Protective and Anti-Inflammatory Effects of Protegrin-1 on Citrobacter Rodentium Intestinal Infection in Mice

Protective and Anti-inflammatory Effects of Protegrin-1 on Citrobacter rodentium Intestinal Infection in Mice

Celina N. Osakowicz

A Thesis

presented to

The University of Guelph

In partial fulfilment of requirements

for the degree of

Master of Science

Animal Biosciences

Guelph, Ontario, Canada

ABSTRACT

PROTECTIVE AND ANTI-INFLAMMATORY EFFECTS OF PROTEGRIN-1 ON Citrobacter rodentium INTESTINAL INFECTION IN MICE

Celina N. Osakowicz Advisor: University of Guelph, 2018 Dr. Julang Li

Intestinal disorders and colitis affect millions of humans and food-animals world-wide.

Antimicrobial peptides (AMPs) and their broad-spectrum antimicrobial activity represent a valuable potential therapeutic solution. Specifically, the potent pig-originated protegrin-1 (PG-1) has previously been shown to reduce the pathological effects of chemically induced digestive tract inflammation (colitis) along with modulation of immune responses and tissue-repair. This study aimed to extend these findings by investigating the potential protective effects of PG-1 on pathogen-induced intestinal colitis. We found that oral administration of PG-1 reduced Citrobacter rodentium intestinal infection in mice evidenced by; reduced histopathologic change in the colon, prevention of body weight loss, milder clinical signs of disease, and ultimately more effective clearance of bacterial infection relative to challenged mice.

Additionally, PG-1 treatment altered the expression of various inflammatory mediators during infection to resolve inflammation and re-establish intestinal homeostasis. Interestingly, PG-1 administered in its mature form was most effective relative to the pro-form (ProPG-1).

Acknowledgements

I would like to acknowledge a great number of individuals who have been invaluable to this research project and whom the completion of this degree would not be possible without their support.

First, I would like to express my utmost appreciation towards my advisor, Dr. Julang Li, for her countless support, guidance, and encouragement as my mentor throughout this stage of my academic career. During my masters, she always challenged me to do my best and supported my pursuit of knowledge and learning. I am truly grateful for having had the opportunity to work under such a passionate and dedicated researcher.

I would like to thank my committee members, Dr. Jeff Caswell and Dr. Robert Friendship, for their valuable input and advice towards improving the research project and thesis. Additional thanks to Dr. Jeff Caswell for his assistance and expertise in the histological components of this project.

To the members of the Li lab - past and present - thank you for all your support, thoughtful discussions, and friendship that helped foster a supportive lab environment. A special thank you goes out to Dr. Evanna Huynh and Dr. Jenna Penney, for their time and effort in enhancing my scientific learning and skills, and to James Haskins for his encouragement in navigating through the ups and downs of graduate studies.

I would like to express my thanks to the Central Animal Facility (CAF) and Isolation Facility staff for their research animal expertise, care, and assistance in all my animal trials. A special thank you goes out to Veronique Carson and our undergraduate student Erin Miehe for their help and patience during the long hours of tissue collection.

To my amazing parents, family, and friends, thank you for believing in me and for always being a constant source of love and encouragement. I am truly privileged to have you all in my life. Lastly, I would like to thank my partner, Michael Osborn, who has been there from the very beginning. You always managed to turn my bad days into good ones. Thank you for seeing me through this chapter of my life.

It has been such a rewarding and exciting experience, and I look forward to what the future holds.

Thank you all.

iii

Table of Contents

ABSTRACT ...... ii

Acknowledgements ...... iii

Table of Contents ...... iv

List of Tables ...... vi

List of Figures ...... vii

List of Abbreviations ...... viii

1. Introduction ...... 1

2. Literature Review ...... 2

2.1 Overview of AMPs ...... 2 2.1.1 AMPs and their Structure ...... 2 2.1.2 Mechanisms of Action ...... 4 1. Bactericidal ...... 4 2. Immunomodulatory Mechanism ...... 5 2.1.3 Multifunctional Roles of AMPs ...... 7 2.1.4 AMPs in IBD ...... 8 2.2 Major Families of AMPs ...... 9 2.2.1 Defensins ...... 9 2.2.2 Cathelicidins ...... 10 1. Protegrins ...... 11 2.3 Protegrin-1 ...... 11 2.3.1 Mechanism and Structure ...... 11 2.3.2 In vivo Applications of PG-1 ...... 13 2.4 Recombinant PG-1 Production and Expression in Pichia pastoris ...... 14 2.5 Citrobacter rodentium ...... 18

3. Rationale and Research Objectives ...... 19

4. Materials and Methods ...... 20

4.1 Construction and Transformation of ProPG-1-pUC Shuttle Vector into E. coli DH5 ...... 20 4.2 Excision of ProPG-1 from pUC Shuttle Vector ...... 20 4.3 Ligation of ProPG-1 into pD915-GAP Vector ...... 21 4.4 Confirmation of Ligation into pD915-GAP Vector ...... 21 4.5 Transformation into P. Pastoris by Electroporation ...... 22 4.6 Transformation Screening and Selection ...... 22 4.7 Shake-Flask Fermentation ...... 23 4.8 Bioreactor Fermentation & Optimization of Recombinant ProPG-1 ...... 23 4.9 Protein Sample Preparation ...... 24 4.10 SDS-PAGE & Western Blot Analysis ...... 25 4.11 Ethics Statement...... 25 4.12 Bacterial Preparation ...... 25 4.13 Animals & Induction of Colitis ...... 26 iv

4.14 Measurement of Bacterial Load ...... 26 4.15 Fecal Water Content ...... 26 4.16 Disease Activity Index Score ...... 27 4.17 Tissue Collection ...... 27 4.18 Histomorphology ...... 28 4.19 Real-Time Quantitative PCR ...... 30 4.20 Statistical Analysis ...... 30

5. Results ...... 31

5.1 Recombinant ProPG-1 Production in P. pastoris using GAP Promoter ...... 31 5.2 PG-1 Reduced C. rodentium-Induced Intestinal Infection ...... 36 5.3 PG-1 Decreased C. rodentium-Induced Histologic Lesions in the Colon...... 42 5.4 PG-1 Modulates Intestinal Gene Expression during C. rodentium Infection ...... 45

6. Discussion...... 50

7. Conclusion ...... 60

References ...... 61

Appendix ...... 69

List of Tables

Table 1. Disease activity index scoring criteria ...... 27

Table 2. Definition of categories and general criteria for histomorphological scoring ...... 29

Appendix Table A1. Optimal bioreactor media composition for recombinant PG-1 expression in Pichia pastoris ...... 69

Appendix Table A2. Primer sequences used for RT-qPCR ...... 70

List of Figures

Figure 1. Sources of AMPs ...... 3

Figure 2. Topological diagram of PG-1’s antimicrobial domain ...... 12

Figure 3. Schematic representation of PG-1 and its domains ...... 13

Figure 4. Comparison of AOX1 and GAP promoter for protein cultivation ...... 17

Figure 5. Expression vector for ProPG-1 production in P. pastoris ...... 33

Figure 6. Restriction enzyme digests for confirmation of ProPG-1 ligation into pD915-GAP ...... 34

Figure 7. Western blot analysis of rProPG-1 transformants and confirmation of expression from fermentation ...... 35

Figure 8. Animal study overview ...... 38

Figure 9. Effects of PG-1 on body weight and DAI...... 39

Figure 10. PG-1 reduced C. rodentium infection ...... 40

Figure 11. Effects of PG-1 on intestinal length and weight ...... 41

Figure 12. Representative H&E stained colon sections ...... 43

Figure 13. Effects of PG-1 on colon histomorphology ...... 44

Figure 14. Effects of PG-1 on AMP gene expression within the colon ...... 47

Figure 15. Effects of PG-1 on self-protecting gene expression within the colon...... 48

Figure 16. Effects of PG-1 on apoptotic, innate immune response, and cytokine gene expression within the colon ...... 49

Figure 17. Proposed working model of mPG-1 regulating the gene expression of intestinal factors during inflammation through NF-κB ...... 59

Appendix Figure A1. Pilot animal study #1 overview and main results ...... 71

Appendix Figure A2. Pilot animal study #2 overview; body weight and DAI ...... 72

Appendix Figure A3. Pilot animal study #2 results ...... 73

Appendix Figure A4. Representative H&E stained colon sections illustrate colonic crypt hyperplasia .. 74

Appendix Figure A5. Effects of PG-1 on lamina propria width and tunica muscularis width of the colon .

...... 75

vii

List of Abbreviations

A/E Attaching and effacing AKT Protein kinase B AMPs Antimicrobial peptides ANOVA Analysis of variance Ang4 Angiogenin 4 AOX1 Alcohol oxidase I BAX BCL associated X BCL3 B-cell lymphoma 3 BLAST Basic local alignment search tool BMGY Buffered glycerol-complex medium BMMY Buffered methanol-complex medium BSA Bovine serum albumin BSM Basal salt medium BW Body weight Camp/mCRAMP Mouse cathelicidin-related antimicrobial peptide CASP1 Caspase-1 CD Crohn's disease CD40 Cluster of differentiation 40 cDNA Complementary DNA CFU Colony forming units COX-2 Cyclooxygenase-2 DAI Disease activity index DC Dendritic cells DEFA Alpha-defensin DO Dissolved Oxygen DSS Dextran sodium sulfate EDTA Ethylenediaminetetraacetic acid EGR1 Early growth response factor 1 EHEC Enterohemorrhagic Escherichia coli EK Enterokinase EPEC Enteropathogenic Escherichia coli ERK Extracellular signal-regulated kinase viii

FT Fermentation Time GAP Glyceraldehyde 3-phosphate dehydrogenase promoter GAPDH Glyceraldehyde 3-phosphate dehydrogenase GI Gastrointestinal tract H&E Hematoxylin and eosin hBD2 Human beta-defensin 2 HDPs Host defense peptide HIF1α Hypoxia inducible factor 1 α subunit HNP Human neutrophil peptide IBD Inflammatory bowel disease IFN- Interferon alpha IGF-1R Insulin-like growth factor type 1 receptor IL-8 Interleukin-8 IP Intraperitoneal injection IRF7 Interferon regulatory factor 7 IV Intravenous injection JNK Jun amino-terminal kinases LB Luria-Bertani broth LPS Lipopolysaccharide LRR Leucin-rich repeats LTA Lipoteichoic acid MAPK Mitogen-activated protein kinases MCP-1 Monocyte chemoattractant protein 1 mPG-1 Mature protegrin-1 MRSA Methicillin-resistant Staphylococcus aureus Muc1 Mucin-1 Mut- Methanol utilization minus Mut+ Methanol utilization plus Muts Methanol utilization slow NF-κB Nuclear factor κB NK Natural killer cell PAMPs Pathogen-associated molecular patterns PBS Phosphate buffered saline

PG-1 Protegrin-1 PI3K Phosphatidylinositol 3-kinase ProPG-1 Pro-form protegrin-1 PRRs Pattern-recognition receptors PTM1 Pichia trace metal salts medium PUFAs Polyunsaturated fatty acids RT-qPCR Quantitative reverse transcription polymerase chain reaction Reg3β Regenerating islet-derived protein 3β Reg3γ Regenerating islet-derived protein 3γ RTD Rhesus macaque theta-defensin SCY2 Scygonadin 2 SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis SEM Standard error of the mean SOCS3 Suppressor of cytokine signaling 3 sPLA2 Secretory phospholipase A2 T3SS Type III secretion system TFF3 Trefoil factor 3 TLR Toll-like receptor TNF- Tumor necrosis factor alpha UC Ulcerative colitis VCAM-1 Vascular cell adhesion molecule VEGF Vascular endothelial growth factor VREF Vancomycin-resistant Enterococci faecalis WT Wild-type YPD Yeast extract peptone dextrose

1. Introduction

Intestinal Infection & Antibiotic Resistance

In Canada, approximately 233,000 individuals suffer from inflammatory bowel disease (IBD), costing around $1.2 billion in medical expenses each year [1]. Inflammatory bowel disease is classified as the chronic recurrence of inflammation in any segment of the intestinal mucosa and includes Crohn’s disease (CD) and ulcerative colitis (UC) [2]. The pathogenesis of these disorders is not fully understood but is speculated to be a combination of genetics and abnormal immune responses to environmental factors or intestinal bacteria that leave individuals susceptible to uncontrolled intestinal inflammation [3]. Disruption of the intestinal barrier is a hallmark for IBD that may be due to a loss of intestinal homeostasis, overall function, or ability for mucosal repair [4]. Importantly, studies have indicated that a dysregulation of antimicrobial peptides (AMPs) may be a significant factor as they are expressed in the intestinal lining and are essential for mucosal maintenance [2, 5]. Scientists have been investigating their expression during intestinal inflammation in CD and UC, but their full contribution has yet to be identified (reviewed in [6]).

Given their significance, investigating their expression and their role during intestinal infection may provide insights and offer alternative therapeutics against colitis.

Similarly, other species are affected by intestinal disorders and colitis. Early weaning of piglets causes incomplete gastrointestinal tract development and abnormal intestinal function, which leads to a post-weaning growth lag and an increased susceptibility to intestinal infection [7, 8]. The common practice to utilize in-feed growth-promoting antibiotics in animal production to combat these intestinal issues has added to the significant rise in antibiotic resistant bacteria in animals and humans [9]. Prolonged, low-dose antibiotic treatment and non-therapeutic use in food animals drives selective pressure for resistant strains which may be spread to humans via direct or indirect contact or consumption [9]. This has influenced the

Canadian government’s decision to impose further restrictions regarding the use of in-feed antibiotics and has prompted a need for anti-infective alternative therapies [10]. Antimicrobial peptides, with their unique 1

multidirectional mechanisms of action and broad-spectrum antimicrobial activity would lessen the chance for resistance and offer a significant advantage over conventional antibiotics.

It has previously been found that recombinant protegrin-1 (PG-1), a potent pig-originated AMP, reduces the pathological effect of chemically induced colitis in mice, suggesting a potential interaction of

PG-1 with intestinal epithelial cells to modulate inflammation and enhance tissue-repair (Huynh E., unpublished [11]). However, a high dose of 10 mg/kg body weight was used, limiting its application in a production setting. Thus, this thesis aims to enhance the recombinant production and expression of PG-1 using the yeast Pichia pastoris (P. pastoris) to increase yield and to investigate the immune-protective potential and tissue repair stimulation in the colon using a challenged animal model.

2. Literature Review

2.1 Overview of Antimicrobial Peptides (AMPs)

2.1.1 AMPs and their Structure

Antimicrobial peptides are gene-encoded natural innate immune defense molecules that are present in prokaryotes, like bacteria, and eukaryotes, such as protozoan, fungi, plants, insects and animals [12]. A total of 2,884 AMPs have been identified to date, with the three major sources coming from animals (2,184), plants (342), and bacteria (333) (Figure 1) (Antimicrobial Peptide Database, http://aps.unmc.edu/AP/main.php). Antimicrobial peptides have a vital role in providing an initial mechanism of host defense, and thus have also been referred to as host defense peptides (HDPs) [13]. These peptides have a broad-spectrum of antimicrobial activity which they are capable of exerting against bacteria, fungi, enveloped viruses, and parasites at the nanomolar to micromolar range [13, 14].

Antimicrobial peptides vary greatly in structure, but most share many key features. They are generally small, between 12 and 100 amino acids long, maintain an amphipathic structure and have an overall positive net charge (generally +2 to +9) [15]. These peptides can be divided into structural groups

based on their amino acid compositions and secondary structures: -helical, β-sheet, extended and loop (as reviewed in [16]). Among these groups, the -helical and β-sheet groups are the most common structures found in AMPs but the most studied peptides to date are -helical. Some examples of -helical AMPs are the human LL-37 cathelicidin, cecropins, and magainins. The β-sheet structures are stabilized by two or four disulfide bridges and some prime examples are the human - and β-defensins [17]. The extended structures, such a indolicin, are highly rich in glycine, proline, tryptophan, arginine, and histidine [13].

Lastly, the loop structures are stabilized by one disulfide bridge and a prime example is the bovine bactenecin AMP [13].

Figure 1. Sources of AMPs. A total of 2,884 AMPs have been identified from all six kingdoms of life. The three major sources are animals (2,184), plants (342), and bacteria (333). (Antimicrobial Peptide Database, http://aps.unmc.edu/AP/main.php).

2.1.2 Mechanisms of Action

1. Bactericidal Mechanism

Antimicrobial peptides have two main classifications of their mechanisms of action. The first mechanism is the bactericidal action mechanism where AMPs induce cell membrane damage. Most AMPs are cationic and amphipathic, which increases their affinity to interact with the negatively charged microbial membranes through electrostatic forces [18]. However, a few AMPs such as amphibian maximim H5 as well as the human dermcidin, are anionic and the mechanisms behind their interaction with the negatively charged membrane is unclear. It is thought that these peptides need to be in the presence of a cationic co- factor, such as positively charged zinc, for them to accumulate at the anionic microbial surface [19]. Not all peptides inflict membrane damage through the same mechanism due to their molecular properties and composition of the membrane itself. Antimicrobial peptides can disrupt the physical integrity of the bilayer through membrane thinning, disruption of the barrier function or by creating a pore; all of which result in cell lysis [20, 21]. Some models explaining the mechanisms of AMP membrane disruption are the barrel- stave, toroidal and carpet models. The barrel-stave model proposes that the peptides form a transmembrane pore via insertion into the target membrane lipid core [20]. Similarly, the toroidal model proposes that the peptides insert themselves into the membrane forming a bundle. This bundle causes the lipid monolayers to become interspaced as they bend through the pore, straining the bilayer [20]. Protegrin is one example of a toroidal pore-forming AMP [22]. Lastly, the carpet model demonstrates that the architecture of the membrane surface is affected by the accumulation of AMPs on the surface. Cell lysis occurs as the membrane is disintegrated once the AMPs reach a certain concentration threshold [23]. These mechanisms make it more challenging for microbes to develop resistance against, and thus provide a significant advantage over conventional antibiotics.

