From Antimicrobial Activity to Zinc Binding: an In-Depth Analysis of the Tunicate Host Defense Peptide Clavanin A

University of Connecticut OpenCommons@UConn

Doctoral Dissertations University of Connecticut Graduate School

2-3-2020

From Antimicrobial Activity to Zinc Binding: An In-Depth Analysis of the Tunicate Host Defense Peptide Clavanin A

Samuel A. Juliano IV University of Connecticut - Storrs, [email protected]

Follow this and additional works at: https://opencommons.uconn.edu/dissertations

Recommended Citation Juliano, Samuel A. IV, "From Antimicrobial Activity to Zinc Binding: An In-Depth Analysis of the Tunicate Host Defense Peptide Clavanin A" (2020). Doctoral Dissertations. 2421. https://opencommons.uconn.edu/dissertations/2421 From Antimicrobial Activity to Zinc Binding: An in-depth Analysis of the Tunicate Host Defense Peptide Clavanin A

Samuel A. Juliano IV, PhD

The University of Connecticut, 2020

Due to the overuse of conventional antibiotics, coupled with the slow rate of development of novel treatments, antibiotic resistant bacteria have been on the rise. Antimicrobial peptides

(AMPs), especially those that bind to biologically relevant metal cations, present an excellent source of potential antibacterial compounds with a low incidence of resistance development. This dissertation explores one such metal binding AMP, the tunicate peptide clavanin A. In the past, clavanin A has been fairly well studied for its activity against the bacterial membrane; however, a fortuitous discovery led us to study clavanin A in the presence of metal ions. We discovered that, in the presence of Zn2+ ions, clavanin A is capable of cleaving bacterial chromosomal DNA, an activity never before reported for a zinc binding AMP. This mechanism of action also has the benefit of improving the antibacterial activity of clavanin A 16-fold. These promising results led to the exploration of AMP mechanisms of action through the application of bacterial cytological profiling. Our study represents the first use of this technique for the study of AMPs. Bacterial cytological profiling allowed us to identify that clavanin A has three distinct mechanisms of action: a membrane-based mechanism at neutral pH that is similar to that of the piscidin peptides, another membrane-based activity at low pH involving strong membrane interaction, and the zinc-based nuclease activity discussed above. The results presented in this dissertation combine to show that not only is clavanin A an interesting AMP system for study, but it is also a prime candidate for the modifications necessary to develop novel peptide-based antibiotic treatments to fend off antibiotic resistant bacteria.

From Antimicrobial Activity to Zinc Binding: An in-depth Analysis of the Tunicate Host Defense Peptide Clavanin A

Samuel A. Juliano IV

B.S., University of Scranton, 2013

A Dissertation

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

at the

University of Connecticut

2020

Samuel A. Juliano IV

2020

APPROVAL PAGE

Doctor of Philosophy Dissertation

From Antimicrobial Activity to Zinc Binding: An in-depth Analysis of the Tunicate Host Defense Peptide Clavanin A

Presented by

Samuel A. Juliano IV

Major Advisor ______Alfredo M. Angeles-Boza

Associate Advisor ______Mark Peczuh

Associate Advisor ______Jessica Rouge

University of Connecticut

2020

iii

This work is dedicated to my wonderful wife and best friend Meaghan Daly. Thank you for always standing beside, behind, or before me… wherever I needed you most at the time.

Acknowledgements

I want to start by thanking my advisor, the inimitable Alfredo Angeles-Boza. Despite the fact that we got off to a rough start, with you convinced I was an undergraduate you just didn’t have the space for, I have always felt that you have had my best interests at heart, making sure I gained the skills I wanted and needed. I am a better scientist and teacher as a result of the opportunities you have afforded me. I would also like to acknowledge my committee members: Mark Peczuh, Jessica Rouge, Edward Neth, and Marcy Balunas. Thank all of you for bearing with me as I scoffed at deadlines and the traditional graduate school timeline. I hope I wasn’t too much of a problem. Special thanks to Dr. Balunas for the help recovering tunicate hemolymph for my first paper. Also, I would like to thank Dr. Neth specifically for his constant reminders to get my general exam done, I might’ve forgotten otherwise. To my lab mates past and present: Dr. Daben Libardo, Dr. Taylor Schneider, Dr. John Nganga, Jasmin Portelinha, IC Duay, Scott Pierce, Maria Heredia-Chavez, Marvin Naing, Kara Heilemann, Murphy Jennings, and Lindley Bower, Thank you all for making the lab just a little more fun. Special thanks to Scott for making me go outside every once and a while, that was more helpful than you could possibly understand. To the staff of the Chemistry Department’s Teaching Laboratory Services (The Teaching Stockroom): Eric, Laurie, Noreen, Sam, and Holly, thank you for accepting me as one of your own for the last semester and a half. The knowledge I gained while working with you will serve me well going forward, and the fact that I have been able to remain employed is invaluable. I hope I was half as useful to you as you were to me. I also want to thank Emilie Hogrebe and the Main Office staff for smoothing the background processes of the department. I would like to thank all of my UConn Chemistry friends, especially Dr. Shannon Poges, Dr. Steve Murphy, Alyssa Hartmann and Melissa Limbacher. Thanks for being there when I needed to vent, and for knowing when the best solution to a problem was to go for a walk. I also want to thank my friends from New Jersey, especially Alex, Mauricio, Kim, and Mike, for reminding me that life isn’t always about science. Also, thanks for not coming to my thesis defense, you’ll never know how much your absence means to me. To all of my students over the years, at UConn Storrs, UConn Hartford, Sacred Heart University and The University of Hartford. Thank you for bearing with me as I figured out how to teach. I have become better through you. I would like to acknowledge Brandon Sanderson, Isaac Stewart and Dragonsteel Entertainment for the use of the Feruchemical symbol for zinc that I used as chapter icons. Next, I would like to thank my family, beginning with my parents, Sam and PJ Juliano, for never letting me believe that good enough was good enough. Without your constant commitment to my academic achievement I would not have become the man I am today. This is the culmination of all of that hard work. Thank you, also, for reading all of my writing, even if you couldn’t always understand it, I knew I always had you in my corner. I want to thank my brother Nicholas Juliano for always believing in me, and for taking an interest in astrophysics, so that we would never have

v to compete with respect to our scientific careers. Family isn’t just the people you are born with, and when I got married I gained the best family of in-laws a man could ask for. Judy, Rich, Richard, Olivia, Joseph, Mary, Edward, Anne, and the extended Daly and McCaughey clans, thank you so much for accepting me wholeheartedly into the fold. It means so much to have this huge second family backing me up by choice, rather than obligation. Lastly, I would like to thank my wife, Meaghan Daly, for everything she has done for me. Honestly, if I were to write all of it here, my dissertation would be twice as long. Suffice it to say that having you by my side all of these years is the only thing that gave me the courage and strength to attempt all that I have. Without your support I don’t know who I would be today.

Oh, and a big thank you to tunicates everywhere. I couldn’t have done it without you!

Table of Contents

Acknowledgements ...... v

List of Figures ...... x

List of Tables ...... xii

Chapter 1 – Introduction: On the Shoulders of Giants ...... 1

1.1 – The Problem: Antibiotic Resistance ...... 2

1.2 – Antimicrobial peptides/Host defense peptides ...... 7

1.3 – Strategies of Metal Binding Host Defense Peptides ...... 14

1.4 – Tunicates and the metal-rich marine environment ...... 18

1.5 – Clavanin A and the Clavanins ...... 20

1.6 – Dissertation overview ...... 22

1.7 – References ...... 24

Chapter 2 – Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide

Clavanin A ...... 30

2.1 – Introduction ...... 31

2.2 – Results ...... 33

2.3 – Discussion ...... 56

2.4 – Methods ...... 59

2.5 – References ...... 66

vii

Chapter 3 – Clavanin A, Zinc metallonuclease effective against gram negative pathogens

...... 72

3.1 – Introduction ...... 73

3.2 – Results ...... 73

3.3 – Discussion ...... 86

3.4 – Methods ...... 86

3.5 – References ...... 93

Chapter 4 – Three Strategies in Motion: Bacterial Cytological Profiling Elucidates the

Mechanisms of Action for Clavanin A ...... 97

4.1 – Introduction ...... 98

4.2 – Results ...... 98

4.3 – Discussion ...... 114

4.4 – Methods ...... 118

4.5 – References ...... 120

Chapter 5 – Continuing the Narrative ...... 126

5.1 – Future Directions and Newly Generated Questions ...... 127

5.1.1 – The Clavanins ...... 127

5.1.2 – Clavanin Synergy ...... 130

5.1.3 – Expanding the Utility of Bacterial Cytological Profiling ...... 132

5.2 – Conclusions ...... 137

viii

5.3 – References ...... 138

List of Figures

Figure 1-1 The Development of Antibiotic Resistance ...... 3

Figure 1-2 Gram-positive and Gram-negative Cell Walls ...... 6

Figure 1-3 Strategies to Overcome Antibiotic Resistance ...... 7

Figure 1-4 The Variety of AMP Structures ...... 9

Figure 1-5 AMPs in Clinical Trials ...... 11

Figure 1-6 Peptoids and β-peptides...... 13

Figure 1-7 Metal Binding HDPs ...... 16

Figure 1-8 The ATCUN Motif ...... 17

Figure 1-9 Tunicate Anatomy ...... 19

Figure 2-1 CD of Clavanin A and Mutants ...... 37

Figure 2-2 Clavanin A Zinc Concentration Dependence ...... 39

Figure 2-3 CD of Clavanin A with Zn2+ ...... 41

Figure 2-4 Time Kill Kinetics ...... 42

Figure 2-5 Clavanin A Membrane Depolarization ...... 44

Figure 2-6 β-galactosidase Leakage...... 46

Figure 2-7 Potassium Leakage ...... 47

Figure 2-8 Synthesis of Clavanin A-TMR ...... 49

Figure 2-9 Clavanin A-TMR Micrographs ...... 51

Figure 2-10 DNA Cleavage Gels and Graphs...... 55

Figure 3-1 TUNEL Results ...... 80

Figure 3-2 Clavanin A Structure ...... 82

Figure 3-3 RMSFs of the Clavanin A-Zn2+ Complex ...... 82

Figure 3-4 Clavanin A-Zn2+ - DNA Complex ...... 83

Figure 3-5 Mechanism of DNA Hydrolysis...... 86

Figure 3-6 MD Conformations of Clavanin A-Zn2+-DNA Complex ...... 91

Figure 4-1 Clavanin A BCP Micrographs ...... 102

Figure 4-2 PCA Plot...... 105

Figure 4-3 Other HDP BCP Micrographs ...... 109

Figure 4-4 CD Spectra with Lipid Vesicles ...... 112

Figure 4-5 XRD and SPR/EIS Results...... 113

Figure 4-6 Clavanin A Mechanisms of Action ...... 116

Figure 5-1 Clavanin C ITC Data ...... 129

Figure 5-2 Clavanin Family PCA Plot ...... 134

List of Tables

Table 1-1 Clavanin Sequences ...... 21

Table 2-1 MICs of Clavanin A with Metal Ions ...... 35

Table 2-2 MICs with and without Zn2+ ...... 38

Table 2-3 MIC of Clavanin A-TMR ...... 48

Table 2-4 MICs against ΔznuABC E. coli ...... 53

Table 2-5 MICs against ΔnikABC E. coli ...... 54

Table 3-1 MICs against Gram-negative Pathogens ...... 77

Table 3-2 Clavanin A – Zinc Antimicrobial Synergy ...... 78

Table 4-1 Summary of the Clavanin A Mechanisms of Action ...... 104

Table 4-2 Average Bacterial Membrane Morphology Measurements ...... 106

Table 4-3 Average Bacterial Nucleoid Morphology Measurements ...... 107

Table 4-4 Clavanin A Helical Content...... 111

Table 5-1 MICs of the Clavanins with and without Zn2+ ...... 128

Table 5-2 Clavanin Synergy Results...... 132

Table 5-3 Average Bacterial Membrane Morphology Measurements ...... 135

Table 5-4 Average Bacterial Nucleoid Morphology Measurements ...... 136

xii

Chapter 1 – Introduction: On the Shoulders of Giants

“If I have seen further it is by standing on the shoulders of Giants.”

-Sir Isaac Newton, 1675

“Everyone shush now…I’m about to be impressive.”

-Sebastien de Castell, Tyrant’s Throne

1.1 – The Problem: Antibiotic Resistance

While many people may look to climate change, war, and cancer as the biggest threats mankind will face over the next thirty years, one problem grows silently, year after year, without much public recognition. Antibiotic resistant bacteria represent a significant threat to the health and wellbeing of mankind, so much so that the World Health Organization lists it as one of its top ten threats to global public health.1 In 2013, the Centers for Disease Control reported that over 2 million individuals were infected with antibiotic resistant infections, resulting in the deaths of about 23,000 people in the United States alone.2 In the last six years these numbers have increased dramatically such that nearly 3 million incidences of infection occur, leading to over 35,000 deaths.3 In Europe, the European Center for Disease Prevention and Control reports

33,000 annual deaths in the EU due to these difficult to treat infections.4 The situation in developing nations is even worse, with 88% of Nigerian Staphylococcus aureus infections resistant to methicillin, and 95% of Indian adults carrying β-lactam resistant bacteria (compared with about 10% for New York City).5 The World Health Organization claims that the world is poised to enter a “post-antibiotic era” unless something is done to stem the tide.6 Some frightening projections estimate that by 2050, the death toll will reach over 2.4 million individuals with costs in excess of $3 billion per year.7

While it is certainly necessary to develop new antimicrobial treatments that are less likely to develop resistance, it is also important to be aware of the factors that lead to resistance in the first place. In his 1945 Nobel Lecture, Alexander Fleming predicted the possibility of bacterial resistance to penicillin, stating “It is not difficult to make microbes resistant to penicillin in the laboratory by exposing them to concentrations not sufficient to kill them…”8 He then concludes his lecture with the admonition “If you use penicillin, use enough.” Even in the earliest days of

2 antibiotics it was known that people will often stop taking their medication because they no longer experience the symptoms that lead them to turn to antibiotics in the first-place, resulting in under-dosed bacterial pathogens. This is one of the leading causes of antimicrobial resistance today, though far from the only one. Figure 1-1 shows a representation of how antibiotic resistance occurs and spreads.

Figure 1-1

The development of antibiotic resistance: 1) A person or animal becomes infected with a bacterial

infection. A small portion of those bacteria are resistant (orange). 2) The patient is treated with an

antibiotic, killing all of the non-resistant bacteria, often including the host’s beneficial bacteria,

and sparing the resistant bacteria. 3) The patient stops taking their antibiotic, allowing the bacteria

to repopulate. The proportion of resistant bacteria is now greater, increasing the likelihood of

transmission of resistant pathogens. 4) Resistance can be transmitted to non-resistant bacteria

through horizontal gene transfer.

While Fleming assumed there was no danger in antibiotic overuse, we have learned in the years since his discovery that this is not the case. In developing countries, in particular, antibiotic use is largely unrestricted, with many countries selling antibiotics as over the counter treatments.9 This availability results in sick individuals self-medicating with antibiotics, even for conditions that do not require them, such as viral infections or malaria. In other countries, such

3 as China, healthcare providers and hospitals are provided with financial incentives for each antibiotic prescription written, resulting in cases of gross over-prescription.5

Aside from antibiotic use and abuse in humans and natural mutations, one of the largest sources of antibiotic resistance mutations to bacteria is the use of antibiotics on animals, typically in an agricultural setting. Commercially farmed animals, including cows, chickens, sheep, and pigs, are often dosed with continuous, subtherapeutic quantities of antibiotic in order to promote growth and prevent disease.10 In the US alone this amounts to over 13 million kilograms of antibiotics annually, a whopping 80% of the antibiotic use for the country.11 Much like in humans who under-dose their bacterial infections, the application of subtherapeutic concentrations of antibiotics in these animals has led to the development of a large reservoir of antibiotic resistant bacteria (Figure 1-1). These bacteria can then be transferred to humans through the direct transfer of bacteria via contact with the animals, improperly handled meat, or materials contaminated with feces. Antibiotic resistance can also arise in bacterial populations through horizontal gene transfer from bacteria specific to livestock to those that can infect humans.12

In developed nations the most common means to acquire an antibiotic resistant infection is through hospital acquired, or nosocomial, infection. These infections are responsible for the deaths of over 1 million patients annually in the US, many of which come from antibiotic resistant bacteria. The most prevalent offenders when it comes to nosocomial infections are the so called ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species).13,

14 A growing number of isolates of these bacteria collected from hospitals and doctor’s offices contain antibiotic resistance genes, which is largely to account for the increased mortality

4 reported above.15 Of particular interest are the gram negative members of this group, as they are more difficult to kill than gram positive bacteria due to their outer membranes and efflux pumps

(Figure 1-2).16, 17 Several isolates of Acinetobacter spp. and Pseudomonas spp. have been identified that have resistance genes to all known classes of antibiotics.18 Additionally, Klebsiella derived carbapenem resistance genes isolated in India have been found in various species of bacteria world-wide, highlighting the dangers of horizontal gene transfer.19 With so many bacteria showing resistance, even to some of the antibiotics of last resort, the need for novel therapeutic compounds to combat the problem is greater than ever.

Figure 1-2

Representative of gram-positive and gram-negative cell walls. The outer most layer of a gram-

positive bacteria (left) is a cell wall composed of peptidoglycan (blue and green), followed by

the cytoplasmic membrane (yellow). Gram-negative cells (right) are surrounded by a layer of

lipopolysaccharide (purple) embedded in the outer membrane (yellow). The outer membrane

contains efflux pumps (red) which can remove antibiotics and AMPs from the periplasmic

space (gray), which contains the peptidoglycan cell wall (blue and green). The final layer of

defense for a gram-negative cell is the inner (cytoplasmic) membrane (yellow).

All of this information points to the need for the development of new, more potent antimicrobials. The current antibiotic development paradigm, the modification of current small molecule antibiotics to overcome resistance mechanisms is falling behind the ability of bacteria to mutate resistance genes.20, 21 The three qualities that should be sought for when considering new molecules as novel antibiotics are i) localized targeting with minimal interaction with host tissue, ii) low incidence of resistance development, and iii) protecting native flora whenever

6 possible (Figure 1-3).12 The current generation of antibiotics based on small molecules is capable of meeting the former criteria, but fails miserably with regard to the latter two.

Figure 1-3

Conceptual diagram outlining the three criteria (left) necessary for any strategy designed to

circumvent the recognized modes of antibiotic resistance (right). Figure repurposed from

reference 12.

1.2 – Antimicrobial peptides/Host defense peptides

Many of the antibiotics we utilize today are derived from natural small molecules produced by a wide variety of organisms in order to defend themselves from infection by bacterial pathogens. The major example of this is the antibiotic penicillin, which was isolated from the penicillium mold in the late 1920s.22 Since then, several antibiotics have been synthesized based on the structure of this influential molecule, including ampicillin, amoxicillin and methicillin.23-25 Entire classes of antibiotics only exist because of discoveries made studying the natural defenses of various organisms. The aminoglycosides were first discovered in 1943

7 with the isolation of streptomycin from the soil bacteria Streptomyces griseus.26 This discovery lead to the isolation of many more compounds with potent antimicrobial activity, including kanamycin, erythromycin and vancomycin. Our past reliance on molecules originating from natural sources can inform a solution to the current predicament of antibiotic resistant bacteria.

We should be returning to natural systems as a means to develop new treatments to stop the growing scourge.

Antimicrobial peptides (AMPs) represent a large and particularly effective pool of molecules with potent antimicrobial activity that we could utilize in order to develop novel means to combat bacterial infections. These evolutionarily ancient molecules can be found as innate immune components across nearly all domains of life, earning them the moniker host defense peptides (HDPs).27, 28 For the most part, AMPs are short, cationic peptide sequences, 10-

50 amino acids in length, which challenge invading bacteria either through membrane disruption or the targeting of intracellular macromolecules like DNA.29 The majority of AMPs discovered form amphipathic α-helices, with hydrophobic and hydrophilic residues on opposite faces of the helix, though other secondary structures have been observed including β-sheets and intrinsically disordered structures (Figure 1-4).30

Figure 1-4

Structures of several (a) α-helical, (b) β-sheet, and (c) random coil AMPs. Figure repurposed

from reference 29.

The vast majority of AMPs are broad spectrum, exhibiting potent antimicrobial activity against gram-negative and gram-positive bacteria. Their ability to target bacteria rather than host cells has been attributed to electrostatic interactions between the cationic peptide and the anionic bacterial membrane.30 Eukaryotic cells more often contain zwitterionic membrane components, decreasing the interaction between their membranes and AMPs. Once an AMP interacts with the membrane, many will exert antimicrobial activity through disruption of the membrane, resulting

9 in leakage of cellular components or disruption of fundamental ion gradients. Several AMPs have also been discovered that will pass through the membrane via non-lytic means and instead target internal components, such as DNA or ribosomes. A growing group of AMPs have been identified which are capable of exerting antimicrobial activity via more than one mechanism.31

These multi-hit AMPs are particularly interesting as they present the best means for the development of novel antibiotics with reduced resistance development, as target promiscuity would require bacteria to acquire multiple mutations in order to be able to survive all possible

AMP mechanisms of action. In addition, many of the targets for AMPs are nonspecific (e.g. targeting membrane phospholipids rather than a specific protein) making the evolution of bacterial resistance to AMPs even less likely. It should be noted, however, that while typical bacterial antibiotic resistance mechanisms are ineffective against AMPs, it is possible for bacteria to develop resistance to AMPs through the expression of efflux pumps and proteases, though these methods often require more generations to develop relative to small molecule resistance mechanisms.32, 33

Due to the advantageous nature of AMPs as potential novel therapeutic compounds, several natural and naturally derived AMPs have made it to clinical and preclinical trials.34

Outside of AMPs and HDPs, peptides have gained FDA approval for the treatment of some cancers and metabolic diseases, paving the way for a future with clinically available AMPs to treat bacterial infections, though currently daptomycin (Figure 1-5A) is the only FDA approved

AMP.35 As of May 2019, 27 AMPs are being tested in clinical trials, with six in phase I, fifteen in phase II, and a promising six in phase III trials (Figure 1-5B).34 Of the AMPs in phase three trials, the vast majority are either synthetic analogs of naturally occurring AMPs or AMPs designed to have a certain function, with only one, ramoplanin, that is natural.

Figure 1-5

(A) Structure of the clinically approved AMP antibiotic, Daptomycin; (B) Status of AMPs

in clinical trials. Adapted from reference 34.

While AMPs and HDPs seem to be exactly what the doctor ordered in terms of a replacement for the current generation of antimicrobials, there are some fairly significant drawbacks to their use in their current form. In addition to the slow development of resistance mentioned above, some AMPs show activities that are sensitive to varied conditions of pH or salt concentrations due to the need for electrostatic interactions between the cationic peptides and the anionic membrane.36, 37 Additionally, the presence of serum proteins can present issues due to

AMP inactivation as a result of protein binding.38 However, the largest drawback to the utilization of AMPs in a clinical setting is the low in vivo stability often observed for AMPs. Due to the presence of proteases and peptidases, AMPs are rapidly degraded in the blood-stream and mucosa, and they are unable to survive harsh conditions found in the stomach.39, 40 While alternate routes of administration are available (e.g. intravenous or intramuscular injections), for

11 the most part, topical applications are explored in order to deal with these limitations. On top of all of these limitations, peptides cost significantly more to produce than conventional small molecule antibiotics, though new production methods are in development with the aim of reducing costs.41

Despite the fact that all of these limitations seem to signal the death knell for AMPs as clinically available antibiotics, there are several strategies that can be employed to overcome these deficiencies. Many of the AMPs in clinical trials contain chemical modifications, which are far and away the most commonly employed strategies for improving in vivo survival of AMPs.12,

42-45 The chemical modifications employed by AMPs in clinical trials include partial or complete replacement of L-amino acids with their D counterparts, macrocyclization and incorporation of noncanonical amino acids.34 Each of these modifications increases stability in vivo by decreasing the rate of proteolysis. Larger modifications are also being explored, including glycosylation, lipidation and halogenation. Other options for increasing stability include the utilization of peptidomimetics, such as peptoids, β-peptides, and peptide-peptoid hybrids (Figure 1-6).46

Figure 1-6

Structure of peptoids, with R groups found on backbone nitrogen instead of carbon, and β-

peptides, with an extra backbone carbon in each amino acid.

The development of peptide based or peptide derived novel antimicrobial compounds requires a strong understanding of what it is exactly about an AMP that makes it function the way it does. For small molecule antibiotics, there are only so many positions that can be modified, thus limiting the number of possibilities for more active modifications. This, coupled with the relative ease by which crystal structures of these molecules bound to their target can be obtained, makes modification of small molecules a relatively targeted process. On the other hand, AMPs have between 10 and 50 amino acids, each of which can contribute different properties depending on their location relative to other residues. It is important to have a well- studied AMP from which to derive novel treatments. For example, two AMPs in clinical trials, omiganan and pexiganan are derived from indolicidin and magainin 2, respectively, each of which has been studied in over 200 publications. A greater understanding of a larger variety

AMPs is necessary in order for them to serve as templates for further antibiotic development.

1.3 – Strategies of Metal Binding Host Defense Peptides

Metal ions are ubiquitous in biological systems, and this holds true for HDPs as well.

Many HDPs can bind to metal ions for a variety of reasons, including metal sequestration, metal activation, and redox chemistry. The vast majority of these metal ion binding peptides utilize multiple histidine residues as part of their binding motif.

Organisms will utilize metal sequestration in order to limit bacterial access to essential micronutrients or metal ions necessary for virulence factors.47, 48 Bacterial hosts can excrete metal binding HDPs, including microplusin, and the S100 peptides psoriasin and calprotectin, which can sequester essential metal ions, limiting bacterial growth in order for the host to eliminate the infection. Microplusin is found in several locations within ticks, and is composed of five α-helical domains with disordered C- and N-termini.49-51 While the exact binding site has not been identified, microplusin exerts its activity against gram positive bacteria as well as fungi through sequestration of Cu2+ ions essential for cellular respiration and other bacterial processes.51 S100 proteins have been known for many years to bind Ca2+ and activate the immune response in mammals, though more recently, they have also been discovered to sequester essential metal ions from bacterial infections.52 Psoriasin, an S100 peptide found on human skin, has been found to contain a Zn2+ binding motif containing three histidines and an aspartate.53 When pre-incubated with Zn2+, psoriasin loses its potency against E. coli, pointing to sequestration as its primary mechanism of action.54 Another S100 peptide, calprotectin, goes one step further, sequestering not only Zn2+ but also Mn2+. This sequestration activity not only serves to deny microbes of essential metallic micronutrients, but also increases stability against proteolysis, a common issue for AMPs.55-57

While metal sequestering HDPs gain their antimicrobial activity from denying bacteria access to micronutrients, HDPs that require metal activation are only useful when bacteria are exposed to the peptide in the metallated state. It is commonly believed that binding metal cofactors allows these HDPs to adopt a favorable conformation for binding to their bacterial targets.58 Bacitracin A (Figure 1-7A), produced by several Bacillus species, can bind to a variety of metal ions including Zn2+, Mn2+ and Na+.59, 60 Metal binding allows for bacitracin A to inhibit cell wall biosynthesis by binding to undecaprenyl pyrophosphate.61 Dermcidin is an HDP secreted in sweat glands which is proteolytically cleaved to the active form DCD-1.62 Crystal structures of DCD-1 (Figure 1-7B) show that, in the presence of Zn2+ ions, it forms antiparallel dimers, stabilized by the zinc ions. The dimers then form a trimer in a cylindrical pattern, allowing for the formation of a transmembrane ion channel, though the means by which this channel inserts into the membrane remains unknown.63, 64 The acidic conditions and high Zn2+ concentrations found in sweat allow DCD-1 to bind Zn2+ and oligomerize in order to kill bacteria.

Figure 1-7

Crystal structures of metal-bound (A) Bacitracin A, and (B) DCD-1.

