lifestyle in cultured cells as well, where approximately 1-

10% of all cells are infected. S. aureus has been shown to be able to bind to host cells to trigger bacterial internalization via fibronectin-binding proteins (Schröder et al 2006). Moreover, given that S. aureus can actively trigger this uptake process, it presents the possibility that other organisms can also be taken up together with S. aureus. This is especially alarming, especially in the context of polymicrobial wounds, where other bacteria species are present. There could be additive effects of immune modulation or enhanced survival within the intracellular niche and could possibly explain why polymicrobial chronic wounds are so recalcitrant to treatment.

Page | 150

Future work should therefore explore these possibilities, using a mixture of both 2D, 3D-tissue culture models in addition to animal models.

Together, our studies have shown that the presence of Enterococci in infected wounds are a significant cause for concern and should not be ignored in the clinic. Our work provides evidence of cooperative interactions of Enterococci with other co-infecting microbial species outside of host cells in the wounds as well as demonstrating that it has a dual extracellular and intracellular lifestyle inside host cells. Our work will be useful to clinicians in managing persistent wound infections and could potentially result in the development of improved treatment strategies or therapeutics.

Page | 151

5.3 References for Chapter 5

Abu Kwaik Y, Bumann D (2015). Host Delivery of Favorite Meals for Intracellular Pathogens. PLoS Pathog 11: e1004866.

Abusleme L, Dupuy AK, Dutzan N, Silva N, Burleson JA, Strausbaugh LD et al (2013). The subgingival microbiome in health and periodontitis and its relationship with community biomass and inflammation. The ISME journal 7: 1016-1025.

Arias CA, Murray BE (2012). The rise of the Enterococcus: beyond vancomycin resistance. Nature reviews Microbiology 10: 266-278.

Belda-Ferre P, Alcaraz LD, Cabrera-Rubio R, Romero H, Simon-Soro A, Pignatelli M et al (2012). The oral metagenome in health and disease. The ISME journal 6: 46-56.

Belkaid Y, Segre JA (2014). Dialogue between skin microbiota and immunity. Science 346: 954-959.

Chong KKL, Tay WH, Janela B, Yong AMH, Liew TH, Madden L et al (2017). Enterococcus faecalis modulates immune activation and slows healing during wound infection. J Infect Dis.

Chu W, Zere TR, Weber MM, Wood TK, Whiteley M, Hidalgo-Romano B et al (2012). Indole production promotes Escherichia coli mixed-culture growth with Pseudomonas aeruginosa by inhibiting quorum signaling. Appl Environ Microbiol 78: 411-419.

Costerton JW, Lewandowski Z, Caldwell DE, Korber DR, Lappin-Scott HM (1995). Microbial biofilms. Annu Rev Microbiol 49: 711-745.

Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA et al (2008). Survey of bacterial diversity in chronic wounds using pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiol 8: 43.

Flemming HC, Wingender J, Szewzyk U, Steinberg P, Rice SA, Kjelleberg S (2016). Biofilms: an emergent form of bacterial life. Nat Rev Microbiol 14: 563-575.

Flores-Mireles AL, Walker JN, Caparon M, Hultgren SJ (2015). Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat Rev Microbiol 13: 269-284.

Giacometti A, Cirioni O, Schimizzi AM, Del Prete MS, Barchiesi F, D'Errico MM et al (2000). Epidemiology and microbiology of surgical wound infections. J Clin Microbiol 38: 918-922.

Giubergia S, Phippen C, Gotfredsen CH, Nielsen KF, Gram L (2016). Influence of Niche-Specific Nutrients on Secondary Metabolism in Vibrionaceae. Appl Environ Microbiol 82: 4035-4044.

Page | 152

Griffen AL, Beall CJ, Campbell JH, Firestone ND, Kumar PS, Yang ZK et al (2012). Distinct and complex bacterial profiles in human periodontitis and health revealed by 16S pyrosequencing. The ISME journal 6: 1176-1185.

Hall-Stoodley L, Costerton JW, Stoodley P (2004). Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2: 95-108.

Hall-Stoodley L, Stoodley P (2009). Evolving concepts in biofilm infections. Cell Microbiol 11: 1034- 1043.

Hamon MA, Ribet D, Stavru F, Cossart P (2012). Listeriolysin O: the Swiss army knife of Listeria. Trends Microbiol 20: 360-368.

Ingersoll MA, Albert ML (2013). From infection to immunotherapy: host immune responses to bacteria at the bladder mucosa. Mucosal Immunol 6: 1041-1053.

James GA, Swogger E, Wolcott R, Pulcini E, Secor P, Sestrich J et al (2008). Biofilms in chronic wounds. Wound Repair Regen 16: 37-44.

Johnson M, Cockayne A, Williams PH, Morrissey JA (2005). Iron-responsive regulation of biofilm formation in staphylococcus aureus involves fur-dependent and fur-independent mechanisms. J Bacteriol 187: 8211-8215.

Johnson M, Cockayne A, Morrissey JA (2008). Iron-regulated biofilm formation in Staphylococcus aureus Newman requires ica and the secreted protein Emp. Infect Immun 76: 1756-1765.