Alternatively, AMPs can translocate across the membrane directly into the cytoplasm and inhibit intracellular function. The peptides act on internal targets and can inhibit cell-wall synthesis, DNA, RNA, nucleic acid, and protein synthesis [22].

2. Immunomodulatory Mechanism

The second mechanism of AMP action is the immune modulation mechanism which can be divided into two categories: a chemotactic approach and the regulation of immune-relevant gene expressions. Upon injury or the stimulation of a foreign invader within the host, AMPs are released from local cells and are able to modulate host immunity through direct recruitment and activation of immune cells [24]. Most commonly, these peptides recruit leukocytes, cytokines and chemokines including monocyte chemoattractant protein (MCP-1), interleukin 8 (IL-8) and interferon alpha (IFN- ) [24, 25]. In turn, this recruitment indirectly stimulates effector cells such as neutrophils, monocytes, macrophages, dendritic and

T cells. For example, two human -defensins called human neutrophil peptide 1 and 2 (HNP-1, HNP-2) have chemotactic activity in moderating the enlistment of monocytes towards the site of inflammation [26].

Similarly, human beta-defensin 2 (hBD2) is able to mobilize mast cells through the G protein-phospholipase

C signaling pathway while hBD3 and 4 mobilize monocytes and macrophages using G proteins as well as p38 mitogen-activated protein kinase and c-Jun N-terminal protein kinase (JNK) as signal transducers [27,

28]. Additionally, AMPs are capable of indirect chemotactic activity through prompting cytokine and chemokine secretion via receptor-dependent methods from a wide range of cell types. hBD3 can stimulate cluster of differentiation 40, 80 and 86 (CD40, CD80, CD86) protein molecules through toll-like receptors

1 and 2 (TLR1, TLR2) on dendritic cells and monocytes leading to myeloid differentiation primary response gene 88 (MyD88) signaling. MyD88 signaling in turn allows IL-1 receptor-associated kinase-1 phosphorylation [29]. hBD3, along with hBD2 and 4 can also increase the gene expression and protein assembly of IL-6, IL-10, and MCP-1 through G-protein coupled receptors in human keratinocytes [30].

Lastly, β-defensins and cathelicidins promote inflammatory cytokine production by activating the ERK

mitogen activated protein kinase (MAPK) pathway in epithelial cells [31]. Ultimately, the effects of the direct and indirect chemotactic activity enhance bacterial clearance and controls inflammation as well as sepsis within the host.

The second category of immune modulating mechanism is the regulation of immune-relevant gene expression through signaling mediated by pattern-recognition receptors (PRRs). Evolutionary conserved,

PRRs are vital in recognizing microbial pathogens as they distinguish host from pathogen material [32].

Specifically, these receptors distinguish moieties conserved within a pathogen class known as pathogen- associated molecular patterns (PAMPs) that lead to AMP products and cytokine expression through the initiation of intracellular signaling cascades [32, 33]. Induction of AMP expression may involve the toll- like-receptor (TLR) pathway, a heavily studied class of PRRs. These integral glycoproteins are structurally characterized by an extracellular domain or a luminal ligand-binding domain consisting of leucine-rich repeat (LRR) motifs [34]. Ligands tether to TLRs via PAMP-TLR interplay which stimulate receptor oligomerization, leading to intracellular signaling transduction [33]. Antimicrobial peptides are capable of either blocking or activating TLR signaling to regulate immunomodulation [33, 35]. Lipopolysaccharide

(LPS) and lipoteichoic acid (LTA) are bacterial cell wall components that are proficient in triggering cells of the immune, inflammatory, and vascular systems, which can advance sepsis [36]. Two ligand recognizing receptor proteins, TLR2 and TLR4, are located on cell surfaces and are required for LPS and

LTA-induced intracellular signaling [36]. Some AMPs block TLR2 and TLR4, thereby preventing LPS and

LTA-induced intracellular signaling, deterring the overproduction of inflammatory mediators, which results in suppression of pathogen-induced inflammation. For example, the mouse cathelicidin-related AMP

(mCRAMP) has been shown to inhibit TLR4 activation as well as the induction of cytokine release in dendritic cells (DC) and macrophages, by altering membrane function in a mouse model of allergic contact dermatitis [35]. Specifically, researchers observed an enhancement of allergic contact response with the deletion of the murine cathelicidin gene Cnlp, and an inhibited allergic response with the administration of mCRAMP. Inhibition of sensitization via mCRAMP is dependent on the presence of a functional TLR4 6

receptor [35]. On the other hand, receptor proteins TLR3, 7, 8, and 9 are located inside cell compartments and are efficient in recognizing nucleic acids derived from viruses and pathogens [33, 34]. Antimicrobial peptides enhance the recognition of the viral nucleic acids and activate these receptors leading to the production of inflammatory cytokines and interferons to help combat infection. For example, a murine cathelicidin encoded by the mouse gene Camp, has been shown to enhance a key receptor for the recognition of DNA, TLR9, in DC and macrophages [37]. It was determined that murine plays an intracellular role in the enhancement of TLR9 based on the observations that the absence of Camp decreased TLR9 response in vivo and in vitro but showed enhanced TLR9 response when Camp was present [37]. Ultimately these functions favor the resolution of infection and aim to reverse potentially harmful inflammation.

2.1.3 Multifunctional Roles of AMPs

Antimicrobial peptides have multiple functions in host defense and can generate an array of responses in host innate immune cells. However, some AMPs seem to be more multifaceted as they perform an assortment of functions beyond microbial killing and immune modulation. A cathelicidin proline- and arginine -rich peptide called PR-39 has been shown to stimulate angiogenesis. This pig originated AMP inhibits the ubiquitin-proteasome-dependent degradation of hypoxia-inducible factor 1a protein leading to accelerated formation of vascular structures [38]. Furthermore, PR-39 is a chemoattractant agent for neutrophils which further promotes angiogenesis via the release of vascular endothelial growth factor

(VEGF) [38]. Another important function AMPs have is their ability to promote wound healing. hBD2 acts like a pro-angiogenic factor by eliciting a chemotactic effect to wounds in the absence of any growth factors

[39]. Specifically, hBD2 allows for accelerated wound closure by encouraging the migration, proliferation, and tube development of wound endothelial cells essential for neovascularization [39]. Additionally, LL-

37 also plays a role in promoting chemotaxis of epithelial cells and the reorganization of the extracellular matrix through the production of metalloproteinases [25]. These calcium-dependent zinc-containing endopeptidases play a major role in cell proliferation, migration, angiogenesis, and apoptosis [25]. Both the

stimulation of angiogenesis and the ability to enhance wound healing are vital requirements for tissue repair and regeneration. As such, researchers have been able to demonstrate that the human LL-37 AMP enhanced epithelium restoration and granulation tissue development through its application to excisional wounds in mice [25]. Lastly, some AMPs have been shown to have a role in tumor surveillance. Cathelicidin has recently been found to be abundantly expressed in tumor-infiltrating human natural killer (NK) cells and seems to have a strong role in cell-mediated tumor suppression [40]. Researchers have discovered that NK cells absent of cathelicidin showed impaired cytotoxity activity towards tumor targets and subsequently noticed enhanced growth of tumors [40].

Recently, researchers identified a new AMP called Scygonadin 2 (SCY2) that was able to exert reproductive immune function in the reproductive tract of mating mud crabs [41]. It was discovered that

SCY2, which is only expressed in the ejaculatory duct of male crabs, was significantly decreased after mating (104 times less) but was highly abundant in the female spermathecal after mating. These results suggest that the SCY2 may be transferred to female crabs during mating to sustain a sterile environment which would lead to successful fertilization [41]. In summary, this interesting observation demonstrates a potential new role of AMPs in reproductive immunity.

2.1.4 AMPs in IBD

Inflammatory bowel disease is characterized by chronic inflammation in the colon and small intestine and is grouped into ulcerative colitis (UC) or Crohn’s disease (CD) based on clinical and histopathological features [2]. These diseases seem to be linked to abnormal intestinal mucosal responses to bacteria, however the pathogenesis is not entirely understood. The gastrointestinal tract (GI) is under constant exposure to a variety of microorganisms, and to ensure proper mucosal barrier function to combat these organisms, intestinal epithelial cells produce various AMPs [5, 42]. For this reason, AMPs have been used as disease markers for infection to help advance disease detection as well as therapeutic response. One example would be the secreted glycoprotein lactoferrin found in secondary granules of polymorphonuclear 8

neutrophils, which play a role in the inflammatory response. Lactoferrin contains a variety of AMPs that are subsequently released upon hydrolysis from proteases to aid in protection against pathogens. It was found that lactoferrin was significantly increased in IBD patients and thus is now frequently tested for in fecal collections as a non-invasive indicator of potential disease. [43]. However, studies have also indicated that a dysregulation of AMPs that assist in innate immunity occurs in IBD (reviewed in [44]). An example would be the decrease of Paneth cell -defensins DEFA5 and DEFA6 in patients with Crohn’s, due to inflammation [45]. Whether the irregular AMP expression in IBDs are a secondary event due to inflammation or are the causative factor for abnormal intestinal microbial clearance is still under heavy debate [42]. Regardless, their expression and role in IBD requires further clarification. Interestingly, Koon et al., found that cathelicidin deficient Camp -/- mice displayed severe colitis in a dextran sodium sulfate

(DSS)-induced colitis model compared to wildtype mice [46]. This protection to inflammatory infection demonstrates that cathelicidins may have therapeutic potential in IBDs.

2.2 Major Families of AMPs

Since AMPs are found in all organisms, and any given organism can have dozens of different peptides, there is a vast diversity of sequences. Similarities of such sequences are often only found within defined groups of AMPs from closely related species [47]. However, two mammalian groups of AMPs called the defensins and cathelicidins are evolutionary distinct and both highly conserved [48].

2.2.1 Defensins

Defensins are an evolutionarily related vertebrate family of AMPs that are found in all mammals, are cationic, and microbicidal. Although defensins have much variation within their sequences, all are rich in arginine and are unique in the sense that they contain six conserved cysteine residues that form three sets of intramolecular disulfide bonds [49]. They are stored in neutrophil granules and Paneth cells but can be produced by keratinocytes, macrophages and their precursors, or various mucosal epithelial cells [50].

Defensins can be divided into three subfamilies: , β, and θ. -defensins are only found in higher eukaryotes such as primates, mice, rats, and pigs, while β-defensins are expressed in many species but are the only defensins found in birds (reviewed in [51]). Both the - and β-defensins have a triple-stranded β-sheet core but differ in the arrangement of the disulphide bonds [39, 49]. An example of an -defensin would be the human neutrophil peptide 3 (HNP3) and the human β-defensin 3 (hBD3) is an example of a β-defensin. θ

-defensins are circular peptides solely found from mammalian lineages [52]. They evolved from primates but are no longer active in humans due to mutations that encode premature stop codons [52]. An example would be the rhesus macaque θ -defensin (RTD) where this subfamily was first identified [53].

2.2.2 Cathelicidins

Cathelicidins are cationic peptides that are found in a variety of invertebrates as well as vertebrates such as birds, reptiles, fish, and mammals (reviewed in [54]). Cathelicidins are typically 12 – 80 amino acids long and have a broad antimicrobial activity pH range. They are named on their well conserved N- terminal region, the cathelin domain, which is capable of inhibiting the cathepsin-L protease [55]. Two disulphide bonds between C85-C96 and C10-C124 residues, stabilize the cathelin protein [56]. The full functional role of the cathelin domain or any uncleaved precursors has yet to be determined but has been speculated to be involved in immune modulation [57]. These peptides are produced by inflammatory as well as epithelial cells and are stored as inactive precursors in macrophages and/or neutrophil granules until degranulation is stimulated through an immune response to release the peptides [48, 54, 58]. The peptides must then be processed by proteases at their proteolytic processing site to cleave and release the antimicrobial C-terminal domain from the N-terminal cathelin domain. The C-terminal or mature AMP domain may then become active and elicit its microbial response. Cathelicidins also have numerous other roles in addition to their direct antimicrobial function such as regulating inflammatory mediators, stimulating cell proliferation, wound healing and angiogenesis (reviewed in [59]). There are approximately

30 cathelicidin family members that have been found in mammals, however LL-37 is the only human cathelicidin identified [54].

1. Protegrins

An important class of cathelicidins are the protegrins. These porcine leukocyte isolates are arginine and cysteine-rich peptides that contain 16 to 18 amino acid residues and are naturally produced by neutrophils [54]. They assume a rigid β-hairpin structure that is stabilized by two disulphide bridges between four cysteine residues [60]. To date, a total of five highly homologous porcine protegrins have been described (PG-1 to PG-5) (reviewed in [61]) and display similar structural features of rhesus macaque

θ-defensins (RTD) and a horseshoe crab AMP tachyplesin [53, 62]. Importantly, protegrins have been shown to maintain their antimicrobial activity in physiological salt solutions, which enhances their therapeutic potential, as these solutions resemble conditions found in extracellular fluids and serum [63].

Furthermore, protegrins are active against Gram-positive and Gram-negative bacteria, making them excellent candidates against a broad spectrum of microorganisms.

2.3 Protegrin-1

2.3.1 Mechanism and Structure

One of the most heavily studied AMPs is the first protegrin isolate known as protegrin-1 (PG-1).

Protegrin-1’s antimicrobial domain is 18 amino acids (RGGRLCYCRRRFCVCVGR) long and contains four cysteine residues at positions 6 to 15 and 8 to 13 that stabilize the β-hairpin structure through their formation of two disulphide bonds [64] (Figure 2). Positively charged regions flank the sides of the collection of hydrophobic residues that accumulate in the center of the hairpin structure. Inactive pro- protegrin is specifically stored in neutrophil granules until required at the site of infection. Once released through degranulation, the peptide can then be processed to a mature peptide by the serine protease, elastase, to separate the cathelin propiece from the active antimicrobial PG-1 domain (mature PG-1) [65] (Figure 3).

Its bactericidal effect is due to the peptides ability to take on an amyloid-like fibril conformation forming toroidal pores in bacterial cell membranes leading to structural disruption, ion leakage, and ultimately lysis

[66-68]. Bolintineanu et al., demonstrated PG-1’s rapid bactericidal effect in vitro on E. coli after they observed between 10 and 100 pores per cell were sufficient to rupture the bacteria and cause cell death in only a few minutes [69]. Most notably, PG-1 has also been shown to exert its effects on antibiotic resistant bacteria including methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococci faecalis (VREF) and Pseudomonas aeruginosa.

Figure 2. Topological diagram of PG-1’s antimicrobial domain. PG-1 is 18 amino acids long and contains two disulfide bonds between Cys 6 & 15, and 8 & 13 (solid black lines) that stabilize the β- hairpin structure. Dotted black lines show the backbone hydrogen bonds. Positively charged (Arg) residues are denoted in blue, hydrophobic residues are in white, and polar residue (Tyr) and Gly are in green. The first and last residue is indicated by the number 1 and 18, respectively. Modified from [70] .

Figure 3. Schematic representation of PG-1 and its domains. The pro-form of PG-1 can be cleaved at its enzyme cleavage site to release the mature peptide domain (mature PG-1) from the cathelin domain.

2.3.2 In vivo Applications of PG-1

Protegrin-1 is considered as one of the most potent AMPs, due to its impeccable ability to eliminate multi-drug resistant microorganisms. Its activity has been widely demonstrated in vitro and in vivo. An important study by Steinberg et al. (1997) showed the effects PG-1 had in vivo against mice challenged with MRSA, VREF, and P. aeruginosa. Immunocompetent mice that were inoculated intraperitoneally (IP) with P. aeruginosa or MRSA had a mortality rate of 93-100% compared to mice that were injected with a single IP injection of 0.5 mg/kg of PG-1 with a 0-27% mortality rate. In the same study, mice infected with

MRSA through an intravenous (IV) injection and single IV dose of 5mg/kg PG-1 limited mortality to 33% in comparison to 93% for the control group without the PG-1 injection. Lastly, mortality was reduced from

87% to 33% in leukopenic mice with IV infections of VREF and 2.5 mg/kg PG-1 in contrast to the control group without PG-1 [60]. These data show that PG-1 can protect mice from lethal challenges of multiple types of bacterium, indicating a future potential for PG-1 in the treatment of local or systemic infections from pathogens. Furthermore, Cheung et al. (2008) were able to demonstrate enhanced bacterial infection resistance to Actinobacillus suis (A. suis) in ectopically expressing PG-1 transgenic mice. 87% of the PG-

1 transgenic mice challenged with 1.0 x 108 to 5.9 x 108 CFU/ml of A. suis survived infection compared to only 37% of the wild-type (WT) mice. Along with a significantly higher survival rate, the transgenic mice 13

had a greater number of neutrophils, reduced pulmonary pathological lesions, and lower bacterial load in their lungs compared to the WT [71]. These results demonstrate the potential for PG-1 to be used therapeutically to enhance resistance of bacterial infections in animals.

Although studies have shown beneficial effects of PG-1 in vivo, they are limited in the sense that they focus solely on the mature PG-1 domain. Previously our lab investigated the protective effects of recombinant pro-form (ProPG-1), cathelin (pro-region of PG-1), and mature PG-1 using a DSS mouse model of colitis. E. Huynh (unpublished [11]) demonstrated that oral-administration of the peptides resulted in reduced body weight loss, improved disease activity index (DAI) scores, histomorphological improvements, and decreased inflammatory response compared to challenged control mice. Interestingly,

ProPG-1 decreased pro-inflammatory tumor necrosis factor alpha (TNF ) and cyclooxygenase-2 (COX-2) while increasing the epithelial tissue repair trefoil factor 3 (TFF3) [11]. ProPG-1 provided evidence of its ability to modulate inflammation and enhance tissue-repair, demonstrating its curative potential. However, a high dose of 10 mg/kg body weight (BW) was used, necessitating higher efficiency and more cost- effective production of the recombinant peptide for clinical application.

