Enzymatic Synthesis of Peptidoglycan in Methicillin-Resistant Staphylococcus Aureus and Its Inhibition by Beta-Lactams

Enzymatic Synthesis of Peptidoglycan in Methicillin-Resistant Staphylococcus Aureus and Its Inhibition by Beta-lactams

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Srisuknimit, Veerasak. 2019. Enzymatic Synthesis of Peptidoglycan in Methicillin-Resistant Staphylococcus Aureus and Its Inhibition by Beta-lactams. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:42029540

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA

Enzymatic Synthesis of Peptidoglycan in Methicillin-Resistant Staphylococcus aureus and its Inhibition by Beta-lactams

A dissertation presented

Veerasak Srisuknimit

The Department of Chemistry and Chemical Biology

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

in the subject of

Chemistry

Harvard University

Cambridge, Massachusetts

May 2019

Dissertation Advisors: Professor Daniel Kahne Veerasak Srisuknimit Professor Suzanne Walker

Enzymatic Synthesis of Peptidoglycan in Methicillin-Resistant Staphylococcus aureus and its Inhibition by Beta-lactams

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) is responsible for more than half of the deaths due to antibiotic resistant bacterial infection in the United States, killing over ten thousand people a year. Beta-lactams, our first-line treatment against Staph infection, target the transpeptidase domain of penicillin-binding proteins (PBPs), the enzymes involved in the final steps of peptidoglycan biosynthesis, rendering the bacteria unable to synthesize the cell wall.

These clinically important drugs however are ineffective against MRSA because MRSA acquired PBP2a, a non-native PBP, that can still crosslink peptidoglycan when S. aureus native

PBPs are inhibited. Despite the importance of PBPs as antibiotic targets, their transpeptidase activities had not been reconstituted in vitro and there was no method to directly evaluate beta- lactam inhibition of the PBP activity. One major hurdle to study the PBPs was the inaccessibility of their substrate, Lipid II.

In this thesis, I will describe a simple and versatile strategy for isolating native Lipid II from S. aureus and other bacteria. A chemical probe triggers Lipid II accumulation and a two- step extraction yields a practical amount of clean Lipid II. I used the isolated Lipid II to reconstitute peptidoglycan synthesis by S. aureus PBP2 and PBP2a. I also reported a direct assay of transpeptidase inhibition by beta-lactams. Comparing the substrate preference of S. aureus

PBPs, I found that PBP2a can unexpectedly crosslink peptidoglycan strands bearing a triglycine branch but not a monoglycine one.

iii This work offers biochemical tools to gain a better understanding of cell wall biosynthesis and especially transpeptidases which are important drug targets. The methods reported here will be useful for developing the next generation of antibiotics and also potentiator compounds that restore sensitivity of MRSA to the currently available beta-lactams.

iv Table of Contents

Abstract iii Table of Contents v List of Figures vii List of Tables viii Glossary of Abbreviations ix Acknowledgement xi Chapter 1: Introduction to methicillin-resistant Staphylococcus aureus and its 1 peptidoglycan biosynthesis 1.1 Antibiotic-resistant bacterial infection is a major global health problem 2 1.2 Introduction to Staphylococcus aureus 3 1.2.1 Discovery of S. aureus 3 1.2.2 S. aureus and human diseases 3 1.2.3 S. aureus versus penicillin 5 1.2.4 Outlook on S. aureus infection treatment and prevention 7 1.3 Introduction to S. aureus cell wall biosynthesis 9 1.3.1 Park nucleotide 10 1.3.2 Synthesis of lipid-linked precursors 12 1.3.3 Peptidoglycan assembly and penicillin-binding proteins 17 1.4 Biology of methicillin-resistant S. aureus 23 1.4.1 SCCmec 23 1.4.2 Beta-lactam-resistant PBP2a 26 1.4.3 Factors essential for methicillin resistance 27 1.5 References 28

Chapter 2: Preparation of S. aureus Lipid II and analogues 44 2.1 Introduction 45 2.1.1 Current approaches to Lipid II preparation 45 2.1.2 Isolation of Lipid II from cell 47 2.1.3 Lipid II accumulates when PG biosynthesis is inhibited in the late stage 49 2.2 A simple method to prepare bacterial native Lipid II 50 2.2.1 Each bacteria requires a different accumulation method 51 2.2.2 Two-step extraction method produces clean Lipid II 54 2.2.3 Lipid II quantification 59 2.3 Vancomycin bead for Park nucleotide extraction 61 2.3.1 Vancomycin bead synthesis 62 2.3.2 Park nucleotide enrichment 65 2.3.3 Enzymatic Lipid II synthesis 70 2.4 Conclusions 71 2.5 Materials and methods 72 2.6 References 82

Chapter 3: Reconstitution of S. aureus Peptidoglycan Biosynthesis 87 3.1 Introduction 88 3.1.1 Existing knowledge of the peptidoglycan transpeptidases 88 3.1.2 Transpeptidase activity assays 92 3.1.3 D-amino acid incorporation by transpeptidases 94 3.2 Purification of PBP2 96 3.3 Reconstitution of PBP2 PGT and TP activities 97 3.4 Purification of PBP2a 100 3.5 Reconstitution of PBP2a TP activity 101 3.6 Direct assay of transpeptidase inhibition by beta-lactams 103 3.7 Substrate specificity of S. aureus PBPs 110 3.8 Conclusions 113 3.9 Materials and methods 115 3.10 References 127

Chapter 4: Conclusions and Future Directions 133 4.1 Conclusions 134 4.2 Future directions 135

Appendix 137

vi List of Figures

Figure 1.1 Chemical structures of four beta-lactams 5 Figure 1.2 Peptidoglycan biosynthesis in S. aureus 11 Figure 1.3 Chemical structure of Park nucleotide 12 Figure 1.4 Chemical structures of lipid-linked intermediates 14 Figure 1.5 Chemical structure of S. aureus native Lipid II 16 Figure 1.6 Mechanism of peptidoglycan glycosyl transferases 18 Figure 1.7 Mechanism of transpeptidases 19 Figure 1.8 Penicillin-binding proteins in S. aureus 21 Figure 1.9 Regulation of penicillin binding protein 2a expression 24 Figure 1.10 Crystal structure of S. aureus PBP2a 27 Figure 2.1 S. aureus PBP4 enables Lipid II labeling 49 Figure 2.2 The stem peptide of Lipid II varies among bacteria 51 Figure 2.3 Chemical probes accumulate Lipid II in bacteria 52 Figure 2.4 A strategy to accumulate Lipid II in E. coli 54 Figure 2.5 Large quantities of native S. aureus Lipid II can be isolated with 55 good purity by two-step extraction Figure 2.6 Isolated S. aureus Lipid II has the correct modification 57 Figure 2.7 Isolated S. aureus Lipid II is relatively clean 58 Figure 2.8 Isolation of S. aureus Lipid II with truncated a truncated glycine 59 branch from S. aureus mutants. Figure 2.9 Quantification strategies for isolated Lipid II 60 Figure 2.10 Vancomycin beads for Park nucleotide purification 63 Figure 2.11 Lipid II purification by vancomycin beads. 64 Figure 2.12 Park nucleotide elution with organic solvent 66 Figure 2.13 Park nucleotide elution with competitive binder 67 Figure 2.14 Optimized Park nucleotide elution 68 Figure 2.15 Vancomycin bead purifies Park nucleotide in one simple step 69 Figure 3.1 Schematic of peptidoglycan synthesis in methicillin-resistant S. 92 aureus Figure 3.2 Schematic of transpeptidase reactions in S. aureus 95 Figure 3.3 Purification of S. aureus PBP2 [59-716] 97 Figure 3.4 Reconstitution of peptidoglycan biosynthesis by PBP2 98 Figure 3.5 PBP2 active serine residue in the transpeptidase domain is 99 necessary for peptidoglycan crosslinking. Figure 3.6 Purification of S. aureus PBP2a [24-668] 100 Figure 3.7 Reconstitution of PBP2a crosslinking and inhibition 102 Figure 3.8 Direct transpeptidase activity assay enables comparison of beta- 105 lactam inhibition of S. aureus PBP2. Figure 3.9 Direct transpeptidase activity assays demonstrated the effect of 107 point mutation on PBP2 susceptibility to a beta-lactam. Figure 3.10 SgtB wild type produces nascent glycan strands that are too long 108 Figure 3.11 Direct transpeptidase activity assay enables comparison of beta- 109 lactam inhibition of S. aureus PBP2a.

vii Figure 3.12 S. aureus PBPs exhibit different tolerance for the change of the 111 glycine branch length in the substrate. Figure 3.13 In cells, Gly3-peptidoglycan undergoes hydrolysis by PBP4 yet 112 can be crosslinked by PBP2a Suppl. Fig. 1 Sufficient quantities of Lipid II can be isolated from bacteria 139 Suppl. Fig. 2 Quantification of Lipid II using PBP5 degradation and Edman 140 reagent Suppl. Fig. 3 Reconstitution of crosslinked peptidoglycan by S. aureus PBP2 141 using native Lipid II Suppl. Fig. 4 LC/MS extracted ion chromatogram of S. aureus PBP2 K406A 142 Suppl. Fig. 5 A direct transpeptidase activity assay enables characterization of 143 inhibitory potencies of different beta-lactams Suppl. Fig. 6 Time-course analysis of PBP2 reaction with Gly5-Lipid II in 144 TGase buffer

List of Tables

Table 1.1 Classification of PBPs in S. aureus 23 Table 2.1 Notable reports of Lipid II preparation 48 Suppl. Table 1 Primers used in this study 137 Suppl. Table 2 Bacterial strains used in this study 138 Suppl. Table 3 The ratio of crosslinking to hydrolysis for PBP2a, PBP2, and 138 PBP4 with three substrates in all three buffer conditions

viii Glossary of Abbreviations

Ac Acetyl ACN Acetonitrile Ala Alanine Asn Asparagine Asp Aspartic acid BDL Biotinylated-D-Lysine BuOH Butanol CHAPS 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate DMSO Dimethyl sulfoxide EIC Extracted ion chromatogram Fos Fos FPLC Fast protein liquid chromatography GlcNAc N-acetyl-glucosamine Gln Glutamine Glu Glutamic acid Gly Glycine HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid His Histidine HPLC High pressure liquid chromatography LC/MS Liquid chromatography mass spectroscopy LDAO N,N-dimethyldodecylamine N-oxide Lys Lysine m-DAP meso-diaminopimelic acid MeOH Methanol MoeA Moenomycin A MS Mass spectrometry MS/MS Tandem mass spectrometry MurNAc N-acetyl-muramic acid PBP Penicillin-binding protein PenG Penicillin G PG Peptidoglycan PGT Peptidoglycan glycosyltransferase PyAc Pyridinium acetate TFA Trifluoroacetic acid TLC Thin-layer chromatography TM Transmembrane TP Transpeptidases Tris Tris(hydroxymethyl)aminomethane UDP Uridine diphosphate UMP Uridine monophosphate Vanco Vancomycin wt Wild-type

For my mom and dad, who had to leave schools too early but never stop learning.

x Acknowledgement

I am very fortunate enough to have Professor Dan Kahne as my advisor. I am quite sure quite sure I would have left graduate school half way if not for Dan. Dan cares very much about mentoring. In the past six years, he guided me to be a better scientist. He trained me to troubleshoot when experiments failed multiple times in a row, taught me to express my thoughts clearly in speech and on paper, and helped me to step back and look at the big picture when my views were too narrow. At the same time, Dan also deeply cares for me as a person. He listened patiently when I am depressed and frustrated. I greatly appreciate his wisdom not only in biology and chemistry but also in life. I hope to become a great scientist and mentor like Dan.

I am also lucky to have Professor Suzanne Walker as my co-advisors. I learned much from writing scientific papers with Suzanne. Suzanne kept me from getting by lost by helping me plan the step I need to take to reach my goal. She also shortened many six-hour meetings with

Dan to just two or three.

I would like to thank my committee Professor Tim Mitchison and Professor Tom

Bernhardt and Professor Nathaniel Gray for their guidance and alternative viewpoints over the year. I thank Jennifer X Wang at the Bauer Core for her help with LC/MS/MS analysis.

I want to thank you everyone in Kahne and Walker labs for their friendship and support. I enjoyed our lunches and game nights together. I thank Helen Corriero, Mike Quinn and Rebecca

Stillo for keeping our labs running and for their support. Joe Wzorek mentored in me in my rotation and convinced me to join the Kahne lab. Katie Schaeffer taught me chemistry reactions and answered me many questions about the cell wall. Many of my projects would be stuck without her. Fred is like a big brother whom I enjoy chatting about chemistry and life. I greatly value his advices and the time he spent to help me with my projects (and laser cutting various

xi items). I would like to thank Alan Yang and Annie Aindow for working with me on the PBP projects. I was not a good mentor and I really appreciate their patience for bearing with me. I enjoyed my discussion with Wonsik Lee, Truc Do and other cell wall team members in the

Walker lab. Most importantly, I would like to Yuan Qiao mentored me for three years and taught me practically everything I know about biology. Yuan graciously allowed me to join her on her projects, patiently guided me through them, and gently supported me when I failed short. I really appreciate her friendship.

I would like to thank David Vosburg, my undergraduate mentor for his unchanging care and guidance - even now many years after my graduation from Harvey Mudd College. I thank

Ajarn Eed Sasinee Ungkanont for instilling her love of chemistry in me during my time at

Mahidol Wittayanusorn School where she was basically my second mom.

I am fortunate to have friends in Cambridge and a far who help keep me sane during graduate school. Thank you for all the good food, good chats, and good time playing board games together.

Thank you Jennifer for believing in me when I do not believe in myself. Thank you for the TLC - that is not thin-layered chromatography. Graduate school is a lot more fun with you by my side.

Finally, I would like to thank my mom and my dad for everything they had done for me.

Although they did not receive formal education beyond grade school, they love learning very much and inoculated in me their deep passion of learning. Dad, I wish you were here. Mom, thank you for your love and your sacrifice. You work harder than anyone else I know. Sorry I won’t be a doctor that save lives as you wished me to be. I hope a doctor of philosophy is okay and I hope my research will save a life one day.

xii

CHAPTER ONE: Introduction to methicillin-resistant Staphylococcus aureus and its peptidoglycan biosynthesis

1 1.1 Antibiotic-resistant bacterial infection is a major global health problem

Bacteria are microscopic single-celled organisms that thrive in all kind environments including human. Some bacteria could proliferate in or on the human body and cause deadly bacterial infection. Many of the devastating diseases in human history are caused by bacteria. such as plague (Yersinia pestis), cholera (Vibrio cholera), and tuberculosis (Mycobacterium tuberculosis).1-3 The discovery of antibiotics, chemical compounds that could kill bacteria, led to a revolution in modern medicine, saving millions of lives from bacterial infection.4

Unfortunately, bacteria evolve and develop resistance to antibiotics. For some bacteria strains, we have no effective treatment at all.5 The antibiotic resistance crisis, if left unchecked, is estimated to cause 10 millions death annually by 2050 and could lead us back the pre-antibiotic era where a single cut by a thorn could kill a person.6

Methicillin-resistant Staphylococcus aureus (MRSA) is currently one of the most important clinical pathogens. In the United States, MRSA is responsible for over half of the deaths due to antibiotic-resistant bacterial infection.7 In 2017, nearly 120,000 S. aureus bloodstream infections were reported and almost 20,000 people died.8

Beta-lactam antibiotics, our first-line treatment against typical S. aureus infection, target penicillin-binding proteins, (PBPs) which are enzymes that synthesize the bacterial cell wall.9

Beta-lactams include penicillin, the first antibiotic discovered, and amoxicillin, one of the most commonly prescribed antibiotics today.10 They make up an important class of antibiotics because they have broad spectrum activity (killing many species of bacteria) and are generally safe.11

Beta-lactams however are not effective against MRSA which acquired drug-resistant PBP called

PBP2a.12

2 In this chapter, I will talk about the S. aureus and its importance in the healthcare setting.

I will recount the arm race between S. aureus and beta-lactams through the years. I will then examine S. aureus cell wall biosynthesis, especially the PBPs which are the targets of beta- lactams.