Beyond the structural and electrostatic benefits of metal binding, HDPs have also been shown to utilize metal ions for their ability to participate in redox reactions which can oxidize bacterial components. The HDPs that utilize the redox properties of metals do so through binding Cu2+ with the amino terminal copper and nickel (ATCUN) binding motif (Figure 1-8), comprised of any two amino acids (with the exception of proline) followed by a histidine in the third position. This motif, discovered in serum albumins, allows for the binding of Cu2+ in a modified square planar arrangement utilizing the terminal nitrogen, the next two backbone nitrogens and the unprotonated histidine nitrogen.65 In addition to ATCUN proteins, many HDPs also utilize this motif for its oxidative ability when bound to Cu2+, including several histatins, ixosin, and piscidins 1 and 3. Both histatin 3 and histatin 5 can bind to Cu2+ through their

ATCUN motifs, using Fenton-like chemistry to generate reactive oxygen species (ROS).66, 67

These peptides can also bind to Zn2+ utilizing an HEXXH motif, though it is unknown what

16 effect this binding has on the antimicrobial activity.66 Ixosin, found in the salivary glands of several tick species, utilizes its GLH ATCUN motif in order to oxidize membrane phospholipids.68, 69 This oxidation allows for synergism with the closely related HDP ixosin B.68

The piscidin family of HDPs, found in fish mast cells, were originally believed to exert their antimicrobial activity on bacterial membranes.70, 71 More recently, ATCUN containing piscidin 1 and piscidin 3 were found to act on both the membrane and DNA as an intracellular target, though piscidin 1 favored the membrane while piscidin 3 acted more strongly against DNA.72

Both peptides utilize their ATCUN motifs for this activity, with H3A mutations drastically reducing their activity. In addition to cleaving DNA intracellularly, piscidin 3 was found to cleave extracellular DNA, disrupting the formation of biofilms.73

Figure 1-8

The ATCUN motif with a bound Cu2+ atom.

1.4 – Tunicates and the metal-rich marine environment

Due in part to both natural sources and human activity, sea water presents a large source of available metal ions, especially in regions near harbors and marinas. The Sea Urchins for

Education group at Stanford University report iron, copper and zinc to be among the most abundant transition metals found in the oceans surrounding the US.74

As seen with the piscidins, marine life presents an excellent resource for the discovery of novel AMPs with the ability to utilize these ubiquitous metal ions. Unlike land animals, marine animals are constantly surrounded by concentrations of dissolved metals and minerals which either need to be sequestered, so as not to harm the organism in question, or otherwise enriched in order to promote some vital function. HDPs are a part of the innate immune system of many lower organisms which lack adaptive immunity found in higher order creatures. Tunicates provide an interesting organism for HDP study because their lack of an adaptive immune system, despite being members of the phylum chordata, to which all vertebrates belong. There is much interest in studying the fairly primitive immune system of tunicates as a means to understand how more complex immune systems, such as our own, came to be.75, 76

As previously mentioned, tunicates are members of the phylum chordata, despite being invertebrates, due to the fact that in their early development stages they contain features linked to all chordates, including a notochord, pharyngeal gill slits and a post-anal tail (Figure 1-9A).

Tunicates start life a free swimming, tadpole-like organisms, though after metamorphosis they become sessile filter-feeders, not unlike sponges (Figure 1-9B).77 One offshoot of metamorphosing into a sessile form is that they become unable to change their local environment, meaning that they need to adapt to the dissolved metal concentrations in their local

18 environment rather than being able to leave the area. With respect to metal ions, this means that tunicates can either sequester and excrete these ions, or employ them as a part of their regular function, including utilizing them for host defense in conjunction with HDPs. In the tunicate

Styela clava zinc concentrations have been reported to be 26.7±7.1 μM, which amounts to roughly two orders of magnitude enrichment over the surrounding seawater.74, 78 This organism, then, presents an excellent opportunity for the discovery of metal-binding HDPs as a component of its innate immune system.

Figure 1-9

Anatomy of (A) a planktonic tunicate larva, and (B) a sessile adult tunicate. Images

courtesy of Jon Houseman.

1.5 – Clavanin A and the Clavanins

The clavanin family of cationic, α-helical HDPs was discovered in the hemocytes

(analogues of red and white blood cells found in invertebrates) of S. clava.79 The family contains five members with about 80% sequence homology, and a sixth member with only ~50% similarity that was discovered through the homology of the prepropeptide region of its genome

(sequences in Table 1-1.79-81 Clavanin A is by far the most studied member of the family, appearing in at least 15 publications across five different laboratories, while each of the other members of the family have only appeared in the publication in which they were first presented.78, 79, 82-84 Clavanin B is nearly identical to clavanin A, with the only difference being an arginine in place of a lysine at position 7, though this is a conserved mutation as both residues are positively charged at biologically relevant pH values. Clavanins C and D were discovered to be the most active members of the clavanin family against not only gram-negative and gram- positive bacteria, but also fungi as well. Despite this potency and broad spectrum activity, these

HDPs have never been studied in depth, likely due to the modified tyrosine residue in the sequence, which was originally believed to be a methyl-tyrosine, though the discovery of the styelin family of peptides has indicated that the modified tyrosine may actually be a dihydroxyphenylalanine (DOPA) residue.79, 85 Additionally, clavanin C contains an ATCUN motif which is likely capable of binding Cu2+ and forming ROS, though this activity has never been explored. Clavanin E and clavaspirin were not discovered in the initial clavanin study, instead appearing as a result of exploration into the S. clava genome.80, 81

Table 1-1

Sequences of the clavanin family of HDPs

Peptide Sequence

Clavanin A79 VFQFLGKIIHHVGNFVHGFSHVF

79 Clavanin B VFQFLGRIIHHVGNFVHGFSHVF

Clavanin C79 VFHLLGKIIHHVGNFVÝGFSHVF a

Clavanin D79 AFKLLGRIIHHVGNFVÝGFSHVF a

Clavanin E80 LFKLLGKIIHHVGNFVHGFSHVF

Clavaspirin81 FLRFIGSVIHGIGHLVHHIGVAL a Ý represents a post translationally modified tyrosine residue

As a result of the lack of general attention for the other members of the clavanin family, clavanin A is the only member for which anything is known beyond relative antibacterial activity. Clavanin A was discovered to be membrane active, with two distinctly different activities depending on the pH of the peptide.86, 87 This pH dependence is not surprising given the relatively high quantity of histidine residues found in its sequence (Table 1-1). Beyond just histidine, the membrane based activity of clavanin A was found to also rely on its abundant glycine and phenylalanine residues.82, 88 Clavanin A has been found in both the phagosomes of macrophage-like hemocytes and the granules of neutrophil-like hemocytes, indicating that the

HDP is utilized both intracellularly and extracellularly.89

As a result of these preliminary studies, the potential for the use of clavanin A as a novel antibiotic has been well explored. It has been studied in vitro in murine wound infection models, wherein clavanin A treated S. aureus wounds showed improved healing and eradication of

21 bacteria from the wound.90 Additionally, in systemic infection models, clavanin A treatment resulted in increased survival for both E. coli and S. aureus infected mice. Clavanin A has also proven effective against biofilms, preventing the formation of E. coli and S. aureus biofilms and perturbing preformed K. pneumoniae biofilms.91 In addition to the study of unmodified clavanin

A as a potential therapeutic, it has also been conjugated to the surface of iron oxide nanoparticles

84 and coated onto central venous catheters. Clavanin A-Fe3O4 coated catheters were capable of reducing adhesion of both gram positive (~50%) and gram negative (70-90%) bacteria and modestly reducing the amount of planktonic bacteria as well, with up to 20% reduction. While more study of clavanin A is certainly necessary, it appears to be an excellent framework for the development of novel, peptide-based antimicrobials in the future.

1.6 – Dissertation overview

This dissertation explores the metal binding characteristics of clavanin A with the aim of elucidating the biological activity of this HDP, so that it can be utilized as a framework for peptidic antibiotics. Chapter two presents the hypothesis that clavanin A is binding to Zn2+ and first reports the possibility of nucleolytic activity. Chapter three further explores clavaninA-Zn2+ as a nuclease, presenting a novel mechanism of action for zinc binding AMPs. In chapter four, bacterial cytological profiling is applied to the study of AMPs for the first time with the aim of exploring the apparent disconnect between the literature reporting clavanin A as a membrane active peptide, and our results which indicate a zinc-based DNA cleavage mechanism of action.

Finally, chapter five presents some of the questions generated as a result of the answers obtained throughout this work, as well as some future directions for the study of not only clavanin A, but the entire clavanin family.

The sum of this work provides novel information related to the future development of antimicrobial interventions. Research resulting in the development of drugs that are less likely to develop resistance will allow for longer lasting, more effective antimicrobial activity.

1.7 – References

1. WHO Ten threats to global health in 2019. https://www.who.int/emergencies/ten-threats-to- global-health-in-2019 (accessed December 3). 2. CDC Antibiotic Resistance Threats in the United States, 2013; U.S. Department of Health and Human Services: Atlanta, GA, 2013. 3. CDC Antibiotic Resistance Threats in the United States, 2019; U.S. Department of Health and Human Services: Atlanta, GA, 2019. 4. Cassini, A.; Högberg, L. D.; Plachouras, D.; Quattrocchi, A.; Hoxha, A.; Simonsen, G. S.; Colomb-Cotinat, M.; Kretzschmar, M. E.; Devleesschauwer, B.; Cecchini, M.; Ouakrim, D. A.; Oliveira, T. C.; Struelens, M. J.; Suetens, C.; Monnet, D. L.; Strauss, R.; Mertens, K.; Struyf, T.; Catry, B.; Latour, K.; Ivanov, I. N.; Dobreva, E. G.; Tambic Andraševic, A.; Soprek, S.; Budimir, A.; Paphitou, N.; Žemlicková, H.; Schytte Olsen, S.; Wolff Sönksen, U.; Märtin, P.; Ivanova, M.; Lyytikäinen, O.; Jalava, J.; Coignard, B.; Eckmanns, T.; Abu Sin, M.; Haller, S.; Daikos, G. L.; Gikas, A.; Tsiodras, S.; Kontopidou, F.; Tóth, Á.; Hajdu, Á.; Guólaugsson, Ó.; Kristinsson, K. G.; Murchan, S.; Burns, K.; Pezzotti, P.; Gagliotti, C.; Dumpis, U.; Liuimiene, A.; Perrin, M.; Borg, M. A.; de Greeff, S. C.; Monen, J. C. M.; Koek, M. B. G.; Elstrøm, P.; Zabicka, D.; Deptula, A.; Hryniewicz, W.; Caniça, M.; Nogueira, P. J.; Fernandes, P. A.; Manageiro, V.; Popescu, G. A.; Serban, R. I.; Schréterová, E.; Litvová, S.; Štefkovicová, M.; Kolman, J.; Klavs, I.; Korošec, A.; Aracil, B.; Asensio, A.; Pérez-Vázquez, M.; Billström, H.; Larsson, S.; Reilly, J. S.; Johnson, A.; Hopkins, S., Attributable deaths and disability-adjusted life-years caused by infections with antibiotic-resistant bacteria in the EU and the European Economic Area in 2015: a population-level modelling analysis. Lancet Infect Dis 2019, 19 (1), 56- 66. 5. Reardon, S., Antibiotic resistance sweeping developing world. Nature 2014, 509 (7499), 141- 2. 6. World Health, O., Antimicrobial resistance: global report on surveillance. World Health Organization: Geneva, 2014. 7. OECD, Stemming the Superbug Tide. 2018. 8. Fleming, A., Penicillin. In Nobel Lecture, Nobel Media: 1945. 9. Ayukekbong, J. A.; Ntemgwa, M.; Atabe, A. N., The threat of antimicrobial resistance in developing countries: causes and control strategies. Antimicrob Resist Infect Control 2017, 6 (1), 47. 10. Mellon, M.; Fondriest, S., Union of Concerned Scientists. Hogging it: estimates of animal abuse in livestock. Nucleus 2001, 23, 1-3. 11. Hollis, A.; Ahmed, Z., Preserving antibiotics, rationally. N Engl J Med 2013, 369 (26), 2474- 6. 12. Brooks, B. D.; Brooks, A. E., Therapeutic Strategies to Combat Antibiotic Resistance. Adv Drug Delivery Rev 2014, 78, 14-27. 13. Klevens, R. M.; Edwards, J. R.; Richards, C. L., Jr.; Horan, T. C.; Gaynes, R. P.; Pollock, D. A.; Cardo, D. M., Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep 2007, 122 (2), 160-166. 14. Rice, L. B., Federal Funding for the Study of Antimicrobial Resistance in Nosocomial Pathogens: No ESKAPE. J Infect Dis 2008, 197 (8), 1079-1081.

15. Santajit, S.; Indrawattana, N., Mechanisms of Antimicrobial Resistance in ESKAPE Pathogens. Biomed Res Int 2016, 2016, 2475067-2475067. 16. Zgurskaya, H. I.; López, C. A.; Gnanakaran, S., Permeability Barrier of Gram-Negative Cell Envelopes and Approaches To Bypass It. ACS Infect Dis 2015, 1 (11), 512-522. 17. Murata, T.; Tseng, W.; Guina, T.; Miller, S. I.; Nikaido, H., PhoPQ-mediated regulation produces a more robust permeability barrier in the outer membrane of Salmonella enterica serovar typhimurium. J Bacteriol 2007, 189 (20), 7213-7222. 18. Livermore, D. M., The need for new antibiotics. Clin Microbiol Infect 2004, 10 Suppl 4, 1-9. 19. Munoz-Price, L. S.; Poirel, L.; Bonomo, R. A.; Schwaber, M. J.; Daikos, G. L.; Cormican, M.; Cornaglia, G.; Garau, J.; Gniadkowski, M.; Hayden, M. K.; Kumarasamy, K.; Livermore, D. M.; Maya, J. J.; Nordmann, P.; Patel, J. B.; Paterson, D. L.; Pitout, J.; Villegas, M. V.; Wang, H.; Woodford, N.; Quinn, J. P., Clinical epidemiology of the global expansion of Klebsiella pneumoniae carbapenemases. Lancet Infect Dis 2013, 13 (9), 785-96. 20. D. Bhatt, M. K., C.K. Nisha, Y. Mahajan, Nanotechnology solutions to combat superbugs. Nanowerk 2013. 21. Spellberg, B.; Blaser, M.; Guidos, R. J.; Boucher, H. W.; Bradley, J. S.; Eisenstein, B. I.; Gerding, D.; Lynfield, R.; Reller, L. B.; Rex, J.; Schwartz, D.; Septimus, E.; Tenover, F. C.; Gilbert, D. N., Combating antimicrobial resistance: policy recommendations to save lives. Clin Infect Dis 2011, 52 Suppl 5, S397-428. 22. Fleming, A., On the Antibacterial Action of Cultures of a Penicillium, with Special Reference to their Use in the Isolation of B. influenzæ. Br J Exp Pathol 1929, 10 (3), 226-236. 23. Lockey, E.; Eaton, B. R.; Compston, N., Preliminary bacteriological and clinical studies with ampicillin (BRL 1341). Br J Clin Prac 1962, 16, 13-7. 24. Handsfield, H. H.; Clark, H.; Wallace, J. F.; Holmes, K. K.; Turck, M., Amoxicillin, a new penicillin antibiotic. Antimicrob Agents Chemother 1973, 3 (2), 262-5. 25. Rutenburg, A. M.; Greenberg, H. L.; Schweinburg, F. B., Clinical experiences with 2, 6- dimethoxyphenyl penicillin monohydrate in staphylococcal infections. N Engl J Med 1960, 263, 1174-8. 26. Jones, D.; Metzger, H. J.; Schatz, A.; Waksman, S. A., Control of gram-negative bacteria in experimental animals by streptomycin. Science 1944, 100 (2588), 103-5. 27. Zasloff, M., Antimicrobial peptides of multicellular organisms. Nature 2002, 415 (6870), 389- 395. 28. Hancock, R. E.; Haney, E. F.; Gill, E. E., The immunology of host defence peptides: beyond antimicrobial activity. Nat Rev Immunol 2016, 16 (5), 321-34. 29. Nguyen, L. T.; Haney, E. F.; Vogel, H. J., The expanding scope of antimicrobial peptide structures and their modes of action. Trends Biotechnol 2011, 29 (9), 464-72. 30. Hancock, R. E.; Sahl, H. G., Antimicrobial and host-defense peptides as new anti-infective therapeutic strategies. Nat Biotechnol 2006, 24 (12), 1551-7. 31. Peschel, A.; Sahl, H. G., The co-evolution of host cationic antimicrobial peptides and microbial resistance. Nature reviews. Microbiology 2006, 4 (7), 529-36. 32. Anaya-Lopez, J. L.; Lopez-Meza, J. E.; Ochoa-Zarzosa, A., Bacterial resistance to cationic antimicrobial peptides. Crit Rev Microbiol 2013, 39 (2), 180-95. 33. LaRock, C. N.; Nizet, V., Cationic antimicrobial peptide resistance mechanisms of streptococcal pathogens. Biochim Biophys Acta 2015, 1848 (11 Pt B), 3047-54. 34. Koo, H. B.; Seo, J., Antimicrobial peptides under clinical investigation. Peptide Science 2019, 111 (5), e24122.

35. Malanovic, N.; Lohner, K., Antimicrobial Peptides Targeting Gram-Positive Bacteria. Pharmaceuticals 2016, 9 (3). 36. Falanga, A.; Lombardi, L.; Franci, G.; Vitiello, M.; Iovene, M. R.; Morelli, G.; Galdiero, M.; Galdiero, S., Marine antimicrobial peptides: nature provides templates for the design of novel compounds against pathogenic bacteria. Int J Mol Sci 2016, 17 (5), 785. 37. Mohamed, M. F.; Abdelkhalek, A.; Seleem, M. N., Evaluation of short synthetic antimicrobial peptides for treatment of drug-resistant and intracellular Staphylococcus aureus. Sci Rep 2016, 6, 29707. 38. Li, J.; Koh, J.-J.; Liu, S.; Lakshminarayanan, R.; Verma, C. S.; Beuerman, R. W., Membrane active antimicrobial peptides: translating mechanistic insights to design. Front Neurosci 2017, 11, 73. 39. Mahlapuu, M.; Håkansson, J.; Ringstad, L.; Björn, C., Antimicrobial peptides: an emerging category of therapeutic agents. Front Cell Infect Microbiol 2016, 6, 194. 40. Arenas, I.; Villegas, E.; Walls, O.; Barrios, H.; Rodríguez, R.; Corzo, G., Antimicrobial activity and stability of short and long based arachnid synthetic peptides in the presence of commercial antibiotics. Molecules 2016, 21 (2), 225. 41. Seo, M.-D.; Won, H.-S.; Kim, J.-H.; Mishig-Ochir, T.; Lee, B.-J., Antimicrobial peptides for therapeutic applications: a review. Molecules 2012, 17 (10), 12276-12286. 42. Mojsoska, B.; Jenssen, H., Peptides and peptidomimetics for antimicrobial drug design. Pharmaceuticals 2015, 8 (3), 366-415. 43. E Greber, K.; Dawgul, M., Antimicrobial peptides under clinical trials. Curr Topics Med Chem 2017, 17 (5), 620-628. 44. Fosgerau, K.; Hoffmann, T., Peptide therapeutics: current status and future directions. Drug Discovery Today 2015, 20 (1), 122-128. 45. Libardo, M. D.; Paul, T. J.; Prabhakar, R.; Angeles-Boza, A. M., Hybrid peptide ATCUN- sh-Buforin: Influence of the ATCUN charge and stereochemistry on antimicrobial activity. Biochimie 2015, 113, 143-55. 46. Avan, I.; Hall, C. D.; Katritzky, A. R., Peptidomimetics via modifications of amino acids and peptide bonds. Chem Soc Rev 2014, 43 (10), 3575-3594. 47. Andreini, C.; Bertini, I.; Cavallaro, G.; Holliday, G. L.; Thornton, J. M., Metal ions in biological catalysis: from enzyme databases to general principles. J Biol Inorg Chem 2008, 13 (8), 1205-18. 48. Ma, L.; Terwilliger, A.; Maresso, A. W., Iron and zinc exploitation during bacterial pathogenesis. Metallomics 2015, 7 (12), 1541-54. 49. Fogaca, A. C.; Lorenzini, D. M.; Kaku, L. M.; Esteves, E.; Bulet, P.; Daffre, S., Cysteine- rich antimicrobial peptides of the cattle tick Boophilus microplus: isolation, structural characterization and tissue expression profile. Dev Comp Immunol 2004, 28 (3), 191-200. 50. Esteves, E.; Fogaca, A. C.; Maldonado, R.; Silva, F. D.; Manso, P. P.; Pelajo-Machado, M.; Valle, D.; Daffre, S., Antimicrobial activity in the tick Rhipicephalus (Boophilus) microplus eggs: Cellular localization and temporal expression of microplusin during oogenesis and embryogenesis. Dev Comp Immunol 2009, 33 (8), 913-9. 51. Silva, F. D.; Rezende, C. A.; Rossi, D. C.; Esteves, E.; Dyszy, F. H.; Schreier, S.; Gueiros- Filho, F.; Campos, C. B.; Pires, J. R.; Daffre, S., Structure and mode of action of microplusin, a copper II-chelating antimicrobial peptide from the cattle tick Rhipicephalus (Boophilus) microplus. J Biol Chem 2009, 284 (50), 34735-46.

52. Kligman, D.; Hilt, D. C., The S100 protein family. Trends Biochem Sci 1988, 13 (11), 437- 443. 53. Brodersen, D. E.; Nyborg, J.; Kjeldgaard, M., Zinc-Binding Site of an S100 Protein Revealed. Two Crystal Structures of Ca2+-Bound Human Psoriasin (S100A7) in the Zn2+-Loaded and Zn2+-Free States. Biochemistry 1999, 38 (6), 1695-1704. 54. Glaser, R.; Harder, J.; Lange, H.; Bartels, J.; Christophers, E.; Schroder, J.-M., Antimicrobial psoriasin (S100A7) protects human skin from Escherichia coli infection. Nat Immunol 2005, 6 (1), 57-64. 55. Brophy, M. B.; Hayden, J. A.; Nolan, E. M., Calcium ion gradients modulate the zinc affinity and antibacterial activity of human calprotectin. J Am Chem Soc 2012, 134 (43), 18089-100. 56. Gagnon, D. M.; Brophy, M. B.; Bowman, S. E.; Stich, T. A.; Drennan, C. L.; Britt, R. D.; Nolan, E. M., Manganese binding properties of human calprotectin under conditions of high and low calcium: X-ray crystallographic and advanced electron paramagnetic resonance spectroscopic analysis. J Am Chem Soc 2015, 137 (8), 3004-16. 57. Nakashige, T. G.; Stephan, J. R.; Cunden, L. S.; Brophy, M. B.; Wommack, A. J.; Keegan, B. C.; Shearer, J. M.; Nolan, E. M., The Hexahistidine Motif of Host-Defense Protein Human Calprotectin Contributes to Zinc Withholding and Its Functional Versatility. J Am Chem Soc 2016, 138 (37), 12243-51. 58. Economou, N. J.; Cocklin, S.; Loll, P. J., High-resolution crystal structure reveals molecular details of target recognition by bacitracin. Proc Natl Acad Sci 2013, 110 (35), 14207. 59. Ming, L. J.; Epperson, J. D., Metal binding and structure-activity relationship of the metalloantibiotic peptide bacitracin. J Inorg Biochem 2002, 91 (1), 46-58. 60. Selzer, G. B., Heavy metals in antibiotics. Antibiotic Chemother 1956, 6 (8), 498-9. 61. Storm, D. R.; Strominger, J. L., Complex formation between bacitracin peptides and isoprenyl pyrophosphates. The specificity of lipid-peptide interactions. J Biol Chem 1973, 248 (11), 3940- 5. 62. Schittek, B.; Hipfel, R.; Sauer, B.; Bauer, J.; Kalbacher, H.; Stevanovic, S.; Schirle, M.; Schroeder, K.; Blin, N.; Meier, F.; Rassner, G.; Garbe, C., Dermcidin: a novel human antibiotic peptide secreted by sweat glands. Nat Immunol 2001, 2 (12), 1133-1137. 63. Paulmann, M.; Arnold, T.; Linke, D.; Ozdirekcan, S.; Kopp, A.; Gutsmann, T.; Kalbacher, H.; Wanke, I.; Schuenemann, V. J.; Habeck, M.; Burck, J.; Ulrich, A. S.; Schittek, B., Structure- activity analysis of the dermcidin-derived peptide DCD-1L, an anionic antimicrobial peptide present in human sweat. J Biol Chem 2012, 287 (11), 8434-43. 64. Song, C.; Weichbrodt, C.; Salnikov, E. S.; Dynowski, M.; Forsberg, B. r. O.; Bechinger, B.; Steinem, C.; de Groot, B. L.; Zachariae, U.; Zeth, K., Crystal structure and functional mechanism of a human antimicrobial membrane channel. Proc Natl Acad Sci 2013, 110 (12), 4586-4591. 65. Harford, C.; Sarkar, B., Amino Terminal Cu(II)- and Ni(II)-Binding (ATCUN) Motif of Proteins and Peptides: Metal Binding, DNA Cleavage, and Other Properties. Acc Chem Res 1997, 30 (3), 123-130. 66. Brewer, D.; Lajoie, G., Evaluation of the metal binding properties of the histidine-rich antimicrobial peptides histatin 3 and 5 by electrospray ionization mass spectrometry. Rapid Commun Mass Spectrometry 2000, 14 (19), 1736-1745. 67. Helmerhorst, E. J.; Troxler, R. F.; Oppenheim, F. G., The human salivary peptide histatin 5 exerts its antifungal activity through the formation of reactive oxygen species. Proc Natl Acad Sci 2001, 98 (25), 14637-42.

68. Libardo, M. D. J.; Gorbatyuk, V. Y.; Angeles-Boza, A. M., Central Role of the Copper- Binding Motif in the Complex Mechanism of Action of Ixosin: Enhancing Oxidative Damage and Promoting Synergy with Ixosin B. ACS Infect Dis 2016, 2 (1), 71-81. 69. Yu, D.; Sheng, Z.; Xu, X.; Li, J.; Yang, H.; Liu, Z.; Rees, H. H.; Lai, R., A novel antimicrobial peptide from salivary glands of the hard tick, Ixodes sinensis. Peptides 2006, 27 (1), 31-5. 70. Silphaduang, U.; Noga, E. J., Peptide antibiotics in mast cells of fish. Nature 2001, 414 (6861), 268-9. 71. Campagna, S.; Saint, N.; Molle, G.; Aumelas, A., Structure and mechanism of action of the antimicrobial peptide piscidin. Biochemistry 2007, 46 (7), 1771-8. 72. Hayden, R. M.; Goldberg, G. K.; Ferguson, B. M.; Schoeneck, M. W.; Libardo, M. D.; Mayeux, S. E.; Shrestha, A.; Bogardus, K. A.; Hammer, J.; Pryshchep, S.; Lehman, H. K.; McCormick, M. L.; Blazyk, J.; Angeles-Boza, A. M.; Fu, R.; Cotten, M. L., Complementary Effects of Host Defense Peptides Piscidin 1 and Piscidin 3 on DNA and Lipid Membranes: Biophysical Insights into Contrasting Biological Activities. J Phys Chem B 2015, 119 (49), 15235- 46. 73. Libardo, M. D. J.; Bahar, A. A.; Ma, B.; Fu, R.; McCormick, L. E.; Zhao, J.; McCallum, S. A.; Nussinov, R.; Ren, D.; Angeles-Boza, A. M.; Cotten, M. L., Nuclease activity gives an edge to host-defense peptide piscidin 3 over piscidin 1, rendering it more effective against persisters and biofilms. FEBS J 2017, 284 (21), 3662-3683. 74. Stanford, U. Mineral Makeup of Seawater. https://seaurchineducation.stanford.edu/mineral (accessed 12/7/19). 75. Rinkevich, B., Primitive immune systems: Are your ways my ways? Immunol Rev 2004, 198 (1), 25-35. 76. Dishaw, L.; Flores-Torres, J.; Mueller, M. G.; Karrer, C.; Skapura, D.; Melillo, D.; Zucchetti, I.; De Santis, R.; Pinto, M. R.; Litman, G., A Basal Chordate Model for Studies of Gut Microbial Immune Interactions. Front Immunol 2012, 3, 96. 77. Lemaire, P., Evolutionary crossroads in developmental biology: the tunicates. Development 2011, 138 (11), 2143. 78. Juliano, S. A.; Pierce, S.; deMayo, J. A.; Balunas, M. J.; Angeles-Boza, A. M., Exploration of the Innate Immune System of Styela clava: Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide Clavanin A. Biochemistry 2017, 56 (10), 1403-1414. 79. Lee, I. H.; Zhao, C.; Cho, Y.; Harwig, S. S.; Cooper, E. L.; Lehrer, R. I., Clavanins, alpha- helical antimicrobial peptides from tunicate hemocytes. FEBS Lett 1997, 400 (2), 158-62. 80. Zhao, C.; Liaw, L.; Lee, I. H.; Lehrer, R. I., cDNA cloning of Clavanins: antimicrobial peptides of tunicate hemocytes. FEBS Lett 1997, 410 (2-3), 490-2. 81. Lee, I. H.; Zhao, C.; Nguyen, T.; Menzel, L.; Waring, A. J.; Sherman, M. A.; Lehrer, R. I., Clavaspirin, an antibacterial and haemolytic peptide from Styela clava. J Peptide Res 2001, 58 (6), 445-56. 82. van Kan, E. J.; van der Bent, A.; Demel, R. A.; de Kruijff, B., Membrane activity of the peptide antibiotic clavanin and the importance of its glycine residues. Biochemistry 2001, 40 (21), 6398-405. 83. Saude, A. C.; Ombredane, A. S.; Silva, O. N.; Barbosa, J. A.; Moreno, S. E.; Araujo, A. C.; Falcao, R.; Silva, L. P.; Dias, S. C.; Franco, O. L., Clavanin bacterial sepsis control using a novel methacrylate nanocarrier. Int J Nanomedicine 2014, 9, 5055-69.