2.4 Recombinant PG-1 Production and Expression in Pichia pastoris

Recombinant technology for protein production has become an increasingly popular and inexpensive technique in the last decade, turning into a multibillion-dollar market. A vast array of host expression systems has been reported, with yeast being a popular choice for their rapid growth and genetic manipulation ease [72]. Chemically synthesized peptide production develops proteins with good purity and yield but is better suited for small proteins (<50 amino acids), as cost for lengthy proteins or large quantities of proteins would be impractical [73]. Recombinant technology allows for the inclusion of secretory pathways for successful protein folding and assembly, as well as post-translational modifications without altering peptide function [73]. This would be advantageous for the use of proteins that require modification to render biological activity, such as PG-1. 14

The most common yeast strains used in recombinant protein production are Saccharomyces cerevisiae (S. cerevisiae) and Pichia pastoris (P. pastoris). The methylotrophic P. pastoris is more suitable than S. cerevisiae with respect to their protein glycosylation methods. Hyperglycosylation of N-linked high- mannose oligosaccharides from S. cerevisiae decreases protein activity and leads to immunological problems making it unsuitable for mammalian use [74]. The P. pastoris expression system allows for high cell densities of active recombinant protein to be secreted directly into the minimal culture medium [74]. A potential drawback would be that expression level limitations are dependent on a suitable culture and may require modifications to composition, as well as bioreactor parameters like temperature, pH, oxygen levels, and feeding [75].

One of the most successful and widely applied promoters for heterologous protein production in P. pastoris would be the methanol-induced alcohol oxidase 1 (AOX1), as it utilizes methanol as a carbon source to control enzyme expression for methanol metabolism [76]. Three phenotypes of P. pastoris host strains in terms of methanol utilization exist. The first, methanol utilization plus (Mut+) has both AOX1 and

AOX2 genes intact and consume methanol actively where the second, methanol utilization slow (MutS) has a disruption in AOX1 and consumes methanol slowly as it relies on AOX2 for methanol metabolism. The third, methanol utilization minus (Mut-) has both AOX genes knocked out and is incapable of growth solely on methanol as a carbon source [77]. Although AOX1 is a powerful and effective system, there are a few disadvantages (reviewed in [78]). The first being that two carbon sources (glucose/glycerol and methanol) are required and must be switched at a specific timepoint for successful production. The second disadvantage would be the use of methanol itself for its flammable nature, presenting a fire and safety hazard in large-scale production where vast quantities are stored at hand [72]. Lastly, methanol poses as a considerable health risk as it is derived from petrochemicals and may not be suitable for recipients of therapeutic applications [78]. As a result, other promoters such as the constitutive glyceraldehyde-3- phosphate dehydrogenase (GAP) promoter have gained considerable attention as alternatives. The GAP

system expresses heterologous proteins with glucose or glycerol as a carbon source, removing methanol- associated safety concerns. In addition, GAP allows for simultaneous generation of protein biomass and synthesis without the limitation of methanol availability for induction using the AOX1 promoter, reducing strict cultivation controls [75] (Figure 4). It also would solve the accumulation of by-products, like formaldehyde and hydrogen peroxide, of methanol metabolism concerns associated with AOX1 [79].

Interestingly, the GAP promoter has not been a popular choice due to concerns about cytotoxic effects of constitutive production in P. pastoris [72]. Nevertheless, studies have not been able to demonstrate evidence of cytoxicity and have shown recombinant protein production levels to be comparable to those using the

AOX1 promoter [72].

Figure 4. Comparison of AOX1 and GAP promoter for protein cultivation. Flow chart representation of the cultivation process of recombinant proteins for the methanol-inducible AOX1 promoter versus the constitutive GAP promoter.

2.5 Citrobacter rodentium

The Gram-negative mucosal pathogen Citrobacter rodentium (C. rodentium) is a murine specific bacterium that shares 67% of its genes with the attaching and effacing (A/E) human pathogens enteropathogenic E. coli (EPEC) and enterohemorrhagic E. coli (EHEC) [80, 81]. As such, it is commonly used to study enteric bacterial infections and examine host-pathogen interactions of colitis in vivo. These

A/E pathogens attach intimately to the intestinal epithelium to create a non-invasive pedestal-like formation on host cells and use a type III secretion system (T3SS) to inject effectors into them to alter cell signaling and normal host immune response [82]. Specifically, infection with C. rodentium occurs through the fecal- oral route and leads to colonization of the cecal patch, before migrating to the primary infection site in the distal colon a few days after infection [83]. Oral challenge with C. rodentium causes brush-border destruction, intestinal crypt hyperplasia, immune cell infiltration and mucus-secreting goblet cell depletion

[84, 85]. As a result, immunocompetent mice have shown transient weight loss and diarrhea [80]. However, a robust adaptive immune response directly after infection can clear the pathogen in 3 to 4 weeks [84].

Nevertheless, genetic composition of mouse strains and age influences the severity of disease as some mice develop fatal infection [86, 87]. The C. rodentium infection model allows for further insight into the understanding of infectious colitis pathogenesis and provides the potential for investigating and developing preventative as well as therapeutic treatments. Thus, it has been extensively used to replicate intestinal disorders such as Crohn’s disease, IBD, ulcerative colitis and colon tumorigenesis [88, 89].

3. Rationale and Research Objectives

Intestinal disorders and colitis affects millions of humans and food-animals world-wide. The pathogenesis behind this chronic inflammation is complex and not entirely understood. Furthermore, the significant rise in antibiotic-resistant bacteria and increase in animal-to-human transmission has warranted an urgent need for alternative anti-infective therapies to remedy these issues. Antimicrobial peptides offer a valuable potential solution due to their 1) their unique multifunctional methods of action, 2) broad- spectrum antimicrobial activity, and importantly 3) protective role in GI tract maintenance. Recently in our lab, it has been found that recombinant PG-1 reduced the pathological effects of chemically-induced colitis in mice suggesting a protective role through inflammation modulation and enhanced tissue repair. However, a high dose of 10 mg/kg body weight was used. Clearly for clinical application, higher efficiency and more cost-effective production of the recombinant peptide is needed. In addition, the effect of PG-1 in a pathogen- challenged colitis model has not been investigated. We hypothesized that: 1) the recombinant production of PG-1 using the GAP promoter will increase yield to improve its application in a production setting, and

2) the administration of recombinant PG-1 will decrease C. rodentium-induced colitis in a mouse model.

Therefore, this thesis aims to:

Objective 1: Examine if utilization of a GAP promoter will enhance the recombinant production of PG-1

in the yeast P. pastoris system.

Objective 2: Investigate the protective, anti-inflammatory role of recombinant ProPG-1 and mature PG in

the colon using a C. rodentium-challenged mouse model.

4. Materials & Methods

4.1 Construction and Transformation of ProPG-1-pUC Shuttle Vector into E. coli DH5

The ProPG-1 expression cassette was codon optimized for optimal expression in P. pastoris with the modification of the replacement of the native neutrophil elastase site separating the cathelin and mature

PG-1 domain to an enterokinase cleavage site (DDDDK) as previously described (Huynh E., unpublished

[11]). Synthetic lyophilized ProPG-1 was ordered in a pUCIDT (ProPG-1-pUC) shuttle vector with an ampicillin selection marker from Integrated DNA Technologies (IDT; Skokie, IL, USA). The vector was reconstituted with nuclease-free water (Thermo Fisher Scientific; Canada) to a concentration of 40 ng/µl.

Competent E. coli DH5 were generated using a one-step preparation method as previously described [90] with a transformation efficiency of 2.4 x 106 CFU/µg DNA. The ProPG-1-pUC shuttle vector was transformed into competent E. coli DH5 using a bacterial heat-shock method. Briefly, 3 µL of plasmid

(120 ng/µL) was added to 100 µL of E. coli DH5 and iced for 10 minutes before being placed in a 42°C water bath for 45 seconds. Immediately after heat shock, the cells were placed on ice for 5 minutes to reduce cell damage. A total of 900 µl of nutrient-rich SOC (0.5% yeast extract, 2% tryptone, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4, 20 mM glucose) media was added for cell recovery before cultures were incubated at 37°C for 45 minutes at 200 rpm. A 100 µL sample of the culture was plated onto low-salt LB agar plates (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 1% agar) supplemented with 100 ng/mL ampicillin (Sigma Aldrich, Canada) and grown at 37°C overnight.

4.2 Excision of ProPG-1 from pUC Shuttle Vector

Successful transformants were grown in 3 mL of LB media and 50 ng/mL ampicillin at 37°C at

200 rpm overnight. Plasmids were isolated using an EZ-10 Spin Column Plasmid DNA Miniprep Kit (Bio

Basic; Markham, ON, Canada) according to manufacturer’s instructions with an additional elution step.

The restriction enzyme SapI (Sense: 5’-GCTCTTC(N)1 -3’, Anti-sense: 3’-CGAGAAG(N)4 -5’) (New

England Biolabs Inc.; Canada) was used for plasmid digestion overnight at 37°C. Products were run on a

0.7% agarose DNA gel for 30 minutes at 135 volts, and bands were visualized using the ChemiDocTM MP

System (Biorad, CA, USA) to confirm successful digestion. A duplicate gel was run simultaneously and exposed briefly to a long-wave UV light for visualization of the appropriate bands to be excised. Excised

ProPG-1 expression cassette was purified using an EZ-10 Spin Column DNA Gel Extraction Kit (Bio Basic;

Markham, ON, Canada) following manufacturer’s instructions and run on a 0.7% agarose gel for 30 minutes at 135 V to verify product size.

4.3 Ligation of ProPG-1 into pD915-GAP Vector

Purified digested ProPG-1 fragments were ligated into the pD915 P. pastoris expression vector from ATUM (Newark, CA, USA) with a GAP promoter in placement of the traditional AOX1 promoter.

Ligation occurred for 1 hour at room temperature and then for 18 hours at 16°C with a 1:3 vector to insert ratio using T4 DNA Ligase (Promega; Madison, WI, USA) according to supplier’s instructions. ProPG-1-

GAP was subsequently transformed into E. coli DH5 as previously stated. A 100 µl sample of the ligation reaction was plated onto low-salt LB agar (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 1% agar) plates supplemented with 25 µg/mL light-sensitive zeocin (InvivoGen; San Diego, CA, USA) as a selection marker. Plates were incubated in the dark overnight at 37°C.

4.4 Confirmation of Ligation into pD915-GAP Vector

Positive transformants were selected and grown in 3 mL of LB media with 25 µg/mL zeocin at

37°C and 200 rpm overnight, before plasmids were isolated using an EZ-10 Spin Column Plasmid DNA

Miniprep Kit (Bio Basic; Markham, ON, Canada) following manufacturer’s instructions with an additional elution step. Diagnostic restriction enzyme double digests with BsrGI (New England Biolabs Inc.; Canada) and High-Fidelity EcoRI® (New England Biolabs Inc.; Canada) following manufacturers procedures were used to ensure proper integration. Digests were incubated for 1 hour at 37°C before being heat inactivated

at 65°C for 20 minutes. Products were run on a 1% agarose gel at 135 V for 30 minutes along with a

HighRanger 1kb DNA ladder (Norgen Biotek Corporation; Thorold, ON Canada) to determine product size. Additionally, the AAC Genomics Facility (University of Guelph, ON, Canada) performed sequence analysis using the anti-sense complement to the forward pD915-GAP promoter

(ATAATAGCGGGCGGACGCAT). The Basic Local Alignment Search Tool (BLAST) (National Center for Biotechnology Information, U.S National Library of Medicine; Bethesda, MD, USA) was utilized to confirm sequencing results.

4.5 Transformation into P. Pastoris by Electroporation

GAP: ProPG-1 confirmed sequence plasmids were linearized using SwaI (New England Biolabs

Inc.; Canada) for 1 hour at room temperature and then overnight at 30°C. Samples were run on a 0.7% agarose gel for 30 minutes at 135 V to confirm linearization before phenol chloroform purification and ethanol precipitation of the DNA was conducted. DNA was resuspended in 15 µL of nuclease-free water

(Thermo Fisher Scientific; Canada). Competent Wild-type X33 (Invitrogen; Canada) and Muts (Invitrogen;

Canada) P. pastoris strains were prepared as previously described [91]. Linearized plasmid DNA was dissolved in 40 µL of competent cells and electroporation was carried out using an electroporation cuvette with a 2.0 mm gap width in an Electro Cell Manipulator 399 (BTX; USA) as described by Lin-Cereghino et al., (2005). Transformed cells were plated on YPD agar (1% yeast extract, 2% glucose, 2% peptone, 2% agar,) plates with 100, 250, and 1000 µg/mL concentrations of zeocin and incubated for 3 days at 30°C.

4.6 Transformation Screening and Selection

Single colonies were grown in 1 ml of YPD (1% yeast extract, 2% glucose, 2% peptone) in a shaking incubator for 3 days at 250 rpm, with 50 µL samples taken at 24, 48, and 72 hours. For every 50

µL of sample taken, fresh YPD was replaced into the culture. Collected samples were immunoblotted to check for ProPG-1 expression, with AOX1: ProPG-1 as a comparison control. ProPG-1-expressing

transformants were then subjected to further screening in 10 mL of BMGY medium (1% yeast extract, 2% peptone, 0.1 M KH2PO4 pH 6.0, 1.34 % yeast nitrogen base, 0.00004% biotin, 1% glycerol). For every 100

µL of sample taken at 24, 48, and 72 hours, fresh YPD was replaced into the culture. After 24 hours, AOX1:

ProPG-1 cultures were washed and transferred into BMMY medium (BMGY medium containing 0.5% methanol instead of glycerol) to induce expression. Expression was induced again at 48 hours with 100% methanol at 0.5% (v/v). All samples were spun for 3 minutes at 12, 000 rpm, supernatant was collected into a sterile Eppendorf tube and frozen at -80°C until immunoblot analysis. The top transformants expressing similarly to the AOX1: ProPG-1 control were selected for shake-flask fermentation.

4.7 Shake-Flask Fermentation

Selected transformants were streaked on YPD agar plates with 100 µg/mL of zeocin for further purification. Incubations were done shaking at 250 rpm at 30°C, all collected samples (24, 48, and 72 hours) were spun for 3 minutes at 12, 000 rpm, and supernatant was collected into a sterile Eppendorf tube and frozen at -80°C until immunoblot analysis. After 24 hours incubation in BMGY, AOX1: ProPG-1 control cultures were washed and transferred into BMMY medium to induce expression, and induction again with

100% methanol at 48 hours at 0.5% (v/v). Again, a single fresh colony from each transformant was inoculated into 10 mL of YPD and incubated. The following day 500 µL of overnight culture was inoculated into 50 mL of fresh BMGY media in 250 mL baffled flasks and incubated for 3 days with 1 mL samples taken at selected time points. Immunoblot analysis determined transformants expressing similarly to

AOX1: ProPG-1 control were selected for bioreactor fermentations.

4.8 Bioreactor Fermentation & Optimization of Recombinant ProPG-1

Large-scale fermentations were conducted and optimized in a New Brunswick Scientific

BioFlo/CelliGen 115 Benchtop Bioreactor (Edison, NJ, USA). Selected transformant cultures were inoculated from fresh colonies into 50 mL of YPD media in 250 mL baffled flasks and incubated at 30°C

shaking at 250 rpm for 24 hours before the start of fermentation. The bioreactor was sterilized containing

1 L fed-batch growth medium (Appendix Table A1), cooled to 30°C with a maintained pH of 5.0- 5.5 using

NH4OH (Thermo Fisher Scientific; Canada) before inoculation with the 50 mL overnight culture at 0-hour

FT (fermentation time). Antifoam A (Sigma Aldrich, Canada)) was added manually to control excessive foaming as needed. The agitation speed was gradually increased from 250 rpm to a maximum of 700 rpm to ensure 20% dissolved oxygen (DO) content and an aeration rate of 5.0 L/min at 10 psi was maintained.

The glycerol-fed batch phase would begin after the increase in DO concentration about 24-hour FT. Fifty- percent (v/v) glycerol (Thermo Fisher Scientific, Canada) supplemented with PTMI (Appendix Table A1) was then maintained at 15 mL/L/hr. To further enable protein expression and secretion of recombinant

ProPG-1, the bioreactor temperature was lowered to 26°C at 20 or 24hr FT. Fermentations were ended at

48 or 72 hours FT and cultures were centrifuged for 1 hour at 4°C at 3000 rpm to separate the rProPG-1- containing supernatant from the cell pellets. Supernatants were stored at -20°C for downstream applications.

4.9 Protein Sample Preparation

At approximately 95% purity, standard synthetic lyophilized mature PG-1 (Top-Peptide

Corporation; Shanghai, China) was dissolved in filter-sterilized 0.01% acetic acid supplemented with 0.1% bovine serum albumin (BSA) (Sigma Aldrich, Canada) to decrease non-specific plasticware adsorption, and diluted to working-stock concentrations of 200 ng/µl. Standard mature PG-1 was stored at -80 °C until use. Recombinant ProPG-1 from fermentation supernatant was dialyzed against 25 mM Tris-Cl (pH 7.0) and concentrated through ultracentrifugation using an Amicon® Ultra-15 10 kDa cut-off centrifugal filter

(Millipore Sigma; Darmstadt, Germany) according to manufacturer’s instructions. Ultracentrifugation retentate and supernatant flow-through was filter-sterilized and desalted with 1 x phosphate buffered saline

(PBS) before being stored at -80 °C.

4.10 SDS-PAGE & Western Blot Analysis

Supernatant samples were separated by 11% SDS-PAGE and transferred for 14 minutes at 25 V and 1.0 A onto a 0.45 µm PVDF membrane (BioRad; Hercules, California, USA) using a Trans-Blot Turbo blotting system (BioRad; Hercules, California, USA). Membranes were blocked at room temperature for 1 hour in 5% skim milk TBST (0.3% Tris, 0.8% NaCl, 0.02% KCL, 0.1% Tween 20), before being washed for 10 minutes in TBST three times. Membranes were then incubated at room temperature for 1 hour in

1:1000 polyclonal rabbit anti-cathelin primary antibody (GenScript; Piscataway, New Jersey, USA) and washed for 10 minutes in TBST three times again. After washing, the membrane was incubated at room temperature for 1 hour in 1:2000 horseradish peroxidase conjugated anti-rabbit IgG secondary antibody

(Cell Signaling Technology, Danvers, Massachusetts, USA) before band visualization using ClarityTM

Western ECL Substrate (BioRad; Hercules, California, USA) according to manufacturer’s instructions.