1.2 Introduction to Staphylococcus aureus

1.2.1 Discovery of S. aureus

Around 140 years ago, Alexander Ogston, a Scottish surgeon in at Aberdeen Royal

Infirmary, attended a young man named James Davidson who was suffering from severe infection of the leg.13 He drew some pus from the infection and took it home to look under microscope. There he found “beautiful tangles, tufts, and chains of round organisms, which stood out clear among the pus cells all stained with the aniline-violet stain I had used.”14 Ogston believed that the bacteria was not in the pus by chance. He hypothesized that the abscesses are caused by the bacteria. He tried to grow this bacteria from Davidson’s pus and eventually found hen’s egg to be the best medium.14 Ogston found that he could cause infection in guinea pigs with the bacteria he isolated. Because he observed bacteria grow in cluster like fish roes, Ogston gave the name “staphylococcus” from Greek “staphyle”, meaning a bunch of grapes, and Greek

“kokkos” meaning berry and used to describe spherical bacteria.15

In 1884, a German physician Friedrich Julius Rosenbach was able to grow staphylococci and therefore able to differentiate the color of the colonies.16 He named the bacteria that forms gold colonies “Staphylococcus aureus” from Latin “aurum” meaning gold.16

1.2.2 S. aureus and human diseases

Staphylococcus aureus is a pathobiont, bacteria that normally live peacefully with human but could sometimes cause deadly infection. Approximately half of the human population is

3 colonized with this bacteria, but most of these people do not have any symptom of S. aureus infection.17 While the primary colonization site of S. aureus is in the nasal passage, the bacteria also colonize skin, throat, vagina, and the gastrointestinal tract. About 20% of the population are persistently S. aureus nasal carriers while the other 30% are intermittent carriers.17

While relatively harmless on the skin, once S. aureus gets inside the human body via cuts or abrasions, it starts to express many virulence factors that allow the bacteria to grow inside the body and become harmful to the human host.18 A few of those virulence factors are:

(1) Adherence factors (adhesins) that allow the bacteria cells to latch on to the human cells.

(2) Toxins such as ⍺-toxin that could form pore on red blood cells causing them to rapture.

(3) Protein A, a surface protein that binds to antibodies allowing S. aureus to evade the

immune system.

(4) Catalase, an enzyme that hydrolyze hydrogen peroxide enabling S. aureus to survive when

engulfed by human white blood cells

S. aureus, especially the antibiotic-resistant strains, is a serious threat to public health. It is a leading cause of bacteremia (blood stream infection), infective endocarditis (heart valve infection), osteoarticular (joint and bone infection), skin and soft tissue infections, pneumonia, and medical device-related infections.19 In the USA, approximately 120,000 S. aureus blood stream infections and 20,000 deaths occurred in 2017.8 People with severe diabetes are at high risk of S. aureus infection in the feet that if severe may require lower-limb amputation.20 The

HIV-infected population is 24 times more likely to suffer from S. aureus blood stream infection.19 Other risk factors include illicit drug uses, recent surgery, old age, etc.19

4 1.2.3 S. aureus versus penicillin

Before the discovery of antibiotic, a simple scratch may lead to fatal bacterial infection.

In 1910s, S. aureus infection has incredibly high mortality of over 70%.21 To save someone with a bad wound infection, surgeons drained the pus in the abscesses, disinfect the wound, and hope that the patient’s immune system is robust enough to fight off the infection.21

In 1928, returning from holiday, Alexander Fleming found that his petri dishes containing colonies of S. aureus was contaminated by a mold Penicillium notatum.22 Interestingly, the area around the mold was clear of the bacteria as if the mold produced something that stopped bacteria from growing. He attempted to purify the active chemical but failed. The compound is not stable. Later on Howard Florey, Ernst Chain, and Norman Heatly at Oxford successfully extracted and purified the chemical that inhibit bacteria growth.23 The compound was named penicillin after the Penicillium mold (Figure 1.1).

N O N H H H H N N S N N S H2N N O N S S O N O S O OH O OH O penicillin G ceftaroline

OMe O N H H H H N S N S N H N 2 N OMe O N NH O N S NH O O OH O O O OH ceftobiprole methicillin

Figure 1.1. Chemical structures of four beta-lactams. Penicillin was the first beta-lactam to be discovered. Methicillin was created to fight bacteria expressing beta-lactamases that can degrade penicillin. Ceftaroline and ceftobiprole were recently developed to combat MRSA expressing transpeptidase PBP2a that is resistant to methicillin and other beta-lactams.

5 In late 1940, Albert Alexander, a 43-year old policeman, was accidentally scratched by a rose thorn on his face. That led to a severe infection by Staphylococcus and Streptococcus bacteria causing abscesses on his face and requiring one of his eye to be removed. In February

1941, he became the first person to be given penicillin.22 Within days, his condition improved significantly. Unfortunately the supply of the drug ran out, even after Florey and colleague recovered penicillin from Alexander’s urine. Albert Alexander died a few days after.

Nevertheless, the significance of penicillin’s therapeutic use was noticed and the drug was used on other patients.24 Soon, penicillin were made on a large scale for troops in the World War II battlefields, saving countless lives.25 By 1945, penicillin was commercially available in the pharmacies. Penicillin kills bacteria by inhibiting a group of enzymes called penicillin-binding proteins (PBPs).26 These enzymes are important for making and maintaining bacterial cell wall.

Penicillin and other related drugs form a family called the beta-lactams due to their shared structure of four-membered (beta) cyclic amide (lactam).

By 1942, before penicillin was even widely used, resistance to penicillin was already detected. Charles H. Rammelkamp found that he could teach S. aureus to become resistant to penicillin by growing the bacteria in increasing concentration of penicillin over a long period of time.27 He also discovered four strains of penicillin-resistant S. aureus from patients. In the next few years, more and more S. aureus became resistant to penicillin. By the late 1960s, over 80 percent of S. aureus were penicillin-resistant.28 Most of the resistance is via the production of penicillinase, an enzyme that could degrade penicillin.

To combat penicillin resistance, a semisynthetic analog of penicillin was developed and methicillin was introduced to the clinic in 1960 (Figure 1.1).29 Due to its structural variation, methicillin could not be degraded by penicillinase.30 Unfortunately, the new drug barely halt the

6 spread of antibiotic resistance. In 1961, British scientists identified the first strain of methicillin- resistant S. aureus (MRSA).31 It was not until 1981, that the resistance mechanism was understood. Alexander Tomasz and Barry Hartman at Rockefeller University revealed that S. aureus had acquired a new enzyme penicillin-binding protein 2a (PBP2a) that could build cell wall even when challenged by penicillin.32 PBP2a is not only resistant to penicillin but the entire family of beta-lactam drugs. Even when new drug like oxacillin replaces methicillin in clinical uses, the resistance by PBP2a in MRSA was never overcome. In 2005, more than 50% of S. aureus clinical isolates in the US were methicillin-resistant.33

In recent years, two new members of the beta-lactam drugs were introduced to specifically counter MRSA (Figure 1.1).34-35 Ceftaroline was approved in the US while

Ceftobiprole was approved in Europe. Both drugs could inhibit PBP2a and thus effective against

MRSA. Again, resistance to these drugs were soon identified.36 Some strains have mutations in the PBP2a target while other strains have mutations in the other PBPs, especially PBP4.37

The arm race between antibiotics and antibiotic resistance will continue as it is driven by natural selection. Human are losing to bacteria in this arm race because fewer and fewer new antibiotics are approved each year. We need to continue to search for new drug but also alternative methods for combating antibiotic-resistant bacteria.

1.2.4 Outlook on S. aureus infection treatment and prevention

As the rate of antibiotic discovery has slowed down, alternative methods for treating and controlling S. aureus are urgently needed. Currently, multiple approaches are being studied such as vaccine, decolonization, and bacteriophage.

Presently, there is no vaccine that can prevent S. aureus infection, but not for lack of trying. Multiple S. aureus vaccine development programs are being conducted and in various

7 state of clinical trials. Merch V710 was in phase III clinical trial but was stopped due to a safety issue.38 Pfizer SA4Ag vaccine candidate is in an ongoing phase II trial.39 Other novel strategies are being tested in preclinical studies such as vaccine against PBP2a40 and live vaccine made from auxotrophic S. aureus that cannot synthesize D-amino acids.41 Most failed candidates targeted a single antigen. As our understanding of S. aureus pathogenicity and immune invasion improve, we should have a better idea on which antigens to target for vaccine development.

S. aureus is one of the leading causes of hospital-acquired infections.19 Occasionally outbreak of MRSA occurs in a hospital.42 One approach to prevent S. aureus infection therefore is to prevent S. aureus from entering hospitals.43 Patients, especially ones being admitted to the intensive care units, are screened by nasal swap to identify whether they are S. aureus carrier. At the site outbreak, decolonization program might be conducted. Patients in a particular hospital wards would have their noses treated with antibiotic or diluted breach to remove S. aureus. The

United Kingdom experienced great reduction in MRSA infection since 2006 after introducing revised national guidelines on MRSA intervention that includes screening and decolonization.44

Decolonization however is often temporary. S. aureus usually grow back after a few months.45

Many nasal bacteria are being investigated for their ability to prevent S. aureus colonization.46

Decolonization together with nasal probiotic might be an effective method in removing S. aureus reservoir in the nose.

Bacteriophage, virus that kills bacteria, is also being studied as a tool to fight MRSA.47-48

An institute in Georgia (Europe) offers phage therapy against MRSA.49 Very little is currently known about how effective phages are against MRSA in human. Antimicrobial phage protein lysins are also investigated. A skin care product based on phage is available for treating dermatitis.50 Lysin by Contrafect against MRSA is in phase 1 clinical trial.51

8 1.3 Introduction to S. aureus cell wall biosynthesis

Bacteria cell wall, also known as peptidoglycan, is a large polymer encasing the cytoplasmic membrane of bacteria. It gives bacteria defined shape such as rod or sphere. The cell wall rigid structure protects bacteria from rupturing under high osmotic pressure inside the cell.52-54 In Gram-positive bacteria such as S. aureus, the cell wall is about 20 nm thick55; meanwhile, Gram-negative bacteria such as Escherichia coli has thinner cell wall of about 5 nm along with the outer membrane.56

The cell wall is made of multiple long glycan (sugar) strands alternating between N- acetyl-glucosamine (GlcNAc) and N-acetyl-muramic acid (MurNAc) with a stem peptide attached at the C3-position of MurNAc.56 The stem peptide differs from one bacteria to another.

In S. aureus, the stem peptide is made L-alanine (L-Ala), D-iso-glutamine (D-iso-Gln), L-lysine

(L-Lys), D-alanine (D-Ala), and D-alanine with the N-terminal of the lysine side chain attached

57 with five glycines (Gly5). These long glycan strands are covalently cross-linked to nearby strands by an amide bond between the pentaglycine sidechain and D-Ala at the fourth position.

The cross-links provide rigidity and strength to the cell wall and are necessary to protect the bacteria from osmotic pressure. The peptidoglycan layer also serves as a site for covalent attachment of wall teichoic acids (WTAs), capsular polysaccharide and some surface proteins.58

The biosynthesis of peptidoglycan is long and complicated. Despite how important cell wall is to bacteria, very little was known about how this complex molecule was made until the mid 20th century. The discovery of penicillin spurred many scientists to study the drug’s mechanism of action and along the way the biosynthesis pathway of peptidoglycan was slowly pieced together (Figure 1.2). I briefly summarized the peptidoglycan synthesis in S. aureus in the following sections.

9 1.3.1 Park nucleotide

In 1949 while trying to understand how penicillin works, James T. Park observed that if he treated S. aureus with a non-lethal concentration of penicillin, several nucleotide compounds would accumulate.59-61 The main component of these compounds was later identified as uridine-

5’-pyrophospho-N-acetylmuramyl-L-alanyl-D-glutamyl-L-lysyl-D-alanyl-D-alanine (UDP-

MurNAc-pentapeptide), which is commonly referred to as the Park nucleotide (Figure 1.3). The other nucleotides were found to be biosynthetic precursors to the Park nucleotide. Around this time, the chemical composition of S. aureus cell wall was revealed to contain a few amino acids including alanine, glutamic acid, lysine and glycine. The similarity between the compositions of

Park nucleotide and the cell wall led Park and Jack Strominger to propose that the penicillin interfere with the cell wall synthesis which directly cause the accumulation of Park nucleotide.26

They conclude that Park nucleotide must be a major cell wall precursor.26,62

Park nucleotide is synthesized in the cytoplasm of the bacteria. The first committed step in the synthesis of Park nucleotide start with the conversion of widely available metabolite UDP-

GlcNAc to UDP-MurNAc. MurA transfer phosphoenolpyruvate to the C3 hydroxyl group of

UDP-GlcNAc.63-64 The alkene intermediate is then reduced by MurB reductase using NADPH to yield UDP-MurNAc.63-64 Next, the ligase enzymes MurC-F, which belong to the ATP-dependent amide bond-forming enzyme family, sequentially install L-Ala, D-iso-Glu, L-Lys, and a dipeptide D-Ala-D-Ala to the lactic acid moiety of muramic acid to yield UDP-MurNAc- pentapeptide. 63-64 The D-Ala-D-Ala dipeptide is specially synthesized for the purpose of cell wall synthesis by Ddl, another ATP-dependent ligase, that condense two D-Ala together.65 D-Ala is racemized from L-alanine by Alr alanine racemase.66-67

10 This pathway is import and highly conserved across many bacteria species. It is a target of multiple antibacterial natural product. Fosfomycin which mimics phosphoenolpyruvate is a covalent inhibitor of MurA.68 D-cycloserine, which is a cyclic analogue D-alanine, inhibits both

Alr racemase and Ddl ligase (Figure 1.2).69

Figure 1.2. Peptidoglycan biosynthesis in S. aureus. Peptidoglycan is synthesized in three main stage. First, Park nucleotide is made in the cytoplasm. Second, Park nucleotide is processed into lipid-linked intermediates Lipid I and Lipid II in the inner leaflet of the cytoplasmic membrane. Third, Lipid II is polymerized and crosslinked into peptidoglycan (PG) on the outer leaflet of the cytoplasmic membrane. PG biosynthesis is a target of many antimicrobial natural products (in red text), some of which are used as antibiotics.