84. Ribeiro, K. L.; Frias, I. A. M.; Franco, O. L.; Dias, S. C.; Sousa-Junior, A. A.; Silva, O. N.; Bakuzis, A. F.; Oliveira, M. D. L.; Andrade, C. A. S., Clavanin A-bioconjugated Fe3O4/Silane core-shell nanoparticles for thermal ablation of bacterial biofilms. Colloids and surfaces. B, Biointerfaces 2018, 169, 72-81. 85. Lehrer, R. I.; Lee, I. H.; Menzel, L.; Waring, A.; Zhao, C., Clavanins and styelins, alpha- helical antimicrobial peptides from the hemocytes of Styela clava. Adv Exp Med Biol 2001, 484, 71-6. 86. Lee, I. H.; Cho, Y.; Lehrer, R. I., Effects of pH and salinity on the antimicrobial properties of clavanins. Infect Immun 1997, 65 (7), 2898-903. 87. van Kan, E. J.; Demel, R. A.; Breukink, E.; van der Bent, A.; de Kruijff, B., Clavanin permeabilizes target membranes via two distinctly different pH-dependent mechanisms. Biochemistry 2002, 41 (24), 7529-39. 88. van Kan, E. J.; Demel, R. A.; van der Bent, A.; de Kruijff, B., The role of the abundant phenylalanines in the mode of action of the antimicrobial peptide clavanin. Biochim Biophys Acta 2003, 1615 (1-2), 84-92. 89. Menzel, L. P.; Lee, I. H.; Sjostrand, B.; Lehrer, R. I., Immunolocalization of clavanins in Styela clava hemocytes. Dev Comp Immuno 2002, 26 (6), 505-515. 90. Silva, O. N.; Fensterseifer, I. C.; Rodrigues, E. A.; Holanda, H. H.; Novaes, N. R.; Cunha, J. P.; Rezende, T. M.; Magalhaes, K. G.; Moreno, S. E.; Jeronimo, M. S.; Bocca, A. L.; Franco, O. L., Clavanin A improves outcome of complications from different bacterial infections. Antimicrob Agents Chemother 2015, 59 (3), 1620-6. 91. Silva, O. N.; Alves, E. S.; de la Fuente-Nunez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A.; Andrade, C. A.; Nascimento, J. M.; Fensterseifer, I. C.; Porto, W. F.; Correa, J. R.; Hancock, R. E.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti-Biofilm Activity. Sci Rep 2016, 6, 27128.

Chapter 2 – Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate

Peptide Clavanin A

Adapted from: Juliano, S. A.; Pierce, S.; deMayo, J. A.; Balunas, M. J.; Angeles-Boza, A. M., Exploration of the Innate Immune System of Styela clava: Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide Clavanin A. Biochemistry 2017, 56 (10), 1403-1414.

2.1 – Introduction

The importance of zinc in immunity is well known; however, despite the considerable advances of recent years, many fundamental aspects of the immunomodulatory and antimicrobial properties of zinc at the cellular and molecular levels have remained unanswered.1-2 Zinc occurs as the divalent cation Zn2+ and together with iron and copper is one of the most abundant transition metals in humans.3 Zinc plays a multitude of roles within human immunology but, not surprisingly, the complexity of the human immune system has not allowed researchers to fully unravel all of the biochemical roles of zinc.4-6 Since the complex role of zinc ions in the human immune system probably evolved from the innate immune system of ancestors in the taxonomic scale, understanding the role of zinc in the immune response of these ancestors could provide explanations to its role in humans. Tunicates, like humans, are members of the phylum chordata, which primarily includes vertebrates, and have the capacity for cell mediated immunity and humoral immunity.7-9 Importantly, tunicates have been used as primitive models for understanding fundamental immunological mechanisms by analyzing their processes either in vivo or in vitro.10-11 Thus, it can be hypothesized that by tracing the evolutionary history of zinc interactions at the molecular level it should be possible to elucidate mechanisms of zinc contributions to the immunological response in the complex human immune system.

Key participants in the immune response are also the molecules known as antimicrobial peptides (AMPs) which can be expressed constitutively or induced by the presence of pathogens.12-13 AMPs are evolutionarily ancient weapons and are found in all complex living organisms. In addition to their ability to directly kill microorganisms, AMPs exert immunomodulatory activities – which has led researchers in recent years to refer to AMPs by the more general term host defense peptides (HDPs).14-15 The chemical characteristics of AMPs (i.e.

31 generally cationic and the presence of about 50% hydrophobic amino acids) allow them to strongly interact with cellular membranes, particularly those of bacteria. Therefore, many AMPs were suggested to kill bacteria through fatal interaction with the components of the bacterial membrane.16-21 More recently, internalization of AMPs with the concomitant damage of an intracellular target has been uncovered for some peptides.22-25

Our group has been particularly interested in the effect of the binding of metal ions on the activity of AMPs.26-27 Because zinc and copper ions co-localize with antimicrobial peptides during immune responses,28-35 we have hypothesized that AMPs with metal-binding motifs can form either transient or stable complexes with these metal ions to improve their antimicrobial activity. Herein, we investigate the effect of zinc ions on the activity of an AMP isolated from the tunicate Styela clava. These investigations are strongly motivated by the search for molecular models to characterize fundamental immunological roles of zinc.

S. clava, a solitary tunicate, has been shown to express several AMPs that are important components of its innate immune system.36-39 Extracts from S. clava hemocytes have produced two families of AMPs, styelins and clavanins.40-42 Immunolocalization studies found clavanins in the cytoplasm of macrophages and in eosinophilic granulocytes.39 In vitro assays have shown that one member of the clavanin family, clavanin A (ClavA,

VFQFLGKIIHHVGNFVHGFSHVF-NH2), is active against a broad spectrum of pathogens, including methicillin-resistant Staphylococcus aureus, and its activity is greater at pH 5.5 than at pH 7.4.36 The activity at low pH is biologically relevant since the vacuoles of phagocytic hemocytes from S. clava are likely to have a pH as low as 5 as found in the tunicate Ascidia ceratodes.43 Such conditions should support optimal antimicrobial activity. Unlike most AMPs,

ClavA remains active at elevated NaCl concentrations (up to 0.3 M).36 In addition, ClavA

32 displays a negligible toxicity towards mammalian cells. Because of these properties, ClavA has been tested in vitro in wound and sepsis murine models and was shown to have potent activity in both models.44

Due to its similarity to the magainins in size, primary sequence, and antimicrobial activity, ClavA is thought to act by permeabilizing bacterial membranes.16, 45-46 In the α-helical conformation, ClavA places His17 and His21 on one face of the helix creating a putative Zn2+

47 binding site (henceforth referred to as the HX3H motif). The i, i + 4 spacing between two amino acids within an α-helix have been previously exploited to design synthetic metal-binding motifs.47-49 Specifically, two histidines in positions 7 and 11 in a 16-residue peptide were used to create a synthetic zinc-based nuclease.47 Coincidentally, hemocytes of aquatic invertebrates contain large amounts of metal ions, particularly Zn2+, at concentrations higher than in the surrounding plasma.50-52 For example, in Ostrea edulis, specialized mechanisms allow for hemocyte uptake of Zn2+ ions up to a 1.2 M concentration.53 Taken together, the presence of a potential Zn2+ binding site and the large concentrations of this metal ion in marine invertebrate hemocytes begs the question whether a relationship exists between metal binding and antimicrobial ability in the activities of ClavA.

In this work, we examine the antimicrobial activity of ClavA in the presence of Zn2+ ions.

We determined that the target of this peptide-Zn2+ complex is located within the cytoplasm, in addition to the previously reported membrane activity. Our data supports a model where ClavA-

Zn2+ uses its nucleolytic properties to kill bacteria.

2.2 – Results

The antimicrobial activity of ClavA is potentiated by the addition of Zn2+. In order to examine whether Zn2+ ions affect the activity of ClavA, we determined its MIC against

Escherichia coli in the presence of Zn2+ and various biologically relevant metal ions using the

54 broth microdilution method. Magainin 2 (GIGKFLHSAKKFGKAFVGEIMNS-NH2) and

Clavanin C (ClavC, VFHLLGKIIHHVGNFVYGFSHVF-NH2) were used for comparison as well as positive controls. Table 2-1 summarizes the MIC values for ClavA and ClavA incubated with the various biologically relevant metals. The activity of ClavA is maintained upon addition of Fe2+ and Ni2+ ions, abolished with Ca2+, and improved after adding Cu2+ and Zn2+ ions. As we hypothesized, ClavA in the presence of Zn2+ showed the greatest increase in activity, with a 16- fold increase, effectively making ClavA-Zn2+ as active as ClavC, the most potent peptide in the clavanin family.55 ClavC, on the other hand, was found to remain unaffected by the addition of any of the metal ions reported in this study.

Table 2-1

Minimum inhibitory concentrations for ClavA with various metal ions a

Treatment MIC

ClavA 64 μM

2+ ClavA + Ca >128 μM

ClavA + Fe2+ 64 μM

ClavA + Ni2+ 64 μM

ClavA + Cu2+ 32 μM

ClavA + Zn2+ 4 μM

a All metal ion concentrations are 2X that of the peptide. The MICs of the

metals used have all been found to be >256 μM, the highest concentration employed

in this study.

We hypothesized that the increase in activity for ClavA when combined with Zn2+ ions was the result of the presence of His17 and His21 in its sequence (representing the proposed

HX3H motif). To explore this, mutants of ClavA were synthesized containing substitutions to one or both of these potentially important histidines. The H17A mutant, ClavA-H17A, and the double mutant, ClavA-H17A-H21A, had lower activity in the absence of additional Zn2+ ions, whereas ClavA-H21A remained as active as the wild type. Depending on the position of histidine to alanine substitutions, an increase of the helical content of peptides can be observed,56 which in turn can result in an increase of antimicrobial activity.27 Circular dichroism shows that the three His to Ala mutants have a larger percentage of helicity as compared to ClavA (Figure

2-1), but ClavA-H21A has the lowest increase in helicity. Because ClavA-H17A and ClavA-

H17A-H21A are less active than ClavA-H21A, helicity increase is not the reason for the better antimicrobial activity in ClavA-H21A as compared to the other two mutants. Notably, marked differences were observed when the peptides were combined with zinc ions. Table 2-2 summarizes these results as well as the activities of the other peptides used in this study in the presence and absence of Zn2+. ClavA-H17A and ClavA-H17A-H21A remained inactive, whereas

ClavA-H21A showed a modest two-fold increase in activity in the presence of additional Zn2+.

These results suggest that His17 is more important for the formation of a putative ClavA-Zn2+ complex (this is not necessarily a 1:1 complex as the nomenclature might suggest; however, we prefer to use this notation until we learn more about the nature of this complex).

Figure 2-1

Circular Dichroism spectra of ClavA, ClavA-H17A, ClavA-H21A and ClavA-H17A-

H21A in 50% 2,2,2-trifluoroethanol 50% MES 5.5 to show the increased helicity of the

alanine substituted mutants. The increase in α-helicity can be observed as a decrease in

the minima at 208 nm and 222 nm.

We also determined MBC values. ClavA had an MBC equal to its MIC, which means that the growth inhibition seen at that concentration was due to the peptide bringing about cell death. In the presence of Zn2+, however, the MBC was 8 μM, twice the MIC, meaning that at

MIC concentrations of ClavA-Zn2+ there were still living E. coli cells that were in some way inhibited from growing by the peptide. The 8-fold decrease in MBC between ClavA and ClavA-

Zn2+ indicates an improvement in the overall bactericidal activity of ClavA upon addition of Zn2+ ions.

Table 2-2

Minimum inhibitory concentrations for ClavA, ClavA Mutants, ClavC and Magainin

2 without and with Zn2+ a

Treatment MIC without Zn2+ MIC with Zn2+

ClavA 64 μM 4 μM

ClavA-H17A >128 μM >128 μM

ClavA-H21A 64 μM 32 μM

ClavA-H17A-H21A >128 μM >128 μM

ClavC 4 μM 4 μM

Magainin 2 8 μM 8 μM

a Zinc ion concentrations are 2X that of the peptide. The MIC of Zn2+ has been found to be

>256 μM, the highest concentration employed in this study.

ClavA shows improved activity even at sub-stoichiometric Zn2+ concentrations. In order to look at how the concentration of Zn2+ affects the activity of ClavA, we incubated bacteria with 8 μM ClavA and varied the concentration of Zn2+ present. Figure 2-2 summarizes these results. Percent viability was determined by plating dilutions of each sample, counting colonies, and comparing to an untreated control. All of the concentrations of Zn2+ used in the study, from 16 μM to 2 μM, resulted in clearance of a vast majority of the E. coli, and growth was only present in samples which contained no additional Zn2+ because the concentration of

ClavA used was below the MBC for the peptide alone. These results indicate that even at very

38 low levels of Zn2+, ClavA activity is greatly improved. This could indicate that one Zn2+ ion interacts with several ClavA units, though more study is required for definitive proof of this.

Figure 2-2

Effect of varied Zn2+ concentration on the antimicrobial activity of 8μM ClavA. These data are

plotted with a logarithmic scale on the y-axis. Values represent the average of three trials ±

standard error from mean (*, P≤0.05; N.S. =not significant).

Styela clava hemolymph contains a sufficient concentration of Zn2+ for biological relevance. We next evaluated whether the Zn2+ concentration that was observed to increase the activity of ClavA in the antimicrobial assays was biologically relevant. For this purpose, we harvested live S. clava and determined the concentration of Zn2+ ions in the hemolymph, a fluid found in invertebrates which serves as a blood analogue, using atomic absorption spectroscopy.

The concentration of Zn2+ ions was determined to be 26.7 ± 7.1 μM, indicating that enough Zn2+

39 ions are found in the hemolymph to potentiate the activity of ClavA. Some percentage of these ions are likely to be tightly bound to proteins in the hemolymph or hemocytes; however, considering the fact that the viability of E. coli was <1% at c.a. 10% of the total measured tunicate Zn2+ concentration (Figure 2-2), it is likely that the labile pool of zinc ions is large enough to potentiate ClavA in vivo. As in many aquatic invertebrates, S. clava hemocytes likely contain greater concentrations of Zn2+ ions than the surrounding sea water due to the presence of special mechanisms that accumulate Zn2+.53

Circular dichroism shows that Zn2+ binding results in a conformational change in

ClavA. To determine whether ClavA is binding the Zn2+ or not, circular dichroism (CD) was performed on a series of titrations of a 50 μM solution of ClavA with Zn2+ in MES 5.5. As can be seen in Figure 2-3, there is a distinct concentration dependent change in the CD spectra upon the addition of Zn2+. Overall, it appears that the percent α-helix increases with increasing Zn2+ concentration. These results, coupled with the observed change in activity upon Zn2+ incubation seem to indicate that Zn2+ binding is occurring and some sort of ClavA-Zn2+ complex is being formed.

Figure 2-3

CD spectrum of 50 μM ClavA titrated with Zn2+ in MES buffer (pH = 5.5). The changes

observed upon the addition of Zn2+ seem to indicate that binding of the ion is occurring. The

helicity increases from 4.2 % for ClavA alone to 9.6 % in the presence of 250 μM Zn2+.

ClavA exerts its bactericidal activity faster in the presence of Zn2+. We next sought to determine the rate at which ClavA and ClavA-Zn2+ killed bacteria. To do this, we performed a time-kill kinetics study by counting viable colonies from a bacterial suspension in a solution of

ClavA or ClavA-Zn2+. Figure 2-4 shows that when both ClavA and ClavA-Zn2+ are at 64 μM,

ClavA-Zn2+ (red) causes total clearance of the E. coli after about 2 hr, whereas apo-ClavA (blue) does not completely clear out the culture at four hours of treatment. These data imply that the killing mechanism of ClavA in the presence of Zn2+ is faster than that of the peptide alone.

ClavA with Zn2+ at its MBC appears to be bactericidal at a lower rate than ClavA alone; however, it does so at 12.5% of the concentration of ClavA. ClavA-Zn2+ at the MIC value appears to only slightly inhibit the growth of E. coli over buffer alone. Although a horizontal or

41 downward curve is expected for a ClavA-Zn2+ at its MIC, upward trending curves have been observed for other antimicrobial agents at their MIC values.57-58 It is also plausible that the upward curve for ClavA-Zn2+ at its MIC is due to matrix effects coming from the media. LB media contains tryptone, which is a partial digest of casein by the protease trypsin. This digest leaves behind oligopeptides from casein, which are known to bind to Zn2+,59 and could deprive

ClavA of zinc ions, reducing its overall efficacy.

Figure 2-4

Time kill kinetics data for ClavA, ClavA-Zn2+, and Magainin 2. [ClavA] = 64 μM, the MIC

and MBC values for the peptide. MIC and MBC values for ClavA-Zn2+ are 4 μM and 8 μM,

respectively. Magainin 2 was used at its MIC, 8 μM. MES 5.5 represents an untreated control.

Each point represents mean ± standard error from mean. In some cases, error bars are too small

to be visible.

ClavA-Zn2+ causes membrane depolarization faster and at a lower concentration than ClavA. ClavA has been shown to cause depolarization of the lipid membrane of bacteria.46,

60 In light of this, we next sought to determine the effects of Zn2+ ions on the membrane depolarization activity of ClavA through the use of the membrane potential sensitive probe,

DiSC3(5). When a cell is undamaged, DiSC3(5) is drawn to the membrane and its fluorescence is self-quenched. If the peptide attack depolarizes the cell membrane, the dye is no longer drawn to the membrane and dissociates, causing DiSC3(5) to fluoresce. The amount of fluorescence recovered after treatment is proportional to the amount of membrane depolarization. B. subtilis was used for these experiments because Gram-positive bacteria do not require special pretreatment, whereas Gram-negative bacteria require pretreatment with EDTA which would have chelated the Zn2+. A 50% solution of the nonionic surfactant T-X100 was used as a positive control, causing complete depolarization of the bacterial membrane, resulting in a 12 to 15–fold increase in fluorescence.

We first looked at the kinetics of membrane depolarization (Figure 2-5A) and found that

ClavA-Zn2+ achieves maximum depolarization faster than ClavA alone. Despite the initial rate of depolarization being higher for ClavA-Zn2+ than for the peptide alone, both reach the same amount of depolarization after one hour. We also wanted to explore the concentration dependence of depolarization and potentially see if ClavA-Zn2+ was causing depolarization at a lower concentration. As can be seen in figure 2-5B, ClavA-Zn2+ reaches maximum depolarization between 8 and 16 μM whereas it takes 64 μM ClavA to reach the same level. This result is consistent with MIC and MBC measurements for the ClavA and ClavA-Zn2+. It is possible that the overall improved membrane depolarization activity is due to stronger interactions with the bacterial membrane as a result of Zn2+ histidine binding as described by

Kurut and coworkers.61 When these data are taken in conjunction with the result of the time kill kinetics study, it becomes clear that not only does the addition of Zn2+ improve the activity of

ClavA, it also shortens the time required for ClavA to bring about cell damage.

Figure 2-5

B. subtilis membrane depolarization for ClavA and ClavA-Zn2+. (A) Depolarization as a

function of time. Error bars were removed for clarity. (B) Depolarization as a function of

peptide concentration.

ClavA does not cause pores large enough to allow β-galactosidase leakage. To determine the cause of the observed membrane depolarization we investigated the formation of pores. For this experiment, β-galactosidase was overexpressed in E. coli by exposure to IPTG.

Cells were then incubated with either ClavA, ClavA-Zn2+, or T-X100, and β-galactosidase leakage was measured in the supernatant through cleavage of ONPG, which produces a yellow product.62 ClavA causes about 35% β-galactosidase leakage and ClavA-Zn2+ causes about 20% relative to T-X100 (Figure 2-6). This low-level leakage is most likely due to death of the cells,

44 rather than pore formation, considering cell killing has been observed to occur within one hour

(Figure 2-4). Large stable pores would be required to cause β-galactosidase leakage, and no significant leakage is observed. As a result of these data it can be concluded that stable pore formation is not the cause of the increased depolarization observed with ClavA-Zn2+. However, this does not entirely eliminate membrane poration as a potential cause for the observed depolarization. AMPs have been shown to cause membrane damage through the formation of transient or disordered toroidal pores, which only requires small quantities of peptide, with the pore walls being formed by lipid head groups.63-64 These transient smaller pores have been shown to work similarly to ion channels, causing depolarization by abolishing membrane potential, and their presence can often be detected by an increase in potassium ion concentration in solution.65

Figure 2-6

β-galactosidase leakage from E. coli cells as caused by ClavA and ClavA-Zn2+. T-X100 was set

to 100% for each trial. Each value represents mean ± standard error from mean.

Potassium efflux is not the sole cause of membrane depolarization and does not cause cell death. Previous work performed by van Kan and coworkers provides evidence that

ClavA is capable of causing membrane poration and K+ leakage.46 We aimed to explore the difference in depolarization rates we observed for ClavA and ClavA-Zn2+ and identify whether it was a function of some difference in the K+ leakage between the two treatments. As can be seen in Figure 2-7, both ClavA and ClavA-Zn2+ cause potassium leakage. The fact that the data presented here differ from those published by van Kan et al. likely originate from differences between Lactobacillus sake and B. subtilis. ClavA and ClavA-Zn2+ cause K+ leakage at relatively similar rates despite the fact that the time-dependent depolarization results (Figure 2-

5A) indicate that ClavA-Zn2+ reaches maximum depolarization of the bacterial membrane

46 roughly 30 min faster than ClavA without Zn2+. This would seemingly indicate that some other mechanism of attack is employed by ClavA-Zn2+, causing depolarization separate from the K+ leakage activity. The depolarization and K+ leakage, however, are likely not bringing about cell death because the time kill kinetics data (Figure 4) show that at one hour there is roughly 70% survival in the case of ClavA and that ClavA-Zn2+ is bacteriostatic at the concentration used for

K+ leakage and depolarization.

Figure 2-7

Potassium leakage from B. subtilis cells as caused by ClavA and ClavA-Zn2+ at their respective

MICs. Each sample was sonicated to determine 100% potassium content. Each point represents

mean ± standard error from mean.

ClavA has different localization in E. coli cells with and without Zn2+ present. To help us elucidate the cause of membrane depolarization, we used fluorescence microscopy.

ClavA is proposed to be a membrane active peptide; however, no microscopy studies have been

47 done to either prove or disprove these claims. A derivative of ClavA was synthesized with a

5(6)-carboxytetramethylrhodamine (TMR) label on a lysine residue added to the peptide sequence instead of Phe4 (Figure 2-8), and this derivative (ClavA-TMR) was used for confocal microscopy studies. ClavA-TMR was first confirmed to be functionally similar to ClavA by obtaining MIC values for the fluorescent peptide in the presence and absence of Zn2+. As can be seen in Table 2-3, ClavA-TMR and ClavA-TMR-Zn2+ have MICs which are close to those of the wild type peptide, and in both cases, the peptides incubated with Zn2+ have 16-fold lower MIC values. This provides evidence that the addition of the fluorescent label did not significantly alter the activity of ClavA.

Table 2-3

Minimum inhibitory concentrations for ClavA and ClavA-TMR in the presence

and absence of Zn2+ions a

Treatment MIC without Zn2+ MIC with Zn2+

ClavA 64 μM 4 μM

ClavA-TMR 32 μM 2 μM

a Zinc ion concentrations are 2X that of the peptide. The MIC of Zn2+ has been

found to be >256 μM, the highest concentration employed in this study.

Figure 2-8

Schematic representation of the addition of 5(6)-carboxytetramethylrhodamine (TMR) to ClavA.

A mutant sequence of ClavA (VFQFLGKIIHHVGNFVHGFSHVF-NH2) containing a lysine

residue protected by a 4-methyltrityl group (Mtt) instead of Phe4 was used. We hypothesized that

replacing a hydrophobic residue by the amino acid containing the hydrophobic TMR label will

have the lowest impact on the activity of the peptide. Reaction conditions are as follows: (i) Lys-

Mtt deprotection with 1% trifluoroacetic acid (TFA) in dichloromethane (DCM); (ii) addition of

TMR to the ε-amino group of Lys was performed using four times excess of TMR, 2-(1H-

benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexaﬂuorophosphate (HBTU), and

diisopropylethylamine (DIEA) dissolved in dimethylformamide (DMF); (iii) cleavage of the

peptide from the resin using a mixture of 95% TFA, 2.5% H2O and 2.5% triisopropylsilane (TIS).

The microscopy experiments were performed at 8 μM peptide to avoid saturation of the fluorescence emission from TMR. The micrographs of ClavA-TMR attack (Figure 2-9A and 2-

9B) show E. coli with peptide internalized after 1-hour incubation. This observation in combination with the time kill kinetic results hints at a possible mechanism involving an internal target for ClavA. This is somewhat contrary to literature findings, which strongly suggest that

ClavA targets the bacterial membrane.45-46 In addition to internalization, we also observed some bacterial membrane damage in roughly 10% of the cells observed. For example, in figure 2-9B, the bacterium in the lower right corner appears to have distortion in the membrane (black arrows), an indication of damage, most likely from the initial attack of the peptide. These spots of damage are reminiscent of the blebs caused by human defensin 5, though the resolution of the microscope is not high enough to make them out for certain, and more powerful techniques would be needed to confirm the presence of blebs.18 If membrane damage is occurring, this could explain why previous studies report membrane activity as ClavA damages the membrane on the way to an internal target, likely DNA. Images of ClavA-TMR-Zn2+ attack (Figure 2-9C and 2-

9D) show that, in roughly 75-80% of the E. coli imaged, the peptide accumulates primarily on the membrane (Figure 2-9C). We propose that this accumulation is likely due to the formation of a peptide-Zn2+-lipid interaction and eventually promotes internalization of the ClavA. The remaining 20-25% of the cells incubated with ClavA-TMR-Zn2+ showed internalization of the complex (Figure 2-7D), with images appearing similar to what was seen for ClavA-TMR. This could again point to an internal target for ClavA. ClavA-Zn2+ also has the interesting feature of creating bright loci on the membranes of about 60% of the cells imaged (Figure 2-9C and 2-9D, white arrows). Of the cells that contain these bright loci there is only one locus per cell. These loci appear to be concentrated regions of peptide and not a feature of the bacterial membrane due to the fact that in spots where a bright locus is found there is no evidence of a damaged membrane in the DIC. Overall, the microscopy studies, combined with the leakage studies reported above, indicate that the observed depolarization probably arises from the ClavA-Zn2+- lipid interaction, which is likely not the sole cause of the observed cell death.

Figure 2-9

Fluorescence micrographs of ClavA-TMR and ClavA-TMR-Zn2+ localizing in E. coli cells. Cells

were all grown to ~1x109 cfu/mL and ClavA concentration is 8 μM. (A) ClavA-TMR showing

internalization. (B) ClavA-TMR showing internalization with membrane damage, potentially

bleb formation (black arrows). (C) ClavA-TMR-Zn2+ showing a membrane bound state with

bright loci potentially representing highly concentrated regions of peptide (white arrows). (D)

ClavA-TMR-Zn2+ showing internalization with a bright locus potentially representing a highly

concentrated region of peptide (white arrow).