Blots were imaged using the ChemiDocTM MP System (Biorad, CA, USA).

4.11 Ethics Statement

All experiments and procedures involving the use of animals were approved by the University of

Guelph Animal Care Committee (AUP #3250) in accordance with the Canadian Council on Animal Care guidelines (CCAC, 2017).

4.12 Bacterial Preparation

C. rodentium DBS 100 was kindly provided by Dr. B. Coombs (McMaster University, ON,

Canada), and was cultured in LB broth at 37 °C with agitation at 200 rpm overnight before being centrifuged at 4000 x g for 10 minutes at room temperature. Pelleted bacteria were washed and resuspended in sterile

1 x PBS at a concentration of 2.0 x 109 colony-forming units (CFU) per mL. The viable count of inocula was determined by retrospective plating on MacConkey agar plates (24 hours at 37 °C).

4.13 Animals & Induction of Colitis

Six- to eight-week-old C57BL/6 (C57BL/6NCrl) female mice were purchased from Charles River

Laboratories (St. Constant, QC, Canada). All mice were housed under temperature-controlled Level 2

Biosafety conditions in the Isolation Facility at the University of Guelph. Mice were provided a 12-hour light and 12-hour dark cycle, water and standard 14% protein rodent diet ad libitum. After one week of acclimatization, mice were fasted of food and water for 4 hours prior to initial infection (Day 1), when mice were administered 50 µl of 2.0 x 107, 2.0 x 108, or 2.0 x 109 CFU/mL C. rodentium DBS 100 via orogastric gavage [92, 93]. A second dose was given on Day 3 or Day 5 to ensure pathogen establishment. Oral administration of recombinant ProPG-1 (10 mg/kg body weight) and synthetic mPG-1 (10 mg/kg body weight) via gavage was given beginning on Day 0 and consecutively for the duration of the trial.

Unchallenged mice were orally gavaged with 150 uL of 1x PBS at the same time. All mice were weighed and the Disease Activity Index (DAI; See section 4.16) was measured daily. Mice were euthanized via CO2 asphyxiation on Day 8 or Day 10 for tissue collection.

4.14 Measurement of Bacterial Load

To determine bacterial load in the stool, fresh fecal pellets from individual cages were collected into sterile Eppendorf tubes at designated timepoints throughout the study. Fecal samples were weighed and diluted to 0.1 g of feces per mL in sterile 1x PBS. Samples were homogenized via vortex and serially diluted onto MacConkey agar plates to determine CFU/mL. Viable bacteria were determined as previously described [85, 94], and counted after 24 hours incubation at 37°C.

4.15 Fecal Water Content

As a measurement of diarrhea within the stool, fresh fecal pellets from individual cages were collected into sterile Eppendorf tubes shortly before euthanization (Day 10). Samples were weighed,

incubated for 24 hours at 60°C and weighed again. The percentage of fecal water content was calculated by dividing the difference of wet and dry weight by the wet weight [95].

4.16 Disease Activity Index Score

All animals were assessed daily to determine their disease activity index (DAI) score as previously published by Maxwell et al. [96] to assess the severity of infection. The DAI score was determined as the sum of the stool consistency, stool blood, and percent (%) body weight loss parameters as shown in Table

1.1. The following formula was used to determine the percent body weight loss of the mice in relation to their initial body weight on Day 0 [96].

(퐵표푑푦 푊푒𝑖𝑔ℎ푡 표푛 퐷푎푦 푋 − 퐼푛𝑖푡𝑖푎푙 퐵표푑푦 푊푒𝑖𝑔ℎ푡) % 퐵표푑푦 푊푒𝑖𝑔ℎ푡 퐿표푠푠 = 푥 100 Initial Body Weight

Table 1. Disease Activity Index (DAI) scoring criteria modified from [96].

DISEASE ACTIVITY INDEX (DAI) Stool Consistency Stool Blood Body Weight Loss (%) Score Normal Normal 0-1 % 0 Moist and Sticky Brown/Reddish Colour 1-5 % 1 Soft Visible Blood 5-10 % 2a Diarrhea Rectal Bleeding 10-20 %b 3 a If a mouse scores ≥ 2 in any of the above criteria, supportive therapy must be initiated, and monitoring increased to 4 times per day. b If a mouse scores ≥ 20% body weight loss, the mouse will be euthanized.

4.17 Tissue Collection

Immediately following euthanasia, the entire small and large intestines were removed, separated, weighed, and straightened for length measurements before photographs were taken. Four empty small intestinal samples (~1 cm each) were taken starting from the jejunum (~10 cm distal the pylorus), and four empty colon samples (~1 cm each) were taken starting from the base of the caecum. All tissues were rinsed thoroughly in 1 x phosphate buffered saline (PBS) immediately and then placed into appropriate solutions

for later analysis; 2 x 1 cm sections into liquid nitrogen, 1 x 1 cm section into RNA Later, and 1 x 1 cm section into 10% formalin.

4.18 Histomorphology

Approximately 0.5 cm colon cross-sections were fixed in 10% formalin for 24 hours before being transferred to 70% ethanol. Tissues were processed overnight using an ExcelsiorTM ES Tissue Processor

(Thermo Scientific; MA, USA), then embedded vertically in paraffin before being sectioned at 5 µm and mounted onto FisherbrandTM SuperfrostTM Plus microscope slides (Thermo Fisher Scientific; Canada).

Slides were stained with hematoxylin and eosin (H&E) and coverslipped with FisherfinestTM Premium

Cover Glasses using CytosealTM glue (Thermo Fisher Scientific; Canada). Slides were examined blindly under bright-field using a Leica DM R microscope (Leica Microsystems; Concord, ON, Canada). Crypt depth and lamina propria width were measured at 100x magnification (field diameter of 0.2 mm) in 10 well- defined crypts in each colon cross-section for each treatment group (4 sections for unchallenged, 6 for

ProPG-1, 8 for challenged and mPG-1). Tunica muscularis width was measured from 3 randomly selected regions per cross-section at 10x magnification. The average number of neutrophils per high-power field

(40x magnification, field diameter of 0.45 mm) was determined from 10 randomly selected regions of the cross-section. The average number of goblet cells per 10 well-defined crypts were counted in each cross- section from each treatment group. Measurements were taken in a blind fashion using ImageJ Software

Version 1.51 (National Institutes of Health, USA). Histological damage was assessed single-blindly according to Table 2. High-quality photos were taken with an Olympus BX45 microscope (Olympus

America Incorporated; Melville, NY, USA) using cellSens Standard Software Version 1.12 (Olympus

America Incorporated; USA).

Table 2: Definition of categories and general criteria for histomorphological scoring

CATEGORY CRITERION DEFINITION SCORE VALUE1 Inflammatory Severity Neutrophil density (per 10 40x high-power fields; field diameter of 0.45 mm): Cell Infiltrate 0 = No neutrophils 0 1 = 1 or 2 neutrophils 1 2 = 3 - 5 neutrophils 2 3 = >5 neutrophils 3 Extent Extent of inflammatory cell infiltration: Mucosal 0.5 Mucosal and submucosal 1 Mucosal, submucosal, and transmural 1.5 Epithelial Hyperplasia Percent (%) above height of control/baseline crypt (0.2 mm); visible as crypt elongation: Changes2 No change: 0 0 Mild: 1- 50% 1 Moderate: 51-100% 2 Marked: >100% 3 Goblet Cell Change in goblet cell numbers relative to baseline goblet cell numbers per crypt: Count 0 = No change from baseline goblet cell count 0 1 = Increase or decrease of 1-3 goblet cells from baseline 1 2 = Increase or decrease of 4-8 goblet cells from baseline 2 3 = Increase or decrease of 8 or more goblet cells from baseline 3 Erosion Loss of surface epithelium: 0 = No pathological changes observed 0 1= Attenuation (gaps of 1-10 epithelial cells/lesion) 1 1.5 = Absence of epithelium (gaps of >10 epithelial cells/lesion) 1.5 Dysplasia Ordered arrangement of crypt epithelial cells & crypt epithelial cell nuclei; loss of epithelial polarity Mild: Mostly ordered arrangement, nuclei are in basal location 0 Moderate: Slightly disordered arrangement, some nuclei are near/in apical location 1 Severe: Disordered arrangement, nuclei are in apical location 2 Mucosal Crypt Loss Mucosa devoid of crypts 1 Architecture

1Score Value Max = 15 points; 2From well-oriented samples with longitudinally cut crypts only; Modified from [97] and [98].

4.19 Real-Time quantitative PCR

For assessment of gene expression profiles, RNA was isolated from ~1 cm cross-section segments of frozen colon samples using a dounce for homogenization and the E.Z.N.A® Total RNA Kit II (Omega

Bio-Tek; Norcross, Georgia, USA) according to manufacturer’s instructions. Total RNA was isolated, treated with RNase-free DNase I (Thermo Fisher Scientific; Canada), and first-strand complementary DNA

(cDNA) was synthesized using 5X All-In-One RT MasterMix (Applied Biological Materials; Richmond,

BC, Canada) according to manufacturer’s instructions. Samples were run on a 1% agarose gel at 135 V for

40 minutes to check successful DNase treatment and for RNA degradation. Quantitative RT-PCR (RT- qPCR) was performed using a CFX96 Real-Time PCR Detection System (Bio-Rad; Mississauga, ON,

Canada) with SsoAdvancedTM Universal SYBR® Green Supermix (Bio-Rad; Mississauga, ON, Canada).

Reactions were performed in duplicate under the following conditions: denaturation at 95 °C for 2 minutes, three-step amplification including denaturation at 95 °C for 25 seconds, annealing at 59 °C for 25 seconds, extension at 70 °C for 15 seconds and a subsequent melting curve (65 °C to 95 °C) determination with continuous fluorescence measurement. Mouse primers used were checked for a valid efficiency between

90 - 110 % and are outlined in Table A2 (Appendix). Relative mRNA levels were determined using the comparative Ct method with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and β-actin used as internal controls.

4.20 Statistical Analysis

Statistical analysis was performed using a two-factor analysis of variance (ANOVA) or a one-way repeated measures ANOVA for body weight and DAI over the course of the trials, through either GraphPad

Prism Version 7.04 (La Jolla; CA, USA) or SPSS Version 24 software (Armonk; NY, USA). Data sets were analyzed by Tukey’s test for multiple comparisons to determine statistical differences between groups. Data were deemed significant at a P-value of *p <0.05, **p <0.01, ***p <0.001 compared to control. Results are expressed as a mean ± standard error of the mean (SEM). 30

5. Results

5.1 Recombinant ProPG-1 Production in P. pastoris using GAP Promoter

Complementary DNA for ProPG-1 codon-optimized for expression in P. pastoris was synthesized by Integrated DNA Technologies (IDT) and ligated into the pD915-GAP expression vector (Figure 5-A).

An alpha-mating secretion signal factor was incorporated upstream of the ProPG-1 cDNA to enable recombinant protein expression and secretion into fermentation culture medium. The ProPG-1 expression cassette contained an enterokinase (EK) cleavage site separating the cathelin from the mature PG-1 domain

(Figure 5-B). Diagnostic restriction enzyme double digests using BsrGI (within the ProPG-1 cassette) and

HF-EcoRI (within the vector) (Figure 5-A) were used to verify ligation into the pD915-GAP vector (Figure

6). The double digestion resulted in a 400 bp product which is the correct size between the two restriction enzyme sites. Furthermore, sequencing analysis confirmed successful integration, of the now GAP: ProPG-

1 vector.

GAP: ProPG-1 vector plasmids were linearized with SwaI, transformed into P. pastoris, and screened at small-scale for ProPG-1 expression. A total of 284 colonies were screened, with 207 and 77 coming from Muts and WT P. pastoris strains, respectively. Immunoblot analysis showed 138/207 (66.6%)

Muts and 28/77 (36.4%) WT transformants expressed ProPG-1 (data not shown). These ProPG-1- expressing transformants were subjected to further screening in 10 mls of BMGY medium, before the top transformants expressing similar levels of ProPG-1 to the AOX1: ProPG-1 control were selected for 50- mL shake-flask fermentation. The top four (H11, H13, I1, and J8) candidates from the shake-flask fermentations that expressed similar levels of ProPG-1 to AOX1: ProPG-1 (Figure 7-A) were selected for bioreactor fermentations using an optimized fermentation culture medium (Appendix Table A1) for recombinant PG-1 expression in P. pastoris. Bioreactor fermentation conditions were optimized to attempt to further increase recombinant protein expression yield of ProPG-1. Optimal expression conditions were determined to be a pH of 5.0, and a decrease in bioreactor temperature from 30°C to 26 °C at 20 hours 31

fermentation time (FT). The highest expression of ProPG-1 from the bioreactor fermentations out of the top four candidates came from I1. An expression of 0.95 ± 0.004 g/ L from I1 ProPG-1: GAP was detected

48 hours after induction (Figure 7-B) compared to AOX1: ProPG-1 at 0.8 ± 0.2 g/ L (Huynh E., unpublished

[11]). The harvested ProPG-1-containing supernatant was further concentrated 4-fold, desalted, and filter- sterilized prior to use in the subsequent C. rodentium challenged mouse study.

Figure 5. Expression vector for ProPG-1 production in P. pastoris. (A) Visual schematic of the pD915-GAP expression vector showing the positioning of the ProPG-1 insert (blue block-arrow) as well as the location of the restriction enzymes BsrGI, EcoRI, and SwaI (grey). (B) Schematic representation of the proform of PG-1 (ProPG-1) expression cassette displaying its enterokinase (EK) cleavage site (DDDDK) (orange block) to release the mature peptide domain (purple block) from the cathelin domain (blue block).

Figure 6. Restriction enzyme digests for confirmation of ProPG-1 ligation into pD915-GAP vector. Isolated plasmids were digested with BsrGI, HF-EcoRI, or both for restriction enzyme double digests to confirm successful ligation into the pD915-GAP vector and analyzed by 1% agarose gel. Lane contents in order from left to right are: (1) 1 kB DNA ladder, (2) uncut plasmid, (3) plasmid digestion with BsrGI, (4) digestion with HF-EcoRI, and (5) digestion with both BsrGI and HF-EcoRI yielding a ~400 bp cutout. Molecular masses of standards are denoted in base pairs (bp).

Figure 7. Western blot analysis of rProPG-1 transformants and confirmation of expression from fermentation. (A)Western blot analysis of top screened transformants after 48 hours of shake-flask fermentation. (B) rProPG-1 from the highest expressing bioreactor fermentation transformant (I1) was detected in immunoblot analysis from 24 to 72 hours. Pro44 (AOX1: ProPG-1) served as a positive control with expression induced by methanol after 24 hrs. The wild-type P. pastoris X33 served as a negative fermentation supernatant control. Chemically synthesized ProPG-1 (200 ng; Std) was used to quantify rProPG-1 produced by P. pastoris in the bioreactor fermentation. Molecular masses of standards are reported in kDa.

5.2 PG-1 Reduced C. rodentium-Induced Intestinal Infection

C. rodentium is a commonly used model to investigate enteric bacterial infections and observe host- pathogen interactions of colitis in vivo [99-101]. Pilot animal studies #1 & #2 were conducted to determine the optimal dosage and dosage timepoints of C. rodentium for intestinal inflammation in C57BL/6 mice

(Appendix Figure A1-A3). It was determined that the best concentration for inducing C. rodentium infection was 2.0 x 109 CFU/mL through oral gavage on Day 1 and Day 3 (Appendix Figure A2, & A3).

The animal study is outlined in Figure 8. There was significant weight loss in the challenged mouse group in comparison to unchallenged mice over the course of the trial (p <0.05; Figure 9-A). Similarly, increasing DAI scores (more severe disease signs) in the challenged mouse group compared to unchallenged mice were present over the course of the trial (p <0.001; Figure 9-B). All treatment groups had slightly increased DAI scores on Days 1, 2, and 3 compared to Day 0, attributed to the starvation of food and water for the first dose of C. rodentium and/or initial stress from gavage. However, all challenged groups had increased DAI scores beginning on Day 4, while unchallenged mice returned to levels similar to Day 0. Visually, challenged mice displayed ruffled coats, lower levels of activity, and moist stool/diarrhea. All mice survived until endpoint except for 2 mice within the ProPG-1 group that died due to gavage complications.

As seen over the course of the trial, treatment with mPG-1 inhibited C. rodentium-induced body weight loss (p <0.01; Figure 9-A). mPG-1-treated mice had similar weight gains as unchallenged mice and

5.9% greater weight gains compared to challenged mice (p =0.06), although no statistical differences in body weight on Day 10 were seen between mPG-1-treated mice and challenged mice (Figure 9-A).

Similarly, the DAI of mPG-1-treated C. rodentium-infected mice initially increased but subsequently decreased over the course of the trial (p <0.01), and the DAI were 10-fold lower on Day 10 relative to untreated challenged mice (p<0.01), but comparable (p =0.74) to the DAI of unchallenged mice (Figure 9-

B). Visually, after Day 4, mPG-1-treated mice displayed characteristics of normal mice with normal; coats, 36

activity levels, and hard stool consistency. Mice given ProPG-1 had slightly reduced yet statistically insignificant body weight loss (Figure 9-A) and DAI scores (Figure 9-B) relative to challenged mice from

Day 1 up until Day 8. After Day 8, ProPG-1-treated mice began to increase in body weight bringing them up to initial body weight (Day 0) on Day 10 and had lower DAI scores compared to challenged mice although statistically insignificant (Figure 9-A & -B). Most of the ProPG-1-treated mice displayed observational characteristics of challenged mice during the trial up until Day 8, when they began to display normal coats, increased activity levels, and lower incidences of moist stool/diarrhea.