11 HO O HO O O O O NH AcHN O P O P O O N O HN OH OH O

O OH OH O NH

O HN

NH NH2

O HN

O OH

Figure 1.3. Chemical structure of S. aureus Park nucleotide. Park nucleotide core is the MurNAc sugar with a UDP leaving group. Its pentapeptide stem is composed of L-Ala, iso-D-Glu, L-Lys, D-Ala and D-Ala. The amino acids on the stem peptide varies between bacteria. 1.3.2 Synthesis of lipid-linked precursors

The next phase of peptidoglycan synthesis occurs on the phospholipid bilayer. First on the inner leaflet of the cytoplasmic membrane and then on the outer leaflet to form the PG layer.

The PG precursor must be translocated across the membrane and the next reactions must occur near the membrane surface. To prevent metabolically expensive molecules to simply diffuse away once they are outside, bacteria utilizes lipid carriers. The use of long linear polyprenyl- phospho lipids to facilitate the translocation of a sugar or long glycan across the membrane and to anchor to the membrane is common in other organisms and for other metabolic pathways.70

Human uses lipid carrier to transport sugar outside the membrane of protein glycosylation.70

Bacteria uses the same lipid carrier also for transporting wall lipopolysaccharides O-antigen71, teichoic acids72, capsular polysaccharides73.

12 In S. aureus, the lipid carrier was identified as Z8,E2,⍵-undecaprenyl phosphate, made from eleven isoprene units.74-75 The synthesis of this long complex lipid chain is carried out by just one enzyme, cis-prenyltransferase, UppS.76 The enzyme consecutively condenses eight isopentenyl pyrophosphate (IPP) units, yielding all cis-configurations, onto all-trans- farnesyl-diphosphate. The product undecaprenyl pyrophosphate is then dephosphorylated by BacA and possibly other enzyme(s) to yield the monophosphate lipid carrier.77 The chain length, the stereochemistry of the double bonds and the degree of saturation differs from one bacteria to another. The lipid carrier also exists as undecaprenol or undecaprenyl pyrophosphate. Interestingly the length of this lipid carrier, if stretched out, is longer than the thickness of the cytoplasmic membrane.

The first reaction that convert Park nucleotide into the first lipid-linked intermediate is carried out by an enzyme MraY, which belongs from the polyprenylphosphate N-acetyl exosamine 1-phosphate-transferase (PNPT) super family.78-

79 The PNPT superfamily includes enzymes such as TagO and WecA which transfer aminosugar on to a lipid carrier in the biosynthesis of wall teichoic and o-antigen respectively.80 MraY is a dimeric integral membrane protein that forms a dimer its activity

2+ is dependent on Mg . In 2004, Bouhss et al. purified Bacillus subtilis Mray (Mraybs) and indentified an aspartate as one of the key active residue.81 In 2013, Chung et al. solved the crystal structure of Aquifex aelicus MraY (MraYaa), the first of the PNPT enzymes to have a

80 2+ structure determined. The structure of Mrayaa reveals Mg -binding motif near the active site for binding to the pyrophosphate moiety of Park nucleotide and hydrophobic groove nearby, proposed to be the binding site of the lipid carrier.82 The key aspartate was hypothesized to deprotonate the phosphate moiety of the lipid carrier so that the carrier

13 could carry out a nucleophilic attack on Park nucleotide. The reaction yields undecaprenyl- pyrophate-MurNAc-pentapeptide and UMP as a leaving group. The lipidated monosaccharide product is commonly called ‘Lipid I’ (Figure 1.4). MraY broad substrate specificity in regard to the lipid carrier length was exploited to synthesize analogues of

Lipid I with various lipid chains.83 Several bacteria produces tunicamycin a structural analogue inhibitor of MraY, although this natural product has no therapeutic use as it is too toxic in human because it also inhibits other glycosylation reactions.84

HO HO O HO HO O O HO O O O O HO O AcHN O O O P O P O AcHN AcHN O O P O P O OH OH O HN OH OH 8 2 HN 8 2 O O O NH O NH

HO HO

O O HN HN

O O NH NH2 NH NH2

O O HN HN

O Lipid I O Lipid II OH OH

Figure 1.4. Chemical structures of lipid-linked intermediates. Lipid I is synthesized from Park nucleotide and the undecaprenyl lipid carrier by MraY. MurG attaches GlcNAc to Lipid I to produce Lipid II. The length and stereoisomer of the polyprenyl tail varies between bacteria.

In the next reaction, a membrane-associated enzyme MurG which belongs to the

NDP-glycosyltransferases superfamily, catalyzes the transfer of N-acetyl glucosamine

(GlcNAc) from UDP to the C4 hydroxyl of the muramic acid moiety on Lipid I.85-86 The product is a lapidated disaccharide and thus commonly called Lipid II (Figure 1.4). The reaction proceeds via an ordered Bi-Bi mechanism.87 UDP-GlcNAc binds first and then Lipid

I whose C4 hydroxyl attacks the anomeric carbon to form the β-(1-4)-glycosidic linkage

14 and release UDP as a leaving group. Walker and coworkers first reconstitute the activity of

MurG in 1988 and obtained its structure in apo-form in 200 and with the UDP-GlcNAc substrate in 2003.88-89

In some bacteria the product of MurG reaction is the final precursor to peptidoglycan. In S. aureus, further steps are needed to produce mature Lipid II. S. aureus has a pentaglycine sidechain attached to the stem peptide at the ε-amino group of L-Lys.90

Three enzymes, FemXAB, are non-ribosomal peptidyltransferase that sequentially install five glycine using glycyl-tRNA as the donor. FemX acts on Lipid II to add the first glycine,

FemA adds the second and the third, and lastly, FemB adds the final two glycines. 90 Their activities are not interchangeable.91-92 While FemX is essential for growth, femA and femB are not but they are necessary for antibiotic resistance in MRSA.93 The structure of S. aureus FemA was reported by Benson et al. in 2002, revealing globular protein with long L- shape hydrophobic groove for binding Lipid II and a long hydrophilic arm that is proposed to facilitate the binding of glycyl-tRNA.94 Other Gram-positive bacteria contain homologues of the Fem proteins that also function to install a peptide branch onto the stem peptide of

Lipid II. For example, Enterococcus faecalis has bppA1 and bppA2 that installs L-Ala-L-Ala branch while Streptococcus pneumoniae has MurM that install L-Ser-L-Ala or L-Ala-L-Ala.95

In S. aureus, the stem peptide of peptidoglycan was found to contain glutamine rather than glutamate that is in Park nucleotide. It was only recently that the enzyme complex that catalyze the amidation reaction was identified as GatD and MurT.96-97 The structure of the binary complex was reported in 2018 by Stehle and coworkers.98 GatD is a glutamine amidotransferase while MurT are structurally similar to other Mur ligases containing ATP-binding pocket, stem peptide binding site, and a zinc finger.98 The complex

15 has an open conformation with large space separating the GatD/MurT catalytic triad and the ATP binding site, although the two ends might close up upon Lipid II binding.98 The free carboxylate of D-iso-glutamate is activated by ATP. The active cysteine residue of GatD attacks glutamine to release ammonia which then attack the activated glutamate to release

ADP and create glutamine. In vitro reconstitution of the complex reveals that GatD/MurT could act on all membrane-bound precursors (Lipid I, Lipid II or Gly5-Lipid II).92 femA or femB mutants of S. aureus have lipid II that are amidated. Therefore, it is likely GatD/MurT and the FemXAB enzymes may work on Lipid II in any order.

Figure 1.5. Chemical structure of S. aureus native Lipid II. (a) S. aureus Lipid II with amidation on iso-D-Glu and with the pentaglycine branch on the 𝓔-amine of lysine. (b) A simplified structure of S. aureus Lipid II that will be used in this thesis.

Once matured Gly5-amidated-Lipid II is synthesized (Figure 1.5), it needs to be transported across the cytoplasmic membrane to the outside of the cell. Two proteins are proposed to be the flippase of Lipid II. Ruiz and coworkers in the United States proposed

MurJ, a member of the MOP (multidrug/oligo-saccharidyl-lipid/polysaccharide) exporter superfamily, as the flippase.99 Ruiz used bioinformatic and genetic evidences to support her

16 claims.99 Since then, many papers have been published in support of MurJ as the flippase.100-101 In vivo inhibition of mutant murJ by a chemical probe MTSES result in Lipid

II accumulation inside the inner leaflet of the cytoplasmic membrane.100 Several structure of MurJ homologues have been reported.102-103 MurJ has two conformations inward- and outward- facing.104 The binding of Lipid II to MurJ in the inward facing state triggers conformation change that push Lipid II to the outside.105 Rubino et al. demonstrated that proton gradient is required in priming MurJ back into the inward-facing state again.106

Meanwhile, Breukink and coworkers in the Netherlands identified FtsW, a member of the SEDS (shape, elongation, division, and sporulation) family, as the flippase via biochemical assays using purified proteins and membrane vesicles.107 FtsW and RodA, another protein in the SEDS family, were however later reported to be peptidoglycan glycosyltransferases.108-109

1.3.3 Peptidoglycan assembly and penicillin-binding proteins

Once transported to the outer leaflet of the cytoplasmic membrane, Lipid II is then polymerized by peptidoglycan glycosyltransferases (PGTs) into long linear peptidoglycan strands connecting each unit of disaccharide-pentapeptide-pentaglycine by β-(1,4)-glycosidic bond between the C4 of GlcNac and the C1 of MurNAc of another repeating unit (Figure 1.6).

The polymerization reaction also releases undecaprenyl pyrophosphate Lipid carrier which is dephosphorylated by BacA into undecaprenyl monophosphate.110 How the monophosphate lipid carrier is flipped back inside the cell has not been elucidated.

Figure 1.6. Mechanism of peptidoglycan glycosyl transferases (PGTs). Lipid II is polymerized into a glycan strand by PGT. A catalytic glutamate residue deprotonates 4-OH of the GlcNAc of a monomeric Lipid II which then attacks the C1 of MurNAc of the growing chain releasing pyrophosphate lipid and forming a β-(1,4)-glycosidic bond. These long glycan strands do not provide enough strength and rigidity to the cell. They are like parallel cotton threads requiring cross strands to become a piece of cloth. The cell wall of

S. aureus is highly crosslinked. One strand of peptidoglycan is crosslinked to another via covalent bond between the N-terminal of the pentaglycine branch and the D-Ala at the fourth position of the stem peptide. The exchange of the terminal D-Ala with the pentaglycine branch of another strand is carried out by enzymes called transpeptidases (TPs). Each strand crosslinks to a nearby strand multiple time and to many other nearby strands.

18 H O N OH R L-Lys N H O H O Enz OH Gly N OH 5 R L-Lys N H O NH2 O Gly5 O H H O H N N N OH R L-Lys O Enz R L-Lys N R L-Lys N H H O Gly5 Gly Gly5 5 NH Enz OH NH 2 NH2 2 OH HN O PG Acyl-enzyme intermediate Crosslinked PG

Figure 1.7. Mechanism of transpeptidases (TPs). The active serine of TP attacks the D-Ala-D-Ala amide bond of the stem peptide releasing the terminal D-Ala and creating an acyl-enzyme intermediate. The amine terminus of a pentaglycine branch from a different glycan strand attacks the high-energy acyl-enzyme intermediate to produce crosslinked PG. Although it was known that the peptidoglycan of S. aureus is crosslinked, the site of crosslink attachment was not revealed until 1965.111 Park and coworkers realize that while the

Park nucleotide contain two D-alanines, the stem peptide of the isolated S. aureus cell wall only contains one. He postulated that the transpeptidase enzyme use the energy of the D-Ala-D-Ala bond to drive the transpeptidation reaction which occurs outside of the cell lacking any ATP as an energy source (Figure 1.7).111 When Park analyzed the cell wall content of S. aureus culture treated with penicillin, he found that the stem peptide has more alanine than that from the culture without treatment. Thus, the mechanism of action of penicillin was finally solved. It inhibits the transpeptidation the final step of peptidoglycan synthesis. A few months later, Tipper and

Strominger published a paper confirming the hypothesis and created what is now called “Tipper-

Strominger hypothesis” – penicillin inhibits the transpeptidation by acting as a substrate analogue.112 They realize that the beta-lactam core of penicillin mimic the D-Ala-D-Ala moiety of the stem peptide. Once the TP enzymes attacks penicillin rather than releasing D-Ala and conforming acyl-enzyme intermediate, instead it forms a rather stable penicilloyl-enzyme adduct,

19 inactivating the TP enzyme as a result. The enzymes that bind to penicillin became known as

“penicillin-binding proteins” (PBPs).

Penicillin and other drugs in the beta-lactam family became useful chemical probes in identifying the PBP enzymes that are responsible for the final steps of PG synthesis.

Radiolabeled penicillin was incubated with the bacteria. The proteins from the membrane was then separated on sodium dodecyl sulfate (SDS) gel electrophoresis and detected by fluorography.113-115 In S. aureus, four protein bands were found (Figure 1.8).113 They are PBP1-4 in order of descending molecular weights. Later on, MRSA was found to have an additional

PBP, called PBP2a.32

The PBPs can be classified broadly as high-molecular weight (HMW) or low-molecular weight (LMW) PBPs. In S. aureus, PBP1-3 are HMW and are significantly heavier than the

LMW PBP4 (Table 1.1).113,115 HMW PBPs are also classified further as class A or class B. Class

A has both PGT and TP domains while class B has C-terminal TP activity and an N-terminal domain of unknown function. The N-terminal domain sometimes called “non-penicillin-binding domain” or “dimerization domain”. It was recently shown that this domain is important in forming the complex and controlling the activity of the SEDS PG polymerases.109,116

Figure 1.8. Penicillin-binding proteins (PBPs) in S. aureus. PBPs are labeled with Bocillin FL, a fluorescent beta- lactams and separated by their molecular weights on SDS-PAGE. Class A HMW PBPs has both PGT and TP domains while class B HMW PBPs and LMW PBPs only have a TP domain. This figure is modified from ref 117. PBP1 is an essential HMW Class B PBP.118 When PBP1 is selectively inhibited by cloxacilin, S. aureus cells become larger and have multiple septa although at the correct placement.119 When the gene is place under control of inducible promoter and the enzyme is depleted, similar results were observed.120 When the catalytic residue of PBP1 is mutated, S. aureus is still viable but could not separate after one or two cell division causing cell growth to stop.121 It was recently shown that complex formation of FtsW to PBP1is necessary the polymerase activity of FtsW.109 It is not clear how the TP activity of PBP1 influences the divisome as the TP activity is not required for FtsW polymerase activity.