ClavA is less active against E. coli with diminished cytoplasmic Zn2+ concentration.

If ClavA is localizing in the cytoplasm it is important to look at the relevance of cytoplasmic

Zn2+ concentrations on ClavA activity. To that end we tested the activity of ClavA on three E. coli strains with deletions to the ZnuABC transporter complex genes (Table 2-4). ClavC was included as a control because Zn2+ concentration has no effect on the activity of the peptide

(Table 2-2). The ΔznuA mutant (JW5831) has a deletion in the zinc binding portion of the transporter (ZnuA), leaving the pore (ZnuB) and ATPase (ZnuC) portions intact. For both ClavA and ClavA-Zn2+ the activity decreases, reflected in the MIC increasing to 128 μM (a 2- and 32- fold increase respectively), whereas the activity, for ClavC, our control peptide which does not utilize Zn2+, remains unchanged. For the ΔznuB and ΔznuC strains (JW1848 and JW1847, respectively) the MIC for ClavA alone increases beyond the scope of the experiment, whereas the MIC for ClavA-Zn2+ increases 16-fold. For these strains, the activity of ClavC improves slightly, but this variation could be due to a side effect of the mutation process. The remaining activity of ClavA in the mutants is most likely due to the fact that some zinc is likely to diffuse into ZnuB without the aid of ZnuA. Overall it appears that cytoplasmic zinc is important to the activity of ClavA, especially when the peptide is not pre-exposed to zinc.

Table 2-4

ClavA, ClavA-Zn2+ and ClavC antimicrobial activity against three mutant E. coli strains with deleted zinc ion transporters a

Peptide MIC (JW5831) MIC (JW1848) MIC (JW1847)

ClavA 128 μM >128 μM >128 μM

ClavA-Zn2+ 128 μM 64 μM 64 μM

ClavC 4 μM 2 μM 2 μM a JW5831, JW1848 and JW1847 are deletion mutants of znuA, znuB and znuC respectively.

Each column represents the MIC against each mutant bacteria strain.

In order to prove that the change in activity is due to the decreased internal Zn2+ concentration from the deletion of the zinc transport system, and not just some feature of the deletion of a membrane protein, we measured the activity of ClavA, ClavA-Zn2+ and ClavC against three E. coli strains with deletions to the nickel transport system, NikABC (Table 2-5).

Overall, although the MIC values are not the same, as expected for different E. coli strains, the pattern observed for the wild type remains (i.e. ClavA shows improved activity in the presence of additional Zn2+ ions). These results confirm that the deletion of Zn2+ transporters, and the subsequent decrease in intracellular Zn2+ concentration is what causes the increase in MIC observed above.

Table 2-5

ClavA, ClavA-Zn2+ and ClavC antimicrobial activity against three mutant E. coli strains with deleted nickel ion transporters a

Peptide MIC (JW3441) MIC (JW3442) MIC (JW3444)

ClavA 32 μM 32 μM 32 μM

ClavA-Zn2+ 8 μM 8 μM 8 μM

ClavC 4 μM 2 μM 2 μM a JW3441, JW3442 and JW3444 are deletion mutants of nikA, nikB and nikD respectively.

Each column represents the MIC against each mutant bacteria strain.

ClavA-Zn2+ cleaves supercoiled DNA in vitro. Because cytoplasmic zinc appears to be important to the function of ClavA and the microscopy shows ClavA present in the cytoplasm, it is likely that ClavA-Zn2+ is forming in the cytoplasm and damaging an internal target. While most AMPs are active against bacterial membranes, a small subset have been shown to be internalized and bind to DNA.66-69 Therefore, ClavA-Zn2+ was tested for its in vitro DNA cleavage ability. As shown in figure 2-10A, DNA cleavage by ClavA pre-incubated with Zn2+ ions can be visibly observed after 60 min at room temperature, with the supercoiled and nicked forms decreasing in intensity, along with a concomitant increase in the linearized form. In addition, the linearized band can be observed to form a smear, indicating that not only is the

DNA linearized, it is also being further cleaved into smaller pieces. A steady decrease in supercoiled DNA was observed across all of the time points (Figure 2-10B). The nicked form can be observed to increase at first, but after 30 min it starts to decrease, while at the same time the amount of linearized DNA increased.

Figure 2-10

In vitro DNA cleavage gels representing pUC19 DNA treated with (A) 1 μM ClavA-Zn2+

quantified in (B), (C) 1 μM ClavA and 20 μM DTPA quantified in (D), (E) 2 μM Zn2+

quantified in (F), (G) pUC19 alone quantified in (H), (I) 1 μM ClavA-H17A-Zn2+ quantified

in (J), and (K) 1 μM ClavA-H21A-Zn2+ quantified in (L). S = supercoiled DNA, N = nicked

DNA and L = linear DNA.

To prove that ClavA and Zn2+ together are required for DNA cleavage to occur, several controls were run. Figures 2-10C and 2-10D show that, when exposed to ClavA in the absence of

Zn2+, the relative concentrations of each of the forms of DNA remained constant. The chelator

DTPA is necessary for this system since Zn2+ has a ubiquitous presence in laboratory settings.70

This same unchanging phenotype can also be observed for Zn2+ and DNA (Figures 2-10E and 2-

10F), as well as DNA in the absence of treatment (Figures 2-10G and 2-10H). The level of DNA damage was also found to be dependent on the concentration of ClavA-Zn2+ present in the sample with lower ClavA-Zn2+ concentrations resulting in lower relative nuclease activity (not shown). These results imply that DNA damage could be a part of the mechanism of ClavA, specifically when bound to Zn2+.

In order to further investigate our proposed Zn2+ binding site, we tested for DNA cleavage using our H17A and H21A mutants incubated with Zn2+. As hypothesized, no DNA damage is observable for either of these molecules (Figure 2-8I – Figure 2-8L). These results provide further evidence for the importance of these residues, and the DNA binding ability of the

HX3H motif.

2.3 – Discussion

In these studies, ClavA has been shown to have increased antimicrobial activity in the presence of Zn2+ not only through a 16-fold reduction in MIC, but also through increased rate of cell killing. The work presented above provides insight not only into the activity of ClavA in the presence of Zn2+ ions, but also sheds light on how ClavA interacts with bacteria. The results of the membrane depolarization, β-galactosidase leakage and potassium leakage experiments combine to paint a picture of a peptide that is membrane active and forms small pores. This behavior has been previously described by van Kan and co-workers, who established the important role played by phenylalanine and glycine residues in the interaction of ClavA with the bacterial membrane.46, 60 The question that arises from our results is how Zn2+ ions potentiate the

56 activity of ClavA. It is conceivable that similarly to DCD-1L, a peptide that oligomerizes into a trimer of dimers,71-72 ClavA uses Zn2+ ions to self-assemble into a higher oligomeric state to

2+ destabilize the bacterial membrane. Our ESI-MS data indicates the presence of a ClavA2-Zn (a dimer of ClavA bound through a Zn2+ ion) in the gas phase; however, such a complex might not

2+ exist in solution. On the other hand, higher oligomers, ClavAx-Zn , could also exist in solution, but were not detected under the ionization conditions used in the ESI-MS experiment.

Interestingly, zinc binding to His17 and His21 can result in a nucleation site for α–helix folding once the peptide interacts with the membrane,73 since the CD measurements determined in buffer indicated that upon Zn2+ binding the α–helicity of ClavA increases from 4.2% to 9.6%.

Alternatively, the interaction of Zn2+ with the phospholipid membrane is known to result in less hydrophilic phosphate groups,74 which could favor the binding and internalization of ClavA without the need for an oligomer to form. Both scenarios need to be tested before any conclusions can be reached.

It is clearer that a large component of the synergy between ClavA and zinc ions occurs in the cytoplasmic space. As evidenced by the microscopy images, ClavA and/or a ClavA-Zn2+ complex is internalized before cell killing occurs. Furthermore, E. coli mutants with deleted zinc transporters are less susceptible to the antimicrobial activity of ClavA. It is unlikely that a putative ClavA-Zn2+ complex will use the ABC transporter designed for Zn2+ given the large difference in size between the metal ion and any ClavA-Zn2+ complex. Within the cytoplasmic space, the ClavA-Zn2+ complex can target different biomolecules. Herein, we propose that similarly to other antimicrobial peptides, the ClavA-Zn2+ complex targets DNA.22, 26, 68, 75-77 The proposed nuclease activity of ClavA in combination with Zn2+ ions is a novel mechanism for natural AMPs that bind zinc. Histatin 5, a human salivary peptide binds to copper and zinc ions

57 and has nuclease activity;78-79 however, when the copper binding site is removed, the DNA cleavage activity is abolished despite the presence of Zn2+.79 Other host defense peptides/proteins that interact with Zn2+ ions have been previously reported; however, no DNA cleavage activity has been observed in those cases. For example, the human protein S100A12, a protein expressed and released by neutrophils,80 acts by binding and sequestering Zn2+ ions depriving bacteria from such an essential metal ion.81 The aforementioned DCD-1L, a peptide derived from dermcidin, adopts an α-helix conformation in which the binding of the Zn2+ ions occurs via residues in an i, i

+ 4 configuration, but no DNA cleavage activity has been reported.71 Kappacin, an AMP produced from κ–casein found in bovine milk, enhances its membranolytic effect in the presence of Zn2+ ions.82 Similarly, an antibacterial hexapeptide isolated from human amniotic fluid containing one lysine, two glycine and three glutamic acid residues potentiates its activity by

Zn2+ via an unknown mechanism.83 Interestingly, the presence of Cu2+ and Mg2+ cations were do not have the same effect.83 Brogden et al. have also reported that ovine pulmonary surfactant- associated anionic AMPs require zinc as a cofactor.84 However, not all anionic AMPs increase their activity in the presence of zinc ions. For instance, maximin H5, an anionic AMP isolated from the toad Bombina maxima, does not increase its activity in the presence of up to 1 mM zinc ions. Finally, we can also add to this collection of examples mutants of the Cardin and Weintraub motif peptides containing histidine residues instead of the amino acids Lys and Arg. These peptides become more active in the presence of Zn2+ ions; however, no cytoplasmic activity have been reported. For the mutants of the Cardin and Weintraub motif peptides, it was hypothesized that the metal ion increases the affinity of the peptides towards heparin.85

Overall, we hypothesize that the already recognized membrane activity of ClavA combines with its nuclease activity to create a putative “one-two-punch” of antimicrobial

58 activity. Additional studies will be necessary to elucidate the structure of the ClavA complex with Zn2+. A better understanding of the role of Zn2+ ions in the chemistry of ClavA and other tunicate AMPs is key to obtain a better understanding of zinc trafficking and its role in the immune system of tunicates. In addition, the unique strategy reported here for ClavA will be useful for the development of novel AMP-based strategies for the treatment of infections. This is particularly important since ClavA is being targeted for the development of new peptide-based antimicrobial strategies since the peptide and its nanoformulations have been shown to be effective for the clearing of sepsis infections in mice.86-87

2.4 – Methods

Solid Phase Peptide Synthesis, Purification and Identification. ClavA and all substitution and fluorescent mutants were synthesized purified and quantified using methods previously described by our lab.26 Briefly, peptides were synthesized manually via standard solid-phase peptide synthesis using fluorenylmethyloxycarbonyl (Fmoc) chemistry on Rink amide resin (Matrix Innovation, Quebec, Canada).

Peptides used for fluorescence microscopy were labeled with 5(6)- carboxytetramethylrhodamine (TMR; Sigma) on a modified lysine residue. The phenylalanine in the fourth position of the peptide sequence was replaced with a lysine protected with a 4- methyltrityl group, which can be selectively deprotected at low acid concentrations. The lysine amide group was then used to attach to the carboxylic acid of the TMR. A scheme of this synthesis can be found in Figure S1.

Peptides were then purified via reverse phase-HPLC (Shimadzu LC-20AD, USA) using a

C18 semi-prep column (Grace Davidson, Deerfield, IL, USA) and masses were confirmed via

59 electrospray ionization mass spectrometry in positive ion mode. Purity was determined using reverse phase-HPLC with a C18 analytical column (Thermo). Fractions containing the purified peptide were lyophilized overnight and dissolved in nanopure water. The peptides were then quantified via UV-Vis spectrophotometry using molar extinction coefficients for the amino acid residues and the peptide bond as determined by Kuipers and Gruppen.88 All peptides were stored at 4°C in solution and diluted before each experiment.

Antimicrobial Assay. Antimicrobial activity was determined using the broth microdilution method described by Hancock.54 Three to five colonies of Escherichia coli

(MG1655) or Bacillus subtilis (PS832) were added to Luria-Bertani broth (LB; Amresco, Solon,

OH, USA) containing 2-(N-morpholino) ethanesulfonic acid buffer (MES; Sigma-Aldrich) adjusted to pH 5.5 (LB 5.5), and incubated until mid-logarithmic phase was achieved,

OD600=0.4-0.6 (3-4 hours for E. coli, 4-6 hours for B. subtilis). Bacteria cultures containing 1 x

106 cfu/mL were then prepared in fresh media for MIC determination. A pH of 5.5 was chosen for these experiments because much of the previous research on ClavA shows that its activity is optimal at lower pH values.36, 45-46 Additionally, this pH has been found to be biologically relevant for phagocytic hemocytes. For example, the pH of hemocytes isolated from the tunicate

Ascidia ceratodes were found to be as low as 5.43

Peptide stock solutions were prepared at 2x the maximum experimental concentration in

50 mM MES pH 5.5 (MES 5.5), and 50μL of two-fold serial dilutions were made in a 96-well polypropylene plate (Greiner). To the serial dilutions were added 50μL of a bacterial suspension of 1 x 106 cfu/mL, for a final bacteria concentration of 5 x 105 cfu/mL. Plates were then incubated at 37°C for 18-20 hr and the minimum inhibitory concentration (MIC) was observed

60 as the lowest concentration that visually inhibited bacterial growth. The MIC values reported represent the mode of three trials.

For experiments using metal ions, the 2x peptide dilutions were incubated with solutions containing the metal ions used in the experiments at 2x peptide concentration for about thirty minutes prior to the addition of the bacteria suspension to allow adequate time for binding. The peptide-metal solutions were then used in the same manner stated above. All of the metal ions used showed MICs greater than 256 μM, the maximum metal concentration used in any experiment.

Minimum bactericidal concentration (MBC) was determined after performing the antimicrobial assay by spotting 10 μL of the MIC mixture and the next two higher peptide concentrations onto an agar plate, and incubating overnight. The MBC values reported are the peptide concentrations that resulted in agar plates without bacterial growth.

Varied Zn2+ Concentration Assay. To determine the optimal Zn2+ concentration for running experiments as well as how much Zn2+ is required to potentiate the activity of ClavA, one concentration of ClavA was incubated with several different concentrations of Zn2+ and bacterial growth was monitored. E. coli cultures of 1x106 cfu/mL were prepared using the same procedure as for the antimicrobial assay. These cultures were incubated with solutions containing

ClavA at a final concentration of 8 μM and either 0 μM, 2 μM, 4 μM, 8 μM, or 16 μM Zn2+. The mixtures were then incubated at 37°C for 4 hr, after which dilutions were plated on LB agar and allowed to grow overnight at 37°C. The next morning colonies were counted and percentages were calculated based on a control culture containing neither ClavA nor Zn2+ as 100%.

Tunicate Collection and Hemolymph Harvest. In order to determine biologically relevant Zn2+ concentrations, S. clava tunicate specimens were collected from shallow water dock lines at Avery Point, CT, in January of 2016 (41° 18.975 N, 72° 3.647 W). Tunicates were rinsed in sterile DI water and stored in 10% menthol (w/v) before dissection back at the lab.

The hemolymph was harvested from the tunicates via dissection and heart puncture. The hemolymph was then filtered through a 70 micron mesh nylon filter (Carolina Biological Supply,

Burlington, NC, USA) and stored at -20°C until ready to use.

Atomic Absorption Spectroscopy. Zn2+ concentration was measured using atomic absorption spectroscopy (AAS). Digestion of the samples was required prior to AAS analysis.

Hemolymph samples were first dried via heating at 120°C in an aluminum block heater for about

2 days in glass test tubes with Teflon screw caps removed. Next, 5 mL of 70% nitric acid

(Sigma-Aldrich) was added to each tube and allowed to rest at room temperature for one hour.

Tubes were then capped and heated in an aluminum heating block for 3 hr at 90°C. Samples were then cooled to room temperature, diluted to 50 mL and analyzed by AAS along with standards prepared in a similar process.

Circular Dichroism Spectroscopy. CD spectra both for the purpose of determining α- helical content of ClavA and its mutants, as well as for determining the effect of Zn2+ on ClavA were obtained using a Jasco J-710 spectropolarimeter. All samples contained 50 μM ClavA. The experiments performed on ClavA and the mutants were performed in a 50:50 solution of H2O and 2,2,2-trifluoroethanol (TFA). The Zn2+ titration experiments were performed in MES 5.5 buffer. The molar ellipticity,훩, was calculated from the raw data using the following equation:

푀푅푊 훩 = 휃 × 푙 ∗ 푐

62 where θ is the measured ellipticity, 푀푅푊 is the mean residue molecular weight of the peptide, calculated as the molar mass divided by the number of peptide bonds, 푙 is the pathlength in cm, and 푐 is the concentration of the peptide in mg/mL.89 The percent α-helix reported was calculated using the equation:

훩 − 3000 %훼 ℎ푒푙𝑖푥 = −100 × 222 33000

90 where 훩222 is the molar ellipticity at 222 nm.

Time Kill Kinetics. E. coli cultures of 1 x 106 cfu/mL were prepared using the same procedure as was used for the antimicrobial assay. Aliquots of the diluted culture were then mixed with the indicated treatments in equal volumes, then incubated at 37°C for the indicated time points. At each time point, aliquots of 10 μL were removed, serially diluted to 1:1000, and plated on LB agar and grown for 16-18hr. Colonies were counted and plotted against time. In instances where it was necessary to do so, dilutions were adjusted up or down to obtain plates with an appropriate number of colonies. Cultures were considered cleared when 100 μL of culture plated directly produced no colonies, indicating fewer than 10 cfu/mL.

DiSC3 (5) Depolarization Assay. B. subtilis cultures were grown at 37°C in LB 5.5 to mid-logarithmic phase, OD600=0.4-0.6, washed three times with MES 5.5 and diluted to OD600-

=0.01 in MES 5.5 containing 20mM glucose (Fisher). The bacteria were then incubated for one hour with 0.4μM 3,3'-dipropylthiadicarbocyanine Iodide (DiSC3(5); Chemodex, St. Gallen,

Switzerland), a membrane-potential active fluorophore. The one hour incubation allows time for the fluorophore to be drawn to polarized membranes and self-quench, as evidenced by the >90% decrease in fluorescence. The bacteria solutions were then mixed with serial dilutions of peptide

63 treatments, 50% Triton X-100 (T-X100; Fisher), or MES 5.5. T-X100 was used as a positive depolarization control, and MES 5.5 was the negative control. Fluorescence was monitored every

2 min for 1 hr at an excitation wavelength of 622 nm and an emission wavelength of 670 nm.

Values represent the percent fluorescence recovery averaged over three trials ± standard error.

β-Galactosidase Leakage Assay. The amount of β-galactosidase leakage caused by

ClavA or ClavA-Zn2+ was determined as described previously.26, 91-92 E. coli transformed with a gene for cytoplasmic β-galactosidase production were grown in LB 5.5 at 37°C to OD600=0.4-

0.6, after which overexpression of β-galactosidase was induced for 1 hr with the addition of 1 mM final concentration of isopropyl β-D-1-thiogalactopyranoside (IPTG; Fisher). Cells were then washed three times in MES buffer and suspended in fresh LB 5.5. Aliquots of the culture were then mixed with either ClavA, ClavA-Zn2+, or 50% T-X100. Mixtures were incubated for 1 hr at 37°C, then centrifuged at 4400rpm and 4°C for 10 min and the supernatant, containing any leaked β-galactosidase, was recovered. Ortho-nitrophenyl-β-galactoside (ONPG; Thermo), which β-galactosidase cleaves into the yellow product o-nitrophenol (ONP),62 was then added to the supernatant at a final concentration of 0.8 mg/mL and absorbance was monitored at 405 nm every 5 min for 1 hr. Values reported are an average of the one hour points for three trials; error bars represent standard error.

Potassium Efflux. B. subtilis was grown to mid-logarithmic phase (OD600=0.4-0.6) in

LB 5.5, washed and resuspended at a concentration of 1 x 108 cfu/mL in MES 5.5. Cells were then mixed with ClavA or ClavA-Zn2+ at twice their respective MBCs to yield solutions containing 5 x 108 cfu/mL B. subtilis and peptides at the MBC. The potassium concentration was then monitored every 10 min for 1 hr using a potassium sensitive electrode and reference electrode (MI-442 and MI-402; Microelectrodes, Bedford, NH, USA) connected to a pH meter

(Orion Star A111; Thermo) measuring potential in mV. Cells were then lysed using a combination of freeze thaw cycles and sonication, and the potential of the resulting solution was used as an internal positive control. Percentages were calculated using a blank bacteria sample with no peptide as a negative control.

Confocal Fluorescence Microscopy. E. coli cultures were grown to an OD600~1.0 (1 x

109 cfu/ml) in LB 5.5, then washed and suspended in MES 5.5 containing 20 mM glucose.

Aliquots were incubated with 8 μM of either ClavA-TMR or ClavA-TMR-Zn2+ for 1 hr at room temperature. This concentration was used for both treatments because higher ClavA-TMR concentrations would cause oversaturation of the image. Mixtures were then spun down and resuspended in fresh MES 5.5 with 20 mM glucose prior to imaging to remove any unbound peptide. Cells were imaged on a Nikon A1R spectral confocal microscope using a 60X oil immersion lens.

In Vitro DNA Cleavage. Samples were prepared by mixing 10 μM base pair pUC19 with 1μM ClavA-Zn2+, 1 μM ClavA with 20 μM diethylenetriaminepentaacetic acid (DTPA) or 2

μM Zn2+ in MES 5.5. Mixtures were allowed to stand at room temperature for indicated times, then were stopped using 3x loading dye containing 1 mM EDTA prior to loading on the gel.

After all samples at various time points were prepared, the samples were loaded on a 1% agarose gel containing ethidium bromide, and the gel was run at 100V for 90 min. Gels were imaged using a Bio-Rad GelDoc XR+ Imager and quantified using Image Lab 5.0 software. A correction factor of 1.47 was applied to the data for supercoiled DNA to account for the difficulty of staining with ethylene bromide.93

2.5 – References

1. McDevitt, C. A.; Ogunniyi, A. D.; Valkov, E.; Lawrence, M. C.; Kobe, B.; McEwan, A. G.; Paton, J. C., A Molecular Mechanism for Bacterial Susceptibility to Zinc. PLoS Pathog 2011, 7 (11), e1002357. 2. Pederick, V. G.; Eijkelkamp, B. A.; Begg, S. L.; Ween, M. P.; McAllister, L. J.; Paton, J. C.; McDevitt, C. A., ZnuA and zinc homeostasis in Pseudomonas aeruginosa. Sci Rep 2015, 5, 13139. 3. Ivano, B., Bioinorganic Chemistry. University Science Books: 1994. 4. Fraker, P. J.; King, L. E.; Laakko, T.; Vollmer, T. L., The Dynamic Link between the Integrity of the Immune System and Zinc Status. J Nutrition 2000, 130 (5), 1399S-1406S. 5. Shankar, A. H.; Prasad, A. S., Zinc and immune function: the biological basis of altered resistance to infection. Am J Clin Nutrition 1998, 68 (2 Suppl), 447s-463s. 6. Prasad, A. S., Clinical, immunological, anti-inflammatory and antioxidant roles of zinc. Exp Gerontol 2008, 43 (5), 370-7. 7. Anderson, R. S., Cellular Responses to Foreign Bodies in the Tunicate Molgula manhattensis (DeKay). Biol Bull 1971, 141 (1), 91-98. 8. Kelly, K. L.; Cooper, E. L.; Raftos, D. A., A humoral opsonin from the solitary urochordate Styela clava. Dev Comp Immun 1993, 17 (1), 29-39. 9. Raftos, D. A., Allorecognition and Humoral Immunity in Tunicatesa. Ann NY Acad Sci 1994, 712 (1), 227-244. 10. Rinkevich, B., Primitive immune systems: Are your ways my ways? Immunol Rev 2004, 198 (1), 25-35. 11. Dishaw, L.; Flores-Torres, J.; Mueller, M. G.; Karrer, C.; Skapura, D.; Melillo, D.; Zucchetti, I.; De Santis, R.; Pinto, M. R.; Litman, G., A Basal Chordate Model for Studies of Gut Microbial Immune Interactions. Front Immunol 2012, 3, 96. 12. Tzou, P.; Ohresser, S.; Ferrandon, D.; Capovilla, M.; Reichhart, J.-M.; Lemaitre, B.; Hoffmann, J. A.; Imler, J.-L., Tissue-Specific Inducible Expression of Antimicrobial Peptide Genes in Drosophila Surface Epithelia. Immunity 2000, 13 (5), 737-748. 13. Lamberty, M.; Zachary, D.; Lanot, R.; Bordereau, C.; Robert, A.; Hoffmann, J. A.; Bulet, P., Insect Immunity: Constitutive Expression of a Cysteine-Rich Antifungal and a Linear Antibacterial Peptide in a Termite Insect. J Biol Chem 2001, 276 (6), 4085-4092. 14. Huttner, K. M.; Bevins, C. L., Antimicrobial peptides as mediators of epithelial host defense. Ped Res 1999, 45 (6), 785-94. 15. Hancock, R. E. W.; Haney, E. F.; Gill, E. E., The immunology of host defence peptides: beyond antimicrobial activity. Nat Rev Immunol 2016, 16 (5), 321-334. 16. Zasloff, M., Magainins, a class of antimicrobial peptides from Xenopus skin: isolation, characterization of two active forms, and partial cDNA sequence of a precursor. Proc Natl Acad Sci 1987, 84 (15), 5449-53. 17. Melo, M. N.; Castanho, M. A., The Mechanism of Action of Antimicrobial Peptides: Lipid Vesicles vs. Bacteria. Front Immunol 2012, 3, 236. 18. Chileveru, H. R.; Lim, S. A.; Chairatana, P.; Wommack, A. J.; Chiang, I. L.; Nolan, E. M., Visualizing attack of Escherichia coli by the antimicrobial peptide human defensin 5. Biochemistry 2015, 54 (9), 1767-77. 19. Paulmann, M.; Arnold, T.; Linke, D.; Ozdirekcan, S.; Kopp, A.; Gutsmann, T.; Kalbacher, H.; Wanke, I.; Schuenemann, V. J.; Habeck, M.; Burck, J.; Ulrich, A. S.; Schittek, B., Structure-

66 activity analysis of the dermcidin-derived peptide DCD-1L, an anionic antimicrobial peptide present in human sweat. J Biol Chem 2012, 287 (11), 8434-43. 20. Hancock, R. E. W.; Lehrer, R., Cationic peptides: a new source of antibiotics. Trends Biotechnol 1998, 16 (2), 82-88. 21. Konno, K.; Hisada, M.; Fontana, R.; Lorenzi, C. C. B.; Naoki, H.; Itagaki, Y.; Miwa, A.; Kawai, N.; Nakata, Y.; Yasuhara, T.; Ruggiero Neto, J.; de Azevedo Jr, W. F.; Palma, M. S.; Nakajima, T., Anoplin, a novel antimicrobial peptide from the venom of the solitary wasp Anoplius samariensis. BBA Protein Struct M 2001, 1550 (1), 70-80. 22. Bustillo, M. E.; Fischer, A. L.; LaBouyer, M. A.; Klaips, J. A.; Webb, A. C.; Elmore, D. E., Modular analysis of hipposin, a histone-derived antimicrobial peptide consisting of membrane translocating and membrane permeabilizing fragments. BBA Biomembranes 2014, 1838 (9), 2228- 2233. 23. Elmore, D. E., Insights into buforin II membrane translocation from molecular dynamics simulations. Peptides 2012, 38 (2), 357-62. 24. Pavia, K. E.; Spinella, S. A.; Elmore, D. E., Novel histone-derived antimicrobial peptides use different antimicrobial mechanisms. Biochim Biophys Acta 2012, 1818 (3), 869-76. 25. Rideout, M. C.; Boldt, J. L.; Vahi-Ferguson, G.; Salamon, P.; Nefzi, A.; Ostresh, J. M.; Giulianotti, M.; Pinilla, C.; Segall, A. M., Potent antimicrobial small molecules screened as inhibitors of tyrosine recombinases and Holliday junction-resolving enzymes. Mol Diversity 2011, 15 (4), 989-1005. 26. Libardo, M. D.; Cervantes, J. L.; Salazar, J. C.; Angeles-Boza, A. M., Improved bioactivity of antimicrobial peptides by addition of amino-terminal copper and nickel (ATCUN) binding motifs. ChemMedChem 2014, 9 (8), 1892-901. 27. Libardo, M. D. J.; Gorbatyuk, V. Y.; Angeles-Boza, A. M., Central Role of the Copper-Binding Motif in the Complex Mechanism of Action of Ixosin: Enhancing Oxidative Damage and Promoting Synergy with Ixosin B. ACS Infect Dis 2016, 2 (1), 71-81. 28. Achard, M. E.; Stafford, S. L.; Bokil, N. J.; Chartres, J.; Bernhardt, P. V.; Schembri, M. A.; Sweet, M. J.; McEwan, A. G., Copper redistribution in murine macrophages in response to Salmonella infection. Biochem J 2012, 444 (1), 51-7. 29. Alonso, S.; Pethe, K.; Russell, D. G.; Purdy, G. E., Lysosomal killing of Mycobacterium mediated by ubiquitin-derived peptides is enhanced by autophagy. Proc Natl Acad Sci 2007, 104 (14), 6031-6. 30. Hiemstra, P. S.; van den Barselaar, M. T.; Roest, M.; Nibbering, P. H.; van Furth, R., Ubiquicidin, a novel murine microbicidal protein present in the cytosolic fraction of macrophages. J Leukocyte Biol 1999, 66 (3), 423-8. 31. Rosenberger, C. M.; Gallo, R. L.; Finlay, B. B., Interplay between antibacterial effectors: a macrophage antimicrobial peptide impairs intracellular Salmonella replication. Proc Natl Acad Sci 2004, 101 (8), 2422-7. 32. Stafford, S. L.; Bokil, N. J.; Achard, M. E.; Kapetanovic, R.; Schembri, M. A.; McEwan, A. G.; Sweet, M. J., Metal ions in macrophage antimicrobial pathways: emerging roles for zinc and copper. Biosci Rep 2013, 33 (4). 33. White, C.; Kambe, T.; Fulcher, Y. G.; Sachdev, S. W.; Bush, A. I.; Fritsche, K.; Lee, J.; Quinn, T. P.; Petris, M. J., Copper transport into the secretory pathway is regulated by oxygen in macrophages. J Cell Sci 2009, 122 (Pt 9), 1315-21. 34. White, C.; Lee, J.; Kambe, T.; Fritsche, K.; Petris, M. J., A role for the ATP7A copper- transporting ATPase in macrophage bactericidal activity. J Biol Chem 2009, 284 (49), 33949-56.