Kelvin Lee KW, Hoong Yam JK, Mukherjee M, Periasamy S, Steinberg PD, Kjelleberg S et al (2015). Interspecific diversity reduces and functionally substitutes for intraspecific variation in biofilm communities. The ISME journal.

Keogh D, Tay WH, Ho YY, Dale JL, Chen S, Umashankar S et al (2016). Enterococcal Metabolite Cues Facilitate Interspecies Niche Modulation and Polymicrobial Infection. Cell Host Microbe 20: 493-503.

Lee JH, Cho HS, Kim Y, Kim JA, Banskota S, Cho MH et al (2013). Indole and 7-benzyloxyindole attenuate the virulence of Staphylococcus aureus. Appl Microbiol Biotechnol 97: 4543-4552.

Lee KW, Periasamy S, Mukherjee M, Xie C, Kjelleberg S, Rice SA (2014). Biofilm development and enhanced stress resistance of a model, mixed-species community biofilm. The ISME journal 8: 894- 907.

Lee W-J, Hase K (2014). Gut microbiota–generated metabolites in animal health and disease. Nat Chem Biol 10: 416.

Page | 153

Liss V, Swart AL, Kehl A, Hermanns N, Zhang Y, Chikkaballi D et al (2017). Salmonella enterica Remodels the Host Cell Endosomal System for Efficient Intravacuolar Nutrition. Cell Host Microbe 21: 390-402.

Marsland BJ, Gollwitzer ES (2014). Host-microorganism interactions in lung diseases. Nat Rev Immunol 14: 827-835.

Mellouk N, Enninga J (2016). Cytosolic Access of Intracellular Bacterial Pathogens: The Shigella Paradigm. Frontiers in Cellular and Infection Microbiology 6: 35.

Nicholson JK, Holmes E, Kinross J, Burcelin R, Gibson G, Jia W et al (2012). Host-Gut Microbiota Metabolic Interactions. Science 336: 1262.

Periasamy S, Nair HA, Lee KW, Ong J, Goh JQ, Kjelleberg S et al (2015). Pseudomonas aeruginosa PAO1 exopolysaccharides are important for mixed species biofilm community development and stress tolerance. Frontiers in microbiology 6: 851.

Peters BM, Jabra-Rizk MA, O'May GA, Costerton JW, Shirtliff ME (2012). Polymicrobial Interactions: Impact on Pathogenesis and Human Disease. Clin Microbiol Rev 25: 193-213.

Romling U, Balsalobre C (2012). Biofilm infections, their resilience to therapy and innovative treatment strategies. J Intern Med 272: 541-561.

Rooks MG, Garrett WS (2016). Gut microbiota, metabolites and host immunity. Nature reviews Immunology 16: 341-352.

SanMiguel A, Grice EA (2015). Interactions between host factors and the skin microbiome. Cellular and Molecular Life Sciences 72: 1499-1515.

Scales BS, Huffnagle GB (2013). The microbiome in wound repair and tissue fibrosis. The Journal of pathology 229: 323-331.

Schröder A, Schröder B, Roppenser B, Linder S, Sinha B, Fässler R et al (2006). Staphylococcus aureus Fibronectin Binding Protein-A Induces Motile Attachment Sites and Complex Actin Remodeling in Living Endothelial Cells. Mol Biol Cell 17: 5198-5210.

Tay WH, Chong KK, Kline KA (2016). Polymicrobial-Host Interactions during Infection. J Mol Biol 428: 3355-3371.

Thänert R, Goldmann O, Beineke A, Medina E (2017). Host-inherent variability influences the transcriptional response of Staphylococcus aureus during in vivo infection. Nature communications 8: 14268.

Page | 154

Thurston TL, Wandel MP, von Muhlinen N, Foeglein A, Randow F (2012). Galectin 8 targets damaged vesicles for autophagy to defend cells against bacterial invasion. Nature 482: 414-418. van Rensburg JJ, Lin H, Gao X, Toh E, Fortney KR, Ellinger S et al (2015). The Human Skin Microbiome Associates with the Outcome of and Is Influenced by Bacterial Infection. mBio 6: e01315-01315. Wade WG (2013). The oral microbiome in health and disease. Pharmacological Research 69: 137- 143.

Westermann AJ, Förstner KU, Amman F, Barquist L, Chao Y, Schulte LN et al (2016). Dual RNA-seq unveils noncoding RNA functions in host–pathogen interactions. Nature 529: 496.

Whiteside SA, Razvi H, Dave S, Reid G, Burton JP (2015). The microbiome of the urinary tract--a role beyond infection. Nature reviews Urology 12: 81-90.

Yang L, Barken KB, Skindersoe ME, Christensen AB, Givskov M, Tolker-Nielsen T (2007). Effects of iron on DNA release and biofilm development by Pseudomonas aeruginosa. Microbiology 153: 1318- 1328.

Zou J, Shankar N (2016). The opportunistic pathogen Enterococcus faecalis resists phagosome acidification and autophagy to promote intracellular survival in macrophages. Cell Microbiol 18: 831- 843.

Page | 155