PBP2 is an essential HMW Class A PBP. Its structure was elucidated in 2007 by

Strynadka and coworkers.122 Like other PBPs, PBP2 has a transmembrane helix for anchoring to the lipid membrane. It has an N-terminal PGT domain and a C-terminal TP domain separated by a small linker region. The PGT domain are membrane bound while the TP domain is solvent- exposed. The TP activity of PBP2 is essential for cell growth in methicillin-sensitive S. aureus, although the addition of MRSA PBP2a TP activity could rescue the cells.123 The PGT activity of

21 PBP2 is not essential as the polymerase activity can be supplemented by non-essential monofunctional PGT, SgtB but is required for PBP2a to confer resistance.124 Bacterial two- hybrid assay also show that both SgtB and another monofunctional PGT SgtA may work in collaboration with PBP2.125

PBP3 is a non-essential HMW Class B PBP. Inactivation of PBP3 cause no detectable change in cell wall composition and had little effect on cell growth.126 In rare cases, cells with

PBP3 deletion show two septa. PBP3 plays a supporting role in cell division and separation as the division of PBP3-deletion strain is more heavily affected by methicillin compared wild-type strains.126 It is unclear whether PBP3 has a SEDS partner like PBP1. This is possible as clinical strains with resistance to methicillin and oxacillin and do not have PBP3 at all were observed.

The bacteria rather removes PBP3 completely than leaving it acylated by an antibiotic.126

PBP4 is the only LMW PBP in S. aureus. Its structure was determined by Gopal and coworkers in 2009.127 It is composed an N-terminal TP domain, a small C-terminal domain of unknown function and a transmembrane helix. Interestingly, the TP domain of PBP4 was reported to have various activity ranging from canonical transpeptidation, carboxypeptidase activity, and beta-lactamase activity. It is not essential for viability, but it has little effect on cell growth.128 ∆PBP4 mutant shows great reduction in PG crosslinking and increase sensitivity to beta-lactams. Its transpeptidation activity has been reconstituted and is used as a biochemical tool for D-amino acid exchange in vitro and in vivo for labeling peptidoglycan. PBP4 expression is essential for beta-lactam resistance in community-acquired MRSA.129 Its carboxypeptidase activity may have physiological relevance as deletion of PBP4 lead to intermediate vancomycin resistance.130-131 Its beta-lactamase activity has only been shown in vitro and is unlikely to be biologically relevance.127 In the absence of PBP2a, mutations in PBP4 could also confer high

22 beta-lactam resistance comparable to the level of resistance from PBP2a. This fact is only recently appreciated. More work is needed to figure out how PBP4 mutant provides the resistance and whether PBP4 requires a complex formation with any PGT enzymes.

Table 1.1. Classification of PBPs in S. aureus. S. aureus PBPs Class Cellular Roles Essentiality Coordination HMW Cell division and Work together with PBP1 Yes (Class B) separation FtsW Major PBP in PG HMW Work together with PBP2 polymerization and Yes (Class A) PBP2a or alone crosslinking HMW Beta-lactam Work together with PBP2a No (Class B) resistance PBP2 Unclear HMW (Possibly cell PBP3 No Not determined (Class B) division and separation) Secondary PG PBP4 LMW No No crosslinking

1.4 Biology of methicillin-resistant S. aureus

Methicillin-resistance S. aureus is resistant not to only methicillin, an old drug that is no longer used due to toxicity, but it is also resistant to virtually all drugs in the beta-lactams family.12 Methicillin-sensitive S. aureus (MSSA) becomes methicillin-resistance S. aureus

(MRSA) upon an acquisition of a mobile genetic element. In the next few sections, I will briefly summarize the molecular mechanism of methicillin-resistance in S. aureus.

1.4.1 SCCmec

Staphylococcal cassette chromosome mec (SCCmec) is a unique mobile genetic element about 21-67 kilobasepairs in length.12 It carries the gene mecA enconding beta-lactam resistant enzyme PBP2a. The expression of PBP2a is regulated by a repressor protein MecI and signal transducer protein MecR1 (Figure 1.9).132 Both proteins have high protein sequence homology with the protein BlaR1 and BlaI from the beta-lactamase induction blaZ system, implying that

23 the mec genes may acquire regulatory genes from the blaZ system.132 In fact, MecI and BlaI are functionally interchangeable.133 MecI is a repressor that binds to the promoter region mecR and mecA. MecR is a membrane protein with penicillin-binding domain on the outer leaflet of the membrane and a zinc metalloprotease signaling domain on the inside. The binding of beta- lactams to MecR causes conformation change inducing autocatalytic cleavage.134 This cleavage leads to degradation of MecI.135 How that occurs is currently unclear. Cell wall fragments such as iso-D-Glu-L-Lys that may be generated upon beta-lactam treatment could also form a complex with MecI preventing it from repressing mecR and mecA and promoting MecI degradation.136 Once MecI is removed, mecA could be transcribed and PBP2a could be expressed.

Figure 1.9. Regulation of penicillin binding protein 2a expression. Under a normal condition, repressor MecI prevents the transcription of mecA and mecR1. When MRSA is threatened by beta-lactams, MecR1 senses and binds to beta-lactams resulting in an autolytic cleavage which by yet to be revealed mechanism leads to the degradation of MecI and expression of PBP2a. Cell wall fragments could also binds to MecI promoting its proteolytic cleavage. This figure is adapted from ref 137.

24 SCCmec also carries cassette chromosome recombinase genes (ccr) which helps the mobile genetic element excise from the chromosome and reintergrate at a specific site on the chromosome called bacterial chromosomal attachment site attBSCC located near S. aureus origin of replication.138 The SCCmec gene is flanked specific inverted repeats and direct repeats that are recognized by the ccr recombinases.

SCCmec sequence is used for epidemiological surveillance of MRSA. SCCmec elements are classified into different types based on the combination of the specific mec and ccr genes.

There are eight types of the ccr complex and five classes of mec complex, resulting in 11 types of SCCmec elements.139 The cassette also contain J regions harboring non-essential components of the cassette that provides resistance to other antibiotics and heavy metals. The J regions are used to differentiate SCCmec into multiple subtypes. Types II and III are large SCCmec elements that also carries genes that confer resistance to other classes of antibiotics such as kanamycin, erythromycin, and tetracycline.140 Type XI contains mecC, encoding PBP2C a homologue of PBP2a.141 Some SCCmec types are associated with hospital-acquired MRSA while others with community-acquired MRSA.

SCCmec elements are not native to S. aureus. It was proposed that SCCmec originated long ago, before the widespread use of beta-lactams, in Staphylococcus sciuri group, the most primitive group of Staphylococcus species. SCCmec evolution occurred through several steps involving at least three species: Staphylococcus sciuri (S. sciuri), Staphylococcus vitulinus (S. vitulinus), and Staphylococcus fleurettii (S. fleuretttii). Rolo et al. proposed that mecA first emerged in S. fleuretttii, combined with its regulatory genes mecR and mecI into the mec complex in S. vitulinus and integrated into the SCC mobile genetic element in S. sciuri.142

25 1.4.2 Beta-lactam-resistant PBP2a

Penicillin-binding protein 2a (PBP2a) is a non-essential HMW class B PBP that has a transpeptidase domain for peptidoglycan crosslinking. When MRSA expressing PBP2a is treated with beta-lactams at the concentration high enough to inhibit the native transpeptidases, the bacteria could still grow and synthesize PG although with lower degree of PG crosslinking.124 In competition binding assay PBP2a displays low-binding affinity to beta-lactams.32 Kinetic studies reveal that PBP2a has low binding affinity to beta-lactams (high Kd) and slower acylation rate

143-144 (low k2).

Lim and Strynadka reported the crystal structure of PBP2a in 2002.122 The enzyme contains an N-terminal transmembrane helix, a non-penicillin-binding domain and a C-terminal transpeptidase. The crystal structure reveals a closed active site that is inaccessible to the beta- lactams without conformational change. This explains the high Kd. PBP2a active serine residue is hidden in a tight groove that requires conformation change for nucleophilic attack explaining the low k2. This closed active site however needs to be reconciled with the activity of PBP2a, crosslinking extremely large peptidoglycan molecules.

Shariar Mobashery proposed that PBP2a contains an allosteric site that modulate the structure of the active site.145 First, he found that PBP2a has higher beta-lactam acylation rate in the presence of peptidoglycan fragment.146 Then, in 2013, Mobashery and coworkers obtained a crystal structure of PBP2a and ceftaroline, a fifth generation beta-lactam that is active against

MRSA.145 Two ceftaroline molecules were found in the structure (Figure 1.10). One acylated by the serine active residue and the second bound to the non-penicillin-binding domain. In a different crystal complex, a peptidoglycan fragment was found binding to the same domain, now called allosteric domain. 145 Although the allosteric domain is over 60 A away from the active

26 site, the binding of ceftaroline to the allosteric domain triggers a large conformational change via a long network of salt bridge that results in expanding the volume of active site from 500 A3 to

1300 A3, opening up the active residue to be reached by not just beta-lactam but also the substrate PG.145 One clinal isolate of MRSA that became resistance to ceftaroline contains

PBP2a with N146K and E150K mutations in the allosteric site that disrupt the salt bridge network.145

Figure 1.10. Crystal structure of S. aureus PBP2a. This structure depicts PBP2a with two ceftaroline (CFT1 and CFT2) bound. PBP2a has a transpeptidase domain with a closed active site (where CFT1 binds) preventing most beta-lactams from inactivating it. Mobashery and coworkers proposed that a binding of peptidoglycan fragment to the allosteric site (where CFT2 is) can triggers an opening of the TP active site allowing crosslinking.145 Ceftaroline is proposed to mimic the PG fragment. One molecule binds to the allosteric domain to trigger TP opening while another molecule covalently inactivates the TP active serine residue. This figure is adapted from ref 145. 1.4.3 Factors essential for methicillin resistance

The expression of PBP2a is necessary for beta-lactam resistance in MRSA, but it is not sufficient. There was no correlation between PBP2a expression level and beta-lactam resistance.

Other additional genes are also necessary. These genes are named fem genes (for factor essential for methicillin resistance) or aux genes (for auxiliary genes).147 These include genes involved in the synthesis of peptidoglycan and its precursors (such as pbp1, pbp2, pbp4, murA-F, murJ,

27 gatD/murT and femXAB), genes involved in the synthesis of teichoic acids (such as tarO, tarS, ltaS) and global regulators (such as agr and sar).148 Inactivation of these enzymes resensitize

MRSA to beta-lactam, providing a new strategy to fight MRSA.

1.5 References

(1) Zietz, B. P.; Dunkelberg, H. The history of the plague and the research on the causative agent Yersinia pestis. International journal of hygiene and environmental health 2004, 207 (2), 165-178.

(2) Barberis, I.; Bragazzi, N. L.; Galluzzo, L.; Martini, M. The history of tuberculosis: from the first historical records to the isolation of Koch's bacillus. J Prev Med Hyg 2017, 58 (1), E9-E12.

(3) Oladokun, M. O.; Okoh, I. A. Vibrio cholerae: A historical perspective and current trend. Asian Pacific Journal of Tropical Disease 2016, 6 (11), 895-908.

(4) Aminov, R. I. A brief history of the antibiotic era: lessons learned and challenges for the future. Front Microbiol 2010, 1, 134-134.

(5) Magiorakos, A. P.; Srinivasan, A.; Carey, R. B.; Carmeli, Y.; Falagas, M. E.; Giske, C. G.; Harbarth, S.; Hindler, J. F.; Kahlmeter, G.; Olsson-Liljequist, B.; Paterson, D. L.; Rice, L. B.; Stelling, J.; Struelens, M. J.; Vatopoulos, A.; Weber, J. T.; Monnet, D. L. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clinical Microbiology and Infection 2012, 18 (3), 268-281.

(6) O’Neill, J. The Review on Antimicrobial Resistance; Tackling Drug-Resistnt Infections Globally: Final Report and Recommendations. 2016. Available at: https://amr- review.org/sites/default/files/160525_Final%20paper_with%20cover.pdf. Accessed May 1, 2019.

(7) Centers for Disease Control and Prevention Office of Infectious Disease Antibiotic resistance threats in the United States, 2013. 2013. Available at: https://www.cdc.gov/drugresistance/threat-report-2013/pdf/ar-threats-2013-508.pdf. Accessed May 1, 2019.

28 (8) Kourtis, A. P.; Hatfield, K.; Baggs, J.; et al. Vital Signs: Epidemiology and Recent Trends in Methicillin-Resistant and in Methicillin-Susceptible Staphylococcus aureus Bloodstream Infections - United States. MMWR. Morbidity and mortality weekly report 2019, 68 (9), 214-219.

(9) Fernandes, R.; Amador, P.; Prudêncio, C. β-Lactams: chemical structure, mode of action and mechanisms of resistance. Reviews in Medical Microbiology 2013, 24 (1), 7-17.

(10) Centers for Disease Control and Prevention Outpatient antibiotic prescriptions - United States, 2015.

(11) Owens, R. C., Jr. An overview of harms associated with beta-lactam antimicrobials: where do the carbapenems fit in? Crit Care 2008, 12 Suppl 4 (Suppl 4), S3-S3.

(12) Peacock, S. J.; Paterson, G. K. Mechanisms of Methicillin Resistance in Staphylococcus aureus. Annual review of biochemistry 2015, 84 (1), 577-601.

(13) Wilson, L. G. The early recognition of streptococci as causes of disease. Medical history 1987, 31 (4), 403-414.

(14) Newsom, S. W. B. Ogston's coccus. Journal of Hospital Infection 2008, 70 (4), 369-372.

(15) Ogston, A. Micrococcus Poisoning. Journal of anatomy and physiology 1882, 17 (Pt 1), 24-58.

(16) Licitra, G. Etymologia: Staphylococcus. Emerg Infect Dis 2013, 19 (9), 1553-1553.

(17) Wertheim, H. F. L.; Melles, D. C.; Vos, M. C.; van Leeuwen, W.; van Belkum, A.; Verbrugh, H. A.; Nouwen, J. L. The role of nasal carriage in Staphylococcus aureus infections. The Lancet Infectious Diseases 2005, 5 (12), 751-762.

(18) Bien, J.; Sokolova, O.; Bozko, P. Characterization of Virulence Factors of Staphylococcus aureus: Novel Function of Known Virulence Factors That Are Implicated in Activation of Airway Epithelial Proinflammatory Response. J Pathog 2011, 2011, 601905-601905.

29 (19) Tong, S. Y. C.; Davis, J. S.; Eichenberger, E.; Holland, T. L.; Fowler, V. G., Jr. Staphylococcus aureus infections: epidemiology, pathophysiology, clinical manifestations, and management. Clinical microbiology reviews 2015, 28 (3), 603-661.

(20) Dunyach-Remy, C.; Ngba Essebe, C.; Sotto, A.; Lavigne, J.-P. Staphylococcus aureus Toxins and Diabetic Foot Ulcers: Role in Pathogenesis and Interest in Diagnosis. Toxins 2016, 8 (7), 209.

(21) Skinner, D.; Keefer, C. S. Significance of bacteremia caused by Staphylococcus aureus. JAMA Internal Medicine 1941, 68 (5), 851-875.

(22) American Chemical Society International Historic Chemical Landmarks Discovery and Development of Penicillin. Available at: http://www.acs.org/content/acs/en/education/whatischemistry/landmarks/flemingpenicilli n.html. Accessed May 10, 2019.

(23) Chain, E.; Florey, H. W.; Gardner, A. D.; Heatley, N. G.; Jennings, M. A.; Orr-Ewing, J.; Sanders, A. G. PENICILLIN AS A CHEMOTHERAPEUTIC AGENT. The Lancet 1940, 236 (6104), 226-228.