35. Wolschendorf, F.; Ackart, D.; Shrestha, T. B.; Hascall-Dove, L.; Nolan, S.; Lamichhane, G.; Wang, Y.; Bossmann, S. H.; Basaraba, R. J.; Niederweis, M., Copper resistance is essential for virulence of Mycobacterium tuberculosis. Proc Natl Acad Sci 2011, 108 (4), 1621-6. 36. Lee, I. H.; Cho, Y.; Lehrer, R. I., Effects of pH and salinity on the antimicrobial properties of clavanins. Infect Immun 1997, 65 (7), 2898-903. 37. Zhao, C.; Liaw, L.; Lee, I. H.; Lehrer, R. I., cDNA cloning of Clavanins: antimicrobial peptides of tunicate hemocytes. FEBS Lett 1997, 410 (2-3), 490-2. 38. Lehrer, R. I.; Lee, I. H.; Menzel, L.; Waring, A.; Zhao, C., Clavanins and styelins, alpha-helical antimicrobial peptides from the hemocytes of Styela clava. Adv Exp Med Biol 2001, 484, 71-6. 39. Menzel, L. P.; Lee, I. H.; Sjostrand, B.; Lehrer, R. I., Immunolocalization of clavanins in Styela clava hemocytes. Dev Comp Immunol 2002, 26 (6), 505-515. 40. Lee, I. H.; Cho, Y.; Lehrer, R. I., Styelins, broad-spectrum antimicrobial peptides from the solitary tunicate, Styela clava. Comp Biochem Physiol B: Biochem Mol Biol 1997, 118B (3), 515- 521. 41. Taylor, S. W.; Craig, A. G.; Fischer, W. H.; Park, M.; Lehrer, R. I., Styelin D, an extensively modified antimicrobial peptide from ascidian hemocytes. J Biol Chem 2000, 275 (49), 38417- 38426. 42. Lehrer, R. I.; Andrew, T. J.; Taylor, S. W.; Menzel, L. P.; Waring, A. J., Natural Peptide antibiotics from tunicates: structures, functions and potential uses. Integr Comp Biol 2003, 43 (2), 313-22. 43. Lee, S.; Nakanishi, K.; Kustin, K., The intracellular pH of tunicate blood cells: Ascidia ceratodes whole blood, morula cells, vacuoles and cytoplasm. BBA Gen Subj 1990, 1033 (3), 311- 317. 44. Silva, O. N.; Fensterseifer, I. C.; Rodrigues, E. A.; Holanda, H. H.; Novaes, N. R.; Cunha, J. P.; Rezende, T. M.; Magalhaes, K. G.; Moreno, S. E.; Jeronimo, M. S.; Bocca, A. L.; Franco, O. L., Clavanin A improves outcome of complications from different bacterial infections. Antimicrob Agents Chemother 2015, 59 (3), 1620-6. 45. van Kan, E. J.; van der Bent, A.; Demel, R. A.; de Kruijff, B., Membrane activity of the peptide antibiotic clavanin and the importance of its glycine residues. Biochemistry 2001, 40 (21), 6398- 405. 46. van Kan, E. J.; Demel, R. A.; Breukink, E.; van der Bent, A.; de Kruijff, B., Clavanin permeabilizes target membranes via two distinctly different pH-dependent mechanisms. Biochemistry 2002, 41 (24), 7529-39. 47. Fitzsimons, M. P.; Barton, J. K., Design of a Synthetic Nuclease: DNA Hydrolysis by a Zinc- Binding Peptide Tethered to a Rhodium Intercalator. J Am Chem Soc 1997, 119 (14), 3379-3380. 48. Zaykov, A. N.; Popp, B. V.; Ball, Z. T., Helix Induction by Dirhodium: Access to Biocompatible Metallopeptides with Defined Secondary Structure. Chem Eur J 2010, 16 (22), 6651-6659. 49. Ball, Z. T., Designing Enzyme-like Catalysts: A Rhodium(II) Metallopeptide Case Study. Acc Chem Res 2013, 46 (2), 560-570. 50. Deb, S. C.; Fukushima, T., Metals in aquatic ecosystems: mechanisms of uptake, accumulation and releaseΓÇÉEcotoxicological perspectives. Int J Environ Stud 1999, 56 (3), 385-417. 51. Hawkins, C. J.; Parry, D. L.; Pierce, C., Chemistry of the Blood of the Ascidian Podoclavella moluccensis. Biol Bull 1980, 159 (3), 669-680. 52. Eisler, R., Compendium of trace metals and marine biota. 1st ed.; Elsevier: Amsterdam ; Boston ; Oxford, 2010.

53. Ruddell, C. L.; Rains, D. W., The Relationship Between Zinc, Copper and the Basophils of Two Crassostreid Oyster, C. gigas and C. virginica. Comp Biochem Physiol A 1975, 51 (3), 585- 591. 54. Wiegand, I.; Hilpert, K.; Hancock, R. E., Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat Protoc 2008, 3 (2), 163- 75. 55. Lee, I. H.; Zhao, C.; Cho, Y.; Harwig, S. S.; Cooper, E. L.; Lehrer, R. I., Clavanins, alpha- helical antimicrobial peptides from tunicate hemocytes. FEBS Lett 1997, 400 (2), 158-62. 56. Nick Pace, C.; Martin Scholtz, J., A Helix Propensity Scale Based on Experimental Studies of Peptides and Proteins. Biophys J 1998, 75 (1), 422-427. 57. Yarlagadda, V.; Akkapeddi, P.; Manjunath, G. B.; Haldar, J., Membrane active vancomycin analogues: a strategy to combat bacterial resistance. J Med Chem 2014, 57 (11), 4558-68. 58. McGrath, D. M.; Barbu, E. M.; Driessen, W. H.; Lasco, T. M.; Tarrand, J. J.; Okhuysen, P. C.; Kontoyiannis, D. P.; Sidman, R. L.; Pasqualini, R.; Arap, W., Mechanism of action and initial evaluation of a membrane active all-D-enantiomer antimicrobial peptidomimetic. Proc Natl Acad Sci 2013, 110 (9), 3477-82. 59. Harzer, G.; Kauer, H., Binding of zinc to casein. Am J Clin Nutr 1982, 35 (5), 981-987. 60. van Kan, E. J.; Demel, R. A.; van der Bent, A.; de Kruijff, B., The role of the abundant phenylalanines in the mode of action of the antimicrobial peptide clavanin. Biochim Biophys Acta 2003, 1615 (1-2), 84-92. 61. Kurut, A.; Henriques, J.; Forsman, J.; Skepö, M.; Lund, M., Role of histidine for charge regulation of unstructured peptides at interfaces and in bulk. Proteins: Structure, Function, and Bioinformatics 2014, 82 (4), 657-667. 62. Steffen, H.; Rieg, S.; Wiedemann, I.; Kalbacher, H.; Deeg, M.; Sahl, H. G.; Peschel, A.; Götz, F.; Garbe, C.; Schittek, B., Naturally Processed Dermcidin-Derived Peptides Do Not Permeabilize Bacterial Membranes and Kill Microorganisms Irrespective of Their Charge. Antimicrob Agents Chemother 2006, 50 (8), 2608-2620. 63. Sengupta, D.; Leontiadou, H.; Mark, A. E.; Marrink, S. J., Toroidal pores formed by antimicrobial peptides show significant disorder. Biochim Biophys Acta 2008, 1778 (10), 2308-17. 64. Leveritt, J. M., 3rd; Pino-Angeles, A.; Lazaridis, T., The structure of a melittin-stabilized pore. Biophys J 2015, 108 (10), 2424-6. 65. Langham, A. A.; Ahmad, A. S.; Kaznessis, Y. N., On the Nature of Antimicrobial Activity: A Model for Protegrin-1 Pores. J Am Chem Soc 2008, 130 (13), 4338-4346. 66. Libardo, M. D.; Paul, T. J.; Prabhakar, R.; Angeles-Boza, A. M., Hybrid peptide ATCUN-sh- Buforin: Influence of the ATCUN charge and stereochemistry on antimicrobial activity. Biochimie 2015, 113, 143-55. 67. Hayden, R. M.; Goldberg, G. K.; Ferguson, B. M.; Schoeneck, M. W.; Libardo, M. D.; Mayeux, S. E.; Shrestha, A.; Bogardus, K. A.; Hammer, J.; Pryshchep, S.; Lehman, H. K.; McCormick, M. L.; Blazyk, J.; Angeles-Boza, A. M.; Fu, R.; Cotten, M. L., Complementary Effects of Host Defense Peptides Piscidin 1 and Piscidin 3 on DNA and Lipid Membranes: Biophysical Insights into Contrasting Biological Activities. J Phys Chem B 2015, 119 (49), 15235- 46. 68. Uyterhoeven, E. T.; Butler, C. H.; Ko, D.; Elmore, D. E., Investigating the nucleic acid interactions and antimicrobial mechanism of buforin II. FEBS Lett 2008, 582 (12), 1715-8.

69. Joyner, J. C.; Hodnick, W. F.; Cowan, A. S.; Tamuly, D.; Boyd, R.; Cowan, J. A., Antimicrobial metallopeptides with broad nuclease and ribonuclease activity. Chem Commun 2013, 49 (21), 2118-2120. 70. Graham, A. I.; Hunt, S.; Stokes, S. L.; Bramall, N.; Bunch, J.; Cox, A. G.; McLeod, C. W.; Poole, R. K., Severe Zinc Depletion of Escherichia coli: roles for high affinity zinc binding by ZinT, zinc transport and zinc-independent proteins. J Biol Chem 2009, 284 (27), 18377-18389. 71. Song, C.; Weichbrodt, C.; Salnikov, E. S.; Dynowski, M.; Forsberg, B. r. O.; Bechinger, B.; Steinem, C.; de Groot, B. L.; Zachariae, U.; Zeth, K., Crystal structure and functional mechanism of a human antimicrobial membrane channel. Proc Natl Acad Sci 2013, 110 (12), 4586-4591. 72. Becucci, L.; Valensin, D.; Innocenti, M.; Guidelli, R., Dermcidin, an anionic antimicrobial peptide: influence of lipid charge, pH and Zn2+ on its interaction with a biomimetic membrane. Soft Matt 2014, 10 (4), 616-626. 73. Silva, O. N.; Alves, E. S. F.; de la Fuente-Núñez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A. R. B.; Andrade, C. A. S.; Nascimento, J. M.; Fensterseifer, I. C. M.; Porto, W. F.; Correa, J. R.; Hancock, R. E. W.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti- Biofilm Activity. Sci Rep 2016, 6, 27128. 74. Binder, H.; Arnold, K.; Ulrich, A. S.; Zschornig, O., Interaction of Zn2+ with phospholipid membranes. Biophys Chem 2001, 90 (1), 57-74. 75. Sim, S.; Wang, P.; Beyer, B. N.; Cutrona, K. J.; Radhakrishnan, M. L.; Elmore, D. E., Investigating the Nucleic Acid Interactions of Histone-Derived Antimicrobial Peptides. FEBS Lett 2017, n/a-n/a. 76. Libardo, M. D. J.; Paul, T. J.; Prabhakar, R.; Angeles-Boza, A. M., Hybrid peptide ATCUN- sh-Buforin: Influence of the ATCUN charge and stereochemistry on antimicrobial activity. Biochimie 2015, 113 (0), 143-155. 77. Cowan, J. A.; Cowan, A. S. Metallodrugs comprising an antimicrobial peptide and/or an antibiotic covalently bound to a metal binding moiety with a bound metal having improved pharmacological properties and methods of manufacture and use thereof. WO2013109927A2, 2013. 78. Melino, S.; Rufini, S.; Sette, M.; Morero, R.; Grottesi, A.; Paci, M.; Petruzzelli, R., Zn2+ ions selectively induce antimicrobial salivary peptide histatin-5 to fuse negatively charged vesicles. Identification and characterization of a zinc-binding motif present in the functional domain. Biochemistry 1999, 38 (30), 9626-9633. 79. Melino, S.; Gallo, M.; Trotta, E.; Mondello, F.; Paci, M.; Petruzzelli, R., Metal-Binding and Nuclease Activity of an Antimicrobial Peptide Analogue of the Salivary Histatin 5. Biochemistry 2006, 45 (51), 15373-15383. 80. Guignard, F.; Mauel, J.; Markert, M., Identification and characterization of a novel human neutrophil protein related to the S100 family. Biochem J 1995, 309 (2), 395-401. 81. Cunden, L. S.; Gaillard, A.; Nolan, E. M., Calcium ions tune the zinc-sequestering properties and antimicrobial activity of human S100A12. Chem Sci 2016, 7 (2), 1338-1348. 82. Dashper, S. G.; O'Brien-Simpson, N. M.; Cross, K. J.; Paolini, R. A.; Hoffmann, B.; Catmull, D. V.; Malkoski, M.; Reynolds, E. C., Divalent metal cations increase the activity of the antimicrobial peptide kappacin. Antimicrob Agents Chemother 2005, 49 (6), 2322-2328. 83. Schlievert, P.; Johnson, W.; Galask, R. P., Isolation of a low-molecular-weight antibacterial system from human amniotic fluid. Infect Immun 1976, 14 (5), 1156-1166.

84. Brogden, K. A.; De Lucca, A. J.; Bland, J.; Elliott, S., Isolation of an ovine pulmonary surfactant-associated anionic peptide bactericidal for Pasteurella haemolytica. Proc Natl Acad Sci 1996, 93 (1), 412-16. 85. Rydengård, V.; Andersson Nordahl, E.; Schmidtchen, A., Zinc potentiates the antibacterial effects of histidine-rich peptides against Enterococcus faecalis. FEBS J 2006, 273 (11), 2399-2406. 86. Silva, O. N.; Fensterseifer, I. C. M.; Rodrigues, E. A.; Holanda, H. H. S.; Novaes, N. R. F.; Cunha, J. P. A.; Rezende, T. M. B.; Magalhaes, K. G.; Moreno, S. E.; Jeronimo, M. S.; Bocca, A. L.; Franco, O. L., Clavanin A improves outcome of complications from different bacterial infections. Antimicrob Agents Chemother 2015, 59 (3), 1620-1626. 87. Saude, A. C. M.; Ombredane, A. S.; Silva, O. N.; Barbosa, J. A. R. G.; Moreno, S. E.; Guerra, A. C.; Falcao, A. R.; Silva, L. P.; Dias, S. C.; Franco, O. L., Clavanin bacterial sepsis control using a novel methacrylate nanocarrier. Int J Nanomed 2014, 9, 5055-5069, 15 pp. 88. Kuipers, B. J.; Gruppen, H., Prediction of molar extinction coefficients of proteins and peptides using UV absorption of the constituent amino acids at 214 nm to enable quantitative reverse phase high-performance liquid chromatography-mass spectrometry analysis. J Agric Food Chem 2007, 55 (14), 5445-51. 89. Juban, M. M.; Javadpour, M. M.; Barkley, M. D., Circular Dichroism Studies of Secondary Structure of Peptides. In Antibacterial Peptide Protocols, Shafer, W. M., Ed. Humana Press: Totowa, NJ, 1997; pp 73-78. 90. McLean, L. R.; Hagaman, K. A.; Owen, T. J.; Krstenansky, J. L., Minimal peptide length for interaction of amphipathic .alpha.-helical peptides with phosphatidylcholine liposomes. Biochemistry 1991, 30 (1), 31-37. 91. Maisetta, G.; Vitali, A.; Scorciapino, M. A.; Rinaldi, A. C.; Petruzzelli, R.; Brancatisano, F. L.; Esin, S.; Stringaro, A.; Colone, M.; Luzi, C.; Bozzi, A.; Campa, M.; Batoni, G., pH-dependent disruption of Escherichia coli ATCC 25922 and model membranes by the human antimicrobial peptides hepcidin 20 and 25. FEBS J 2013, 280 (12), 2842-2854. 92. Pruul, H.; Reynolds, B. L., Interaction of Complement and Polymyxin with Gram-Negative Bacteria. Infect Immun 1972, 6 (5), 709-717. 93. Joyner, J. C.; Reichfield, J.; Cowan, J. A., Factors Influencing the DNA Nuclease Activity of Iron, Cobalt, Nickel, and Copper Chelates. J Am Chem Soc 2011, 133 (39), 15613-15626.

Chapter 3 – Clavanin A, Zinc metallonuclease effective against gram negative pathogens

Adapted from: Juliano, S. A.; et. al., A Potent Host Defense Peptide Triggers DNA Damage and is Active against Multi-Drug Resistant Gram-Negative Pathogens. Manuscript Submitted

3.1 – Introduction

The worldwide emergence of antibiotic-resistant bacteria has endangered the efficacy of antibiotics. Some frightening projections estimate that antibiotic-resistant bacteria will result in the death of 2.4 million people annually by 2050, costing in excess of $3 billion per year.1 Hospital acquired infections are responsible for the death of over 1 million patients per year in the US alone, with the primary offenders including the so called

ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species).2-3 Isolates of each of these bacterial species have been reported to contain multiple mechanisms for antibiotic resistance, setting the stage for the fulfilment of these projections.4 Gram-negative bacteria, in particular, have shown an alarming prevalence of antibiotic resistance. For example, particular species of Acinetobacter and Pseudomonas are showing resistance to all available antibiotics, and carbapenem resistance genes from

Klebsiella spp. have spread to gram-negative bacteria worldwide.5-6 With so many bacteria showing resistance, even to some of the antibiotics of last resort, the need for novel therapeutic compounds to combat the problem is greater than ever.

In recent years, host defence peptides (HDPs) have become of great interest as tools to supplement our antibiotic arsenal. Their beneficial qualities include broad spectrum activity and low incidence of resistance development. Clavanin A (ClavA,

VFQFLGKIIHHVGNFVHGFSHVF-NH2), a 23-residue HDP isolated from the hemocytes of the solitary tunicate Styela clava, has shown great promise as a novel therapeutic against a wide range of bacterial pathogens. In vitro studies of murine wound infection models showed improved healing and eradication of bacteria, while in systemic infection models,

ClavA treatment resulted in increased survival for both Escherichia coli and

Staphylococcus aureus infected mice.7 In addition to this, ClavA has also proven effective against biofilms, preventing the formation of E. coli and S. aureus biofilms and perturbing preformed biofilms of K. pneumoniae.8

In our own studies of ClavA, we found evidence for a Zn(II)-regulated intracellular mechanism of action (MoA), which we hypothesized to involve damage to bacterial DNA.9

This hypothesis was supported by evidence of in vitro DNA cleavage by ClavA-Zn(II), and inhibition of ClavA activity in E. coli strains lacking the ability to transport Zn(II) ions.

These results represent a novel MoA for zinc-binding HDPs, most of which have been reported to sequester Zn2+ to keep this essential micronutrient away from invading pathogens.10

Zinc, the second most abundant trace metal in human cells after iron, is an important micronutrient for many cellular processes in mammalian cells including DNA and RNA synthesis, and membrane stabilization.11-13 More importantly, Zn(II) deficiencies have been found to be important in several chronic diseases, including diabetes, and neurodegenerative diseases, as well as increasing susceptibility to many infectious diseases, including HIV, malaria, and bacterial infections.14 Important immune functions, such as phagocytosis and intracellular killing are affected by zinc deficiency.15 Zinc(II) ions are a common weapon used by eukaryotes to combat bacteria. For example, zinc levels in blood serum of infected mice can reach up to 900 μM.16 Zinc is also trafficked into the phagosome of murine macrophages infected with mycobacteria. When infected with Mycobacterium tuberculosis, phagosomal zinc concentrations reach of up to 459 μM, after 24 hr post infection.17 Although the concentration of zinc(II) ions reach very high micromolar levels,

74 even higher concentrations seem to be needed for it to have a bactericidal effect. Zinc concentrations of 1 mM have been shown to abolish the growth of wild-type Streptococcus pneumoniae, while 250 μM Zn(II) only delays its growth. In S. clava hemolymph the zinc concentration was found to be 26.7 ± 7.1 μM, though it is likely that the local concentration within phagocytic cells is much higher.9

As a result of elevated zinc concentrations, ClavA based treatments could enhance the body’s natural infection response, allowing for treatments that work in tandem with our bodies’ natural defences. It is important, however, to first understand exactly how ClavA and zinc ions are working together to eliminate bacterial infections, as such knowledge will allow for further optimization of this natural framework.

3.2 – Results

Zn(II) potentiates the antimicrobial activity of ClavA against Gram-negative pathogens

We have previously shown that when ClavA is combined with Zn(II), the activity of the peptide against E. coli is increased. In order to determine if this activity potentiation is unique to

E. coli, we tested the activity of ClavA and ClavA-Zn(II) against the gram-negative ESKAPE pathogens, including some multi-drug resistant clinical isolates. Gram-negative pathogens are of great interest due to the difficulty of penetrating their double membrane structure, with the outer membrane containing LPS and lipid A which aid in defending the bacteria against cationic

AMPs.18-19 Additionally, gram-negative bacteria contain efflux pumps that not only remove small molecules from the cytoplasm and periplasm, but also many AMPs. The results of testing ClavA and ClavA-Zn(II) against gram-negative bacteria are summarized in Table 3-1. In every case we observed a decrease in the minimum inhibitory concentration (MIC) upon incubation of the peptide with Zn(II) at double the peptide concentration, which correlates to an increase in activity. This

75 improvement in activity can even be seen in multi-drug resistant ESKAPE pathogen clinical isolates and multi-drug resistant E. coli, though the Zn(II)-based potentiation against E. coli AR

Bank #0114 is not quite as pronounced as for the wild type E. coli.20 The activity of ClavA against

P. aeruginosa and K. pneumoniae was above the limit of detection of the experiment, so they can only be said to show at least 8-fold and 4-fold improvements, respectively, while activity against

A. baumannii and Enterobacter cloacae each improved 4-fold upon the addition of Zn(II). While none of these quite reached the 16-fold improvement seen against E. coli, the modest activity potentiation observed is still very promising for the effectiveness of ClavA-Zn(II) against other gram-negative pathogens. Additionally, the fact that ClavA-Zn(II) showed improvement against multi-drug resistant clinical isolates provides further evidence of how effective ClavA is as an antimicrobial compound.

Table 3-1

Clavanin A minimum inhibitory concentrations against various wild type and drug resistant bacteria.

Bacterial Strain ClavA MIC ClavA-Zn(II) MIC

E. coli MG1655 a 64 μM 4 μM

E. coli TD172 8 μM 2 μM

E. coli AR Bank #0114 b 64 μM 8μM

K. pneumoniae AR Bank #0113 b >256 μM 128 μM

A. baumannii Naval-17 b 32 μM 8 μM

P. aeruginosa PA01 >256 μM 64 μM

P. aeruginosa AR Bank #0229 b >256 μM 64 μM

E. cloacae AR Bank #0136 b 256 μM 64 μM a Wild type, Taken from reference 9 b Multi-drug resistant clinical isolate These results indicate a strong synergy between ClavA and zinc ions when coupled with the MIC values for Zn(II) against these pathogens (Table 3-2). The lowest MIC recorded for Zn(II) was 1mM against E. coli MG1655 and A. baumannii Naval-17, which is still double the highest zinc concentration used in any of the ClavA-Zn(II) MIC measurements. In many cases, the MIC of Zn(II) with ClavA is a full order of magnitude lower than the MIC of Zn(II) alone. This improvement in activity clearly indicates strong synergism between ClavA and Zn(II) ions.

Table 3-2

ClavA Zn(II) Antimicrobial Synergy

Bacterial Strain Zn2+ MIC FIC Definition a

E. coli MG1655 1024 μM 0.070 Synergy

E. coli TD172 4096 μM 0.251 Synergy

E. coli AR Bank #0114 2048 μM 0.133 Synergy

K. pneumoniae AR Bank #0113 8192 μM 0.281 Synergy

A. baumannii Naval-17 1024 μM 0.266 Synergy

P. aeruginosa PA01 2048 μM 0.187 Synergy

P. aeruginosa AR Bank #0229 2048 μM 0.187 Synergy

E. cloacae AR Bank #0136 2048 μM 0.313 Synergy a Synergy is defined as an FIC ≤ 0.5

Clavanin A shows nuclease activity in cellulae

Although we have shown evidence that ClavA-Zn(II) can damage DNA in an in vitro assay,9 evidence for this damage in cellulae is absent. In order to prove that this activity occurs within living bacteria, we employed E. coli TD172 (ΔrecA::kan) in which the gene for the DNA repair protein RecA has been deleted. Treatments which lead to DNA damage will thus be more potent as the ΔrecA strain is less capable of repairing the damage.

Both ClavA and ClavA-Zn(II) show improved activities against E. coli TD172 as compared with wild type E. coli (Table 3-1), though the two-fold increase in activity for ClavA-Zn(II) does not represent a significant change. It is likely that the range of 2-4 μM represents the

78 absolute minimum amount of ClavA-Zn(II) required to inhibit growth through DNA cleavage. The fact that both ClavA and ClavA-Zn(II) approach this minimum amount in the ΔrecA E. coli provides more concrete evidence that DNA damage is important to the activity of ClavA.

While MICs against E. coli TD172 provide evidence that E. coli which are incapable of repairing DNA damage are more susceptible to ClavA and ClavA-Zn(II), these results do not definitively prove that DNA damage is occurring, nor can they quantify the amount of damage.