Fecal samples were serially diluted, and plated on MacConkey agar plates on Day 0, 4, 7, and 10 throughout the trial, to determine bacterial load. C. rodentium was not detected in any fecal samples from unchallenged mice for the duration of the trial, whereas all other treatment groups had an average bacterial load of 9.0 log10 CFU/mL by Day 4 (Figure 10-A). While both ProPG-1 and challenged mice had an average bacterial load of 8.0 log10 CFU/mL on Day 7 and Day 10, C. rodentium was not detected in the mPG-1 group on either day. Although there were numerical increases in water consumption in the challenged group, no statistical differences were observed (Figure 10-B). Another index of the severity of colitis is amount of water content within the stool. By Day 10 post-infection, both the unchallenged and mPG-1 groups had undetectable fecal water content while ProPG-1 had a 3.03% fecal water content, which was still less than the 5% fecal water content measured from the challenged group (Figure 10-C). However, no significant differences were observed for small intestine, colon, and caecum length and weight among all treatment groups (Figure 11-A-F).

Figure 8. Animal Study Overview. Visual representation of the (A) trial timeline showing induction of colitis, (B) treatment groups, and (C) number of mice used per treatment. Two mice within the ProPG-1 group died due to gavage complications (8 mice -2 = 6 mice). BW = body weight.

Figure 9. Effects of PG-1 on body weight and DAI. Protegrin treatments reduced body weight loss (A) and reduced disease activity index (B) compared to challenged mice. Higher DAI score represents more severe disease signs. **p <0.01 indicates DAI on day 10 is significantly different than in challenged control mice. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one-way repeated measures ANOVA post hoc Tukey test over the course of the trial.

A Bacterial Load 10 Unchallenged

8 Challenged L

m ProPG-1 /

U 6 mPG-1

0 4

g o

L 2

0 0 2 4 6 8 10 Day

B Average Daily Water Consumption C Fecal Water Content on Day 10

15 ) 6

(

)

t m

n 4

( 10

s a

o 5 2

c e F 0 Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1

Challenged Challenged

Figure 10. PG-1 reduced C. rodentium infection. (A) Bacterial load of C. rodentium was measured on Day 0, 4, 7, and 10. Fecal samples were collected, resuspended in 1 x PBS and serially diluted onto

MacConkey agar plates. Colony counts are expressed as Log10 CFU/mL. (B) Average daily water consumption per treatment was calculated and is presented as mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one-way ANOVA post hoc Tukey test. (C) Fecal samples from individual cages were collected shortly before euthanization, weighed, and incubated for 24 hours at 60°C, and re- weighed. The percent of fecal water content was calculated by dividing the difference of the wet and dry weight by the wet weight.

A B S m a ll In te s tin e L e n g th S m a ll In te s tin e W e ig h t

4 0 1 .5

3 0

) )

m 1 .0

(

h 2 0

e n

e 0 .5 W L 1 0

U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d

C D C o lo n L e n g th C o lo n W e ig h t

7 .5 0 .4

0 .3

) )

m 5 .0

(

h 0 .2

n e

2 .5 W L 0 .1

U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d E F C a e c u m L e n g th C a e c u m W e ig h t

3 0 .3

) )

m 2 0 .2

(

n e

e 1 0 .1

W L

U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d

Figure 11. Effects of PG-1 on intestinal length and weight. Intestinal tissues were weighed and the length was measured. Protegrin treatments did not have any significant effects on small intestine length (A) or weight (B), colon length (C) or weight (D), or caecum length (E) or weight (F). Data represent the mean ± SEM.

5.3 PG-1 Decreased C. rodentium-Induced Histologic Lesions in the Colon

Colonic crypt hyperplasia, dysplasia, and immune cell infiltration were observed in H&E-stained sections of colon from mice challenged with C. rodentium (Figure 12, Appendix Figure A4). In contrast, unchallenged mice showed no significant lesions (Figure 12, Appendix Figure A4). The administration of

ProPG-1 and mPG-1 resulted in less severe mucosal lesions in challenged mice, and the colonic morphology was similar to unchallenged mice with decreased dysplasia and immune cell infiltrations

(Figure 12).

Colonic lesions were semi-quantitatively scored for crypt depth, goblet cell number, dysplasia, number of neutrophils, and summed for an overall histologic score. Treatment with mPG-1 had a 2.2-fold lower score of dysplasia, an epithelial change characteristic of colitis, compared to untreated challenged mice (p <0.05; Figure 13-A), while ProPG-1 had a 1.7-fold decrease although statistically insignificant.

Neutrophil counts were significantly lower in ProPG-1- and mPG-1-treated challenged mice compared to that of the untreated challenged mice (p <0.001; Figure 13-B). Similar trends were observed for crypt depth and goblet cell count although the differences between groups were not statistically significant (Figure 13-

C & -D). No significant differences in lamina propria width (Appendix Figure A5-A) or tunica muscularis width (Appendix Figure A5-B) were found among treatment groups. Compared to unchallenged mice, overall histological scores were increased in untreated challenged (p <0.01) and ProPG-1-treated challenged (p <0.05) groups based on the criteria in Table 2, while treatment with mPG-1 reduced the histological score to a level that was not significantly different from the unchallenged or challenged group

(Figure 13-E).

Figure 12. Representative hematoxylin and eosin stained colon sections. Histological evaluation of colonic sections on Day 10 demonstrate noticeable alterations to the mucosal morphology of the colon in C. rodentium challenged mice compared to unchallenged mice. This is manifested by immune cell infiltration (black arrows), colonic crypt hyperplasia (Appendix Figure A4), and dysplasia (not displayed). Both ProPG-1- and mPG-1-treated challenged mice resulted in improvements compared to challenged mice, but mPG-1-treated challenged mice prevented more of the C. rodentium-induced dysplasia and immune cell infiltration. Magnification: 40 x.

A B Average Total Number of Neutrophils Dysplasia Score *** *** ** * *** ***

0.75 15

e p

r t

c 0.50 10

r p

0.25 e 5

0 Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1

Challenged Challenged C D

Average Crypt Depth Goblet Cell Count

400 20

)

n m

300 u

(

p e

e 200 10

y r

100 o

C G

Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1

Challenged Challenged E Overall Histological Score 7.5 * **

5.0

c S 2.5

Unchallenged Challenged mPG-1 ProPG-1 Challenged

Figure 13. Effects of PG-1 on colon histomorphology. Compared to untreated mice challenged with C. rodentium, mPG-1-treated challenged mice had a decreased (A) dysplasia score, while both ProPG-1- and mPG-1-treated challenged mice had a decrease in (B) average total number of neutrophils. No significant differences were observed among treatments for (C) average crypt depth or (D) goblet cell counts, but mPG-1-treated challenged mice had a reduced overall histological score (E) based on the criteria in Table 2, relative to challenged mice although statistically insignificant. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 compared to controls; one-way ANOVA post hoc Tukey test.

5.4 PG-1 Modulates Intestinal Gene Expression during C. rodentium Infection

To investigate the protective effects of PG-1 on C. rodentium infection, RT-qPCR was utilized to examine intestinal expression of genes associated with self-defense and inflammation. Intestinal AMP gene expressions were examined as bacteria can enhance their progression by altering innate host responses and homeostasis through the modulation of host AMP expression [102]. The expression of epithelial regenerating islet-derived protein 3β (Reg3β) and regenerating islet-derived protein 3γ (Reg3γ) were both markedly up-regulated in challenged mice compared to unchallenged mice, and this induced expression was suppressed by mPG-1 and ProPG-1 (p <0.001; Figure 14-A &-B). No difference was observed in the expression of secretory phospholipase A2 (sPLA2) among the groups of unchallenged, challenged, and mPG-1- treated mice. However, treatment of ProPG-1 resulted in an increase (p< 0.001) in sPLA2 expression compared to challenged and unchallenged mice (Figure 14-C).

Mucins contribute to epithelial barrier protection, among other self-protecting mechanisms, as part of normal intestinal homeostasis [103]. The membrane-bound mucin Mucin-1 (Muc1) was increased by

2.5-fold in mPG-1 (p <0.05) and 3.2-fold in ProPG-1 groups (p <0.01) compared to the challenged group, while the opposite was observed for the primary mucin component, mucin-2 (Muc2), in which the induced expression by infection was significantly decreased in mPG-1 and ProPG-1 groups to a level comparable to unchallenged mice (Figure 15-A & -B). The expression of vascular cell adhesion molecule 1 (VCAM-1) was significantly suppressed by C. rodentium challenge, and this suppression was completely reversed by mPG-1 but not ProPG-1 (Figure 15-C). Interestingly, mPG-1 did not have any significant effects on early growth response 1 (EGR1) expression compared to challenged mice, while ProPG-1 increased its expression level to that of the unchallenged control (p <0.001; Figure 15-D). Additionally, C. rodentium infection suppressed the expression of stress-response hypoxia inducible factor 1 α subunit (HIF1α), which mPG-1 significantly reversed, while ProPG-1 had no significant effect on expression compared to all other treatment groups (Figure 15-E).

No difference was observed among the unchallenged, challenged, and mPG-1-treated mice on the expression of apoptotic regulator BCL Associated X (BAX), although its level was down-regulated by

ProPG-1 (Figure 16-A). The expression of anti-apoptotic B Cell leukemia/lymphoma 3 (BCL3) was up- regulated by C. rodentium infection, but this was reversed to levels comparable to the unchallenged control by both mPG-1 and ProPG-1 (p< 0.001; Figure 16-B). Furthermore, TLR expression was investigated as they are involved in innate immune responses due to stimulation by PAMPs, which are expressed by microbial pathogens [104]. The expression of Toll-like receptor 2 (TLR2) was significantly up-regulated by C. rodentium infection but was down-regulated by mPG-1 and ProPG-1 to levels in unchallenged mice

(p< 0.001; Figure 16-C). Treatment of mPG-1 had no effect on Toll-like receptor 6 (TLR6) expression relative to both controls, although its expression was suppressed in the ProPG-1 group compared to both challenged and unchallenged groups (Figure 16-D). Lastly, C. rodentium induced expression of the suppressor of cytokine signaling 3 (SOCS3) which was reversed by mPG-1 but not ProPG-1 to levels comparable to unchallenged mice (Figure 16-E). No gene expression changes were observed for cathelicidin-related AMP (Camp), angiogenin 4 (ANG4), interferon regulatory factor 7 (IRF7), and caspase-1 (CASP1) (Data not shown).

Reg3 A B Reg3 *** *** *** ***

n *** ***

n i

1.0 o

s i

s 1.0

E e

0.5

v e

i 0.5

R R

Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1 Challenged Challenged

C sPLA2 ***

4 ***

o i

s 3

r p

x 2

t a

l 1

e R

Unchallenged Challenged mPG-1 ProPG-1 Challenged

Figure 14. Effects of PG-1 on AMP gene expression within the colon. Colonic gene expression levels of the AMPs Reg3β (A), Reg3γ (B), and sPLA2 (C) were determined by RT-qPCR with values normalized against GAPDH and β-actin. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 compared to controls calculated by a one-way ANOVA post hoc Tukey test. All gene expression analysis is compared to challenged mice set to a relative expression of 1.

B A Muc1 * Muc2 ** 4 ***

* *** ***

o i

i 1.0 s

s 3

x x

2 E

e 0.5

a l

l 1

R R

Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1

Challenged Challenged

C VCAM-1 D

* EGR1 5 * 4 ***

***

n o

o ** i

4 * i s

s 3

e r

3 r

p p x

x 2

E E

2 e

i i

t t

a a l

l 1 e

1 e

R R

0 Unchallenged Challenged mPG-1 ProPG-1 Unchallenged Challenged mPG-1 ProPG-1 Challenged Challenged E HIF1

5 ***

n o

i 4

s e

r 3

e 2

a l

e 1 R

Unchallenged Challenged mPG-1 ProPG-1 Challenged

Figure 15. Effects of PG-1 on self-protecting gene expression within the colon. Protegrin treatments differentially affect gene expression of Muc1 (A) and Muc2 (B), cell adhesion molecule VCAM-1 (C), cell proliferation gene EGR1 (D), and the stress response factor HIF1α (E). Expression was determined by RT-qPCR with values normalized against GAPDH and β-actin. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 compared to controls calculated by a one-way ANOVA post hoc Tukey test. All gene expression analysis is compared to challenged mice set to a relative expression of 1.

A B A X B B C L 3

* * * *

* * * * * * * * *

o i

i 1 .0 1 .0

e 0 .5 0 .5

R R

U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d

T L R 6 T L R 2 C D * * * * * * * * 2 .0

n * * *

* * * n

i s

s 1 .5

s 1 .0

x x

E 1 .0

e e

v 0 .5

a l

l 0 .5

R R

0 .0 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d

E S O C S 3 * * *

* * * * * *

n o

i 1 .0

e 0 .5

e R

U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d

Figure 16. Effects of PG-1 on apoptotic, innate immune response, and cytokine gene expression within the colon. Protegrin treatments differentially affect gene expression of the apoptotic markers BAX (A) and BCL3 (B), the innate immune response genes TLR2 (C) and TLR6 (D), and the cytokine regulator SOCS3 (E). Expression was determined by RT-qPCR with values normalized against GAPDH and β-actin. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 compared to controls calculated by a one-way ANOVA post hoc Tukey test. All gene expression analysis is compared to challenged mice set to a relative expression of 1.

Discussion

The first objective of this thesis was to examine the utilization of a GAP promoter in placement of the methanol-inducible AOX1 promoter to determine if recombinant production of PG-1 in the yeast P. pastoris expression system would be enhanced. Although a widely applied and successful system, the use of AOX1 has various disadvantages that have been gaining greater concern in the production industry.

Using a GAP promoter instead of AOX1 relieves many of these disadvantages as it does not require continuous induction of flammable hazardous methanol as a carbon source, does not allow the accumulation of harmful by-products, and provides less strict cultivation controls during production. In the current study, we successfully generated and expressed recombinant ProPG-1 using the GAP promoter with a yield of

0.95 ± 0.004 g/ L after 48 hours of induction. This yield is comparable to the AOX1: ProPG-1 previously generated in our lab at 0.8 ± 0.2 g/ L (Huynh E., unpublished [11]). Although protein production was only slightly higher than AOX1-driven expression system, the constitutive expression using GAP: ProPG-1 allowed for faster production over the AOX1: ProPG-1, as high yields were obtained by 48 hours compared to 72 hours fermentation time, respectively. This is consistent with a previous study by Kittl et al., showing that the production of the fungal laccase using the GAP promoter was 517 mg/ L versus AOX1 at 495 mg/L, with sufficient product within 48 hours of fermentation using the GAP promoter [105]. The fact that the target proteins are produced in a shorter time, further demonstrates the beneficial use of the GAP promoter in the recombinant protein production industry [106]. Many successful recombinant proteins have already been expressed using the P. pastoris GAP expression system such as the human chitinase at 300 mg/L/day

[107], human serum albumin at 1.4 g/L [108], and an insulin precursor at 1.5 g/L [109]. However, high protein yields are only attained after cultivation conditions such as media components, pH, oxygen levels, and temperature regulation are optimized. Thus, further improvement to the bioreactor processes should be explored as it may provide increased protein concentrations that would be suitable for downstream therapeutic potential.

The murine-specific bacterium, C. rodentium, is commonly used to model enteric bacterial infections as it induces intestinal inflammation (colitis). A dosage of 2 x 109 CFU/mL was selected based on preliminary trials conducted and is consistent with former studies as this dosage has been standardized for this infection model [110, 111]. Also consistent with previous research, C57BL/6 mice orally challenged with C. rodentium displayed transient weight loss and diarrhea [80], which was measured daily using the

DAI criteria in Table 1. The in vivo antimicrobial role of PG-1 was first identified in the study by Steinberg et al., where immunocompetent mice were intraperitoneally (IP) inoculated with P. aeruginosa; a single IP injection of 0.5 mg/kg PG-1 decreased the mortality rate to 0-27%, compared to control mice with a 93-

100% mortality rate [60]. Additionally, transgenic mice that ectopically expressed PG-1 had enhanced resistance to challenge with A. suis. Eighty-seven percent of the PG-1 transgenic mice survived infection compared to only 37% of the wild-type mice [71]. Furthermore, the PG-1 transgenic mice had a reduced bacterial load in their lungs compared to the wild-type mice. The effects of PG-1 on digestive tract inflammation have been studied previously using a DSS-induced colitis mouse model. In that study, all three recombinantly produced protegrin domains (ProPG-1, cathelin, and mPG-1) partially prevented body weight loss and improved overall DAI scores compared to the challenged control group (Huynh E., unpublished [11]). Our study further extended this finding and revealed that oral administration of mPG-1;

1) decreased C. rodentium-induced body weight loss, 2) decreased diarrhea, and 3) reduced or eliminated

C. rodentium bacterial load. Taken together, these results demonstrate that protegrin treatments, especially mPG-1, help to reduce infection characteristics of colitis and are capable of clearing C. rodentium infection.

To our knowledge, this is the first study showing the protective effects of PG-1 on pathogen-induced intestinal colitis.

One of the histological hallmarks of C. rodentium infection is intestinal crypt hyperplasia, which can result from intestinal inflammation and damage, triggering epithelial cell proliferation that is visible as crypt elongation [84, 85]. Additional epithelial changes such as dysplasia and a depletion in mucus- secreting goblet cells were also reported in this colitis model [84, 86]. In our study, C. rodentium infection 51

increased overall histological scores compared to the unchallenged control mice, while mPG-1 reduced this histological change. A significant decrease in dysplasia by mPG-1 and mucosal neutrophils by mPG-1 and

ProPG-1 was demonstrated compared to challenged mice. This is consistent with our previous finding in the DSS mouse model of colitis, in which recombinant ProPG-1 and mPG-1 administration improved histomorphological changes with a decrease in mucosal epithelial damage and inflammation (Huynh E., unpublished [11]) Interestingly, in the same study it was demonstrated that ProPG-1 and mPG-1 increased goblet cell counts by 47% and 53% respectively, compared to DSS-challenged mice. However, no loss in goblet cell numbers was observed in the current study, regardless of infection or treatments. It appears that this cellular response to C. rodentium is not significant.