(24) Abraham, E. P.; Chain, E.; Fletcher, C. M.; Gardner, A. D.; Heatley, N. G.; Jennings, M. A.; Florey, H. W. FURTHER OBSERVATIONS ON PENICILLIN. The Lancet 1941, 238 (6155), 177-189.

(25) Quinn, R. Rethinking antibiotic research and development: World War II and the penicillin collaborative. Am J Public Health 2013, 103 (3), 426-434.

(26) Park, J. T.; Strominger, J. L. Mode of action of penicillin. Science (New York, N.Y.) 1957, 125 (3238), 99-101.

(27) Rammelkamp, C. H.; Maxon, T. Resistance of Staphylococcus aureus to the Action of Penicillin. Proceedings of the Society for Experimental Biology and Medicine 1942, 51 (3), 386-389.

(28) Lowy, F. D. Antimicrobial resistance: the example of Staphylococcus aureus. The Journal of clinical investigation 2003, 111 (9), 1265-1273.

(29) Knox, R. A new penicillin (BRL 1241) active against penicillin-resistant staphylococci. British medical journal 1960, 2 (5200), 690-693.

30 (30) Barber, M.; Waterworth, P. M. PENICILLINASE-RESISTANT PENICILLINS AND CEPHALOSPORINS. British medical journal 1964, 2 (5405), 344-349.

(31) Jevons, M. P. “Celbenin” - resistant Staphylococci. British medical journal 1961, 1 (5219), 124-125.

(32) Hartman, B.; Tomasz, A. Altered penicillin-binding proteins in methicillin-resistant strains of Staphylococcus aureus. Antimicrobial agents and chemotherapy 1981, 19 (5), 726-735.

(33) Styers, D.; Sheehan, D. J.; Hogan, P.; Sahm, D. F. Laboratory-based surveillance of current antimicrobial resistance patterns and trends among Staphylococcus aureus: 2005 status in the United States. Annals of clinical microbiology and antimicrobials 2006, 5, 2.

(34) Duplessis, C.; Crum-Cianflone, N. F. Ceftaroline: A New Cephalosporin with Activity against Methicillin-Resistant Staphylococcus aureus (MRSA). Clin Med Rev Ther 2011, 3, a2466.

(35) Kisgen, J.; Whitney, D. Ceftobiprole, a Broad-Spectrum Cephalosporin With Activity against Methicillin-Resistant Staphylococcus aureus (MRSA). P T 2008, 33 (11), 631- 641.

(36) Chan, L. C.; Basuino, L.; Diep, B.; Hamilton, S.; Chatterjee, S. S.; Chambers, H. F. Ceftobiprole- and Ceftaroline-Resistant Methicillin-Resistant Staphylococcus aureus. Antimicrobial agents and chemotherapy 2015, 59 (5), 2960.

(37) Chan, L. C.; Gilbert, A.; Basuino, L.; da Costa, T. M.; Hamilton, S. M.; dos Santos, K. R.; Chambers, H. F.; Chatterjee, S. S. PBP 4 Mediates High-Level Resistance to New- Generation Cephalosporins in Staphylococcus aureus. Antimicrobial agents and chemotherapy 2016, 60 (7), 3934-3941.

(38) McNeely, T. B.; Shah, N. A.; Fridman, A.; Joshi, A.; Hartzel, J. S.; Keshari, R. S.; Lupu, F.; DiNubile, M. J. Mortality among recipients of the Merck V710 Staphylococcus aureus vaccine after postoperative S. aureus infections: an analysis of possible contributing host factors. Human vaccines & immunotherapeutics 2014, 10 (12), 3513-3516.

(39) Begier, E.; Seiden, D. J.; Patton, M.; Zito, E.; Severs, J.; Cooper, D.; Eiden, J.; Gruber, W. C.; Jansen, K. U.; Anderson, A. S.; Gurtman, A. SA4Ag, a 4-antigen Staphylococcus

31 aureus vaccine, rapidly induces high levels of bacteria-killing antibodies. Vaccine 2017, 35 (8), 1132-1139.

(40) Haghighat, S.; Siadat, S. D.; Sorkhabadi, S. M. R.; Sepahi, A. A.; Mahdavi, M. A novel recombinant vaccine candidate comprising PBP2a and autolysin against Methicillin Resistant Staphylococcus aureus confers protection in the experimental mice. Molecular immunology 2017, 91, 1-7.

(41) Moscoso, M.; García, P.; Cabral, M. P.; Rumbo, C.; Bou, G. A D-Alanine auxotrophic live vaccine is effective against lethal infection caused by Staphylococcus aureus. Virulence 2018, 9 (1), 604-620.

(42) Rubin, I. M.; Hansen, T. A.; Klingenberg, A. M.; Petersen, A. M.; Worning, P.; Westh, H.; Bartels, M. D. A Sporadic Four-Year Hospital Outbreak of a ST97-IVa MRSA With Half of the Patients First Identified in the Community. Front Microbiol 2018, 9, 1494- 1494.

(43) Buehlmann, M.; Frei, R.; Fenner, L.; Dangel, M.; Fluckiger, U.; Widmer, A. F. Highly effective regimen for decolonization of methicillin-resistant Staphylococcus aureus carriers. Infection control and hospital epidemiology 2008, 29 (6), 510-516.

(44) Edgeworth, J. D. Has decolonization played a central role in the decline in UK methicillin-resistant Staphylococcus aureus transmission? A focus on evidence from intensive care. Journal of Antimicrobial Chemotherapy 2011, 66 (suppl_2), ii41-ii47.

(45) Robicsek, A.; Beaumont, J. L.; Thomson, R. B., Jr.; Govindarajan, G.; Peterson, L. R. Topical therapy for methicillin-resistant Staphylococcus aureus colonization: impact on infection risk. Infection control and hospital epidemiology 2009, 30 (7), 623-632.

(46) Zipperer, A.; Konnerth, M. C.; Laux, C.; Berscheid, A.; Janek, D.; Weidenmaier, C.; Burian, M.; Schilling, N. A.; Slavetinsky, C.; Marschal, M.; Willmann, M.; Kalbacher, H.; Schittek, B.; Brötz-Oesterhelt, H.; Grond, S.; Peschel, A.; Krismer, B. Human commensals producing a novel antibiotic impair pathogen colonization. Nature 2016, 535, 511.

(47) Jensen, K. C.; Hair, B. B.; Wienclaw, T. M.; Murdock, M. H.; Hatch, J. B.; Trent, A. T.; White, T. D.; Haskell, K. J.; Berges, B. K. Isolation and Host Range of Bacteriophage with Lytic Activity against Methicillin-Resistant Staphylococcus aureus and Potential Use as a Fomite Decontaminant. PLOS ONE 2015, 10 (7), e0131714.

32 (48) Verstappen, K. M.; Tulinski, P.; Duim, B.; Fluit, A. C.; Carney, J.; van Nes, A.; Wagenaar, J. A. The Effectiveness of Bacteriophages against Methicillin-Resistant Staphylococcus aureus ST398 Nasal Colonization in Pigs. PLoS One 2016, 11 (8), e0160242.

(49) Phage Therapy Center Medical Conditions that We Treat. Available at: https://www.phagetherapycenter.com/pii/PatientServlet?command=static_staph. Accessed May 2, 2019.

(50) Totte, J. E. E.; van Doorn, M. B.; Pasmans, S. Successful Treatment of Chronic Staphylococcus aureus-Related Dermatoses with the Topical Endolysin Staphefekt SA.100: A Report of 3 Cases. Case reports in dermatology 2017, 9 (2), 19-25.

(51) Fischetti, V. A. Development of Phage Lysins as Novel Therapeutics: A Historical Perspective. Viruses 2018, 10 (6), 310.

(52) Weibull, C. The isolation of protoplasts from Bacillus megaterium by controlled treatment with lysozyme. Journal of bacteriology 1953, 66 (6), 688-695.

(53) Salton, M. R. J. Studies of the bacterial cell wall: IV. The composition of the cell walls of some gram-positive and gram-negative bacteria. Biochimica et Biophysica Acta 1953, 10, 512-523.

(54) Salton, M. R. J. Cell Wall of Micrococcus Lysodeikticus as the Substrate of Lysozyme. Nature 1952, 170 (4331), 746-747.

(55) Matias, V. R.; Beveridge, T. J. Native cell wall organization shown by cryo-electron microscopy confirms the existence of a periplasmic space in Staphylococcus aureus. J Bacteriol 2006, 188 (3), 1011-1021.

(56) Vollmer, W.; Blanot, D.; de Pedro, M. A. Peptidoglycan structure and architecture. FEMS microbiology reviews 2008, 32 (2), 149-167.

(57) Strominger, J. L.; Park, J. T.; Thompson, R. E. Composition of the Cell Wall of Staphylococcus aureus: Its Relation to the Mechanism of Action of Penicillin. Journal of Biological Chemistry 1959, 234 (12), 3263-3268.

(58) Silhavy, T. J.; Kahne, D.; Walker, S. The bacterial cell envelope. Cold Spring Harbor perspectives in biology 2010, 2 (5), a000414.

33 (59) Park, J. T. Uridine-5'-pyrophosphate derivatives. II. Isolation from Staphylococcus aureus. The Journal of biological chemistry 1952, 194 (2), 877-884.

(60) Park, J. T. Uridine-5'-pyrophosphate derivatives. II. A structure common to three derivatives. The Journal of biological chemistry 1952, 194 (2), 885-895.

(61) Park, J. T. Uridine-5'-pyrophosphate derivatives. III. Amino acid-containing derivatives. The Journal of biological chemistry 1952, 194 (2), 897-904.

(62) Strominger, J. L.; Park, J. T.; Thompson, R. E. Composition of the cell wall of Staphylococcus aureus: its relation to the mechanism of action of penicillin. The Journal of biological chemistry 1959, 234, 3263-3268.

(63) Bugg, T. D. H.; Walsh, C. T. Intracellular steps of bacterial cell wall peptidoglycan biosynthesis: enzymology, antibiotics, and antibiotic resistance. Natural Product Reports 1992, 9 (3), 199-215.

(64) Barreteau, H.; Kovac, A.; Boniface, A.; Sova, M.; Gobec, S.; Blanot, D. Cytoplasmic steps of peptidoglycan biosynthesis. FEMS microbiology reviews 2008, 32 (2), 168-207.

(65) Zawadzke, L. E.; Bugg, T. D.; Walsh, C. T. Existence of two D-alanine:D-alanine ligases in Escherichia coli: cloning and sequencing of the ddlA gene and purification and characterization of the DdlA and DdlB enzymes. Biochemistry 1991, 30 (6), 1673-1682.

(66) Scaletti, E. R.; Luckner, S. R.; Krause, K. L. Structural features and kinetic characterization of alanine racemase from Staphylococcus aureus (Mu50). Acta crystallographica. Section D, Biological crystallography 2012, 68 (Pt 1), 82-92.

(67) Esaki, N.; Walsh, C. T. Biosynthetic alanine racemase of Salmonella typhimurium: purification and characterization of the enzyme encoded by the alr gene. Biochemistry 1986, 25 (11), 3261-3267.

(68) Marquardt, J. L.; Brown, E. D.; Lane, W. S.; Haley, T. M.; Ichikawa, Y.; Wong, C. H.; Walsh, C. T. Kinetics, stoichiometry, and identification of the reactive thiolate in the inactivation of UDP-GlcNAc enolpyruvoyl transferase by the antibiotic fosfomycin. Biochemistry 1994, 33 (35), 10646-10651.

(69) Walsh, C. T. Suicide substrates: mechanism-based enzyme inactivators with therapeutic potential. Trends in Biochemical Sciences 1983, 8 (7), 254-257.

34 (70) Manat, G.; Roure, S.; Auger, R.; Bouhss, A.; Barreteau, H.; Mengin-Lecreulx, D.; Touzé, T. Deciphering the metabolism of undecaprenyl-phosphate: the bacterial cell-wall unit carrier at the membrane frontier. Microb Drug Resist 2014, 20 (3), 199-214.

(71) Valvano, M. A. Export of O-specific lipopolysaccharide. Frontiers in bioscience : a journal and virtual library 2003, 8, s452-471.

(72) Brown, S.; Santa Maria, J. P., Jr.; Walker, S. Wall teichoic acids of gram-positive bacteria. Annual review of microbiology 2013, 67, 313-336.

(73) Whitfield, C. Biosynthesis and assembly of capsular polysaccharides in Escherichia coli. Annual review of biochemistry 2006, 75, 39-68.

(74) Higashi, Y.; Strominger, J. L.; Sweeley, C. C. Structure of a lipid intermediate in cell wall peptidoglycan synthesis: a derivative of a C55 isoprenoid alcohol. Proc Natl Acad Sci U S A 1967, 57 (6), 1878-1884.

(75) Higashi, Y.; Strominger, J. L.; Sweeley, C. C. Biosynthesis of the peptidoglycan of bacterial cell walls. XXI. Isolation of free C55-isoprenoid alcohol and of lipid intermediates in peptidoglycan synthesis from Staphylococcus aureus. The Journal of biological chemistry 1970, 245 (14), 3697-3702.

(76) Apfel, C. M.; Takács, B.; Fountoulakis, M.; Stieger, M.; Keck, W. Use of genomics to identify bacterial undecaprenyl pyrophosphate synthetase: cloning, expression, and characterization of the essential uppS gene. Journal of bacteriology 1999, 181 (2), 483- 492.

(77) El Ghachi, M.; Bouhss, A.; Blanot, D.; Mengin-Lecreulx, D. The bacA gene of Escherichia coli encodes an undecaprenyl pyrophosphate phosphatase activity. The Journal of biological chemistry 2004, 279 (29), 30106-30113.

(78) Ikeda, M.; Wachi, M.; Jung, H. K.; Ishino, F.; Matsuhashi, M. The Escherichia coli mraY gene encoding UDP-N-acetylmuramoyl-pentapeptide: undecaprenyl-phosphate phospho- N-acetylmuramoyl-pentapeptide transferase. Journal of Bacteriology 1991, 173 (3), 1021.

(79) Boyle, D. S.; Donachie, W. D. mraY Is an Essential Gene for Cell Growth in Escherichia coli. Journal of Bacteriology 1998, 180 (23), 6429.

35 (80) Chung, B. C.; Zhao, J.; Gillespie, R. A.; Kwon, D.-Y.; Guan, Z.; Hong, J.; Zhou, P.; Lee, S.-Y. Crystal structure of MraY, an essential membrane enzyme for bacterial cell wall synthesis. Science (New York, N.Y.) 2013, 341 (6149), 1012-1016.

(81) Bouhss, A.; Crouvoisier, M.; Blanot, D.; Mengin-Lecreulx, D. Purification and characterization of the bacterial MraY translocase catalyzing the first membrane step of peptidoglycan biosynthesis. The Journal of biological chemistry 2004, 279 (29), 29974- 29980.