With that aim in mind, we employed the terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay, which fluorescently labels strand breaks to allow for quantification. As can be seen in Figure 3-1, the DNA damage caused by ClavA-Zn(II) is steady across all time points at about 75% higher than baseline, whereas ClavA alone takes more time to reach the same level, with the two showing statistically identical levels of damage after 4 hours. It is likely that ClavA-

Zn(II) is causing immediate DNA damage due to the abundance of Zn(II) ions available for binding, while ClavA alone takes longer to achieve statistically similar levels due to the need to scavenge zinc ions from the E. coli themselves. These results provide proof that the ClavA-Zn(II) complex is capable of damaging DNA in intact bacterial cells.

Figure 3-1

TUNEL assay results for ClavA and ClavA-Zn showing the fold increase in DNA

damage over the untreated control.

Further evidence for DNA as the target of ClavA-Zn(II) can be observed in the MIC values obtained using the all D-amino acid form of ClavA. The MICs observed for D-ClavA and D-

ClavA-Zn(II) against E. coli MG1655 were both found to be 32 μM, showing none of the potentiation usually observed for the zinc containing peptide. Often, HDPs with achiral targets, such as the membrane, will show minimal change in MIC when D-amino acids are used, whereas those with chiral targets, like DNA, will see diminished activity.21-24 The change in MIC from 4

μM to 32 μM for ClavA-Zn(II) is indicative of a peptide with a chiral target, further promoting the likelihood of DNA as the target molecule.

QM/MM allows for the determination of the nuclease mechanism of ClavA-Zn(II)

The electrostatic surface potential and helical wheel projection of ClavA displaying its amphipathic character are shown in Figure 3-2. In the α-helical conformation, the hydrophilic and

80 lipophilic residues are located on opposite sides of the peptide. Every pose provided by molecular docking show that the positively charged hydrophilic side of ClavA binds to the negatively charged major groove of DNA, while the hydrophobic side of ClavA is exposed the water solvent. In the lowest energy docked pose, ClavA occupies the major groove of DNA and interacts through multiple hydrogen bonds and CH-π interactions and the Zn(II) ion is located ~8 Å away from the closest phosphate group of DNA. During the MD simulations using this pose, the Zn(II) ion rapidly approaches the phosphate group of DNA and binds the negatively charged oxygen atom in a monodentate fashion. In the equilibrated structure, residues Lys7, His10, His11, Asn14, His17 and

Ser20 of ClavA are responsible for retaining its position in the major groove of DNA through hydrogen bonding with phosphate groups, while His10 and His17 associate through CH-π interactions with cytosine base pairs. These interactions also help to maintain the alpha helicity of the fragment of ClavA that is inserted in the major groove of DNA. The small root-mean-square- fluctuations (RMSFs) support the overall stability of the α-helical structure of ClavA (Figure 3-3).

In the most representative structure derived from MD simulations, the Zn(II) ion is found to interact with His17, His21, phosphate group (in the monodentate form) and three solvent water molecules (Figure 3-4).

Figure 3-2

(a) Electrostatic surface potential and (b) helical wheel projection of ClavA showing the

amphipathicity of the peptide.

Figure 3-3

Root-mean-square-fluctuations (RMSFs) in the ClavA-Zn complex.

Figure 3-4

The MD equilibrated structure of the ClavA-Zn complex used for the QM/MM (ONIOM)

calculations. Some hydrogen atoms and solvent water molecules were omitted for clarity.

Nitrogen in blue, phosphorus in orange and oxygen in red.

This structure was subsequently utilized to investigate the mechanism of DNA hydrolysis by ClavA through hybrid two-layer (QM/MM) ONIOM calculations. In the mechanism, the

Zn(II) ion is proposed to be involved in the Lewis acid activation, nucleophile activation and generation of a good leaving group25-29. In the optimized structure of the reactant (R), Zn(II) is coordinated to two histidine residues (His17 and His21), one hydroxyl ion (-OH), a water molecule and a phosphate group of DNA (Figure 3-5). The binding of a water molecule to the

Zn(II) ion and interactions with another water and phosphate groups significantly lowers its pKa value. This lowering of the pKa of water to generate a hydroxyl group through metal coordination and hydrogen bonding is a common feature of metallohydrolases30-31. Additionally, two water molecules (W1 and W2) are also present at the active site. These water molecules form a hydrogen bonding network between the Zn(II) bound OH group and the phosphate groups including Cy10 and Cyt11 of DNA (Figure 3-4). Furthermore, due to the Lewis acidity of the Zn

83 ion, the scissile P-O bond (1.53 Å) of DNA in R is elongated by 0.05 Å in comparison to the bond (1.48 Å) in its free form. In the first step, the Zn(II) bound OH functions as a base and abstracts a proton from W2 through the bridging W1 water, i.e. the W2→W1→OH(Zn) pathway.

This chain of proton transfer causes the formation of a nucleophile (-OH) close to the phosphate group, which simultaneously attacks the electrophilic phosphorus (P) atom and cleaves the scissile phosphoester [P-O(Cyt10)] bond (Figure 3-5). This concerted process occurs with a barrier of 30.2 kcal/mol and the optimized transition state (T1) is shown in Figure 3-5. This is the rate-limiting step of the entire mechanism. The reason for this high barrier is the low acidity of W2 in the major grove. The removal of W1 in the model significantly increases the barrier for this process by 6.2 kcal/mol i.e. 36.4 kcal/mol from the corresponding reactant. Additionally, a direct attack by the Zn(II) coordinated -OH nucleophile, without the assistance of both W1 and

W2, takes place with a barrier further higher by 11.1 kcal/mol i.e. 41.3 kcal/mol from the reactant. These results show that in this ligand environment the Zn(II) bound -OH is a weak nucleophile. Furthermore, formation of the pentavalent phosphorane intermediate created by the direct Zn(II)-OH nucleophilic attack is endothermic by 45 kcal/mol. In the intermediate (In1), formed through T1, one of the DNA strands is broken to create a negatively charged cytosine

(Cyt10O-) fragment (Figure 3-5). Since this charged fragment is not a good leaving group, the formation of In1 is endothermic by 20.7 kcal/mol from R. Due to the bulkiness of the Cyt10O- fragment, it cannot move and accept a proton from the -OH nucleophile that is now coordinated to the P atom. Therefore, the neutralization of this fragment requires two additional steps. In the next step, from In1 the proton from the nucleophile through W1 is transferred to the non- esterified O1 atom. This step occurs with a small barrier of 1.7 kcal/mol from In1 and the optimized transition state (T2) is shown in Figure 3-5. The intermediate (In2) formed in this step

84 is exothermic by 23.2 kcal/mol from the reactant R. In In2, the Cyt10 fragment is ideally located to abstract the O1H proton. In the final step, this transfer takes place with an extremely low barrier of 0.5 kcal/mol from In2 i.e. 22.7 kcal/mol from R (Figure 3-5). The product (P) with the neutral form of the Cyt10 fragment is exothermic by 30.7 kcal/mol.

Figure 3-5

The proposed mechanism for the DNA hydrolysis by the ClavA-Zn complex via activation

of secondary coordination sphere water molecules.

3.3 – Discussion

In this study we were able to more definitively prove a hypothesis we presented in our previously published work,9 namely that not only is the antimicrobial activity of the HDP ClavA

86 potentiated by the presence of zinc ions, but also that ClavA-Zn(II) is capable of functioning as a metallonuclease within live bacteria. We were also able to propose a probable mechanism for the cleavage of DNA through the application of QM/MM. This mechanism is particularly novel, as most metallonucleases cleave DNA via activated water molecules in the primary coordination sphere, whereas ClavA-Zn(II) utilizes a secondary coordination sphere water molecule to cleave the phosphoester bond in the DNA backbone.

In addition to all of the mechanistic work on ClavA, we were also able to expand the scope of the Zn(II) based potentiation of ClavA to encompass not only E. coli, but also all of the gram-negative ESKAPE pathogens, including some multi-drug resistant clinical isolates. These results seem to imply that ClavA-Zn(II) will show a similar potentiation against all gram- negative pathogens. This finding is of particular importance as gram-negative bacteria are particularly difficult to treat with conventional methods due to their double membrane structure.18 Gram-negative bacteria have a significantly lower success rate than gram-positive bacteria when it comes to drug screening tests, with P. aeruginosa reporting up to 1000-fold fewer successes than gram-positive pathogens.32 The successful application of ClavA-Zn(II) to the treatment of these difficult pathogens could help to encourage the development of peptidic antibiotics for the treatment of gram-negative infections, as well as the development of new strategies for bypassing the particularly resilient gram-negative membrane.

Many transition metals, when presented in sufficient quantities, have proven to be toxic to bacteria, and zinc is no exception. We have shown that the MIC values for Zn(II) against the bacteria used in this study are all in the low millimolar range, while in the MIC experiments for

ClavA-Zn(II) the highest concentration of Zn(II) used is 512μM. When fractional inhibitory concentration (FIC) is calculated it is clear that the interaction of ClavA with Zn(II) ions is

87 synergistic, with bacterial growth inhibition occurring at much lower concentrations than for each alone.

While we have established that ClavA-Zn(II) is able to function as a nuclease in cellulae, there are still some questions that still need answers. It is unclear from the results of this study and previous work exactly how ClavA is able to cross the membrane.9, 33 This information will be important if peptide-based antibiotics are to be engineered for future use. Additionally, there is still a discrepancy regarding the mechanism of action for ClavA, as publications from three independent laboratories identify the membrane as the target for this HDP.33-35 Despite all that has already been discovered for this remarkable system, it is clear that there is more to be done before we can close the book on ClavA.

3.4 – Methods

Bacterial Culture Conditions

Bacteria treated under low pH conditions were grown in LB media containing 50 mM

MES buffer at pH 5.5. For the neutral pH experiments, bacteria were grown and treated in standard MHB media at pH 7.4. Cells were incubated at 37℃ and 234 rpm to mid-log phase

(OD600 = 0.4-0.6).

Minimum Inhibitory Concentration Determination

Minimum Inhibitory Concentrations (MICs) for ClavA and ClavA-Zn(II) were determined via the broth microdilution method, which has been previously described.36 Briefly,

6 bacteria were grown to mid-log phase (OD600= 0.4-0.6), diluted to a final concentration of 1x10 cfu/mL and treated with serially diluted peptide for 18 hr. All of the growth media used contained 50mM MES buffer at pH 5.5. MIC was determined as the concentration of peptide at

88 which no growth could be visually observed. The data presented represent the mode of three trials. For all experiments containing ClavA-Zn(II), the peptides were incubated with ZnCl2 for

30 min at room temperature at 2x the maximum concentration used, with the maximum zinc concentration of any of these assays being 512 μM.

When determining MIC values for Zn(II), a similar process was utilized. Diluted bacteria were treated with ZnCl2, and MICs correspond to the lowest concentration at which no growth could be observed visually. Values were measured at pH 5.5 and 7.4, and were found to be identical. Results represent the mode of three trials.

TUNEL Assay

The TUNEL assay has been described previously.37 Briefly, an overnight culture of E. coli was washed in PBS. The cells were then treated with MIC concentrations of ClavA and ClavA-

Zn(II), with aliquots removed at 1 hr, 2 hr and 4 hr. A negative control was performed with PBS and a positive control with 1.5% H2O2. After incubation, cells were fixed and dyed, then propidium iodide and fluorescein labeled cells were measured via flow cytometry. Data presented represent the average of three trials, and the error bars are standard error from mean.

Computational Details

Molecular Dynamics (MD) simulations

The structure of ClavA was obtained from the protein data bank (PDB ID: 6C41).38 The initial ClavA-Zn structure in which His17 and His21 were coordinated to the Zn ion was build and energy minimized using the YASARA software.39 The CAAGACGGCCCCGGCCC nucleotide sequence40 of E. coli TD172 DNA was built using the Avogadro software41. The HDOCK webserver was used to explore the binding poses of the ClavA-Zn complex to DNA.42 In all the

89 docked structures, ClavA-Zn was located in the major groove of DNA. Two different ClavA-Zn binding poses were used as the starting structures for molecular dynamics (MD) simulations i.e.

(1) ClavA-Zn is bound in a perpendicular manner inside the major groove of DNA and (2) ClavA-

Zn is bound parallel to the major groove of DNA (Figure 3-6). All 200 ns all-atom MD simulations in aqueous solution were performed using the GROMACS-4.5.643 program utilizing the

CHARMM3644 force field. This force field was reported to reproduce the key structural features of ClavA and ClavA-Zn in our previous work.45 The ClavA-Zn--DNA complex was placed in the cubic box of 10 × 10 × 10 nm dimension. The shortest distance from the edge of the box to the surface of the complex was 1.0 nm. Particle Mesh Ewald method46 was used to compute the electrostatic interactions and for both Coulombic and van der Waals interactions 1.2 nm cutoff distance was maintained. TIP3P47 water was used as a solvent and Na+ and Cl- were added to neutralize and to maintain the physiological ion concentration (154 mM) of the system. The starting structures for MD simulations were energy minimized for 3000 steps using the steepest descent method. The simulations were performed using a constant number of particles (N), pressure (P) and temperature (T) i.e. NPT ensemble. SETTLE48 algorithm was used to constrain the bond length and angle of the water molecules and LINCS49 algorithm was employed to constrain the remaining bond lengths. The trajectories were computed for each model with a time step of 2 fs. Cluster analysis was utilized to obtain the most representative ClavA-Zn--DNA structure from the MD simulations. The ClavA-Zn and DNA binding energies were calculated by utilizing the Molecular Mechanics Poisson–Boltzmann Surface Area (MM-PBSA) method.50 A comparison of the binding energies shows that which shows that the structure in which ClavA was bound in a perpendicular fashion inside the major groove of DNA is 37.0 kcal/mol lower in energy than the second structure (Figure 3-6).

Figure 3-6

Two MD equilibrated conformations of the ClavA-Zn--DNA complex. The

perpendicular arrangement has a lower energy.

QM/MM simulations

All hybrid two-layer QM/M (ONIOM) calculations were performed using the

Gaussian 09 (ONIOM) software package.51 ONIOM utilizes a subtractive method in which the MM energy [EMM (model)] of the QM (model) part is subtracted from the sum of the

QM energy of the model [EQM (model)] and MM energy [EMM (real)] of the whole (real) system, i.e., EONIOM (QM/MM) = EQM (model) + EMM (real) ‐ EMM (model). This

91 subtraction method corrects the artefacts introduced by the link atom.52 The ONIOM optimization procedure uses macro/microiterations53 and the electrostatic interactions between the QM and the MM part were treated by mechanical embedding (ME). The QM region comprised of Zn(II) ion, His17, His21, two cytosine nucleotides, a hydroxyl ion and one to three water molecules (depending on the reaction mechanism) as shown in Figure 3-

4. The remaining part of the system was included in the MM layer. The ClavA-Zn--DNA complex was surrounded with explicit water molecules (2139 water molecules; MM layer) to stabilize the highly negative charge of DNA. The QM layer was optimized without any geometric constraints using the hybrid B3LYP54 exchange-correlation functional. The

Lanl2dz basis set (for the Zn atom) with the corresponding Hay-Wadt effective core potential (ECP)55 and the 6-31G(d) basis set (for the remaining atoms in the QM layer) were used. The MM layer was modeled using the AMBER force field.56-57 The energies of the optimized structures were further improved by performing single-point calculations using the large basis sets e.g. Lanl2tz for Zn and triple-zeta 6-311+G(d,p) for other atoms.

Hessians were calculated at the same level of theory as the optimizations to confirm the nature of the stationary points along the reaction coordinate. The transition states were confirmed to have only one imaginary frequency corresponding to the reaction coordinates.

The transition states were confirmed to have only one imaginary frequency corresponding to the reaction coordinates. Zero-point vibrational, thermal and entropy corrections were added to the final energies (at 298.15 K and 1 atm). Helical wheel projection of ClavA was obtained from NetWheel online webserver.58 VMD59, YASARA39, ChemDraw and

Chimera60 programs were used for the visualization of the MD trajectories and preparation of the figures used in this study.

3.5 – References

1. OECD, Stemming the Superbug Tide. 2018. 2. Klevens, R. M.; Edwards, J. R.; Richards, C. L., Jr.; Horan, T. C.; Gaynes, R. P.; Pollock, D. A.; Cardo, D. M., Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep 2007, 122 (2), 160-166. 3. Rice, L. B., Federal Funding for the Study of Antimicrobial Resistance in Nosocomial Pathogens: No ESKAPE. J Infect Dis 2008, 197 (8), 1079-1081. 4. Santajit, S.; Indrawattana, N., Mechanisms of Antimicrobial Resistance in ESKAPE Pathogens. Biomed Res Int 2016, 2016, 2475067-2475067. 5. Livermore, D. M., The need for new antibiotics. Clin Microbiol Infect 2004, 10 Suppl 4, 1-9. 6. Munoz-Price, L. S.; Poirel, L.; Bonomo, R. A.; Schwaber, M. J.; Daikos, G. L.; Cormican, M.; Cornaglia, G.; Garau, J.; Gniadkowski, M.; Hayden, M. K.; Kumarasamy, K.; Livermore, D. M.; Maya, J. J.; Nordmann, P.; Patel, J. B.; Paterson, D. L.; Pitout, J.; Villegas, M. V.; Wang, H.; Woodford, N.; Quinn, J. P., Clinical epidemiology of the global expansion of Klebsiella pneumoniae carbapenemases. Lancet Infect Dis 2013, 13 (9), 785-96. 7. Silva, O. N.; Fensterseifer, I. C.; Rodrigues, E. A.; Holanda, H. H.; Novaes, N. R.; Cunha, J. P.; Rezende, T. M.; Magalhaes, K. G.; Moreno, S. E.; Jeronimo, M. S.; Bocca, A. L.; Franco, O. L., Clavanin A improves outcome of complications from different bacterial infections. Antimicrob Agents Chemother 2015, 59 (3), 1620-6. 8. Silva, O. N.; Alves, E. S. F.; de la Fuente-Núñez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A. R. B.; Andrade, C. A. S.; Nascimento, J. M.; Fensterseifer, I. C. M.; Porto, W. F.; Correa, J. R.; Hancock, R. E. W.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti-Biofilm Activity. Sci Rep 2016, 6, 27128. 9. Juliano, S. A.; Pierce, S.; deMayo, J. A.; Balunas, M. J.; Angeles-Boza, A. M., Exploration of the Innate Immune System of Styela clava: Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide Clavanin A. Biochemistry 2017, 56 (10), 1403-1414. 10. Alexander, J. L.; Thompson, Z.; Cowan, J. A., Antimicrobial Metallopeptides. ACS Chem Biol 2018. 11. Gammoh, N. Z.; Rink, L., Zinc in Infection and Inflammation. Nutrients 2017, 9 (6). 12. Vallee, B. L.; Falchuk, K. H., The biochemical basis of zinc physiology. Phys Rev 1993, 73 (1), 79-118. 13. Maret, W.; Li, Y., Coordination dynamics of zinc in proteins. Chem Rev 2009, 109 (10), 4682- 707. 14. Sigel, A.; Sigel, H.; Sigel, R. K., Interrelations between essential metal ions and human diseases. Springer: 2013. 15. Prasad, A. S., Zinc in Human Health: Effect of Zinc on Immune Cells. Mol Med 2008, 14 (5), 353-357. 16. McDevitt, C. A.; Ogunniyi, A. D.; Valkov, E.; Lawrence, M. C.; Kobe, B.; McEwan, A. G.; Paton, J. C., A Molecular Mechanism for Bacterial Susceptibility to Zinc. PLoS Pathog 2011, 7 (11), e1002357. 17. Wagner, D.; Maser, J.; Lai, B.; Cai, Z.; Barry, C. E.; Höner zu Bentrup, K.; Russell, D. G.; Bermudez, L. E., Elemental Analysis of Mycobacterium avium-, Mycobacterium tuberculosis-, and Mycobacterium

93 smegmatis-Containing Phagosomes Indicates Pathogen-Induced Microenvironments within the Host Cell’s Endosomal System. J Immunol 2005, 174 (3), 1491. 18. Zgurskaya, H. I.; López, C. A.; Gnanakaran, S., Permeability Barrier of Gram-Negative Cell Envelopes and Approaches To Bypass It. ACS Infect Dis 2015, 1 (11), 512-522. 19. Murata, T.; Tseng, W.; Guina, T.; Miller, S. I.; Nikaido, H., PhoPQ-mediated regulation produces a more robust permeability barrier in the outer membrane of Salmonella enterica serovar typhimurium. J Bacteriol 2007, 189 (20), 7213-7222. 20. CDC & FDA Antibiotic Resistance Isolate Bank. CDC: Atlanta (GA), 2018. 21. Wade, D.; Boman, A.; Wåhlin, B.; Drain, C. M.; Andreu, D.; Boman, H. G.; Merrifield, R. B., All-D amino acid-containing channel-forming antibiotic peptides. Proc Natl Acad Sci 1990, 87 (12), 4761-4765. 22. Won, A.; Khan, M.; Gustin, S.; Akpawu, A.; Seebun, D.; Avis, T. J.; Leung, B. O.; Hitchcock, A. P.; Ianoul, A., Investigating the effects of L- to D-amino acid substitution and deamidation on the activity and membrane interactions of antimicrobial peptide anoplin. BBA Biomembranes 2011, 1808 (6), 1592-1600. 23. Casteels, P.; Tempst, P., Apidaecin-Type Peptide Antibiotics Function through a Nonporeforming Mechanism Involving Stereospecificity. Biochem Biophys Res Commun 1994, 199 (1), 339-345. 24. Bulet, P.; Urge, L.; Ohresser, S.; Hetru, C.; Otvos Jr, L., Enlarged Scale Chemical Synthesis and Range of Activity of Drosocin, an O-Glycosylated Antibacterial Peptide of Drosophila. Eur J Biochem 1996, 238 (1), 64-69. 25. Chin, J., Developing artificial hydrolytic metalloenzymes by a unified mechanistic approach. Acc Chem Res 1991, 24 (5), 145-152. 26. Bashkin, J. K.; Jenkins, L. A., The Role of Metals in the Hydrolytic Cleavage of DNA and RNA. Comment Inorg Chem 1994, 16 (1-2), 77-93. 27. Fothergill, M.; Goodman, M. F.; Petruska, J.; Warshel, A., Structure-Energy Analysis of the Role of Metal Ions in Phosphodiester Bond Hydrolysis by DNA Polymerase I. J Am Chem Soc 1995, 117 (47), 11619-11627. 28. Williams, N. H.; Takasaki, B.; Wall, M.; Chin, J., Structure and Nuclease Activity of Simple Dinuclear Metal Complexes: Quantitative Dissection of the Role of Metal Ions. Acc Chem Res 1999, 32 (6), 485-493. 29. Das, P.; Chandar, N. B.; Chourey, S.; Agarwalla, H.; Ganguly, B.; Das, A., Role of Metal Ion in Specific Recognition of Pyrophosphate Ion under Physiological Conditions and Hydrolysis of the Phosphoester Linkage by Alkaline Phosphatase. Inorg Chem 2013, 52 (19), 11034-11041. 30. Prince, R. H.; Woolley, P. R., Metal Ion Function in Carbonic Anhydrase. 1972, 11 (5), 408- 417. 31. Groves, J. T.; Olson, J. R., Models of zinc-containing proteases. Rapid amide hydrolysis by an unusually acidic Zn2+-OH2 complex. Inorg Chem 1985, 24 (18), 2715-2717. 32. Silver, L. L., Challenges of antibacterial discovery. Clin Microbiol Rev 2011, 24 (1), 71-109. 33. van Kan, E. J.; Demel, R. A.; Breukink, E.; van der Bent, A.; de Kruijff, B., Clavanin permeabilizes target membranes via two distinctly different pH-dependent mechanisms. Biochemistry 2002, 41 (24), 7529-39. 34. Lee, I. H.; Cho, Y.; Lehrer, R. I., Effects of pH and salinity on the antimicrobial properties of clavanins. Infect Immun 1997, 65 (7), 2898-903. 35. Silva, O. N.; Alves, E. S.; de la Fuente-Nunez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A.; Andrade, C. A.; Nascimento, J. M.; Fensterseifer, I. C.; Porto, W.

F.; Correa, J. R.; Hancock, R. E.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti-Biofilm Activity. Sci Rep 2016, 6, 27128. 36. Wiegand, I.; Hilpert, K.; Hancock, R. E., Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat Protoc 2008, 3 (2), 163- 75. 37. Vilcheze, C.; Hartman, T.; Weinrick, B.; Jacobs, W. R., Jr., Mycobacterium tuberculosis is extraordinarily sensitive to killing by a vitamin C-induced Fenton reaction. Nat Commun 2013, 4, 1881. 38. Silva, O. N.; Alves, E. S.; De La Fuente-núñez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A.; Andrade, C. A.; Nascimento, J. M., Structural studies of a lipid- binding peptide from tunicate hemocytes with anti-biofilm activity. Sci Rep 2016, 6, 27128. 39. Krieger, E.; Vriend, G., YASARA View—molecular graphics for all devices—from smartphones to workstations. Bioinformatics 2014, 30 (20), 2981-2982. 40. Blattner, F. R.; Plunkett, G.; Bloch, C. A.; Perna, N. T.; Burland, V.; Riley, M.; Collado-Vides, J.; Glasner, J. D.; Rode, C. K.; Mayhew, G. F., The complete genome sequence of Escherichia coli K-12. Science 1997, 277 (5331), 1453-1462. 41. Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.; Hutchison, G. R., Avogadro: advanced semantic chemical editor, visualization, and analysis platform. J Cheminformatics 2012, 4 (1), 17. 42. Yan, Y.; Zhang, D.; Zhou, P.; Li, B.; Huang, S.-Y., HDOCK: a web server for protein–protein and protein–DNA/RNA docking based on a hybrid strategy. Nucleic Acid Res 2017, 45 (W1), W365-W373. 43. Hess, B.; Kutzner, C.; Van Der Spoel, D.; Lindahl, E., GROMACS 4: algorithms for highly efficient, load-balanced, and scalable molecular simulation. J Chem Theory Comput 2008, 4 (3), 435-447. 44. Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I., CHARMM general force field: A force field for drug‐like molecules compatible with the CHARMM all‐atom additive biological force fields. J Comput Chem 2010, 31 (4), 671-690. 45. Duay, S. S.; Sharma, G.; Prabhakar, R.; Angeles-Boza, A. M.; May, E. R., Molecular Dynamics Investigation into the Effect of Zinc (II) on the Structure and Membrane Interactions of the Antimicrobial Peptide Clavanin A. J Phys Chem B 2019, 123 (15), 3163-3176. 46. Darden, T.; York, D.; Pedersen, L., Particle mesh Ewald: An N⋅ log (N) method for Ewald sums in large systems. J Chem Phys 1993, 98 (12), 10089-10092. 47. Price, D. J.; Brooks III, C. L., A modified TIP3P water potential for simulation with Ewald summation. J Chem Phys 2004, 121 (20), 10096-10103. 48. Miyamoto, S.; Kollman, P. A., Settle: An analytical version of the SHAKE and RATTLE algorithm for rigid water models. J Comput Chem 1992, 13 (8), 952-962. 49. Hess, B.; Bekker, H.; Berendsen, H. J.; Fraaije, J. G., LINCS: a linear constraint solver for molecular simulations. J Comput Chem 1997, 18 (12), 1463-1472. 50. Kumari, R.; Kumar, R.; Consortium, O. S. D. D.; Lynn, A., g_mmpbsa A GROMACS tool for high-throughput MM-PBSA calculations. J Chem Inf Model 2014, 54 (7), 1951-1962. 51. Frisch, M.; Trucks, G.; Schlegel, H.; Scuseria, G.; Robb, M.; Cheeseman, J.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G., Gaussian 09, revision D. 01. Gaussian, Inc., Wallingford CT: 2009.