Infection with C. rodentium triggers a robust host inflammatory response including the expression of various cytokines, AMPs, and innate immune response genes [83]. Our RT-qPCR analysis results generally suggest that PG-1 modulated the pathogen-induced intestinal expression of these genes. The C- type lectin-like antimicrobial proteins, Reg3β and Reg3γ, are expressed by intestinal epithelial cells upon innate recognition of bacterial components through TLRs, like many AMPs, and play a role in bacterial killing [112]. A study in rodents found that high levels of Reg3 proteins in the ileal mucosa as well as feces during Salmonella infection were associated with an increase in the severity of infection [113]. The gene expression of both Reg3β and Reg3γ were significantly decreased in ProPG-1- and mPG-1-treated mice compared to infected mice, perhaps due to a decreased level of C. rodentium infection. Another epithelial

AMP [114], sPLA2, which penetrates bacterial cell walls through enzymatic degradation [115] and has proinflammatory properties, is known to be expressed in colonic epithelial and goblet cells [116]. During inflammation, sPLA2 is secreted into the intestinal lumen and promotes the synthesis of prostaglandins which aid in the recovery of damaged tissue [116]. Our finding that ProPG-1-treated mice had an increase in sPLA2 gene expression compared to infected mice suggests ProPG-1 may play a role in modulating inflammation during infection to enhance injured tissue repair. Together with the observation that C. rodentium is cleared in mPG-1-treated mice by Day 7, it is possible that the mPG-1-treated mice did not 52

display an increase in sPLA2 gene expression relative to challenged mice, as there may have been less tissue damage.

Muc1 and Muc2 are O-glycosylated proteins that create a mucus layer that coat the gastrointestinal tract to protect the epithelium from direct bacterial interactions. Muc1 is classified as a transmembrane or membrane-bound mucin that assists in cell migration, while the primary mucin component Muc2 is secreted by goblet cells and plays a major role in the organization of the protective intestinal mucus layers [103,

117]. In our study, mPG-1- and ProPG-1-treated mice showed increased Muc1 gene expression relative to challenged mice, suggesting enhanced protection from the infection. This correlates with a previous colitis study which demonstrated that an absence in Muc1 leads to intensification of chronic inflammation [118], as well as body weight loss [119]. These findings may also help to explain why protegrin administered mice had reduced body weight loss compared to challenged mice. Another major mucin, Muc2, has been shown to protect against lethal colitis infections [120]. It is known that Muc2 controls pathogen burden by binding to and removing bacteria that accumulate on the bacterial surface [120], and it has also been shown that C. rodentium directly binds to Muc2/mucus in vitro [121]. This is consistent with our observation that challenged mice showed an increase in Muc2 expression relative to unchallenged mice, suggesting the host uses an increase in mucus to help rid the mucosal surface of the pathogen. Interestingly, our data showed that C. rodentium-induced Muc2 expression was downregulated in both mPG-1- and ProPG-1-treated mice to the level comparable to the unchallenged mice, at this specific timepoint. It may be possible that a reduction of Muc2 expression by protegrin-1, likely via a negative-feedback loop triggered by excessive mucous production, is necessary to reduce mucus hypersecretion; a characteristic of chronic inflammatory diseases involving mucosal surfaces [122].

At a basic level, inflammation is the hosts response to infection and tissue injury, that involves the complex coordinated events of many diverse mediators and regulatory pathways [123]. Acute inflammation is considered beneficial in protecting the host from infection and restoring homeostasis, when effectively controlled [123]. However, dysregulated or prolonged inflammation can have detrimental effects such as septic shock [124]. Adhesion molecules such as VCAM-1 play a role in inflammation after their attachment to endothelial cells allows their infiltration into inflamed areas [125]. Once migrated to the site of infection, these molecules facilitate adhesion of numerous leukocytes such as lymphocytes, monocytes, and eosinophils to help regulate inflammation and stimulate tissue healing [126]. The mPG-1-treated mice showed an increase in VCAM-1 gene expression relative to challenged mice, which suggest it may play a role in stimulating or recruiting inflammatory cells for intestinal repair. Interestingly, ProPG-1-treated mice did not display an increase in the expression of VCAM-1. However, it has been found that differential expression may solely be part of the commencement of leucocyte migration [127], as by Day 10, the level of infection in mPG-1- and ProPG-1-treated mice appears to be different. Moreover, EGR1, a gene involved in cell proliferation [128], can be induced by stress signals such as injury [129]. EGR1 expression was suppressed in the challenged mice compared to unchallenged mice, suggesting that the normal hosts response to injury was altered by infection. The ProPG-1-treated mice showed an up-regulation in the expression of EGR1, relative to challenged mice, comparable to that of the unchallenged control. It is known that in response to bacterial invasion, colonic epithelial cells undergo controlled apoptosis to remove infected or injured cells and to allow the epithelial restoration process to occur [130]. Increased cell proliferation of intestinal lining epithelial cells helps to restore epithelial integrity that is altered during intestinal infection [130]. Therefore, up-regulation of EGR1 in the ProPG-1-treated group suggests it may be working towards epithelial restoration. Interestingly, mPG-1-treated mice did not display a stimulation of EGR1 expression, possibly due to mPG-1 clearing infection by this time point and thus there was less apoptosis and therefore less need for cell proliferative factors. Furthermore, the stress response gene HIF1α, is vital in maintaining intestinal homeostasis by increasing the expression of protective mucosal barrier

genes and initiating defensive innate immune responses [131]. The increased expression of HIF1α in protegrin treatment groups compared to challenged mice, further suggests PG-1 may play a role in enhancing host protection from bacterial infection, which is also reflective at the physiological level.

Apoptosis is a vital component in the regulation of immune responses during inflammation, if properly modulated. BAX, one major pro-apoptotic factor, can be induced by a variety of apoptotic stimuli

[132] and can be regulated directly or indirectly by many factors/mediators including the transcription factor nuclear factor κB (NF-κB) [133]. Our results showed a down-regulation of BAX in the ProPG-1- treated mice but not the mPG-1-treated mice. This suggests that ProPG-1 may play a role in pathways that regulate BAX, to reduce cell death and intestinal sloughing, while mPG-1-treated mice may no longer need to inhibit bacterial-induced apoptosis as they cleared infection by this time. On the other hand, the expression of pro-survival/anti-apoptotic Bcl-3 [134] was up-regulated in challenged mice compared to unchallenged mice, whereas mPG-1- and ProPG-1-treated mice reversed this increase to match that of the unchallenged treatment group. Increased expression in the challenged group may be a result of the host’s response to prevent excessive cell death caused by bacterial infection by promoting cell survival [134]. It is possible that the protegrin treatments may have already assisted in combating infection and thus remain at the level of the unchallenged mice. Consequently, a balance of apoptotic regulators seems critical for reestablishing intestinal homeostasis during inflammation.

TLR2 is up-regulated in bacterial infections as its agonists are microbial cell-wall components such as peptidoglycan and lipoteichoic acid, and its stimulation activates innate immune response cascades to eliminate pathogens [104]. Consistent with our results, challenged mice displayed an up-regulation of TLR2 gene expression while mPG-1- and ProPG-1-treated mice showed a down-regulation to match the unchallenged mice gene expression levels. This suggests that protegrin-1 may have the ability to suppress

C. rodentium-induced immune stimulation and counteract the overstimulation of cytokines which have been shown to induce tissue damage and intestinal barrier destruction [135]. On the other hand, TLR6 gene

expression was up-regulated in unchallenged compared to challenged mice and while mPG-1-treated mice had no effect on TLR6 gene expression, ProPG-1-treated mice seemed to have further suppressed the receptor relative to both challenged and unchallenged mice. These results are conflicting as TLR6 has been shown to be up-regulated in the colon during colitis and stimulates T-helper 1 and 17 (Th1/Th17) to influence the severity of disease [136]. However, the role of TLR6 on mucosal surfaces is still not well understood and no consensus has been established.

The regulation of cytokines is also an important part of the innate immune response to inflammation. Particularly in IBD, cytokines drive intestinal inflammation and associated symptoms, such as diarrhea, as they control multiple aspects of the inflammatory response (reviewed in [137]). SOCS3 is a negative regulator of cytokines that is induced by cytokine stimulation and works by creating a negative feedback loop by interfering with cytokine signaling and helps to regulate downstream signaling by other cytokines [138]. The upregulation in the gene expression level of SOCS3 in challenged and ProPG-1-treated mice, may be due to various cytokines being stimulated under inflammatory conditions, which in turn induces SOCS3 expression to help regulate cytokine levels. While the ProPG-1-treatment was less effective than the mPG-1-treatment, mPG-1-treated mice reduced SOCS3 gene expression level to the level of the unchallenged mice suggesting less inflammation.

Taken together, our results suggest that PG-1 can influence the expression of various AMPs, cytokines, and inflammatory mediators which help modulate inflammation, resolve intestinal infection, and work towards reestablishing homeostasis as demonstrated at the physiological, histological, and gene expression level. However, ProPG-1-treatment appeared to be less effective than the mPG-1-treatment, and this may come down to the AMPs direct bacterial killing ability which seems to be critical in controlling infection. This is apparent as mPG-1-treated mice cleared infection while ProPG-1-treated mice were still infected and reestablishing intestinal homeostasis. A potential reason why the ProPG-1-treatment may not have been as effective as mPG-1, may be the inefficient cleavage between the cathelin and the antimicrobial

mPG-1 domain via the enterokinase (EK) recognition site. The mPG-1 domain must be separated from the cathelin piece to initiate its microbicidal action [65]. One explanation could be a low cleavage efficiency, which may limit the release of the mPG-1 domain and thus its bacterial-killing capability. It is known that the EK protease does not exhibit stringent specificity for the DDDDK sequence, as it has been found that it can preferentially cleave at more accessible external off-target sites [139]. The addition of denaturants, such as urea, may be beneficial as it has been shown to enhance EK specificity, by allowing protein structures to “open” up more to allow easier access to the target site [139]. Another explanation for inefficient cleavage could be a high amount of protease inhibitors that are found to be upregulated in intestinal infections [140]. Natural endogenous protease inhibitors are normally expressed in the GI tract for digestion [140]. However, a dysregulation of proteolytic activity occurs during infection as infiltrating inflammatory cells also produce protease inhibitors, to counteract proteases released by bacteria, creating an excess of these molecules [140]. Despite this inefficient cleavage, the ProPG-1-treatment was still effective in some respects as ProPG-1 displayed similar gene expression modulation of Reg3β, Reg3γ,

Muc1, Muc2, Bcl3, and TLR2, to mPG-1. Although not statistically significant, ProPG-1 also followed similar trends to mPG-1 with a reduction in BW loss, DAI scores, and fecal water content. These effects in modulating infection could be due to the released mPG-1 or the tissue repair function of ProPG-1 as shown previously [11], or both in combination.

For PG-1 to have the effects presented here, it may function through affecting one primary pathway or it may regulate a number of factors involved in different pathways, to help regulate inflammation. The family of inducible transcription factors, NF-κB, serves as a complex, yet vital, mediator in the inflammatory process, as it can regulate numerous factors involved in innate and adaptive immune responses [141]. More specifically, NF-κB regulates the expression of various genes involved in cell differentiation, proliferation, apoptosis, inflammation (i.e. cytokines, chemokines, and adhesion molecules), and immunity (reviewed in [142]). Additionally, NF-κB activation leads to the recruitment and stimulation of effector cells, such as leukocytes, and is essential in the production of antimicrobial peptides 57

[143]. Interestingly, there is extensive evidence that multiple diverse factors can stimulate, either directly or indirectly, this important signaling pathway [141, 142]. In most cases, inflammatory stimuli such as the recognition of bacterial components by TLRs, inflammatory cytokines like TNF- or IL-1, and antigen receptors like T- or B-cell receptor (TCR/BCR), all activate NF-κB [142]. However, other physiological factors, such as oxidative stress, are also capable of inducing this transcription factor [142]. Within our gene expression data, it is apparent that many factors that had altered expression in the mPG-1-treatment group, can be regulated through NF-κB. Particularly, mPG-1 altered host antimicrobial peptides (Reg3β/γ and sPLA2), increased adhesion and cell migratory molecules (VCAM-1, Muc1), and regulated inflammatory/apoptotic factors (HIF1 , BAX, Bcl3); all of which can be downstream genes of NF-κB.

Thus, I propose a working model (Figure 17) that mPG-1 may regulate NF-κB, to facilitate the regulation of various mediators during intestinal inflammation. There is some evidence, with the well-studied human cathelicidin LL-37 AMP, that NF-κB can be stimulated by AMPs through downstream signaling of the insulin-like growth factor type 1 receptor (IGF-1R) pathway [144]. Specifically, Girnita et al., demonstrated that LL-37 binds to IGF-1R, activating extracellular signal-regulated kinase (ERK)[144], which in turn, once phosphorylated, can regulate NF-κB [145]. Recently in our lab, the molecular mechanism behind PG-

1’s action in intestinal cells has been investigated and one potential pathway has been speculated to be the same as LL-37 [146]. Importantly, Penney and Li, also demonstrated that mPG-1 (20 µg/mL) can upregulate the expression of NF-κB (p <0.01) [146]. Therefore, it is possible that mPG-1 may regulate NF-

κB to modulate the gene expression of intestinal mediators during inflammation. ProPG-1 could also function through NF-κB due to the presence of the mPG-1 domain, however the unique and currently uninvestigated cathelin domain of ProPG-1, may regulate additional mediators or pathways. Ultimately, further investigation into the specific mechanisms behind both mPG-1 and ProPG-1 are warranted.

C. rodentium is known to share similarities with other A/E human pathogens such as EPEC and

EHEC and is extensively used to replicate intestinal disorders such as Crohn’s, IBD, and ulcerative colitis

[80, 88, 101]. Using a C. rodentium challenge model, our study demonstrated the protective effects of PG- 58

1 on intestinal infection via its antimicrobial activity, immunomodulatory effects, and tissue repair capabilities. Its ability to stimulate the regulation of inflammation, enhance intestinal protection and repair, and ultimately clear infection suggests that this AMP could potentially serve a therapeutic use in combating colitis.

Figure 17. Proposed working model of mPG-1 regulating the gene expression of intestinal factors during inflammation through NF-κB. Many inflammatory stimuli such as bacterial component recognition via TLRs, and inflammatory cytokines, can activate NF-κB. This is the proposed model that mPG-1 may regulate the gene expression of various intestinal mediators during colitis such as antimicrobial peptides (Reg3β/γ and sPLA2), adhesion and cell migratory molecules (VCAM-1, Muc1), and inflammatory/apoptotic factors (HIF1 , BAX, Bcl3), through NF-κB.

Conclusion

This thesis met the research objectives of evaluating the utilization of a GAP promoter in the recombinant production of PG-1 in the yeast P. pastoris and investigating the protective, anti-inflammatory effects on intestinal health in a challenged mouse model. Specifically, the use of the GAP promoter demonstrated comparable yields to the commonly used AOX1 promoter, emphasizing its use as a safer alternative in the recombinant production industry. To our knowledge this is also the first study demonstrating the protective effects of PG-1 on pathogen-induced colitis. Overall, PG-1 reduced intestinal infection and histopathological changes in the colon caused by C. rodentium and resulted in altered expression of inflammatory mediators expressed during infection. Consequently, PG-1 exhibits great potential as an alternative therapeutic against colitis.

References

1. Angela Rocchi, Charles N Bernstein, Alain Bitton, Brian Feagan, Remo Panaccione, Kevin W Glasgow, Aida Fernandes, Subrata Ghosh, Inflammatory bowel disease: A Canadian burden of illness review. Can J Gastroentero, 2012. 26(11). 2. Xavier, R.J. and D.K. Podolsky, Unravelling the pathogenesis of inflammatory bowel disease. Nature, 2007. 448(7152): p. 427-34. 3. Loftus, E.V., Clinical epidemiology of inflammatory bowel disease: incidence, prevalence, and environmental influences. Gastroenterology, 2004. 126(6): p. 1504-1517. 4. Neurath, M.F. and S.P. Travis, Mucosal healing in inflammatory bowel diseases: a systematic review. Gut, 2012. 61(11): p. 1619-35. 5. Cunliffe, R.N. and Y.R. Mahida, Expression and regulation of antimicrobial peptides in the gastrointestinal tract. J Leukoc Biol, 2004. 75(1): p. 49-58. 6. Samantha Ho, and Hon Wai Koon, Antimicrobial Peptides and Colitis. Curr Pharm Des, 2013. 19: p. 40-47. 7. McLamb, B.L., et al., Early weaning stress in pigs impairs innate mucosal immune responses to enterotoxigenic E. coli challenge and exacerbates intestinal injury and clinical disease. PLoS One, 2013. 8(4): p. e59838. 8. Smith, F., et al., Early weaning stress impairs development of mucosal barrier function in the porcine intestine. Am J Physiol Gastrointest Liver Physiol, 2010. 298(3): p. G352-63. 9. Marshall, B.M. and S.B. Levy, Food animals and antimicrobials: impacts on human health. Clin Microbiol Rev, 2011. 24(4): p. 718-33. 10. Website, G.o.C., Federal action plan on antimicrobial resistance and use in Canada. 2015. Available from: http://healthycanadians.gc.ca/alt/pdf/publications/drugs-products- medicaments-produits/antibiotic-resistance-antibiotique/action-plan-daction-eng.pdf 11. Huynh, E., Recombinant Production of Growth Factors and Antimicrobial Peptides Using Microorganisms to Modulate Tissue Damage Healing. 2016. (Thesis) 12. Bahar, A.A. and D. Ren, Antimicrobial peptides. Pharmaceuticals, 2013. 6(12): p. 1543-75. 13. Hancock, R., and Sahl, Hans-Georg., Antimicrobial and host-defense peptides as new anti- infective therapeutic strategies. Nature Biotechnology, 2006. 24(12): p.1551-1557. 14. Aoki, W. and M. Ueda, Characterization of Antimicrobial Peptides toward the Development of Novel Antibiotics. Pharmaceuticals, 2013. 6(8): p. 1055-81. 15. Yuping L. and Gallo, R.L., AMPed up immunity: how antimicrobial peptides have multiple roles in immune defense. Cell, 2008. 30(3): p. 131-141. 16. Peters, B.M., M.E. Shirtliff, and M.A. Jabra-Rizk, Antimicrobial peptides: primeval molecules or future drugs? PLoS Pathog, 2010. 6(10): p. e1001067. 17. Christopher P. Hill, J.Y., Michael E. Selsted and David Eisenberg, Crystal Structure of Defensin HNP-3, an Amphiphilic Dimer: Mechanisms of Membrane Permeabilization. American Association for the Advancement of Science, 1991. 251: p. 1481-1485. 18. Yeaman, M.R. and N.Y. Yount, Mechanisms of antimicrobial peptide action and resistance. Pharmacol Rev, 2003. 55(1): p. 27-55. 19. Kim A. Brogden, Paul B. McCray, and Kenneth M. Huttner, Differences in the Concentrations of Small, Anionic, Antimicrobial Peptides in Bronchoalveolar Lavage Fluid and in Respiratory Epithelia of Patients with and without Cystic Fibrosis. Infection and Immunity, 1999. 67(8): p. 4256–4259. 20. Ming-Tao Lee, and Huey W. Huang, Energetics of Pore Formation Induced by Membrane Active Peptides. Biochemistry, 2004. 43: p. 3590-3599. 61