(82) Chung, B. C.; Zhao, J.; Gillespie, R. A.; Kwon, D.-Y.; Guan, Z.; Hong, J.; Zhou, P.; Lee, S.-Y. Crystal Structure of MraY, an Essential Membrane Enzyme for Bacterial Cell Wall Synthesis. Science (New York, N.Y.) 2013, 341 (6149), 1012.

(83) Huang, L. Y.; Huang, S. H.; Chang, Y. C.; Cheng, W. C.; Cheng, T. J.; Wong, C. H. Enzymatic synthesis of lipid II and analogues. Angewandte Chemie (International ed. in English) 2014, 53 (31), 8060-8065.

(84) Hakulinen, J. K.; Hering, J.; Branden, G.; Chen, H.; Snijder, A.; Ek, M.; Johansson, P. MraY-antibiotic complex reveals details of tunicamycin mode of action. Nature chemical biology 2017, 13 (3), 265-267.

(85) Salmond, G. P.; Lutkenhaus, J. F.; Donachie, W. D. Identification of new genes in a cell envelope-cell division gene cluster of Escherichia coli: cell envelope gene murG. J Bacteriol 1980, 144 (1), 438-440.

(86) Mengin-Lecreulx, D.; Texier, L.; Rousseau, M.; van Heijenoort, J. The murG gene of Escherichia coli codes for the UDP-N-acetylglucosamine: N-acetylmuramyl- (pentapeptide) pyrophosphoryl-undecaprenol N-acetylglucosamine transferase involved in the membrane steps of peptidoglycan synthesis. Journal of bacteriology 1991, 173 (15), 4625-4636.

(87) Chen, L.; Men, H.; Ha, S.; Ye, X. Y.; Brunner, L.; Hu, Y.; Walker, S. Intrinsic lipid preferences and kinetic mechanism of Escherichia coli MurG. Biochemistry 2002, 41 (21), 6824-6833.

(88) Hu, Y.; Chen, L.; Ha, S.; Gross, B.; Falcone, B.; Walker, D.; Mokhtarzadeh, M.; Walker, S. Crystal structure of the MurG:UDP-GlcNAc complex reveals common structural principles of a superfamily of glycosyltransferases. Proceedings of the National Academy of Sciences 2003, 100 (3), 845.

36 (89) Men, H.; Park, P.; Ge, M.; Walker, S. Substrate Synthesis and Activity Assay for MurG. Journal of the American Chemical Society 1998, 120 (10), 2484-2485.

(90) Niyomporn, B.; Dahl, J. L.; Strominger, J. L. Biosynthesis of the peptidoglycan of bacterial cell walls. IX. Purification and properties of glycyl transfer ribonucleic acid synthetase from Staphylococcus aureus. The Journal of biological chemistry 1968, 243 (4), 773-778.

(91) Bouhss, A.; Trunkfield, A. E.; Bugg, T. D.; Mengin-Lecreulx, D. The biosynthesis of peptidoglycan lipid-linked intermediates. FEMS microbiology reviews 2008, 32 (2), 208- 233.

(92) Schneider, T.; Senn, M. M.; Berger-Bachi, B.; Tossi, A.; Sahl, H. G.; Wiedemann, I. In vitro assembly of a complete, pentaglycine interpeptide bridge containing cell wall precursor (lipid II-Gly5) of Staphylococcus aureus. Molecular microbiology 2004, 53 (2), 675-685.

(93) Hegde, S. S.; Shrader, T. E. FemABX Family Members Are Novel Nonribosomal Peptidyltransferases and Important Pathogen-specific Drug Targets. Journal of Biological Chemistry 2001, 276 (10), 6998-7003.

(94) Benson, T. E.; Prince, D. B.; Mutchler, V. T.; Curry, K. A.; Ho, A. M.; Sarver, R. W.; Hagadorn, J. C.; Choi, G. H.; Garlick, R. L. X-ray crystal structure of Staphylococcus aureus FemA. Structure (London, England : 1993) 2002, 10 (8), 1107-1115.

(95) Bouhss, A.; Josseaume, N.; Severin, A.; Tabei, K.; Hugonnet, J.-E.; Shlaes, D.; Mengin- Lecreulx, D.; van Heijenoort, J.; Arthur, M. Synthesis of the l-Alanyl-l-alanine Cross- bridge of Enterococcus faecalis Peptidoglycan. Journal of Biological Chemistry 2002, 277 (48), 45935-45941.

(96) Figueiredo, T. A.; Sobral, R. G.; Ludovice, A. M.; Almeida, J. M.; Bui, N. K.; Vollmer, W.; de Lencastre, H.; Tomasz, A. Identification of genetic determinants and enzymes involved with the amidation of glutamic acid residues in the peptidoglycan of Staphylococcus aureus. PLoS pathogens 2012, 8 (1), e1002508.

(97) Munch, D.; Roemer, T.; Lee, S. H.; Engeser, M.; Sahl, H. G.; Schneider, T. Identification and in vitro analysis of the GatD/MurT enzyme-complex catalyzing lipid II amidation in Staphylococcus aureus. PLoS pathogens 2012, 8 (1), e1002509.

37 (98) Nöldeke, E. R.; Muckenfuss, L. M.; Niemann, V.; Müller, A.; Störk, E.; Zocher, G.; Schneider, T.; Stehle, T. Structural basis of cell wall peptidoglycan amidation by the GatD/MurT complex of Staphylococcus aureus. Scientific Reports 2018, 8 (1), 12953.

(99) Ruiz, N. Bioinformatics identification of MurJ (MviN) as the peptidoglycan lipid II flippase in Escherichia coli. Proc Natl Acad Sci U S A 2008, 105 (40), 15553-15557.

(100) Sham, L.-T.; Butler, E. K.; Lebar, M. D.; Kahne, D.; Bernhardt, T. G.; Ruiz, N. MurJ is the flippase of lipid-linked precursors for peptidoglycan biogenesis. Science (New York, N.Y.) 2014, 345 (6193), 220.

(101) Butler, E. K.; Davis, R. M.; Bari, V.; Nicholson, P. A.; Ruiz, N. Structure-function analysis of MurJ reveals a solvent-exposed cavity containing residues essential for peptidoglycan biogenesis in Escherichia coli. J Bacteriol 2013, 195 (20), 4639-4649.

(102) Kuk, A. C.; Mashalidis, E. H.; Lee, S. Y. Crystal structure of the MOP flippase MurJ in an inward-facing conformation. Nature structural & molecular biology 2017, 24 (2), 171- 176.

(103) Zheng, S.; Sham, L. T.; Rubino, F. A.; Brock, K. P.; Robins, W. P.; Mekalanos, J. J.; Marks, D. S.; Bernhardt, T. G.; Kruse, A. C. Structure and mutagenic analysis of the lipid II flippase MurJ from Escherichia coli. Proc Natl Acad Sci U S A 2018, 115 (26), 6709- 6714.

(104) Kumar, S.; Rubino, F. A.; Mendoza, A. G.; Ruiz, N. The bacterial lipid II flippase MurJ functions by an alternating-access mechanism. The Journal of biological chemistry 2019, 294 (3), 981-990.

(105) Kuk, A. C. Y.; Hao, A.; Guan, Z.; Lee, S.-Y. Visualizing Conformational Transitions of the Lipid II Flippase MurJ. bioRxiv 2019, 552745.

(106) Rubino, F. A.; Kumar, S.; Ruiz, N.; Walker, S.; Kahne, D. E. Membrane Potential Is Required for MurJ Function. Journal of the American Chemical Society 2018, 140 (13), 4481-4484.

(107) Mohammadi, T.; van Dam, V.; Sijbrandi, R.; Vernet, T.; Zapun, A.; Bouhss, A.; Diepeveen-de Bruin, M.; Nguyen-Disteche, M.; de Kruijff, B.; Breukink, E. Identification of FtsW as a transporter of lipid-linked cell wall precursors across the membrane. The EMBO journal 2011, 30 (8), 1425-1432.

38 (108) Meeske, A. J.; Riley, E. P.; Robins, W. P.; Uehara, T.; Mekalanos, J. J.; Kahne, D.; Walker, S.; Kruse, A. C.; Bernhardt, T. G.; Rudner, D. Z. SEDS proteins are a widespread family of bacterial cell wall polymerases. Nature 2016, 537, 634.

(109) Taguchi, A.; Welsh, M. A.; Marmont, L. S.; Lee, W.; Sjodt, M.; Kruse, A. C.; Kahne, D.; Bernhardt, T. G.; Walker, S. FtsW is a peptidoglycan polymerase that is functional only in complex with its cognate penicillin-binding protein. Nature Microbiology 2019, 4 (4), 587-594.

(110) El Ghachi, M.; Howe, N.; Huang, C.-Y.; Olieric, V.; Warshamanage, R.; Touzé, T.; Weichert, D.; Stansfeld, P. J.; Wang, M.; Kerff, F.; Caffrey, M. Crystal structure of undecaprenyl-pyrophosphate phosphatase and its role in peptidoglycan biosynthesis. Nature Communications 2018, 9 (1), 1078.

(111) Wise, E. M., Jr.; Park, J. T. Penicillin: its basic site of action as an inhibitor of a peptide cross-linking reaction in cell wall mucopeptide synthesis. Proc Natl Acad Sci U S A 1965, 54 (1), 75-81.

(112) Tipper, D. J.; Strominger, J. L. Mechanism of action of penicillins: a proposal based on their structural similarity to acyl-D-alanyl-D-alanine. Proc Natl Acad Sci U S A 1965, 54 (4), 1133-1141.

(113) Suginaka, H.; Blumberg, P. M.; Strominger, J. L. Multiple penicillin-binding components in Bacillus subtilis, Bacillus cereus, Staphylococcus aureus, and Escherichia coli. The Journal of biological chemistry 1972, 247 (17), 5279-5288.

(114) Georgopapadakou, N. H.; Liu, F. Y. Binding of beta-lactam antibiotics to penicillin- binding proteins of Staphylococcus aureus and Streptococcus faecalis: relation to antibacterial activity. Antimicrobial agents and chemotherapy 1980, 18 (5), 834.

(115) Georgopapadakou, N. H.; Liu, F. Y. Penicillin-binding proteins in bacteria. Antimicrobial agents and chemotherapy 1980, 18 (1), 148-157.

(116) Rohs, P. D. A.; Buss, J.; Sim, S. I.; Squyres, G. R.; Srisuknimit, V.; Smith, M.; Cho, H.; Sjodt, M.; Kruse, A. C.; Garner, E. C.; Walker, S.; Kahne, D. E.; Bernhardt, T. G. A central role for PBP2 in the activation of peptidoglycan polymerization by the bacterial cell elongation machinery. PLOS Genetics 2018, 14 (10), e1007726.

39 (117) Asli, A.; Brouillette, E.; Krause, K. M.; Nichols, W. W.; Malouin, F. Distinctive Binding of Avibactam to Penicillin-Binding Proteins of Gram-Negative and Gram-Positive Bacteria. Antimicrobial Agents and Chemotherapy 2016, 60 (2), 752.

(118) Wada, A.; Watanabe, H. Penicillin-binding protein 1 of Staphylococcus aureus is essential for growth. J Bacteriol 1998, 180 (10), 2759-2765.

(119) Okonog, K.; Noji, Y.; Nakao, M.; Imada, A. The possible physiological roles of penicillin-binding proteins of methicillin-susceptible and methicillin- resistantStaphylococcus aureus. Journal of Infection and Chemotherapy 1995, 1 (1), 50- 58.

(120) Pereira, S. F. F.; Henriques, A. O.; Pinho, M. G.; de Lencastre, H.; Tomasz, A. Role of PBP1 in cell division of Staphylococcus aureus. Journal of bacteriology 2007, 189 (9), 3525-3531.

(121) Pereira, S. F.; Henriques, A. O.; Pinho, M. G.; de Lencastre, H.; Tomasz, A. Evidence for a dual role of PBP1 in the cell division and cell separation of Staphylococcus aureus. Molecular microbiology 2009, 72 (4), 895-904.

(122) Lim, D.; Strynadka, N. C. Structural basis for the beta lactam resistance of PBP2a from methicillin-resistant Staphylococcus aureus. Nature structural biology 2002, 9 (11), 870- 876.

(123) Pinho, M. G.; Filipe, S. R.; de Lencastre, H.; Tomasz, A. Complementation of the essential peptidoglycan transpeptidase function of penicillin-binding protein 2 (PBP2) by the drug resistance protein PBP2A in Staphylococcus aureus. J Bacteriol 2001, 183 (22), 6525-6531.

(124) Pinho, M. G.; de Lencastre, H.; Tomasz, A. An acquired and a native penicillin-binding protein cooperate in building the cell wall of drug-resistant staphylococci. Proc Natl Acad Sci U S A 2001, 98 (19), 10886-10891.

(125) Reed, P.; Veiga, H.; Jorge, A. M.; Terrak, M.; Pinho, M. G. Monofunctional transglycosylases are not essential for Staphylococcus aureus cell wall synthesis. Journal of bacteriology 2011, 193 (10), 2549-2556.

(126) Pinho, M. G.; de Lencastre, H.; Tomasz, A. Cloning, Characterization, and Inactivation of the Gene pbpC, Encoding Penicillin-Binding Protein 3 of Staphylococcus aureus. Journal of Bacteriology 2000, 182 (4), 1074.

40 (127) Navratna, V.; Nadig, S.; Sood, V.; Prasad, K.; Arakere, G.; Gopal, B. Molecular Basis for the Role of Staphylococcus aureus Penicillin Binding Protein 4 in Antimicrobial Resistance. Journal of Bacteriology 2010, 192 (1), 134.

(128) Hayes, M. V.; Curtis, N. A. C.; Wyke, A. W.; Ward, J. B. A mutant of Staphylococcus aureus H lacking penicillin-binding protein 4 and transpeptidase activity in vitro. FEMS Microbiology Letters 1980, 9 (4), 263-266.

(129) Memmi, G.; Filipe, S. R.; Pinho, M. G.; Fu, Z.; Cheung, A. Staphylococcus aureus PBP4 is essential for beta-lactam resistance in community-acquired methicillin-resistant strains. Antimicrobial agents and chemotherapy 2008, 52 (11), 3955-3966.

(130) Finan, J. E.; Archer, G. L.; Pucci, M. J.; Climo, M. W. Role of penicillin-binding protein 4 in expression of vancomycin resistance among clinical isolates of oxacillin-resistant Staphylococcus aureus. Antimicrobial agents and chemotherapy 2001, 45 (11), 3070- 3075.

(131) Sieradzki, K.; Pinho, M. G.; Tomasz, A. Inactivated pbp4 in highly glycopeptide-resistant laboratory mutants of Staphylococcus aureus. The Journal of biological chemistry 1999, 274 (27), 18942-18946.

(132) Liu, J.; Chen, D.; Peters, B. M.; Li, L.; Li, B.; Xu, Z.; Shirliff, M. E. Staphylococcal chromosomal cassettes mec (SCCmec): A mobile genetic element in methicillin-resistant Staphylococcus aureus. Microbial Pathogenesis 2016, 101, 56-67.

(133) Rosato, A. E.; Kreiswirth, B. N.; Craig, W. A.; Eisner, W.; Climo, M. W.; Archer, G. L. mecA-blaZ corepressors in clinical Staphylococcus aureus isolates. Antimicrobial agents and chemotherapy 2003, 47 (4), 1460-1463.