52. Senn, H. M.; Thiel, W., QM/MM methods for biomolecular systems. Angew 2009, 48 (7), 1198-1229. 53. Vreven, T.; Morokuma, K.; Farkas, Ö.; Schlegel, H. B.; Frisch, M. J., Geometry optimization with QM/MM, ONIOM, and other combined methods. I. Microiterations and constraints. J Comput Chem 2003, 24 (6), 760-769. 54. Becke, A. D., Density-functional exchange-energy approximation with correct asymptotic behavior. Phys Rev A 1988, 38 (6), 3098. 55. Hay, P. J.; Wadt, W. R., Ab initio effective core potentials for molecular calculations. Potentials for the transition metal atoms Sc to Hg. J Chem Phys 1985, 82 (1), 270-283. 56. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A., Development and testing of a general amber force field. J Comput Chem 2004, 25 (9), 1157-1174. 57. Duan, Y.; Wu, C.; Chowdhury, S.; Lee, M. C.; Xiong, G.; Zhang, W.; Yang, R.; Cieplak, P.; Luo, R.; Lee, T., A point‐charge force field for molecular mechanics simulations of proteins based on condensed‐phase quantum mechanical calculations. J Comput Chem 2003, 24 (16), 1999-2012. 58. Mol, A. R.; Castro, M. S.; Fontes, W., NetWheels: A web application to create high quality peptide helical wheel and net projections. BioRxiv 2018, 416347. 59. Humphrey, W.; Dalke, A.; Schulten, K., VMD: visual molecular dynamics. J Mol Graphics 1996, 14 (1), 33-38. 60. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E., UCSF Chimera—a visualization system for exploratory research and analysis. J Comput Chem 2004, 25 (13), 1605-1612.

Chapter 4 – Three Strategies in Motion: Bacterial Cytological Profiling Elucidates the Mechanisms of Action for Clavanin A

Adapted from: Juliano, S. A.; et. al., A Potent Host Defense Peptide Triggers DNA Damage and is Active against Multi-Drug Resistant Gram-Negative Pathogens. Manuscript Submitted

4.1 – Introduction

Up to this point, our studies of Clavanin A (ClavA, VFQFLGKIIHHVGNFVHGFSHVF-

NH2), the 23-residue, tunicate derived, host defense peptide (HDP), have resulted in a thorough understanding of the Zn2+ based nuclease mechanism of action (MoA), a novel mechanism for zinc binding peptides. This, however, presented some confusion due to the fact that studies from three independent laboratories provided evidence that ClavA exerts its activity through disruption of the bacterial membrane.1-3

ClavA was discovered as a member of the Clavanin family of HDPs isolated from the hemocytes of the tunicate Styela clava. Lehrer and co-workers, who first isolated the Clavanin family, were able to demonstrate a pH dependent MoA for ClavA with the peptide showing greater activity at lower pH values.2 This idea was later expanded upon by van Kan et al. who proposed two mechanisms for ClavA acting at different pH values.1 At neutral pH (7.4) it was concluded that the MoA of ClavA involves nonspecific pore formation driven by hydrophobic effects, while at low pH (5.5) the activity was hypothesized to originate from increased permeability of the membrane to small ions, likely the result of interaction between ClavA and a membrane ionophore or ATP synthase.1

ClavA is a very good candidate for drug discovery and development as in vitro studies in murine wound infection models showed improved healing and eradication of bacteria.5 In systemic infection models, ClavA treatment resulted in increased survival for both Escherichia coli and Staphylococcus aureus infected mice. In addition to this, ClavA has also proven effective against biofilms, preventing the formation of E. coli and S. aureus biofilms and perturbing preformed biofilms of Klebsiella pneumoniae.6 ClavA has also been conjugated to the

98 surface of iron oxide nanoparticles and coated onto central venous catheters.7 ClavA-Iron oxide coated catheters were capable of reducing adhesion of both gram positive (~50%) and gram negative (70-90%) bacteria and modestly reducing the amount of planktonic bacteria (up to 20%) as well.

Despite the apparent suitability of ClavA as a potential therapeutic, our limited understanding of its chemistry makes it impossible to predict what changes in its sequence will result in more potent antimicrobial activity, a problem not limited to ClavA alone.8-16 Thus, there is need for a technique capable of quickly and inexpensively identifying the MoA of HDPs so that further studies can be targeted to obtaining more precise information on how a given peptide kills or inhibits the growth of bacteria. With this information regarding HDP mechanisms of action in hand, it becomes much more likely that antibiotics based on HDPs could help to stem the growing tide of antibiotic resistance currently plaguing society.

If novel antibiotics are going to be designed based off of the framework provided by

HDPs it is important to have a firm understanding of how natural HDPs are capable of fighting bacterial and fungal infections. The current paradigm for identifying the mechanism of action of newly discovered HDPs involves a battery of tests which often result in only being able to determine if a peptide is membrane active or internal targeting. The application of bacterial cytological profiling (BCP) to the study of HDPs will streamline the determination of mechanism of action, resulting in more targeted studies aimed at fine tuning the precise target for the peptide.17 This knowledge will then allow for the future development of peptidic antibiotics to stem the growing tide of multi-drug resistant bacteria. Through the application of BCP to

ClavA we have proposed the most complete picture of the activity of ClavA to date.

4.2 – Results

Phenotypic Analysis Reveals Multiple ClavA Mechanisms of Action

Despite the compelling evidence indicating that ClavA is capable of cleaving DNA in the presence of zinc ions, it cannot be ignored that almost all of the published research on the HDP indicates that it is membrane active.1-3, 18 We have shown previously that ClavA has antimicrobial activity in the absence of zinc ions, which would indicate that ClavA has at least two mechanisms of action depending on the environment of the peptide. In order to further explore the mechanisms employed by ClavA, we hypothesized that bacterial cytological profiling (BCP) will allow for the observation and differentiation of each of the

ClavA mechanisms of action. BCP is based on the idea that antibiotics that act on the same pathway or target will likewise cause the same phenotypic changes in treated bacteria.17

This technique has been previously applied to the study of the MoA for small molecule antibiotics and anticancer metal-ligand complexes.19, 20

The study of the antimicrobial effect of any molecule via BCP begins with phenotypic analysis of treated cells. The first step in this process is to collect and analyse micrographs of E. coli treated with ClavA. The morphology of the bacterial membranes was measured via staining with FM 4-64 (Figure 4-1, red) and its permeability reported by

SYTOX green (Figure 4-1, green), a cell impermeable DNA dye. The cell permeable DNA stain DAPI (Figure 4-1, blue) was also used to report on nucleoid morphology. These images are analysed for size and shape of the membrane and nucleoid as well as the relative intensity of the DNA dyes.

After treating E. coli with ClavA for 60 minutes at pH 5.5, we observed two distinct phenotypes (Figure 4-1C). The first phenotype, indicated by the solid white box, were

100 longer than untreated E. coli (Figure 4-1A), but maintained the same overall appearance with a ring of red for the membrane and blue DNA inside. Cells displaying the second phenotype, indicated by the dashed white box, maintained a similar overall size to untreated cells, but with a diffuse red signal throughout and with DNA stained both blue and green.

We called these two phenotypes Type A1 and Type A2, respectively, due to the acidic pH at which they are observed. Approximately 60% of the cells show Type A1 damage and the remaining 40% Type A2 after 60 minutes of treatment (ClavA damage types are summarized in Table 4-1). Based on these results we conclude that, at low pH, ClavA is actually working via two different mechanisms of action.

101

Figure 4-1

Micrographs of E. coli that have been (A) untreated, or treated with (B) ClavA for 30 min,

for 60 min at neutral pH. Scale bars are 5 μm long. Solid white boxes indicate ClavA Type

A1, while dashed white boxes indicate Type A2.

102

We hypothesized that these two mechanisms of bacterial damage were based on exposure time to ClavA. To test this, we modified the treatment time from the original 60 min to 30 and 120 min, represented by Figure 4-1B and 4-1D, respectively. After 30 min of treatment, the predominant MoA for ClavA is Type A1, with about 80% of cells falling into that category. The ratio flips at 120 min, with 75% of the cells exhibiting the Type A2 phenotype. This would indicate that the initial interaction of ClavA with E. coli results in

Type A1 damage, with Type A2 appearing after, and possibly as a result of, Type A1.

Since ClavA is potentiated by Zn(II) at low pH, we next determined if and how the addition of Zn(II) affected the phenotype of the treated E. coli. As can be seen in Figure 4-

1E, ClavA-Zn(II) treatment results exclusively in damage resembling Type A2. The visual similarities between E. coli treated with ClavA-Zn(II) and Type A2 damage point to the likelihood that these two mechanisms are the same and that the Type A2 MoA results from scavenging Zn(II) ions from E. coli or the media.

Next, we determined if and how these mechanisms differ from the activity of ClavA at neutral pH since van Kan et al. postulated a different MoA under those conditions, namely nonspecific pore formation.1 Figure 4-1F shows cells treated with ClavA at pH 7.4.

It can be observed that these cells are longer than either untreated or ClavA Type A1 damaged E. coli, and that most (~88%) of these cells have two nucleoids (Figure 4-1F).

The MoA that resulted in this visually distinct phenotype is referred to as ClavA Type N, due to the neutral pH at which this phenotype appears.

103

Table 4-1

A Summary of the ClavA Mechanisms of Action.

Damage Type General Features

Slightly longer than untreated E. coli, normal ring-shaped

Type A1 membrane

Diffuse membrane staining, blue and green stained DNA

Type A2

Longer than Type A1, most cells have two nucleoids, normal

Type N membrane

Identification of Mechanisms of Action through Principal Component Analysis

For a more quantitative approach to data analysis, we employed Principal

Component Analysis (PCA), which is a statistical treatment that reduces the dimensionality of a large dataset, while minimizing the loss of variance. In this procedure, the set of measurements of possibly correlated variables are orthogonally transformed into values of uncorrelated variables, called principal components. The first principal component reflects the variable with the highest variance, followed by the second principal component with the second highest variance, and so on. Commonly, the values of the first, second, and third principal components are used for presenting the visual 2D and 3D plots. We have elected to use the 2D plot as the variance in the third dimension was minimal.

104

Figure 4-2 shows a PCA plot of the data pertaining to the various mechanisms of action for ClavA presented above (Data collected for all treatments and used for PCA can be found in Table 4-2 and 4-3). It is clear from the plot that Type A1 and Type N are found in different regions of the plot, while Type A2 and ClavA-Zn(II) cluster together. These results corroborate what was found through phenotypic analysis, namely that Type A1,

Type A2, and Type N mechanisms are all different, and that the MoA corresponding to

Type A2 is the same as that of ClavA-Zn(II).

Figure 4-2

PCA plot of the ClavA treatments with model AMPs. The boxes indicate treatments which cluster

together.

105

Table 4-2

Average morphology measurements of treated bacterial membranes

Area Perimeter Length Width Nucleoids/ Treatment Circularity (μm2) (μm) (μm) (μm) Cell

Clavanin A 3.787 8.532 3.287 1.239 0.658 1.333 Type A1

Clavanin A 3.272 7.933 2.961 1.072 0.667 1.229 Type A2

Clavanin A 5.557 11.336 4.641 1.278 0.550 1.873 Type N

ClavA- 3.443 7.574 2.823 1.173 0.752 1.063 Zn(II)

Buforin II 3.656 8.399 3.328 1.160 0.653 1.000

Indolicidin 4.683 9.566 3.798 1.211 0.652 1.226

Ixosin 2.448 6.687 2.444 1.046 0.692 1.271

Magainin 2 6.602 13.018 5.439 1.285 0.503 1.600

Piscidin 1 5.510 11.785 4.792 1.172 0.516 1.342

Piscidin 3 4.996 10.914 4.436 1.145 0.539 1.449

106

Table 4-3

Average morphology measurements of treated bacterial nucleoids

Corrected Corrected Area Perimeter Length Width Treatment Circularity SYTOX DAPI (μm2) (μm) (μm) (μm) Intensity Intensity

Clavanin A 1.288 4.845 1.967 0.693 0.677 1.285 2.869 Type A1

Clavanin A 1.746 5.641 2.083 0.845 0.688 15.165 3.626 Type A2

Clavanin A 1.425 5.068 1.919 0.775 0.701 1.005 0.842 Type N

ClavA- 2.476 6.643 2.480 0.945 0.702 15.644 2.430 Zn(II)

Buforin II 1.712 5.868 2.339 0.739 0.635 1.280 1.674

Indolicidin 1.633 5.965 2.535 0.609 0.582 1.023 1.929

Ixosin 0.738 3.696 1.364 0.652 0.678 5.096 1.526

Magainin 2 2.173 7.038 2.886 0.743 0.559 1.069 1.349

Piscidin 1 1.816 6.838 2.902 0.675 0.519 1.108 1.299

Piscidin 3 1.852 6.270 2.371 0.735 0.604 1.122 0.929

107

PCA can be more informative when used in conjunction with treatments of a known

MoA. When these established treatments cluster with the treatment of interest, it indicates that their mechanisms of action likely share the same target molecule or pathway. For this study we used six HDPs with well-defined mechanisms of action: (i) magainin 2, a membrane active HDP known to form toroidal pores in the membranes of both gram- positive and gram-negative bacteria;21-23 (ii) buforin II, an HDP that binds to DNA after crossing the cellular membrane of bacteria without damaging it;24-26 (iii) indolicidin, a non- lytic peptide with an internal target capable of creating pores in the inner and outer membranes of gram-negative bacteria;9, 27 (iv) ixosin, an HDP that exerts its activity against the membrane through oxidative damage utilizing bound Cu(II), resulting in membrane perturbation28, 29; and (v-vi) piscidin 1 and 3 which have been shown to bind both the membrane and DNA.30 Piscidin 1 interacts more strongly with the membrane, while piscidin 3 interacts more strongly with DNA. Micrographs for each of the established HDPs can be found in Figure 4-3.

108

Figure 4-3

Representative micrographs of E. coli treated with (A) Buforin II, (B) Indolicidin, (C)

Ixosin, (D) Magainin 2, (E) Piscidin 1, and (F) Piscidin 3. Scale bars represent 5 μm.

109

Based on the varied mechanisms of action presented above, it is no surprise that all of our model peptides localize in different parts of the PCA plot, with the exception of piscidin 1 and 3, which are found close together (Figure 4-2). ClavA Type A1 clusters with indolicidin, indicating that Type A1 damage is likely similar to that inflicted by indolicidin.

ClavA Type N clusters with piscidin 1 and 3, as well as appearing near, but not clustering with, magainin 2. This seems to indicate that while ClavA at neutral pH does interact with the membrane, it is not forming pores as previously believed.1 In contrast with both Type

A1 and Type N, Type A2 does not cluster with any of the established peptide and instead only clusters with ClavA-Zn(II), which serves to emphasize the uniqueness of the MoA proposed for ClavA-Zn(II), and supports the observation that these mechanisms of action are the same from the phenotypic analysis.

Membrane interaction studies reveal lack of pore formation in model membranes

While we have explored in detail the MoA of ClavA TypeA2, it is also important to take a closer look at Type N and Type A1. Considering both of these have been described previously by van Kan et al. as membrane based, and each clusters with HDPs that have membrane activity as a part of their MoA, we decided to explore how ClavA interacts with the membrane at pH 7.4 and 5.5.1 First, we performed circular dichroism (CD) spectroscopy in the presence of lipid vesicles in order to study interaction of the peptide with the membrane, the results of which are summarized in Table 4-4. These results show that the peptide interacts strongly with the membrane at pH 5.5, achieving almost 100% helical content at a peptide to lipid ratio of 1:80 (Figure 4-4A). Combined with the fact that Type

A1 clusters with indolicidin, this would seem to indicate that the MoA for Type A1, and likely the means by which ClavA enters bacterial cells, is similar to the membrane activity

110 of indolicidin. The presence of Zn2+ appears to drastically reduce this interaction, as can be seen by the 70% reduction in α-helical content. At pH 7.4, however, ClavA shows low alpha helical content and largely inconsistent results across three trials (Figure 4-4B), which seems to indicate breakdown of the lipid vesicles. This is somewhat consistent with findings by van Kan et al. that, at neutral pH, ClavA causes more leakage of calcein than at 5.5.1

These results need to be explored further in order to draw meaningful conclusions, though suffice it to say that the membrane interactions at pH 5.5 are much stronger than those at

7.4

Table 4-4

Percent α-helix formed by ClavA in the presence and absence of Zn(II) When exposed to lipid vesicles

% Helical Content P:L Ratio pH 5.5 pH 5.5+Zn2+ pH 7.4 pH 7.4+Zn2+

1:0 9.59 -1.75 17.14 5.49

1:2 24.25 4.80 9.78 3.95

1:4 24.36 14.85 2.66 8.73

1:10 45.68 26.61 22.75 20.62

1:20 71.18 26.72 38.45 23.79

1:40 91.41 27.69 41.03 24.26

1:80 95.09 25.79 41.50 24.24

111

Figure 4-4

Circular Dichroism spectra of ClavA exposed to lipid vesicles at (A) pH 5.5, and (B) pH

7.4. Each line represents a different peptide to lipid ratio.

Many membrane active HDPs have been reported to form pores, especially magainin

2, while others, such as piscidin 1 and 3, can insert into the membrane and kill bacteria without the formation of transmembrane pores. With this in mind, we next sought to determine if nonmetallated ClavA behaves more like the former or the latter. The results of

X-ray diffraction experiments (Figure 4-5A) indicate that at pH 7.4, ClavA appears to remain on the surface of the membrane, resulting in membrane thickening, which is in direct contrast to the membrane thinning observed for magainin 2.31 Additionally, Surface

Plasmon Resonance combined with Electrical Impedance Spectroscopy (SPR/EIS) indicate that ClavA interaction with the membrane does not result in the formation of pores (Figure

4-5B and 4-5C). In fact, ClavA appears to increase the resistivity of the membrane, the opposite effect of what would occur during pore formation. These results together indicate that while ClavA may be interacting with the membrane, it is doing so in a way that is

112 fundamentally different from either magainin 2 or the piscidins, though clustering with piscidin 1 and 3 does indicate some similarity in target or activity. Further study of the effect of ClavA interaction with bacterial membranes is necessary to fully understand what is occurring in the Type N and Type A1 mechanisms.

Figure 4-5

(A) X-ray diffraction for POPC in the presence of ClavA. Lamellar diffraction from aligned

multilayers of POPC without (black) and with Clavanin A (red). There is an increase in the

repeat spacing with Clav A by roughly 1 Å, compared to POPC control. (B) Surface plasmon

resonance and electrical impedance spectroscopy of ClavA interacting with POPC were

measured simultaneously on a tethered bilayer membrane, at pH 7.4 and 5.5. At both pHs, the

changes in the SPR signal is negligible, indicative of weak or no adsorption of ClavA to the

bilayer. The increase in bilayer resistance seen in the EIS spectra may be due to superficial

peptide binding and occlusion of bilayer defects.

113

4.3 – Discussion

We were also able to bridge the gap between our results, which show that ClavA can act on DNA, resulting in growth inhibition, and the results first proposed by van Kan and coworkers that provide evidence for ClavA as a membrane active peptide.1 We accomplished this through the utilization of BCP. As a means of studying mechanisms of action, BCP has been utilized with small molecule antibiotics and anti-cancer organometallic compounds, but never before has BCP been applied to the study of HDPs.17,

20 Through the application of BCP we were able to determine that ClavA has at least three different mechanisms of action which are dependent on pH and zinc concentration.

ClavA is not the first instance of a peptide for which the literature appears to conflict regarding the MoA, especially when the apparent disagreement is between a membrane-based MoA and an internal one. For example, indolicidin, upon first discovery was believed to create membrane pores for its antimicrobial activity.8 It was later discovered that indolicidin can bind to DNA, interfering with replication, transcription, and translation.9 It is likely that with varied treatment conditions, including incubation time and peptide concentration, BCP would result in multiple phenotypes for indolicidin treated cells, much like what we observed for ClavA. Another HDP for which the literature is inconsistent is tachyplesin, which has been shown to target lipopolysaccharide, membrane phospholipids, and the minor groove of DNA.10-13 Tachyplesin would be an excellent candidate for study via BCP in the future to help pin down exactly what is happening during the complicated HDP bacteria interaction. Piscidin 3 has been shown to be membranolytic, as are many members of the piscidin family, and while it has been shown to have some membrane-based activity, the MIC has been shown to be largely controlled by a DNA based

114

MoA.14-16 This is further supported by the fact that while both appear close together in

Figure 4-2, piscidin 1 trends towards magainin 2, while piscidin 3 trends towards the DNA active HDPs. In each of these cases, the application of BCP could streamline the process of determining the MoA, though it is necessary to have a set of peptides with well-defined and agreed upon mechanisms of action, such as buforin II and magainin 2, before this method can be applied on a larger scale.

Together, the data presented in this study and the previous work on ClavA1, 4 allow us to propose the following as the most complete MoA for ClavA to date (Figure 4-6). At neutral pH, ClavA exerts its activity by causing nonspecific membrane damage, likely resulting in the loss of the proton motive force and the leakage of cellular contents (Type

N, Figure 4-6A). While this damage is not the result of the formation of membrane pores

(Figure 4-5B), this activity is supported by the clustering of Type N with the piscidins in the PCA plot (Figure 4-2), the potential breakdown of lipid vesicles (Figure 4-4B), as well as work performed by van Kan et al., which shows that at neutral pH, ClavA can dissipate the transmembrane potential and can cause leakage of dyes from model membranes.1 Based on the elongated appearance of the cells treated at pH 7.4 (figure 4-1F), this activity appears to inhibit the ability of E. coli to properly divide, a common MoA observed for several classes of conventional antibiotics.17

115

Figure 4-6

The MoA of ClavA (A) at neutral pH (Type N) involves membrane disruption (here

represented as the formation of a pore) as previously proposed1, (B and C) Type A1 at pH

5.5 which starts with the proposed binding of ClavA to ionophores, distorting their gating6

followed by indolicidin like DNA synthesis inhibition, and (D) Type A2 at pH 5.5 in which

ClavA-Zn(II) cleaves DNA as previously shown4. Orange helices represent ClavA; green

circles represent Zn(II) ions.

116

When the pH is lowered to 5.5, as typically happens in phagosomes, the mechanism changes to involve binding to membrane embedded ion channels, as presented by van Kan and coworkers (Figure 4-6B).1, 4 Additionally, we have shown via CD that there is strong membrane interaction at pH 5.5 (Figure 4-4A, Table 4-4), which could allow the translocation of ClavA into the cytoplasm. Once there, ClavA binds to DNA and inhibits

DNA synthesis in a similar manner to indolicidin (Type A1, Figure 4-6C), which is supported by Figure 4-2.

The interaction of ClavA with currently unidentified transmembrane ion channels has some effect on their gating, allowing the free exchange of ions between the cytoplasm and the extracellular space. It is this change to ion permeability that likely allows for an increase in the intracellular Zn(II) concentration in bacteria (Figure 4-6B). ClavA, either free floating or DNA bound, can bind to some of the newly increased concentration of

Zn(II) ions, allowing for subsequent DNA hydrolysis (Type A2, Figure 4-6D). Clavanins were isolated from phagocytic cells of S. clava and postulated to be delivered into their phagosomes when required.32-33 Although there are no reports on zinc ion trafficking inside

S. clava phagocytic cells, high micromolar concentrations of free zinc ions have been shown to accumulate inside phagosomes of macrophages in response to certain bacterial pathogens or stimuli.34-35 It is likely that the phagocytic cells in which the Clavanins are found behave in a similar fashion to their murine and human counterparts. ClavA has also been found in cytoplasmic granules, which are commonly secreted at the site of infection.33

In conditions outside of the Zn(II) enriched phagosome, ClavA would need to get Zn(II) from the bacteria themselves, however, the free pool of zinc ions within a bacterium is low, with most of the zinc bound within bacterial enzymes and virulence factors.36-37 One

117 possible explanation for why ClavA exhibits these multiple mechanisms of action is that in order to have free Zn(II) available to bind, ClavA needs to first kill some of the bacteria via a different MoA, thereby increasing the local concentration of zinc ions.

Type A1 on its own would likely be only bacteriostatic, though this cannot be observed because ClavA will eventually scavenge Zn(II) ions from the bacterium or surrounding media, resulting in the Type A2 mechanism. This is supported by the previously reported time kill kinetics data in which ClavA caused culture sterilization in 4 hr, while ClavA-Zn(II) accomplished the same feat in half the time.4 The potentiated activity that is observed when Zn(II) is added to the culture medium comes from the increased availability of those zinc ions. The proposed mechanisms also explain the 16- fold decrease in MIC as normally ClavA needs to get inside and scavenge Zn(II) from the bacterium.

As with all mechanistic studies on HDPs, it is important to note here that one MoA cannot fully characterize the behavior of a population of HDPs. This has been shown recently in a study utilizing bacterial spheroplasts to study the localization of buforin II and magainin 2.38 In this work it is highly possible that in our low pH experiments, some small percentage of the ClavA population may never transition from Type A1 to Type A2. It may mean that the cells treated with that subset survive, or they may be killed through a yet undiscovered mechanism.

4.4 – Methods

Bacterial Culture Conditions

Bacteria treated under low pH conditions were grown in LB media containing 50 mM MES buffer at pH 5.5. For the neutral pH experiments, bacteria were grown and treated in

118 standard MHB media at pH 7.4. Cells were incubated at 37℃ and 234 rpm to mid-log phase

(OD600 = 0.4-0.6).

Confocal Fluorescence Microscopy

E. coli cultures were inoculated in culture media described above and grown to an OD600 of about 0.5 and treated with equal volumes of host defense peptides (HDPs) at a final concentration of each peptide’s MIC value. These were incubated for one hour on a shaking incubator at 37°C, followed by staining with 1 μg/mL FM 4-64, 2 μg/mL DAPI and 0.5 μM

SYTOX Green for 30 min. FM 4-64 selectively stains bacterial membranes while DAPI and

SYTOX Green both stain DNA and are cell permeable and impermeable, respectively.

These changes from the protocol by Nonejuie et al. were required because many HDPs have been shown to completely sterilize cultures at timescales between one and two hours.

Additionally, with several of the HDPs tested few bacteria remained when treated with concentrations of 5x MIC of the peptides.

After staining, the cultures were concentrated via centrifugation at 14000 rpm for 90 s, and resuspended in 1/10 their original volume. Three microliters of the culture was then spotted onto 1.5% agarose pads containing 20% LB medium, prepared using 65 μL gene frame seals on microscope slides. The cells were then imaged using a Nikon A1R spectral confocal microscope with a 60x oil immersion lens.

Bacterial Cytological Profiling

Bacterial Cytological Profiling (BCP) was performed using a procedure adapted from

Nonejuie et al.17 All cellular measurements were performed using ImageJ software, utilizing the

119 freehand selection tool to trace either the membrane or nucleoid. Each was then measured for area, perimeter, and circularity. Length and width were measured separately using the straight- line selection tool. Average DAPI and SYTOX Green intensities were measured using the outline of the membrane, and subtracted from the background intensity. The intensities were then normalized against the intensities from untreated cells measured on the same day with the same microscope and camera settings. In every case, measurements were made on at least 50 cells, and every cell was measured in each image used, excluding any cells which we only partially captured.

Principal Component Analysis

For each antimicrobial treatment, the cell morphology parameters were obtained from at least three independent experiments. For each parameter, the average measurement was also calculated and included in the dataset. The parameters were reduced into three orthogonal variables using Principal Component Analysis (PCA) from our in-house Python scripts. For each treatment, three cells were selected as representative cells by determining three transformed measurements with the shortest Euclidean distance from the transformed average measurement. The cell morphology parameters of the representative cells were used for PCA of various group of treatments. Clustering was performed using the Euclidean distance method from our in-house Python scripts, where centroids having distance of 2.4 units or lower are in the same cluster.

Circular Dichroism

CD spectroscopy was used to monitor titration of ClavA with 3:1 POPC/POPG

LUVs in phosphate buffer at pH 5.5 and pH 7.4. LUVs were prepared at both pHs as

120 follows. Lipid films were made by mixing in a round bottom flask appropriate volumes of

POPC and POPG lipids solvated in chloroform. The chloroform was then evaporated under a flow of nitrogen gas prior to lyophilization overnight. The dried lipid films were hydrated with buffer (buffer concentration 2.5 mM, pH 5.5 or 5 mM, pH 7.4), producing

5 mM lipid suspensions. After several freeze-thaw cycles, the suspensions were extruded using an Avanti Polar Lipids Mini Extruder, resulting in vesicles with a diameter of 200 nm. The LUV suspensions were then diluted to 2 mM, and the vesicles added to aqueous

ClavA (321.00 μM) and buffer to obtain samples with P:L ratios between 1:0 and 1:80, each with a fixed peptide concentration of 20 µM. CD spectra were acquired by averaging three scans collected at 293 K on a J-1500 spectrometer (Jasco Analytical Instruments,

Easton, MD) over a wavelength range of 190-260 nm using a scan speed of 100 nm/min and a 1 nm bandwidth. Spectra for buffer and phospholipid blanks without peptide were obtained and subtracted from the ClavA sample spectra to account for background signal.