21. Vandamme, D., et al., A comprehensive summary of LL-37, the factotum human cathelicidin peptide. Cell Immunol, 2012. 280(1): p. 22-35. 22. Brogden, K.A., Antimicrobial peptides: pore formers or metabolic inhibitors in bacteria? Nat Rev Microbiol, 2005. 3(3): p. 238-50. 23. Ziv Oren, Y.S., Mode of Action of Linear Amphipathic Alpha-Helical Antimicrobial Peptides. Biopolymers (Peptide Science), 1998. 47: p. 451–463. 24. Braff, M.H., et al., Structure-Function Relationships among Human Cathelicidin Peptides: Dissociation of Antimicrobial Properties from Host Immunostimulatory Activities. The Journal of Immunology, 2005. 174(7): p. 4271-4278. 25. Carretero, M., et al., In vitro and in vivo wound healing-promoting activities of human cathelicidin LL-37. J Invest Dermatol, 2008. 128(1): p. 223-36. 26. Mary C. Territo, T.G., Michael E. Selsted, and Robert Lehrer, Monocyte-Chemotactic Activity of Defensins from Human Neutrophils. The Journal of Clinical Investigation, 1989. 84. p. 2017-2020. 27. Soruri, A., et al., Beta-Defensins chemoattract macrophages and mast cells but not lymphocytes and dendritic cells: CCR6 is not involved. Eur J Immunol, 2007. 37(9): p. 2474-86. 28. Franceois Niyonsaba, Hiroshi Matsuda, Hideoki Ogawa and Isao Nagaoka, Epithelial cell-derived human b-defensin-2 acts as a chemotaxin for mast cells through a pertussis toxin-sensitive and phospholipase C-dependent pathway. International Immunlogy, 2002. 14: p. 421-426. 29. Nicholas Funderburg, Zhimin Feng, Michael G. Drage, Julie Jadlowsky, Clifford V. Harding, Aaron Weinberg, and Scott F. Sieg, Human Beta-defensin-3 activates professional antigen-presenting cells via Toll-like receptors 1 and 2. PNAS 2007. 104(47): p.18631-18635.

30. Niyonsaba, F., et al., Antimicrobial peptides human beta-defensins stimulate epidermal keratinocyte migration, proliferation and production of proinflammatory cytokines and chemokines. J Invest Dermatol, 2007. 127(3): p. 594-604. 31. Niyonsaba, F., et al., The Human Beta -Defensins (-1, -2, -3, -4) and Cathelicidin LL-37 Induce IL- 18 Secretion through p38 and ERK MAPK Activation in Primary Human Keratinocytes. The Journal of Immunology, 2005. 175(3): p. 1776-1784. 32. Lester, S.N. and K. Li, Toll-like receptors in antiviral innate immunity. J Mol Biol, 2014. 426(6): p. 1246-64. 33. Mogensen, T.H., Pathogen recognition and inflammatory signaling in innate immune defenses. Clin Microbiol Rev, 2009. 22(2): p. 240-73. 34. Botos, I., D.M. Segal, and D.R. Davies, The structural biology of Toll-like receptors. Structure, 2011. 19(4): p. 447-59. 35. Di Nardo, A., et al., Cathelicidin Antimicrobial Peptides Block Dendritic Cell TLR4 Activation and Allergic Contact Sensitization. The Journal of Immunology, 2007. 178(3): p. 1829-1834. 36. Kathy Triantafilou, M.T., and Russell L. Dedrick, Interactions of Bacterial Lipopolysaccharide and Peptidoglycan with a 70 kDa and an 80 kDa Protein on the Cell Surface of CD14ⴙ and CD14ⴚ Cells.Human Immunology, 2001. 62: p. 50-63. 37. Nakagawa, Y. and R.L. Gallo, Endogenous intracellular cathelicidin enhances TLR9 activation in dendritic cells and macrophages. J Immunol, 2015. 194(3): p. 1274-84. 38. Li, J., et al., PR39, a peptide regulator of angiogenesis. Nature Medicine, 2000. 6(1): p. 49-55. 39. Baroni, A., et al., Antimicrobial human beta-defensin-2 stimulates migration, proliferation and tube formation of human umbilical vein endothelial cells. Peptides, 2009. 30(2): p. 267-72. 40. Buchau, A.S., et al., The host defense peptide cathelicidin is required for NK cell-mediated suppression of tumor growth. J Immunol, 2010. 184(1): p. 369-78.

41. Qiao, K., et al., A new antimicrobial peptide SCY2 identified in Scylla Paramamosain exerting a potential role of reproductive immunity. Fish Shellfish Immunol, 2016. 51: p. 251-262. 42. Arijs, I., et al., Mucosal gene expression of antimicrobial peptides in inflammatory bowel disease before and after first infliximab treatment. PLoS One, 2009. 4(11): p. e7984. 43. Sidhu, R., et al., Faecal lactoferrin--a novel test to differentiate between the irritable and inflamed bowel? Aliment Pharmacol Ther, 2010. 31(12): p. 1365-70. 44. J Wehkamp, M.S., EF Stange, Defensins and other antimicrobial peptides in inflammatory bowel disease. Curr Opin Gastroenterol, 2007. 23(4): p. 370-380. 45. Simms, L.A., et al., Reduced alpha-defensin expression is associated with inflammation and not NOD2 mutation status in ileal Crohn's disease. Gut, 2008. 57(7): p. 903-10. 46. Koon, H.W., et al., Cathelicidin signaling via the Toll-like receptor protects against colitis in mice. Gastroenterology, 2011. 141(5): p. 1852-63 e1-3. 47. Schmitt, P., R.D. Rosa, and D. Destoumieux-Garzon, An intimate link between antimicrobial peptide sequence diversity and binding to essential components of bacterial membranes. Biochim Biophys Acta, 2016. 1858(5): p. 958-70. 48. Fabiano Pinheiro da Silva, and Marcel C. C. Machado, Antimicrobial peptides: Clinical relevance and therapeutic implications. Peptides, 2012. 36: p. 308-314. 49. Enno Kluver, Knut Adermann, and Axel Schulz, Synthesis and structure–activity relationship of β- defensins, multi-functional peptides of the immune system. Journal of Peptide Science, 2005. 12: p. 243-257. 50. Louise A. Duits, Mirjam Rademaker, Pieter S. Hiemstra & Peter H. Nibbering, Expression of b- defensin 1 and 2 mRNA by human monocytes, macrophages and dendritic cells. Immunology, 2002. 106: p. 517-525. 51. Wong, J.H., L. Xia, and T.B. Ng, A review of defensins of diverse origins. Curr Protein Pept Sci, 2007. 8(5): p. 446-59. 52. Selsted, M.E. and A.J. Ouellette, Mammalian defensins in the antimicrobial immune response. Nat Immunol, 2005. 6(6): p. 551-7. 53. Yi-Quan Tang, J.Y., George Ösapay, Klara Ösapay, Dat Tran, Christopher J. Miller, Andre J. Ouellette and Michael E. Selsted, A Cyclic Antimicrobial Peptide Produced in Primate Leukocytes by the Ligation of Two Truncated α-Defensins. Science, 1999. 286: p. 498-502. 54. Kosciuczuk, E.M., et al., Cathelicidins: family of antimicrobial peptides. A review. Mol Biol Rep, 2012. 39(12): p. 10957-70. 55. Tomasinsig, L., and Zanetti, M. The Cathelicidins – Structure, Function and Evolution. Current Protein and Peptide Science, 2005. 6: p.23-34. 56. Jean Frederic Sanchez, F.W., Yin-Shan Yang, Marie-Paule Strub, Jean Marc Strub, and M.M. Alain Van Dorsselaer, Robert Lehrer, Tomas Ganz, Alain Chavanieu, Bernard Calas, and Andre Aumelas, Overexpression and Structural Study of the Cathelicidin Motif of the Protegrin-3 Precursor. Biochemistry, 2002. 41: p. 21-30. 57. Zaiou, M., V. Nizet, and R.L. Gallo, Antimicrobial and protease inhibitory functions of the human cathelicidin (hCAP18/LL-37) prosequence. J Invest Dermatol, 2003. 120(5): p. 810-6. 58. Hua Wu, G.Z., Christopher R. Ross, and Frank Blecha, Cathelicidin Gene Expression in Porcine Tissues: Roles in Ontogeny and Tissue Specificity. Infection and Immunity, 1998: p. 439–442. 59. R. Bals and J. M. Wilsonb, Cathelicidins – a family of multifunctional antimicrobial peptides. CMLS Cellular and Molecular Life Sciences, 2002. 60: p. 711-720. 60. Steinberg, D., et al., Protegrin-1: a Broad-Spectrum, Rapidly Microbicidal Peptide with In Vivo Activity. Antimicrobial Agents and Chemotherapy, 1997. 41: p. 1738-1742.

61. Chengquan Zhao, T.G., Robert I. Lehre, The structure of porcine protegrin genes. FEBS Letters, 1995. 368. p. 197-202. 62. Takanori Nakamura, H.F., Toshiyuki Miyata, Fuminori Tokunaga, Tatsushi Muta, and Sadaaki Iwanag, Tachyplesin, A Class of Antimicrobial Peptide from the Hemocytes of the Horseshoe Crab (Tach ypleus tridentatus. The Journal of Biological Chemistry, 1988. 263(32): p. 16709-16713. 63. Sylvia S. L. Harwig, Hye Jin Yang, Yoon Cho, Leonie Tan Robert I. Lehre, Intramolecular disulfide bonds enhance the antimicrobial and lytic activities of protegrins at physiological sodium chloride concentrations. Eur. J. Biochem, 1996. 240: p. 352-357. 64. Benincasa, M., et al., Fungicidal activity of five cathelicidin peptides against clinically isolated yeasts. J Antimicrob Chemother, 2006. 58(5): p. 950-959. 65. Alexander M. Cole, J.S., Alejandro Ceccarelli, Yong-Hwan Kim, Albert Park, and Tomas Ganz, Inhibition of neutrophil elastase prevents cathelicidin activation and impairs clearance ofbacteria from wounds. Blood 2001. 97(1): p. 297-304.

66. Lin Yang, T.A.H., Thomas M. Weiss, Lai Ding, and Huey W. Huang, Barrel-Stave Model or Toroidal Model? A Case Study on Melittin Pores. Biophysical Journal, 2001. 81: p. 1475-1485. 67. Kin Lok H. Lam, Yuji Ishitsuka, Yishan Cheng, Karen Chien, Alan J. Waring, Robert I. Lehrer, and Ka Yee C. Lee, Mechanism of Supported Membrane Disruption by Antimicrobial Peptide Protegrin-1. J. Phys. Chem. B, 2006. 110: p. 21282-21286. 68. Jang, H., et al., Antimicrobial protegrin-1 forms amyloid-like fibrils with rapid kinetics suggesting a functional link. Biophys J, 2011. 100(7): p. 1775-83. 69. Bolintineanu, D., et al., Antimicrobial mechanism of pore-forming protegrin peptides: 100 pores to kill E. coli. Peptides, 2010. 31(1): p. 1-8. 70. Jang, H., B. Ma, and R. Nussinov, Conformational study of the protegrin-1 (PG-1) dimer interaction with lipid bilayers and its effect. BMC Struct Biol, 2007. 7(21): p. 1-15. 71. Cheung, Q.C., et al., Enhanced resistance to bacterial infection in protegrin-1 transgenic mice. Antimicrob Agents Chemother, 2008. 52(5): p. 1812-1818. 72. Anikó Várnai, C.T., Oskar Bengtsson, Andrew Atterton, Geir Mathiesen and Vincent GH Eijsink, Expression of endoglucanases in Pichia pastoris under control of the GAP promoter. Microbial Cell Factories, 2014. 13(57): p. 1-10. 73. Clement, H., et al., A comparison between the recombinant expression and chemical synthesis of a short cysteine-rich insecticidal spider peptide. J Venom Anim Toxins Incl Trop Dis, 2015. 21(9): p. 1-10. 74. Demain, A.L. and P. Vaishnav, Production of recombinant proteins by microbes and higher organisms. Biotechnol Adv, 2009. 27(3): p. 297-306. 75. Çalık, P., et al., Recombinant protein production in Pichia pastoris under glyceraldehyde-3- phosphate dehydrogenase promoter: From carbon source metabolism to bioreactor operation parameters. Biochemical Engineering Journal, 2015. 95: p. 20-36. 76. Kim, S., et al., Regulation of alcohol oxidase 1 (AOX1) promoter and peroxisome biogenesis in different fermentation processes in Pichia pastoris. J Biotechnol, 2013. 166(4): p. 174-81. 77. Krainer, F.W., et al., Recombinant protein expression in Pichia pastoris strains with an engineered methanol utilization pathway. Microb Cell Fact, 2012. 11(22): p. 1-14. 78. Macauley-Patrick, S., et al., Heterologous protein production using the Pichia pastoris expression system. Yeast, 2005. 22(4): p. 249-70. 79. Zhao, W., et al., Scale-up fermentation of recombinant Candida rugosa lipase expressed in Pichia pastoris using the GAP promoter. J Ind Microbiol Biotechnol, 2008. 35(3): p. 189-95.

80. Steven A. Luperchio, Don J. Brenner, Charles A. Dangler, Arnold G. Steigerwalt, David B. Schauer and Mark D. Schrenzel, Citrobacter rodentium, the Causative Agent of Transmissible Murine Colonic Hyperplasia, Exhibits Clonality: Synonymy of C. rodentium and Mouse-Pathogenic Escherichia coli. Journal of Clinical Microbiology, 2000. 38(12): p. 4343-4350. 81. Schauer, D., and Falkow, S. The eae Gene of Citrobacterfreundii Biotype 4280 Is Necessary for Colonization in Transmissible Murine Colonic Hyperplasia. Infection and Immunity, 1993. 61(11): p. 4654-4661. 82. Wong, A.R., et al., Enteropathogenic and enterohaemorrhagic Escherichia coli: even more subversive elements. Mol Microbiol, 2011. 80(6): p. 1420-38. 83. Crepin, V.F., et al., Citrobacter rodentium mouse model of bacterial infection. Nat Protoc, 2016. 11(10): p. 1851-76. 84. Maaser, C., et al., Clearance of Citrobacter rodentium requires B cells but not secretory immunoglobulin A (IgA) or IgM antibodies. Infect Immun, 2004. 72(6): p. 3315-24. 85. Alipour, M., et al., A balanced IL-1beta activity is required for host response to Citrobacter rodentium infection. PLoS One, 2013. 8(12): p. e80656. 86. Borenshtein, D., et al., Development of fatal colitis in FVB mice infected with Citrobacter rodentium. Infect Immun, 2007. 75(7): p. 3271-81. 87. Diana Borenshtein, David B. Schauer, Utility of the Citrobacter rodentium infection model in laboratory mice. Current Opinion in Gastroenterology, 2008. 24(1): p. 32-37. 88. Lisa M. Higgins, Gill Douce, Citrobacter rodentium Infection in Mice Elicits a Mucosal Th1 Cytokine Response and Lesions Similar to Those in Murine Inflammatory Bowel Disease. Infection and Immunity, 1999. 67(6): p. 3031-3039. 89. Chandrakesan, P., et al., Utility of a bacterial infection model to study epithelial-mesenchymal transition, mesenchymal-epithelial transition or tumorigenesis. Oncogene, 2014. 33(20): p. 2639- 54. 90. C. T. Chung, and Roger H. Miller, One-step preparation of competent Escherichia coli: Transformation and storage of bacterial cells in the same solution. Proc. Natl. Acad. Sci. , 1989. 86: p. 2172-2175,. 91. Joan Lin-Cereghino, See Xiong, William Giang, Linda T. Luong, Jane Vu, Sabrina D. Johnson, and Geoff P. Lin-Cereghino, Condensed protocol for competent cell preparation and transformation of the methylotrophic yeast Pichia pastoris. Biotechniques, 2005. 38: p. 44–48. 92. Reid-Yu, S.A., et al., CXCL9 contributes to antimicrobial protection of the gut during citrobacter rodentium infection independent of chemokine-receptor signaling. PLoS Pathog, 2015. 11(2): p. e1004648. 93. Bhinder, G., et al., The Citrobacter rodentium mouse model: studying pathogen and host contributions to infectious colitis. J Vis Exp, 2013(72): p. e50222. 94. Schauer, D., and Falkow, S. Attaching and Effacing Locus of a Citrobacterfreundii Biotype That Causes Transmissible Murine Colonic Hyperplasia. Infection and Immunity, 1993.61(6): p. 2486- 2492. 95. Spehlmann, M.E., et al., CXCR2-dependent mucosal neutrophil influx protects against colitis- associated diarrhea caused by an attaching/effacing lesion-forming bacterial pathogen. J Immunol, 2009. 183(5): p. 3332-43. 96. Maxwell, J.R., et al., Methods of inducing inflammatory bowel disease in mice. Curr Protoc Pharmacol, 2009. 47(5.58). 97. Ulrike Erben, Christoph Loddenkemper, Katja Doerfel, Simone Spieckermann, Dirk Haller, M Heimesaat, Martin Zeitz and Anja A Kühl, A guide to histomorphological evaluation of intestinal inflammation in mouse models. Int J Clin Exp Patho, 2014. 7(8): p. 4557-4576. 65