(134) Berzigotti, S.; Benlafya, K.; Sépulchre, J.; Amoroso, A.; Joris, B. Bacillus licheniformis BlaR1 L3 Loop Is a Zinc Metalloprotease Activated by Self-Proteolysis. PLOS ONE 2012, 7 (5), e36400.

(135) Llarrull, L. I.; Mobashery, S. Dissection of events in the resistance to β-lactam antibiotics mediated by the protein BlaR1 from Staphylococcus aureus. Biochemistry 2012, 51 (23), 4642-4649.

(136) Amoroso, A.; Boudet, J.; Berzigotti, S.; Duval, V.; Teller, N.; Mengin-Lecreulx, D.; Luxen, A.; Simorre, J.-P.; Joris, B. A Peptidoglycan Fragment Triggers β-lactam Resistance in Bacillus licheniformis. PLoS pathogens 2012, 8 (3), e1002571.

41 (137) Peacock, S. J.; Paterson, G. K. Mechanisms of Methicillin Resistance in Staphylococcus aureus. Annual review of biochemistry 2015, 84, 577-601.

(138) Ericson Sollid, J. U.; Hanssen, A.-M. SCCmec in staphylococci: genes on the move. FEMS Immunology & Medical Microbiology 2006, 46 (1), 8-20.

(139) International Working Group on the Classification of Staphylococcal Cassette Chromosome Elements Classification of Staphylococcal Cassette Chromosome mec (SCCmec): Guidelines for Reporting Novel SCCmec Elements. Antimicrobial agents and chemotherapy 2009, 53 (12), 4961.

(140) Saber, H.; Jasni, A. S.; Jamaluddin, T. Z. M. T.; Ibrahim, R. A Review of Staphylococcal Cassette Chromosome mec (SCCmec) Types in Coagulase-Negative Staphylococci (CoNS) Species. The Malaysian journal of medical sciences : MJMS 2017, 24 (5), 7-18.

(141) Garcia-Alvarez, L.; Holden, M. T.; Lindsay, H.; Webb, C. R.; Brown, D. F.; Curran, M. D.; Walpole, E.; Brooks, K.; Pickard, D. J.; Teale, C.; Parkhill, J.; Bentley, S. D.; Edwards, G. F.; Girvan, E. K.; Kearns, A. M.; Pichon, B.; Hill, R. L.; Larsen, A. R.; Skov, R. L.; Peacock, S. J.; Maskell, D. J.; Holmes, M. A. Meticillin-resistant Staphylococcus aureus with a novel mecA homologue in human and bovine populations in the UK and Denmark: a descriptive study. The Lancet. Infectious diseases 2011, 11 (8), 595-603.

(142) Rolo, J.; Worning, P.; Nielsen, J. B.; Bowden, R.; Bouchami, O.; Damborg, P.; Guardabassi, L.; Perreten, V.; Tomasz, A.; Westh, H.; de Lencastre, H.; Miragaia, M. Evolutionary Origin of the Staphylococcal Cassette Chromosome mec (SCCmec). Antimicrobial agents and chemotherapy 2017, 61 (6), e02302-02316.

(143) Lu, W. P.; Sun, Y.; Bauer, M. D.; Paule, S.; Koenigs, P. M.; Kraft, W. G. Penicillin- binding protein 2a from methicillin-resistant Staphylococcus aureus: kinetic characterization of its interactions with beta-lactams using electrospray mass spectrometry. Biochemistry 1999, 38 (20), 6537-6546.

(144) Fuda, C.; Suvorov, M.; Vakulenko, S. B.; Mobashery, S. The basis for resistance to beta- lactam antibiotics by penicillin-binding protein 2a of methicillin-resistant Staphylococcus aureus. The Journal of biological chemistry 2004, 279 (39), 40802-40806.

(145) Otero, L. H.; Rojas-Altuve, A.; Llarrull, L. I.; Carrasco-López, C.; Kumarasiri, M.; Lastochkin, E.; Fishovitz, J.; Dawley, M.; Hesek, D.; Lee, M.; Johnson, J. W.; Fisher, J. F.; Chang, M.; Mobashery, S.; Hermoso, J. A. How allosteric control of Staphylococcus

42 aureus penicillin binding protein 2a enables methicillin resistance and physiological function. Proceedings of the National Academy of Sciences 2013, 110 (42), 16808.

(146) Fuda, C.; Hesek, D.; Lee, M.; Morio, K.; Nowak, T.; Mobashery, S. Activation for catalysis of penicillin-binding protein 2a from methicillin-resistant Staphylococcus aureus by bacterial cell wall. J Am Chem Soc 2005, 127 (7), 2056-2057.

(147) Berger-Bachi, B.; Barberis-Maino, L.; Strassle, A.; Kayser, F. H. FemA, a host-mediated factor essential for methicillin resistance in Staphylococcus aureus: molecular cloning and characterization. Molecular & general genetics : MGG 1989, 219 (1-2), 263-269.

(148) Berger-Bächi, B.; Rohrer, S. Factors influencing methicillin resistance in staphylococci. Archives of Microbiology 2002, 178 (3), 165-171.

CHAPTER TWO: Preparation of S. aureus native Lipid II and analogues

Work present in this chapter was adapted from:

Qiao, Y.; Srisuknimit, V.; Rubino, F.; Schaefer, K.; Ruiz, N.; Walker, S.; Kahne, D. Lipid II overproduction allows direct assay of transpeptidase inhibition by β-lactams. Nature Chemical Biology. 2017, 13 (7), 793-798. and

Srisuknimit, V.; Qiao, Y.; Schaefer, K.; Kahne, D.; Walker, S. Peptidoglycan Cross-Linking Preferences of Staphylococcus aureus Penicillin-Binding Proteins Have Implications for Treating MRSA Infections. Journal of American Chemical Society, 2017, 139 (29), 9791-9794.

Contributions: VS developed the two-step extraction after noticing the interface layer. VS developed the PBP4 quantification method, LC/MS quantification method, and vancomycin bead. YQ carried out preliminary work on Lipid II build-up, the growth curve, and Lipid II titration. FR helped with the Lipid II accumulation strategy in E. coli and purified MraY.

44 2.1 Introduction

Inaccessibility to Lipid II, the monomer of peptidoglycan, has been a major roadblock that stymied the research on peptidoglycan biosynthetic enzymes. Although it was over five decades ago that PBPs were determined to be the lethal target of penicillin, the glycosyltransferase and transpeptidase activities of these important enzymes had not been studied in detail until the early 2000s when Lipid II was synthesized for the first time.

2.1.1 Current approaches to Lipid II preparation

Multiple approaches to prepare Lipid II have been reported in the past twenty years

(Table 2.1). Our lab reported one of the first syntheses of Lipid II. Our chemoenzymatic route started with three building blocks: MurNAc, the pentapeptide, and the lipid tail. The pentapeptide was coupled to MurNAc with a protected phosphate.1 With 1,1’- carbonyldiimidazole (CDI) coupling method, the MurNAc-pentapeptide-phosphate was conjugated to lipid phosphate forming the pyrophosphate bond and yielding Lipid I. MurG was then used to catalyze the transfer of GlcNAc from UDP-GlcNAc to convert Lipid I into Lipid II.2

Lipid phosphate of various length were converted to different analogues of Lipid I. MurG broad substrate scope then allows preparation of multiple Lipid II analogues. Our lab reported that heptaprenyl (C35)-Lipid II is a better substrate for E. coli PGTs than the native undecaprenyl

(C55)-Lipid II in a membrane assay, likely due to the better solubility of the shorter chain.2

Around the same time, the Schawartz group at Dupont and the Van Nieuwenhze group at Eli

Lily independently reported two total synthetic routes to Lipid II. Both routes similarly couple

GlcNAc-MurNAc-(L-Ala)-phosphate to the rest of the stem peptide and install lipid tail on late- stage via CDI coupling. These syntheses are long and laborious with low yield. Although the synthesized Lipid II are useful for studying the activity of PGTs, they are not suitable for

45 studying TPs as due to the lack of the peptide branch. With a revised synthetic route that allows the installation of different pentapeptide to MurNAc-pyrophospho-lipid, our lab reported the first chemical synthesis of canonical Lipid II from Gram-negative bacteria.3 This Lipid II has meso-diaminopimelic acid (m-DAP) at the third amino acid position of the stem peptide. Lebar et al. showed that E. coli PBPs could polymerize both Lys-Lipid II and m-DAP Lipid II but crosslink only the m-DAP substrate.3 This was the first reconstitution of PBP transpeptidases.

Another widely used method to prepare Lipid II employs right-side-out membrane vesicles (often from Micrococcus flavus) enriched with peptidoglycan biosynthetic enzymes such as MraY and MurG. The vesicles are treated with Park nucleotide extracted in advance from cell treated with antibiotic such as beta-lactam or vancomycin. Park nucleotide is then conjugated by

MraY to the lipid carrier that exists on the vesicle and coupled by MurG to supplemented

GlcNAc to yield Lipid II. This method was first employed by Park to synthesize peptidoglycan from Park nucleotide and the membrane particulate fraction from S. aureus cells.4 Strominger used the same technique to discover Lipid I and Lipid II. One shortcoming of this technique however is the minuscule pool of undecaprenylphosphate, the limiting reagent, that is naturally present in cells and therefore the vesicles. Breukink and coworkers overcame this problem by supplying exogenous undecaprenylphosphate, improving the yield significantly and allowing production of milligram scale of Lipid II.

S. aureus Lipid II with a conical pentapeptide branch was first prepared by Wiedemann and coworkers in 2004. They first prepared Lys-Lipid II with the membrane particulate method.

The Lipid II was then subjected to a complex reaction mixture consisting staphylococcal tRNA pool, purified FemXAB proteins, and purified glycyl-tRNA synthetase.5 Later on Schneider and coworkers further applied purified GatD/MurT complex to amidate this Gly5-Lipid II to yield for

46 the first time the native S. aureus Lipid II with the amidated iso-Glu and the pentaglycine branch.6 Only small amount of Lipid II was produced and no biochemical assay was carried out with the substrate. Dowson group similarly used purified aminoacyl ligase and acyl-tRNA to prepare Streptococcus pneumoniae branched Lipid II.7

In 2014, Wong and coworkers reported a totally enzymatic route to Lipid II. A kinase was employed to generate a wide variety of lipid phosphates from lipid alcohols.8 Park nucleotide was made from multiple enzymes from the pathway. Even UDP-GlcNAc was enzymatically synthesized to reduce cost. Lastly, purified MraY and MurG covert all these reagents into Lipid II. Over all, more than ten purified enzymes were used in this route.

The current methods to obtain Lipid II mentioned above are lengthy and laborious. Very few labs in the world are capable of implementing the total synthetic routes. The membrane particulate methods do not yield the correct Lipid II with the branch peptide. The enzymatic synthesis to produce native S. aureus Lipid II is low yielding and requires multiple purified enzymes and expensive cofactors. A better method to obtain native bacterial Lipid II is needed to accelerate the research in peptidoglycan synthesis and the discovery of new antibiotics.

2.1.2 Isolation of Lipid II from cell

In theory, the simplest method to obtain Lipid II is direct extraction from bacterial cells.

Lipid II however is present in low abundant in cells. As soon as it is made, it is consumed to the

PG synthesis. In S. aureus, the Lipid II pool was estimated to be around 50,000 molecules per cell.9 The pool is even lower in Gram-negative bacteria. In E. coli, only 2,000 copies of Lipid II are present per cell.10 Previous attempt to extract Lipid II directly from cell yield limited quantity of Lipid II.11

47 Table 2.1. Notable reports of Lipid II preparation.

Lipid II variant(s) Year Research Team Crosslinking Amidation Method Reference Published Lipid tail branch of Glu

C55 Walker Chemoenzymatic 2001 Lys C35 No Ref 2 (Princeton) synthesis and others

Schawartz 2001 Lys C55 No Total synthesis Ref 12 (Dupont)

VanNieuwenhze 2002* Lys C55 Yes Total synthesis Ref 13 (Eli Lilly)

Breukink C55 and Membrane 2003 Lys No Ref 14 (Utrecht) others particulate (MP)

Wiedemann 5 2004 Lys-Gly5 C55 No MP Ref (Bonn)

Enzymatic Dowson 2008 Lys-Ser-Ala C55 No synthesis (ES) + Ref 7 (Warwick) MP

Schneider 6 2012 Lys-Gly5 C55 Yes ES + MP Ref (Bonn)

Kahne Chemoenzymatic 2013 m-DAP C35 No Ref 3 (Harvard) synthesis

Wong Lys C55 Enzymatic (Academia 2014 (+ fluorescent No Ref 8 and others synthesis Sinica) label)

* VanNieuwenhze and coworkers claimed the first total synthesis of Lipid II by citing their conference

presentations in 2000 and 2001.13

48 2.1.3 Lipid II accumulates when PG biosynthesis is inhibited in the late stage

Detection of the cellular pool of Lipid II has previously been difficult, mostly due to the low abundance of Lipid II. The lipid has to be extracted from the cell membrane. Its pyrophosphate bond must then be cleaved in order to purify and ionize the muropeptide for mass spectrometry analysis. In 2014, Yuan Qiao from the Kahne/Walker lab published a paper reporting the use of an unusual PBP, S. aureus PBP4, to label cellular Lipid II (Figure 2.1).

Unlike many other PBPs whose TP domains only function on polymeric peptidoglycan, PBP4 could interact with the PG monomer Lipid II. PBP4 serine active residue form an acyl enzyme intermediate with Lipid II, releasing the terminal D-Ala. Normally, this intermediate is then attacked by amine of a pentaglycine branch from a different PG strand resulting in crosslinking.

However, if a copious amount of unnatural D-amino acid is present in the reaction, the D-amino acid could intercept the acyl-enzyme intermediate, resulting in a transpeptidation reaction, namely swapping D-Ala for a different D-amino acid. Using biotin-D-lysine (BDL), Yuan could biotinylated Lipid II, allowing visualization of Lipid II via Western Blot with horse radish peroxidase conjugated-streptavidin (HRP-streptavidin).

Figure 2.1. S. aureus PBP4 enables Lipid II labeling. (a) chemical structure of biotin-D-lysine (BDL). (b) PBP4 exchanges the terminal D-Ala of Lipid II with BDL. Biotinylated-Lipid II can be detected HRP-streptavidin.

49 The new biotinylation method reveals that Lipid II in S. aureus is greatly accumulated or depleted when treated with certain antibiotics. Moenomycin inhibits PGTs preventing Lipid II from consumption by the polymerases. Vancomycin binds to the D-Ala-D-Ala moiety of Lipid II rendering it unusable by the PGTs or TPs. Both of these antibiotics cause S. aureus to accumulate Lipid II. Fosfomycin inhibits MurA ligase, an enzyme early in the Lipid II biosynthesis. Bacitracin binds to the undecaprenyl pyrophosphate lipid carrier removing it from the Lipid II synthesis cycle. Both of these drugs cause depletion of Lipid II in S. aureus.

Meanwhile, kanamycin, a protein synthesis inhibitor, does not cause observable change in the cellular pool of Lipid II.