Samples were made in triplicate using the same batch of LUVs. The molar ellipticity obtained at 222 nm was converted into helicity, assuming an ellipticity of -32,000 deg·cm2/dmol for a perfect helix and -2000 deg·cm2/dmol for the fully disordered peptide.39-40 Helicity was plotted in the form of a binding isotherm and fitted by a nonlinear least-squares minimization in Microsoft Excel 2016 using the Solver Add-in.

Fit parameters included the binding constant, KD and the helical content of peptides in the free and bound state. A binding stoichiometry of 1:1 between peptides and lipids was used.

X-ray Diffraction Spectroscopy

121

Clavanin A was incubated with preformed POPC lipid vesicles at a peptide to lipid molar ration of 1:25. The solution was fused onto glass cover slips and bulk water was allowed to evaporate slowly overnight at room temperature, producing aligned lipid multilayers with incorporated peptide. Before diffraction experiments, the samples were annealed at 98 % relative humidity and 30 C, for at least 12 h. X-ray diffraction measurements were performed on a 3KW Rigaku Smartlab diffractometer located at the

Institute for Bioscience and Biotechnology Research (IBBR), Rockville, MD, as described previously.41

Surface Plasmon Resonance/Electrical Impedance Spectroscopy

Tethered lipid bilayer membranes (tBLM) were prepared as described previously using POPC lipid.42-43 The freshly prepared tBLM was stabilized in either TRIS buffer

(20 mM TRIS, 50 mM NaCl, pH=7.5) or MES buffer (50 mM MES, 50 mM NaCl, pH =

5.5) for SPR/EIS measurements at various pHs. Clavanin A was injected on top of the preform tBLM in the corresponding buffer, at a concentration of 3 μM, allowed to incubate with the bilayer for 10 minutes, followed by a buffer rinse. SPR/EIS spectra were collected simultaneously, as a function of time, as described previously.41

122

4.5 – References

1.van Kan, E. J.; Demel, R. A.; Breukink, E.; van der Bent, A.; de Kruijff, B., Clavanin permeabilizes target membranes via two distinctly different pH-dependent mechanisms. Biochemistry 2002, 41 (24), 7529-39. 2. Lee, I. H.; Cho, Y.; Lehrer, R. I., Effects of pH and salinity on the antimicrobial properties of clavanins. Infect Immun 1997, 65 (7), 2898-903. 3. Silva, O. N.; Alves, E. S.; de la Fuente-Nunez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A.; Andrade, C. A.; Nascimento, J. M.; Fensterseifer, I. C.; Porto, W. F.; Correa, J. R.; Hancock, R. E.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti-Biofilm Activity. Sci Rep 2016, 6, 27128. 4. Juliano, S. A.; Pierce, S.; deMayo, J. A.; Balunas, M. J.; Angeles-Boza, A. M., Exploration of the Innate Immune System of Styela clava: Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide Clavanin A. Biochemistry 2017, 56 (10), 1403-1414. 5. Silva, O. N.; Fensterseifer, I. C.; Rodrigues, E. A.; Holanda, H. H.; Novaes, N. R.; Cunha, J. P.; Rezende, T. M.; Magalhaes, K. G.; Moreno, S. E.; Jeronimo, M. S.; Bocca, A. L.; Franco, O. L., Clavanin A improves outcome of complications from different bacterial infections. Antimicrob Agents Chemother 2015, 59 (3), 1620-6. 6. Silva, O. N.; Alves, E. S. F.; de la Fuente-Núñez, C.; Ribeiro, S. M.; Mandal, S. M.; Gaspar, D.; Veiga, A. S.; Castanho, M. A. R. B.; Andrade, C. A. S.; Nascimento, J. M.; Fensterseifer, I. C. M.; Porto, W. F.; Correa, J. R.; Hancock, R. E. W.; Korpole, S.; Oliveira, A. L.; Liao, L. M.; Franco, O. L., Structural Studies of a Lipid-Binding Peptide from Tunicate Hemocytes with Anti-Biofilm Activity. Sci Rep 2016, 6, 27128. 7. Ribeiro, K. L.; Frias, I. A. M.; Franco, O. L.; Dias, S. C.; Sousa-Junior, A. A.; Silva, O. N.; Bakuzis, A. F.; Oliveira, M. D. L.; Andrade, C. A. S., Clavanin A-bioconjugated Fe3O4/Silane core-shell nanoparticles for thermal ablation of bacterial biofilms. Colloid Surface B 2018, 169, 72-81. 8. Falla, T. J.; Karunaratne, D. N.; Hancock, R. E., Mode of action of the antimicrobial peptide indolicidin. J Biol Chem 1996, 271 (32), 19298-303. 9. Subbalakshmi, C.; Sitaram, N., Mechanism of antimicrobial action of indolicidin. FEMS Lett 1998, 160 (1), 91-6. 10. Nakamura, T.; Furunaka, H.; Miyata, T.; Tokunaga, F.; Muta, T.; Iwanaga, S.; Niwa, M.; Takao, T.; Shimonishi, Y., Tachyplesin, a class of antimicrobial peptide from the hemocytes of the horseshoe crab (Tachypleus tridentatus). Isolation and chemical structure. J Biol Chem 1988, 263 (32), 16709-13. 11. Matsuzaki, K.; Fukui, M.; Fujii, N.; Miyajima, K., Interactions of an antimicrobial peptide, tachyplesin I, with lipid membranes. Biochim Biophys Acta 1991, 1070 (1), 259-64. 12. Ohta, M.; Ito, H.; Masuda, K.; Tanaka, S.; Arakawa, Y.; Wacharotayankun, R.; Kato, N., Mechanisms of antibacterial action of tachyplesins and polyphemusins, a group of antimicrobial peptides isolated from horseshoe crab hemocytes. Antimicrob Agents Chemother 1992, 36 (7), 1460-5. 13. Yonezawa, A.; Kuwahara, J.; Fujii, N.; Sugiura, Y., Binding of tachyplesin I to DNA revealed by footprinting analysis: significant contribution of secondary structure to DNA binding and implication for biological action. Biochemistry 1992, 31 (11), 2998-3004.

123

14. Silphaduang, U.; Noga, E. J., Peptide antibiotics in mast cells of fish. Nature 2001, 414 (6861), 268-9. 15. Perrin, B. S.; Tian, Y.; Fu, R.; Grant, C. V.; Chekmenev, E. Y.; Wieczorek, W. E.; Dao, A. E.; Hayden, R. M.; Burzynski, C. M.; Venable, R. M.; Sharma, M.; Opella, S. J.; Pastor, R. W.; Cotten, M. L., High-Resolution Structures and Orientations of Antimicrobial Peptides Piscidin 1 and Piscidin 3 in Fluid Bilayers Reveal Tilting, Kinking, and Bilayer Immersion. J Am Chem Soc 2014, 136 (9), 3491-3504. 16. Libardo, M. D. J.; Bahar, A. A.; Ma, B.; Fu, R.; McCormick, L. E.; Zhao, J.; McCallum, S. A.; Nussinov, R.; Ren, D.; Angeles-Boza, A. M.; Cotten, M. L., Nuclease activity gives an edge to host-defense peptide piscidin 3 over piscidin 1, rendering it more effective against persisters and biofilms. FEBS J 2017, 284 (21), 3662-3683. 17. Nonejuie, P.; Burkart, M.; Pogliano, K.; Pogliano, J., Bacterial cytological profiling rapidly identifies the cellular pathways targeted by antibacterial molecules. Proc Natl Acad Sci 2013, 110 (40), 16169-16174. 18. van Kan, E. J.; van der Bent, A.; Demel, R. A.; de Kruijff, B., Membrane activity of the peptide antibiotic clavanin and the importance of its glycine residues. Biochemistry 2001, 40 (21), 6398-405. 19. Wu, Y.; Seyedsayamdost, M. R., Synergy and Target Promiscuity Drive Structural Divergence in Bacterial Alkylquinolone Biosynthesis. Cell Chem Biol 2017, 24 (12), 1437-1444.e3. 20. Sun, Y.; Heidary, D. K.; Zhang, Z.; Richards, C. I.; Glazer, E. C., Bacterial Cytological Profiling Reveals the Mechanism of Action of Anticancer Metal Complexes. Mol Pharm 2018, 15 (8), 3404-3416. 21. Zasloff, M., Magainins, a class of antimicrobial peptides from Xenopus skin: isolation, characterization of two active forms, and cDNA sequence of a precursor. Proc Natl Acad Sci 1987, 84 (15), 5449-53. 22. Ludtke, S. J.; He, K.; Heller, W. T.; Harroun, T. A.; Yang, L.; Huang, H. W., Membrane pores induced by magainin. Biochemistry 1996, 35 (43), 13723-8. 23. Matsuzaki, K.; Murase, O.; Fujii, N.; Miyajima, K., An antimicrobial peptide, magainin 2, induced rapid flip-flop of phospholipids coupled with pore formation and peptide translocation. Biochemistry 1996, 35 (35), 11361-8. 24. Park, C. B.; Kim, M. S.; Kim, S. C., A Novel Antimicrobial Peptide fromBufo bufo gargarizans. Biochem Biophys Res Commun 1996, 218 (1), 408-413. 25. Park, C. B.; Kim, H. S.; Kim, S. C., Mechanism of action of the antimicrobial peptide buforin II: buforin II kills microorganisms by penetrating the cell membrane and inhibiting cellular functions. Biochem Biophys Res Commun 1998, 244 (1), 253-7. 26. Kobayashi, S.; Takeshima, K.; Park, C. B.; Kim, S. C.; Matsuzaki, K., Interactions of the antimicrobial peptide buforin 2 with lipid bilayers: proline as a translocation promoting factor. Biochemistry 2000, 39 (29), 8648-54. 27. Marchand, C.; Krajewski, K.; Lee, H. F.; Antony, S.; Johnson, A. A.; Amin, R.; Roller, P.; Kvaratskhelia, M.; Pommier, Y., Covalent binding of the natural antimicrobial peptide indolicidin to DNA abasic sites. Nucleic Acid Res 2006, 34 (18), 5157-65. 28. Yu, D.; Sheng, Z.; Xu, X.; Li, J.; Yang, H.; Liu, Z.; Rees, H. H.; Lai, R., A novel antimicrobial peptide from salivary glands of the hard tick, Ixodes sinensis. Peptides 2006, 27 (1), 31-5.

124

29. Libardo, M. D. J.; Gorbatyuk, V. Y.; Angeles-Boza, A. M., Central Role of the Copper- Binding Motif in the Complex Mechanism of Action of Ixosin: Enhancing Oxidative Damage and Promoting Synergy with Ixosin B. ACS Infect Dis 2016, 2 (1), 71-81. 30. Hayden, R. M.; Goldberg, G. K.; Ferguson, B. M.; Schoeneck, M. W.; Libardo, M. D.; Shrestha, A.; Bogardus, K. A.; Hammer, J.; Pryshchep, S.; Lehman, H. K.; McCormick, M. L.; Blazyk, J.; Angeles-Boza, A. M.; Fu, R.; Cotten, M. L., Complementary Effects of Host Defense Peptides Piscidin 1 and Piscidin 3 on DNA and Lipid Membranes: Biophysical Insights into Contrasting Biological Activities. J Phys Chem B 2015, 119 (49), 15235-46. 31. Ludtke, S.; He, K.; Huang, H., Membrane thinning caused by magainin 2. Biochemistry 1995, 34 (51), 16764-9. 32. Lee, I. H.; Zhao, C.; Cho, Y.; Harwig, S. S.; Cooper, E. L.; Lehrer, R. I., Clavanins, alpha- helical antimicrobial peptides from tunicate hemocytes. FEBS Lett 1997, 400 (2), 158-62. 33. Menzel, L. P.; Lee, I. H.; Sjostrand, B.; Lehrer, R. I., Immunolocalization of clavanins in Styela clava hemocytes. Dev Comp Immunol 2002, 26 (6), 505-515. 34. Wagner, D.; Maser, J.; Lai, B.; Cai, Z.; Barry, C. E., 3rd; Honer Zu Bentrup, K.; Russell, D. G.; Bermudez, L. E., Elemental analysis of Mycobacterium avium-, Mycobacterium tuberculosis-, and Mycobacterium smegmatis-containing phagosomes indicates pathogen-induced microenvironments within the host cell's endosomal system. J Immunol 2005, 174 (3), 1491-500. 35. Botella, H.; Peyron, P.; Levillain, F.; Poincloux, R.; Poquet, Y.; Brandli, I.; Wang, C.; Tailleux, L.; Tilleul, S.; Charrière, G. M.; Waddell, S. J.; Foti, M.; Lugo-Villarino, G.; Gao, Q.; Maridonneau-Parini, I.; Butcher, P. D.; Castagnoli, P. R.; Gicquel, B.; de Chastellier, C.; Neyrolles, O., Mycobacterial p(1)-type ATPases mediate resistance to zinc poisoning in human macrophages. Cell Host Microbe 2011, 10 (3), 248-259. 36. Ma, L.; Terwilliger, A.; Maresso, A. W., Iron and zinc exploitation during bacterial pathogenesis. Metallomics 2015, 7 (12), 1541-54. 37. Chandrangsu, P.; Rensing, C.; Helmann, J. D., Metal homeostasis and resistance in bacteria. Nat Rev Microbiol 2017, 15 (6), 338-350. 38. Wei, L.; LaBouyer, M. A.; Darling, L. E. O.; Elmore, D. E., Bacterial Spheroplasts as a Model for Visualizing Membrane Translocation of Antimicrobial Peptides. Antimicrob Agents Chemother 2016, 60 (10), 6350. 39. Wu, C.-S. C.; Ikeda, K.; Yang, J. T., Ordered conformation of polypeptides and proteins in acidic dodecyl sulfate solution. Biochemistry 1981, 20 (3), 566-570. 40. Marqusee, S.; Robbins, V. H.; Baldwin, R. L., Unusually stable helix formation in short alanine-based peptides. Proc Natl Acad Sci 1989, 86 (14), 5286. 41. Mihailescu, M.; Sorci, M.; Seckute, J.; Silin, V. I.; Hammer, J.; Perrin, B. S.; Hernandez, J. I.; Smajic, N.; Shrestha, A.; Bogardus, K. A.; Greenwood, A. I.; Fu, R.; Blazyk, J.; Pastor, R. W.; Nicholson, L. K.; Belfort, G.; Cotten, M. L., Structure and Function in Antimicrobial Piscidins: Histidine Position, Directionality of Membrane Insertion, and pH-Dependent Permeabilization. J Am Chem Soc 2019, 141 (25), 9837-9853. 42. McGillivray, D. J.; Valincius, G.; Vanderah, D. J.; Febo-Ayala, W.; Woodward, J. T.; Heinrich, F.; Kasianowicz, J. J.; Losche, M., Molecular-scale structural and functional characterization of sparsely tethered bilayer lipid membranes. Biointerphases 2007, 2 (1), 21-33. 43. Vanderah, D. J.; Parr, T.; Silin, V.; Meuse, C. W.; Gates, R. S.; La, H.; Valincius, G., Isostructural self-assembled monolayers. 2. Methyl 1-(3-mercaptopropyl)-oligo(ethylene oxide)s. Langmuir 2004, 20 (4), 1311-6.

125

Chapter 5 – Continuing the Narrative

“The important thing is not to stop questioning.”

-Albert Einstein, 1955

“I may not have gone where I intended to go, but I think I have ended up where I needed to be.”

126

-Douglas Adams, The Long Dark Tea-Time of the Soul

5.1 – Future Directions and Newly Generated Questions

Most people believe that answers will follow logically from questions, but a good scientist knows that this is not the case. For every scientist knows that as we gain answers, what we really end up with is more questions. While this dissertation has answered many questions about the antimicrobial activity of the tunicate HDP clavanin A, it also presents myriad questions still in need of answers.

5.1.1 – The Clavanins

Despite the fact that the whole of this dissertation has been devoted to the study of clavanin A, especially the zinc-based activity of clavanin A, there is still much more to understand about this fascinating HDP. As alluded to at the end of chapter four, the previously held beliefs regarding how clavanin A exerts its antimicrobial activity at pH 7.4 and 5.5 may not be completely accurate, and more study of this topic will certainly uncover more information.1 In chapters 2 and 3 the activity of clavanin A-Zn2+ was studied fairly thoroughly, and while we were able to establish that His17 and His21 are important to that activity, it is not clear what the role of His10 and His11 could be. Additionally, van Kan and coworkers were able to identify that the abundant phenylalanine residues held importance for interaction with the membrane, but nothing is known if those residues may play a role in the zinc based activity or interaction with

DNA.2 It is also unknown exactly how clavanin A is able to cross the membrane, knowledge which will be important for the future development of antibiotics.

Little is known about the antimicrobial activity of clavanin B, though based on its high degree of structural similarity to clavanin A (Table 1-1) it is likely that they will behave in a

127 similar fashion. Preliminary data (Table 5-1) seems to indicate that this is the case, with the activity of clavanin B improving 8-fold upon interaction with Zn2+. A comparison study between clavanin A and clavanin B would provide interesting insight regarding the importance of arginine versus lysine in the clavanins and in AMPs as a whole.

Table 5-1

Antimicrobial activities of the clavanins in the presence and absence of Zn2+

Peptide MIC without Zn2+ MIC with Zn2+

Clavanin A3 64 μM 4 μM

Clavanin B 64 μM 8 μM

Clavanin C 4 μM 4 μM

Clavanin D 2 μM 2 μM

Clavanin E 4 μM 4 μM

Clavaspirin ------

Clavanin C and clavanin D present some of the most interesting questions from the family due to the presence of the DOPA residue in place of His17. This residue is capable of crosslinking proteins, peptides, and DNA in a reaction catalyzed by Fe3+, which could result in hitherto unexplored mechanisms of action for AMPs.4-6 DOPA is also capable of binding divalent metal cations, though the results in Table 5-1 indicate that the addition of Zn2+ has no effect on antimicrobial activity.6 It is possible, however, that metal chelators may reduce the activity of these peptides. Additionally, the VFH ATCUN motif found in clavanin C has not

128 been studied in detail, though preliminary results (Figure 5-1) show that it is Cu2+ binding, making it likely that ROS formation is part of the activity of clavanin C.

Figure 5-1

(A) Isothermal titration calorimetry data for the binding of clavanin C to Cu2+ in MES

buffer, pH 5.5. (B) Mass spectrometry showing copper bound clavanin C.

Clavanin E and Clavaspirin remain the biggest mysteries of the clavanin family. Both of these HDPs have received minimal attention, and their most defining feature is that they were discovered as a result of studying the DNA of S. clava.7-8 Clavanin E is particularly interesting due to the fact that despite having the exact same configuration of histidine residues as clavanin

A and clavanin B (Table 1-1), clavanin E shows no improvement upon addition of Zn2+. This may indicate that a different metal is important to the antimicrobial activity of clavanin E, or some feature of its structure keeps the improvement seen in clavanin A and clavanin B from occurring. Despite its drastic difference in sequence from the other clavanins, Clavaspirin has

129 two HX3H motifs, which strengthens the argument that this is an important motif, as well as presenting the possibility for Clavaspirin to bind Zn2+ or some other metal ion with biological relevance to S. clava.

5.1.2 – Clavanin Synergy

All six clavanins are produced by the same cells in S. clava by genes in the same region of the genome. It stands to reason that if an organism as simple as S. clava was able to evolve one or two effective antimicrobial compounds it would not waste genetic space and valuable resources to produce six related compounds. This line of thought leads to the question: Is there a reason S. clava produces so many related antimicrobial peptides? One possible answer to this question is the idea that some members of the clavanin family may be synergistic with one another. Synergy among AMPs is widely sought after as a means of creating drug cocktails which are not only effective, but also limit the ability of bacteria to develop antibiotic resistance.

To date only a small handful of AMP pairs have been identified that behave synergistically.

Synergy can be quantified by calculating the fractional inhibitory concentration (FIC) for a pair of antimicrobials using the following formula:

푀퐼퐶[퐴] 푀퐼퐶[퐵] 퐹퐼퐶 = + 푀퐼퐶퐴 푀퐼퐶퐵

where MICA and MICB are the MIC values for AMP A and AMP B alone, and MIC[A] and

MIC[B] are the MICs of each component when in a mixture. FIC values less than or equal to 0.5 represent a 4-fold activity increase for each AMP, indicating synergy. While FIC can be most accurately obtained by measuring a wide range of different concentration combinations in a checkerboard assay, looking at one to one mixtures can provide a reasonable approximation. For

130 the clavanins (excluding Clavaspirin) synergy in one to one mixtures can only be observed for clavanin A and clavanin B (Table 5-2), while the other combinations can be described as additive or indifferent. It still remains to be seen how Clavaspirin factors in to the synergism among the clavanins, though potential synergy may explain the drastic difference between Clavaspirin and the other clavanins. It is also possible that there may be synergy with mixtures of three or more clavanins, though those conditions still need to be explored.

131

Table 5-2

Summary of synergy testing on 1:1 mixtures of clavanins A-E

a Peptide MICpeptide 1 MICpeptide 2 MICmixture FIC Definition

A+B 64 μM 64 μM 16 μM 0.25 Synergy

A+C 64 μM 4 μM 8 μM 1.0625 Indifferent

A+D 64 μM 2 μM 4 μM 1.03125 Indifferent

A+E 64 μM 4 μM 4 μM 0.53125 Additive

B+C 64 μM 4 μM 8 μM 1.0625 Indifferent

B+D 64 μM 2 μM 4 μM 1.03125 Indifferent

B+E 64 μM 4 μM 4 μM 0.53125 Additive

C+D 4 μM 2 μM 4 μM 1.5 Indifferent

C+E 4 μM 4 μM 4 μM 1.0 Additive

D+E 2 μM 4 μM 4 μM 1.5 Indifferent a Synergy = FIC≤0.5, Additive = 0.5

Antagonism = FIC>4.0

5.1.3 – Expanding the Utility of Bacterial Cytological Profiling

In chapter 4, BCP was utilized for the first time for the identification of the mechanism of action of an AMP. While this proof of concept is an excellent first step toward wide-spread use of this technique for the study of AMPs, there is still much that need to be done in order for this to come to fruition. One of the biggest hurdles that needs to be overcome is the overall lack of data for AMPs in BCP. Since this is a relatively new technique for the field, it is important to study as many AMPs as possible with BCP in order to establish trends for future use. The work

132 done in chapter 4 with magainin 2, indolicidin, buforin II, ixosin, piscidin 1, and piscidin 3 is certainly a strong start, however, further application of this technique will certainly lead to better data and faster mechanism of action determination for many newly identified or synthesized

AMPs.

Despite the obvious need for more groundwork, BCP can still have utility for studying the clavanin family of HDPs. One of the possible questions mentioned above is why S. clava would waste precious resources producing six antimicrobial compounds when one or two of the more potent examples would suffice. While synergy may be one possible answer to this question, another could be discovered through application of BCP. Preliminary results, summarized in

Figure 5-2, Table 5-3 and Table 5-4, show that the clavanins have a wide range of activities.

Additionally, clavanin B has at least two different mechanisms of action, much like clavanin A, though while one appears very close to clavanin A Type A1, the other is quite different from anything seen for the other clavanins. It is possible, given these preliminary results, that other clavanins will have multiple mechanisms of action under particular conditions, which may lead to a greater understanding of why all of these closely related HDPs would be produced.

133

Untreated A Type 1 A Type 2 A with Zn(II) B Type 1

PCA3 (3.14%) B Type 2 C D E

P C A 2 ( 8 .9 5 % )

PCA1 (86.5%)

Figure 5-2

Three dimensional PCA plot showcasing preliminary findings on the clavanin family of

HDPs.

134

Table 5-3

Average morphology measurements of clavanin treated bacterial membranes

Area Perimeter Length Width Nucleoids/ Treatment Circularity (μm2) (μm) (μm) (μm) Cell

Clavanin A 3.787 8.532 3.287 1.239 0.658 1.333 Type A1

Clavanin A 3.272 7.933 2.961 1.072 0.667 1.229 Type A2

Clavanin A 5.557 11.336 4.641 1.278 0.550 1.873 Type N

ClavA- 3.443 7.574 2.823 1.173 0.752 1.063 Zn(II)

Clavanin B 4.513 9.260 3.471 1.410 0.660 1.667 Type 1

Clavanin B 4.015 8.822 3.116 1.199 0.652 1.364 Type 2

Clavanin C 3.932 8.963 3.320 1.157 0.623 1.469

Clavanin D 4.116 9.660 3.736 1.155 0.569 1.652

Clavanin E 5.314 11.141 4.359 1.180 0.557 1.704

135

Table 5-4

Average morphology measurements of clavanin treated bacterial nucleoids

Corrected Corrected Area Perimeter Length Width Treatment Circularity SYTOX DAPI (μm2) (μm) (μm) (μm) Intensity Intensity

Clavanin A 1.288 4.845 1.967 0.693 0.677 1.285 2.869 Type A1

Clavanin A 1.746 5.641 2.083 0.845 0.688 15.165 3.626 Type A2

Clavanin A 1.425 5.068 1.919 0.775 0.701 1.005 0.842 Type N

ClavA- 2.476 6.643 2.480 0.945 0.702 15.644 2.430 Zn(II)

Clavanin B 1.128 4.400 1.510 0.697 0.736 1.061 2.126 Type 1

Clavanin B 0.799 3.531 1.110 0.811 0.806 8.862 5.264 Type 2

Clavanin C 1.254 4.753 1.496 0.715 0.697 1.065 1.126

Clavanin D 1.236 4.631 1.612 0.765 0.726 1.040 1.311

Clavanin E 1.829 5.837 2.094 0.849 0.682 0.997 1.335

136

5.2 – Conclusions

The results presented in this dissertation serve to elucidate how much we still have to learn about HDPs. Prior to our first study, clavanin A appeared in ten publications which showed that it was a membrane active HDP. Despite this, we were able to discover a new mechanism of action for clavanin A when bound to Zn2+, a mechanism of action that is novel for zinc-binding

AMPs. Results such as these are vital for the future, as these discoveries will pave the way toward peptide based antimicrobials that will, one day, stem the ever growing scourge of antibiotic resistant bacteria.

137

5.3 – References

1. van Kan, E. J.; van der Bent, A.; Demel, R. A.; de Kruijff, B., Membrane activity of the peptide antibiotic clavanin and the importance of its glycine residues. Biochemistry 2001, 40 (21), 6398- 405. 2. van Kan, E. J.; Demel, R. A.; van der Bent, A.; de Kruijff, B., The role of the abundant phenylalanines in the mode of action of the antimicrobial peptide clavanin. Biochim Biophys Acta 2003, 1615 (1-2), 84-92. 3. Juliano, S. A.; Pierce, S.; deMayo, J. A.; Balunas, M. J.; Angeles-Boza, A. M., Exploration of the Innate Immune System of Styela clava: Zn2+ Binding Enhances the Antimicrobial Activity of the Tunicate Peptide Clavanin A. Biochemistry 2017, 56 (10), 1403-1414. 4. Taylor, S. W.; Chase, D. B.; Emptage, M. H.; Nelson, M. J.; Waite, J. H., Ferric Ion Complexes of a DOPA-Containing Adhesive Protein from Mytilus edulis. Inorg Chem 1996, 35 (26), 7572- 7577. 5. Morin, B.; Davies, M. J.; Dean, R. T., The protein oxidation product 3,4- dihydroxyphenylalanine (DOPA) mediates oxidative DNA damage. Biochem J 1998, 330 (Pt 3), 1059-1067. 6. Yang, J.; Cohen Stuart, M. A.; Kamperman, M., Jack of all trades: versatile catechol crosslinking mechanisms. Chem Soc Rev 2014, 43 (24), 8271-98. 7. Zhao, C.; Liaw, L.; Lee, I. H.; Lehrer, R. I., cDNA cloning of Clavanins: antimicrobial peptides of tunicate hemocytes. FEBS Lett 1997, 410 (2-3), 490-2. 8. Lee, I. H.; Zhao, C.; Nguyen, T.; Menzel, L.; Waring, A. J.; Sherman, M. A.; Lehrer, R. I., Clavaspirin, an antibacterial and haemolytic peptide from Styela clava. J Peptide Res 2001, 58 (6), 445-56.

138