98. Buschor, S., et al., Innate immunity restricts Citrobacter rodentium A/E pathogenesis initiation to an early window of opportunity. PLoS Pathog, 2017. 13(6): p. e1006476. 99. MacDonald, T.T., et al., Host defences to Citrobacter rodentium. Int J Med Microbiol, 2003. 293(1): p. 87-93. 100. Kathene C. Johnson-Henry, M.N., Yaron Avitzur, David J. Mitchell, Bo-Yee Ngan, Esther Galindo- Mata, Nicola L. Jones, and Philip M. Sherman, Amelioration of the Effects of Citrobacter rodentium Infection in Mice by Pretreatment with Probiotics. The Journal of Infectious Diseases, 2005. 191: p. 2106-2117. 101. Collins, J.W., et al., Citrobacter rodentium: infection, inflammation and the microbiota. Nat Rev Microbiol, 2014. 12(9): p. 612-23. 102. Heimlich, D.R., A. Harrison, and K.M. Mason, Host Antimicrobial Peptides in Bacterial Homeostasis and Pathogenesis of Disease. Antibiotics (Basel), 2014. 3(4): p. 645-76. 103. Johansson, M.E., J.M. Larsson, and G.C. Hansson, The two mucus layers of colon are organized by the MUC2 mucin, whereas the outer layer is a legislator of host-microbial interactions. Proc Natl Acad Sci U S A, 2011. 108(1): p. 4659-65. 104. Kawai, T. and S. Akira, The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat Immunol, 2010. 11(5): p. 373-84. 105. Kittl, R., et al., Constitutive expression of Botrytis aclada laccase in Pichia pastoris. Bioengineered, 2012. 3(4): p. 232-5. 106. Qin, X., et al., GAP promoter library for fine-tuning of gene expression in Pichia pastoris. Appl Environ Microbiol, 2011. 77(11): p. 3600-8. 107. J. C. Goodrick, M.X., R. Finnegan, B. M. Schilling, S. Schiavi, H. Hoppe, N. C. Wan, High-Level Expression and Stabilization of Recombinant Human Chitinase Produced in a Continuous Constitutive Pichia pastoris Expression System. Biotechnology and Bioengineering, 2001. 74: p. 492-497. 108. Kaoru Kobayashi, S.K., Tomoshi Ohya, Toyoo Ohda, Masao Ohyama, Hideyuki Ohi, Kenji Tomomitsu, Takao Ohmura, High-level expression of recombinant human serum albumin from the methylotrophic yeast pichia pastoris with minimal protease production and activation. Journal of Bioscience and Bioengineering, 2000. 89: p. 55-61. 109. Yan Wang, Z.-H.L., You-Shang Zhang, Shi-Yin Yao, Ying-Gao Xu, Yue-Hua Tang, Shang-Quan Zhu, Da-Fu Cui, You-Min Feng, Human Insulin from a Precursor Overexpressed in the Methylotrophic Yeast Pichia pastoris and a Simple Procedure for Purifying the Expression Product. Biotechnology and Bioengineering, 2001. 73: p. 74-79. 110. Lopez-Saucedo, C., et al., CD40 Ligand Deficient C57BL/6 Mouse Is a Potential Surrogate Model of Human X-Linked Hyper IgM (X-HIGM) Syndrome for Characterizing Immune Responses against Pathogens. Biomed Res Int, 2015. 2015: p. 679850. 111. Symonds, E.L., et al., Involvement of T helper type 17 and regulatory T cell activity in Citrobacter rodentium invasion and inflammatory damage. Clin Exp Immunol, 2009. 157(1): p. 148-54. 112. van Ampting, M.T., et al., Intestinally secreted C-type lectin Reg3b attenuates salmonellosis but not listeriosis in mice. Infect Immun, 2012. 80(3): p. 1115-20. 113. van Ampting, M.T., et al., Ileal mucosal and fecal pancreatitis associated protein levels reflect severity of salmonella infection in rats. Dig Dis Sci, 2009. 54(12): p. 2588-97. 114. Mukherjee, S. and L.V. Hooper, Antimicrobial defense of the intestine. Immunity, 2015. 42(1): p. 28-39. 115. Koprivnjak, T. and A. Peschel, Bacterial resistance mechanisms against host defense peptides. Cell Mol Life Sci, 2011. 68(13): p. 2243-54.

116. Schewe, M., et al., Secreted Phospholipases A2 Are Intestinal Stem Cell Niche Factors with Distinct Roles in Homeostasis, Inflammation, and Cancer. Cell Stem Cell, 2016. 19(1): p. 38-51. 117. Hill, J.P.-V.a.R.L., The Structure and Assembly of Secreted Mucins. The Journal of Biological Chemistry, 1999. 274(45): p. 31751–31754. 118. Nishida, A., et al., The membrane-bound mucin Muc1 regulates T helper 17-cell responses and colitis in mice. Gastroenterology, 2012. 142(4): p. 865-874 e2. 119. Kosuke Kato, Wenju Lu, Kwang Chul Kim, Muc1: The first respiratory mucin with an anti- inflammatory function. Journal of Clinical Medicine, 2017. 6(12). 120. Bergstrom, K.S., et al., Muc2 protects against lethal infectious colitis by disassociating pathogenic and commensal bacteria from the colonic mucosa. PLoS Pathog, 2010. 6(5): p. e1000902. 121. Linden, S.K., T.H. Florin, and M.A. McGuckin, Mucin dynamics in intestinal bacterial infection. PLoS One, 2008. 3(12): p. e3952. 122. Hauber, H.-P., S.C. Foley, and Q. Hamid, Mucin overproduction in chronic inflammatory lung disease. Canadian Respiratory Journal, 2006. 13(6): p. 327-335. 123. Medzhitov, R., Origin and physiological roles of inflammation. Nature, 2008. 454(7203): p. 428- 35. 124. Polat, G., et al., Sepsis and Septic Shock: Current Treatment Strategies and New Approaches. Eurasian J Med, 2017. 49(1): p. 53-58. 125. Trzeciak-Jedrzejczyk, A., et al., The role of adhesion molecules in inflammatory bowel disease in children. Assessment of the possible risk of cardiovascular complications. Prz Gastroenterol, 2017. 12(3): p. 181-185. 126. Miquel Sans, Esther Ardite, J. Ignasi Elizalde, Yolanda Arce, Montserrat Elena, Antonio Palacin, J. Carlos Fernandez–Checa, Donald C. Anderson, Roy Lobb, and Josep M. Pique, VCAM-1 and ICAM-1 Mediate Leukocyte-Endothelial Cell Adhesion in Rat Experimental Colitis. Gastroenterology, 1999. 116: p. 874–883. 127. S C Jones, R.E.B., A Haidar, AJ H Gearing, I K Hemingway, S H Ibbotson, M F Dixon, AT Axon, Adhesion molecules in inflammatory bowel disease. Gut, 1995. 36: p. 724-730. 128. Liu, C., et al., The transcription factor EGR-1 directly transactivates the fibronectin gene and enhances attachment of human glioblastoma cell line U251. J Biol Chem, 2000. 275(27): p. 20315-23. 129. Min, I.M., et al., The transcription factor EGR1 controls both the proliferation and localization of hematopoietic stem cells. Cell Stem Cell, 2008. 2(4): p. 380-91. 130. Kim, J.M., et al., Apoptosis of human intestinal epithelial cells after bacterial invasion. J Clin Invest, 1998. 102(10): p. 1815-23. 131. Shah, Y.M., The role of hypoxia in intestinal inflammation. Mol Cell Pediatr, 2016. 3(1): p. 1. 132. Garner, T.P., et al., An Autoinhibited Dimeric Form of BAX Regulates the BAX Activation Pathway. Mol Cell, 2016. 63(3): p. 485-97. 133. Bentires-Alj, M., et al., Inhibition of the NF-kB transcription factor increases Bax expression in cancer cell lines. Oncogene, 2001. 20: p. 2805 - 2813. 134. Poveda, J., et al., Bcl3: a regulator of NF-kappaB inducible by TWEAK in acute kidney injury with anti-inflammatory and antiapoptotic properties in tubular cells. Exp Mol Med, 2017. 49(7): p. e352. 135. Fan, Y. and B. Liu, Expression of Toll-like receptors in the mucosa of patients with ulcerative colitis. Exp Ther Med, 2015. 9(4): p. 1455-1459.

136. Morgan, M.E., et al., Toll-like receptor 6 stimulation promotes T-helper 1 and 17 responses in gastrointestinal-associated lymphoid tissue and modulates murine experimental colitis. Mucosal Immunol, 2014. 7(5): p. 1266-77. 137. Neurath, M.F., Cytokines in inflammatory bowel disease. Nat Rev Immunol, 2014. 14(5): p. 329- 42. 138. Ben A. Croker, H.K., and Sandra E. Nicholson, SOCS Regulation of the JAK/STAT Signalling Pathway. Semin Cell Dev Biol., 2008. 19: p. 414–422. 139. Shahravan, S.H., et al., Enhancing the specificity of the enterokinase cleavage reaction to promote efficient cleavage of a fusion tag. Protein Expr Purif, 2008. 59(2): p. 314-9. 140. Vergnolle, N., Protease inhibition as new therapeutic strategy for GI diseases. Gut, 2016. 65(7): p. 1215-24. 141. Liu, T., et al., NF-kappaB signaling in inflammation. Signal Transduct Target Ther, 2017. 2. 142. Oeckinghaus, A. and S. Ghosh, The NF-kappaB family of transcription factors and its regulation. Cells, 2009. 1(4): p. a000034. 143. Ghosh, S. and M.S. Hayden, New regulators of NF-kappaB in inflammation. Nat Rev Immunol, 2008. 8(11): p. 837-48. 144. Girnita, A., et al., Identification of the cathelicidin peptide LL-37 as agonist for the type I insulin- like growth factor receptor. Oncogene, 2012. 31(3): p. 352-65. 145. Jiang, B., et al., Temporal control of NF-kappaB activation by ERK differentially regulates interleukin-1beta-induced gene expression. J Biol Chem, 2004. 279(2): p. 1323-9. 146. Penney, J., Li, J., Protegrin-1: Investigating beyond antimicrobial activity. Frontiers in Cellular and Infection Microbiology, 2018. (In review).

Appendix

Table A1. Optimal bioreactor media composition for recombinant PG-1 expression in Pichia pastoris (Huynh E., unpublished [11])

Fermentation Basal Salt Medium (BSM) (1 L) Pichia Trace Metal Salts Medium (PTM1) (1 L)

Calcium sulfate dihydrate 0.47 g Cupric sulfate pentahydrate 6.0 g

Potassium sulfate 9.1 g Sodium iodide 0.08 g

Potassium Hydroxide 2.07 g Manganese sulfate hydrate 3.0 g

Magnesium sulfate heptahydrate 7.45 g Sodium molybdate dihydrate 0.2 g

Ethylenediaminetetraaceticacid (EDTA) 0.6 g Boric acid 0.02 g

Phosphoric Acid (85%) 13.4 mL Cobalt chloride 0.5 g

Glycerol 40.0 g Zinc chloride 20.0 g

Sodium Chloride 0.22 g Ferrous sulfate heptahydrate 65.0 g

PTMI (Trace Elements) 4.35 mL Biotin 0.2 g

Sulfuric acid 5.0 mL

Table A2. Primer sequences used for RT-qPCR

GENE FORWARD (5’ - 3’) REVERSE (5’ - 3’) EFFICIENCY (%)

β-Actin CTTCTTTGCAGCTCCTTCGTT TTCTGACCCATTCCCACCA 96.2

GAPDH CCTCGTCCCGTAGACAAAATG TGAAGGGGTCGTTGATGGC 92.7

Muc1 GGCATTCGGGCTCCTTTCTT TGGAGTGGTAGTCGATGCTAAG 103.1

Muc2 AGGGCTCGGAACTCCAGAAA CCAGGGAATCGGTAGACATCG 105.5

Reg3β ACTCCCTGAAGAATATACCCTCC CGCTATTGAGCACAGATACGAG 102.1

Reg3γ ATGCTTCCCCGTATAACCATCA GGCCATATCTGCATCATACCAG 98.2

sPLA2 TGGCTCAATACAGGACCAAGG GTGGCATCCATAGAAGGCATAG 99.7

BAX TGAAGACAGGGGCCTTTTTG AATTCGCCGGAGACACTCG 101.2

BCL3 CCGGAGGCCCTTTACTACCA GGAGTAGGGGTGAGTAGGCAG 104.4

EGR1 AGACGAGTTATCCCAGCCAAA GGTCGGAGGATTGGTCATGC 105.7

TLR6 TGAGCCAAGACAGAAAACCCA GGGACATGAGTAAGGTTCCTGTT 100.9

TLR2 GCAAACGCTGTTCTGCTCAG AGGCGTCTCCCTCTATTGTATT 102.0

SOCS3 ATGGTCACCCACAGCAAGTTT TCCAGTAGAATCCGCTCTCCT 105.2

VCAM-1 AGTTGGGGATTCGGTTGTTCT CCCCTCATTCCTTACCACCC 96.5

HIF1α ACCTTCATCGGAAACTCCAAAG ACTGTTAGGCTCAGGTGAACT 106.5

Camp GCTGTGGCGGTCACTATCAC TGTCTAGGGACTGCTGGTTGA 102.0

ANG4 GGTTGTGATTCCTCCAACTCTG CTGAAGTTTTCTCCATAAGGGCT 102.1

IRF7 ATGCCAATCACTCGAATGAG TTGTATCGGCCTGTGTGAATG 96.8

CASP1 CTTGGAGACATCCTGTCAGGG AGTCACAAGACCAGGCATATTCT 106.3

Figure A1. Pilot animal study #1 overview & main results. The purpose of this study was to identify the appropriate dosage of C. rodentium that can effectively cause colitis. Visual representation of the (A) trial timeline showing induction of colitis, (B) treatment groups, and (C) number of mice used per treatment. Infection with 2 x 108 and 2 x 109 CFU/mL of C. rodentium induced significant weight loss compared to negative control mice (D) and moderately increased disease activity index (E). Higher DAI scores represent more severe disease signs. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one-way repeated measures ANOVA post hoc Tukey test over the course of the trial. RESULT: Significantly severe signs were not seen within this time frame. The highest dosage of 2 x 109 CFU/mL was selected for the next pilot study.

Figure A2. Pilot animal study #2 overview; body weight and DAI score results. The purpose of this study was to identify the appropriate dosage and infection time points of C. rodentium that effectively cause colitis. Visual representation of the (A) trial timeline showing induction of colitis, (B) treatment groups, and (C) number of mice used per treatment. Infections with C. rodentium significantly induced body weight loss (D) and increased disease activity index (E). Higher DAI score represents more severe disease signs. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one- way repeated measures ANOVA post hoc Tukey test. RESULT: 2 x 109 CFU/mL B (Gavage Day 1 & 3) had an overall greater and more significant effect on body weight loss and DAI scores and was used for the animal study.

A 1 0 .0 B

W a te r C o n s u m p tio n

7 .5 1 5 2 x 1 0 ^ 9 A

L 2 x 1 0 ^ 9 A * *

m * *

r )

/ 2 x 1 0 ^ 9 B 2 x 1 0 ^ 9 B

l a F 1 x P B S

m 1 0 (

C 5 .0

u s

o 5

2 .5 n

C A

0 .0 0 0 1 2 3 4 5 6 7 8 9 1 0 1 1 2 x 1 0 ^ 9 A 2 x 1 0 ^ 9 B 1 x P B S D a y

C F e c a l W a te r C o n te n t D a y 1 0 P I 5 2 x 1 0 ^ 9 A

4 2 x 1 0 ^ 9 B

t )

n 1 x P B S

t %

( 3

r c

r 2

a P

W 1

0 2 x 1 0 ^ 9 A 2 x 1 0 ^ 9 B 1 x P B S

Figure A3. Pilot animal study #2 results. (A) Bacterial load of C. rodentium in feces was measured on Day 0, 3, 7, and 10. Fecal samples were collected, resuspended in 1 x PBS and serially diluted onto

MacConkey agar plates. Colony counts are expressed as Log10 CFU/mL. (B) Daily average water consumption per cage was calculated and is presented as mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one-way ANOVA post hoc Tukey test. (C) Fecal samples from individual cages were collected shortly before euthanization, weighed, and incubated for 24 hours at 60°C, and re-weighed. The percent of fecal water content on Day 10 post-infection (PI) was calculated by dividing the difference of the wet and dry weight by the wet weight. RESULT: 2 x 109 CFU/mL A (Gavage Day 1 & 5) and 2 x 109 CFU/mL B (Gavage Day 1 & 3) had similar effects on C. rodentium numbers in feces, daily water consumption, and fecal water content; 2 x 109 CFU/mL B (Gavage Day 1 & 3) was used for the animal study.

Figure A4. Representative hematoxylin and eosin stained colon sections illustrate colonic crypt hyperplasia. Histological evaluation of colonic sections on Day 10 demonstrate noticeable colonic crypt hyperplasia in C. rodentium challenged mice compared to unchallenged mice. Hyperplasia appears as elongated crypts (black double-sided arrows). Both ProPG-1- and mPG-1-treated challenged mice resulted in a decrease in the hyperplastic response to infection compared to challenged mice, but mPG-1- treated challenged mice prevented more of the C. rodentium-induced colonic crypt hyperplasia although no statistical differences were seen among the treatment groups. Scale bar: 100 µm.

A B

L a m in a P ro p ria W id th T u n ic a M u s c u la ris W id th )

5 0 3 0 0

)

(

t 4 0

t i

d 2 0 0

i W

3 0

r o

2 0 a r

l 1 0 0

a s

n 1 0

a L U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1 U n c h a lle n g e d C h a lle n g e d m P G -1 P ro P G -1

C h a lle n g e d C h a lle n g e d

Figure A5. Effects of PG-1 on lamina propria width and tunica muscularis width of the colon. Protegrin treatments did not have any significant effects on lamina propria width (A) or on tunica muscularis width (B) of the colon. Data represent the mean ± SEM; *p <0.05, **p <0.01, ***p <0.001 calculated by a one-way ANOVA post hoc Tukey test.