The Lipid II biotinylation method was applied to characterize the mechanism of action of lysobactin, a macrocyclic peptide natural product. Wonsik Lee from the Walker lab showed that

S. aureus accumulate Lipid II when treated with lysobactin. With a revised SDS-PAGE protocol, he was able to resolve the thick band of accumulated Lipid II into several bands. I determined that the high bands are the result of PBP4 crosslinking one Lipid II to one or more molecules of

Lipid II. When the labeling reaction product is treated with lysostaphin, an endopeptidase that cleaves the pentaglycine crosslink bridges, the higher bands disappeared. This result in addition with vitro data suggest that lysobactin inhibits PG synthesis by binding to Lipid II.

2.2 A simple method to prepare bacterial native Lipid II

The amount of Lipid II that appear to accumulate as revealed by the biotinylation method came as a surprise to us. We did not expect such a large change. We wondered whether we could extract useful amount of Lipid II directly from S. aureus culture treated with moenomycin. We were also curious about the possibility of obtaining Lipid II with a different structure, possibly from other bacteria. We focus on E. coli and B. subtilis, two of the most widely used model

50 organisms. They both have m-DAP at the third amino position on the branch peptide on Lipid II, although B. subtilis m-DAP is amidated. The glutamate at the second position is also not amidated in both Lipid II structures (Figure 2.2).

Figure 2.2. The stem peptide of Lipid II varies among bacteria. S. aureus has iso-D-glutamine at the second position and pentaglycine branch on the lysine for crosslinking. B. subtilis and E. coli crosslinks via m-DAP but B. subtilis has amidated m-DAP. In this section I will describe a simple method to prepare bacterial native Lipid II from S. aureus, E. coli, B. subtilis. The method is also suitable to preparing S. aureus Lipid II with different glycine branches.

2.2.1 Each bacteria requires a different accumulation method

To build up Lipid II in S. aureus, we decided to use moenomycin as a tool as previously demonstrated. We first looked into how the a small culture (2 mL) responses to moenomycin treatment by plotting a growth curve. We found that S. aureus stopped growing immediately after being treated with moenomycin (0.6 µg/mL; 2 x minimum inhibitory concentration (MIC)) consistent with the known bacteriostatic effect of moenomycin.15 We took aliquot at various time points (15, 60, and 120 minutes) from the culture treated with moenomycin and from the culture that was not treated. When we isolated and visualized Lipid II from these cultures with the previously mentioned biotinylation method, we found that Lipid II accumulates significantly, as early as 15 minutes after treatment (Figure 2.3a). The level of accumulation seems to be stable for at least one hour and appears to drop slightly after two hours. Serial dilution reveals that

51 treatment of moenomycin for 15 minutes increase the amount of Lipid II by roughly 10-fold compared to no treatment (Figure 2.3d).

Figure 2.3. Chemical probes accumulate Lipid II in bacteria. (a, left) Growth curves of S. aureus treated in the presence and absence of moenomycin. (a, right) Western blot analysis of biotinylated-Lipid II taken at the indicated time point. (b) Growth curves of B. subtilis treated with vancomycin and without. (c) Growth curves of E. coli MurJA29C treated with MTSES and without. (d-f) A serial dilution estimation of Lipid II in S. aureus, B. subtilis, and E. coli with chemical probe treatment compared to without treatment.

52 We next explored strategies to accumulate Lipid II in B. subtilis. The baseline level of

Lipid II in B. subtilis was incredibly low. We had to scale up the culture to 10 mL to be able to detect Lipid II with our biotinylation method. Since B. subtilis is naturally resistant to meonomycin due to the activity of non-canonical glycosyltransferase RodA16-17, we decided to use vancomycin to build up Lipid II instead. Vancomycin binds to the D-Ala-D-Ala moiety of

Lipid II preventing from being consumed, however, it could dissociate readily from Lipid II in the right pH and solvent facilitating down-stream work-up. Vancomycin is bacteriocidal in B. subtilis, but cell density remains constant for up to 20 minutes after the antibiotic treatment

(Figure 2.3b). Because we extracted Lipid II from cell pellets, we did not want the cells to lyse before we extract Lipid II. Therefore, we extracted Lipid II at 20 minutes after vancomycin treatment (8 µg/mL; 8 x MIC). We detected a 30-fold increase in B. subtilis Lipid II (Figure

2.3e).

We also explore strategies to accumulate E. coli Lipid II. One main difficulty is the

Gram-negative outer membrane of E. coli which prevents moenomycin and vancomycin from reaching their cellular targets.18 One method to circumvent the outer membrane is to use E. coli mutant strain imp4213 which has an outer membrane defect allowing large sugars and antibiotics to penetrate the OM.19 Treatment of vancomycin to E. coli imp 4213 cause Lipid II accumulation. However, moenomycin treatment does not. We found that this was due to activity of E. coli carboxypeptidases which cleaves the terminal D-Ala disabling Lipid II from our biotinylation. Instead, we accumulated Lipid II by blocking the activity of MurJ, the Lipid II flippase. This blockage would build up Lipid II in the cytoplasm out of reach of carboxypeptidases. To do this, we utilized the MurJ variant (MurJA29C) which has all cysteine mutated to alanine and one alanine facing the outside mutated to cysteine. This MurJ variant in

53 fully functional but is inactivated upon exposure to the thiol-reactive agent MTSES (2- sulfonatoethyl methanethiosulfonate), resulting in the build-up of Lipid II (Figure 2.4). MTSES

(1 mM; 8 x MIC) causes cell lysis after 20 minutes (Figure 2.3c); therefore, we also extracted at that time before the cell density drops. We found 16-fold increase in Lipid II level with MTSES treatment (Figure 2.3f).

Figure 2.4. A strategy to accumulate Lipid II in E. coli. A mutant variant of Lipid II flippase MurJ containing no cysteine except at A29C position replaces the wild type MurJ. This MurJA29C could be covalent blocked with a thiol- reactive agent 2-sulfonatoethyl methanethiosulfonate (MTSES), resulting in Lipid II building up inside the cell.

2.2.2 Two-step extraction method produces clean Lipid II

To isolate practical quantity of accumulated Lipid II, we treated S. aureus culture with moenomycin. After 20 minutes, we centrifuged the culture and collected cell pellets. The lipids were then extracted using a chloroform and methanol (a modified Bligh-Dyer method).

Typically, the phospholipids are in the organic layer (chloroform) while the water-soluble compounds reside in the aqueous layer. However, we consistently observed a thick, white interface layer between the aqueous and the organic layer (Figure 2.5a). We wondered whether

Lipid II which has both hydrophobic and hydrophilic moieties and might not partition cleanly could reside in the interface layer. Therefore, we collected both the interface layer and the organic layer separately, removed solvent, and redissolved the material in DMSO for analysis.

With the PBP4 biotinylation method, we observed that the interface layer contained much more

Lipid II than the organic layer (Figure 2.5b). At the same time, thin-layer chromatography (TLC)

54 revealed that the cellular phospholipids such as phosphatidylglycerol resides mainly in the organic layer and is present only in a very small amount in the interface layer (Figure 2.5c). As a result, we focused on collecting Lipid II from the interface layer.

Figure 2.5. Large quantities of native S. aureus Lipid II can be isolated with good purity by extraction. (a) Schematic showing the two-step extraction protocol. The first extraction (CHCl3/MeOH) produced three phases with a thick, white interface enriched in Lipid II. The second extraction (pyridinium acetate/butanol) separated Lipid II from water-soluble peptidoglycan precursors. (scale bar = 1 cm) (b) Western blot of biotinylated Lipid II in the interface and organic layers after the first extraction showed large amounts of Lipid II in the interface layer. (c) TLC of the interface and organic layers showed that cellular phospholipids had partitioned into the organic layer. CL: cardiolipin; PG: phosphatidylglycerol; LPG: lysyl-phosphatidylglycerol. (d-e) Extracted ion chromatograms (EICs) of the interface layer after the first extraction (d) and the organic layer after the second extraction (e). The two-step extraction effectively separated the Park nucleotide from Lipid II. The samples were treated to remove the lipid on Lipid II prior to LC/MS analysis. (f) Structures of the Park nucleotide and delipidated Lipid II species. We next tried to characterize the structure Lipid II in the interface layer. We could not purify or detect full Lipid II due to the hydrophobic undecaprenyl tail causing poor solubility, purification problems, and low ionizability in mass spectroscopy analysis. Therefore, we boiled

Lipid II in a weakly acidic buffer to cleave the pyrophosphate linkage in Lipid II to remove the lipid tail. Liquid chromatography-mass spectroscopy analysis of the boiled interface layer revealed a small peak corresponding to Lipid II monophosphate (Figure 2.5d), the product of hydrolysis, with mass-to-charge ratio of 1329.5 (Figure 2.5f). However, we also observed a much larger peak with mass-to-charge ratio of 1148.3 which corresponds to Park nucleotide

(Figure 2.5f)., the precursor to Lipid II. Since the lipid carrier is the limiting reagent in PG

55 biosynthesis, we were not surprised that a method that results in Lipid II build-up would also accumulates Park nucleotide. The water-soluble precursor resides mainly in the aqueous layer during this extraction. While collecting the somewhat-solid interface layer manually with a

Pasteur pipette, we unavoidably had to collect a little of the aqueous layer and therefore Park nucleotide. We envisioned Park nucleotide might cause problems in our Lipid II usages since it contains many structural features of Lipid II, so we came up with a plan to remove Park nucleotide.

We subjected the dried interface layer to a second extraction using a mixture of pyridinium acetate, butanol and water that Strominger previously used to isolate Lipid II.20 Lipid

II partitions fully to the organic layer. On the other hand, Park nucleotide stays in the aqueous layer. When we hydrolyzed Lipid II collected from the organic layer of the second extraction, we found only Lipid II monophosphate (Figure 2.5e). MS/MS analysis confirmed the presence of the pentaglycine branch and isoglutamine on the stem peptide (Figure 2.6). We did not observe any other related Lipid II species (e.g. Lipid II tetrapeptide, Gly3-Lipid II or unamidated Lipid

II). TLC analysis of the isolated Lipid II reveals one major spot which migrated with an Rf value comparable to synthetic Lipid II (Figure 2.7). We concluded that the two-step extraction procedure produces Lipid II largely free of cellular phospholipids and Park nucleotide.

Figure 2.6. Isolated S. aureus Lipid II has the correct modifications. (a) Chemical structure of S. aureus muropeptide monophosphate and its fragmentation pattern. (b) LC/MS/MS result showing the fragmented masses.

Figure 2.7. Isolated S. aureus Lipid II is relatively clean. TLC of isolated S. aureus Lipid II compared to synthetic Lys-Lipid II. The two steps extraction method works well not only for S. aureus but also for E. coli and B. subtilis Lipid II extractions despites the difference in Lipid II structures. An interface enriched with Lipid II was observed after the modified Bligh-Dyer in both organisms. For E. coli, only the canonical structure was observed, but for B. subtilis, we observed a minor amount

(10%) of Lipid II variant containing a tripeptide lacking D-Ala-D-Ala terminus. We think that this is due to vancomycin not binding to all available Lipid II and some are hydrolyzed by LD- carboxypeptidases such as LdcB.21

In addition to the canonical Gly5-Lipid II from S. aureus, we could also obtain Lipid II with a truncated glycine branch. To obtain Gly3-Lipid II, we treat moenomycin to a S. aureus

∆femB mutant lacking the FemB enzyme that is responsible for installing the fourth and fifth glycines on the branch (Figure 2.8a). To obtain Gly1-Lipid II, we similarly treat moenomycin to

∆femA mutant lacking the FemA enzyme that put on the second and third glycine (Figure 2.8a).

Both truncated Lipid II were hydrolyzed and analyzed by LC/MS. Their structures were verified via MS/MS analysis (Figure 2.8b). We found that in both cases, the extracted Lipid II is amidated at the second amino acid position like the canonical S. aureus Lipid II. This result confirms that the GatD/MurT complex could amidate Lipid II regardless on the glycine branch length as previously found.6

Figure 2.8. Isolation of S. aureus Lipid II with a truncated glycine branch from S. aureus mutants. (a) Schematic showing the two-step extraction method and a subsequent hydrolysis of the lipid tail for LC/MS analysis. Gly1-Lipid II was obtained from S. aureus ∆femA while Gly3-Lipid II was obtained from S. aureus ∆femB. (b) LC/MS analysis of delipidated Gly1- and Gly3-Lipid II monophosphates. 2.2.3 Lipid II quantification

The extracted lipid II despite its relative purity would be of limited use if we do not know the amount of Lipid II present in each extraction. We had three different approaches for determining the amount of Lipid II from the two-step extraction.

In the first method, we mixed a known but varying concentration of synthetic Lys-Lipid II with unknown but fixed concentration of the extracted Gly5-Lipid II (Figure 2.9a). The sample is then subjected to mild acid cleavage to yield the delipidated species of both Lipid II which was then analyzed by LC/MS. We made a standard curve from the ion count of muropeptide 1 and used that to approximate the amount of muropeptide 2 (Supplementary Figure 1). This method estimated that 300 µg of S. aureus Lipid II was obtained from 1 L of moenomycin-treated S. aureus. This method made many assumptions. For example, both Lipid II species were assumed to ionized equally well and hydrolyzed efficiently.

An alternative quantification method using the PBP4 biotinylation method was also used.

Known concentration of Lys-Lipid II and unknown concentration of S. aureus Lipid II were serially diluted and biotinylated via PBP4. The two biotin-labeled Lipid II species were then visualized and the signal was compared via densitometry (Supplementary Figure 1). This method approximated that about 800 µg of Lipid II could be obtained per 1 L of moenomycin-treated S.

59 aureus culture. This method assumed that both Lipid II are labeled equally well and transferred equally well during Western blot analysis. Both of these methods require synthetic Lys-Lipid II that are not widely available and its quantification is already a rough estimation.

Figure 2.9. Quantification strategies for isolated Lipid II. (a) Schematic depicting an LC/MS method for quantifying S. aureus Lipid II. Unknown concentration of isolated Lipid II was mixed with a known concentration of synthetic Lipid II. Both species are delipidated by acid hydrolysis. The monophosphate products were analyzed by LC/MS. A standard curve of synthetic Lipid II was used to quantify the amount of isolated Lipid II. (b) Schematic depicting a different quantification method. Isolated Lipid II is cleaved by E. coli PBP5 to release the terminal D-Ala which is functionalized by Edman reagent and analyzed by LC/MS. A standard curve of synthetic D-Ala was used to quantify the amount of isolated Lipid II. Our third quantification method is via Edman derivatization. We used purified E. coli

PBP5, a known carboxypeptidase to cleave the terminal D-Ala of Lipid II. D-Ala is then labeled with Edman’s reagent (phenyl isothiocyanate) to make the amino acid more easier to purify densitometry and to install the fluorescent functional group for simpler detection (Figure 2.9b).

We compared the ion count of this released D-Ala to a stand curve made from known concentration of purchased D-Ala (Supplementary Figure 2). Without extracted Lipid II or

PBP5, we did not detect any signal of labeled D-Ala implying that we only detect D-Ala from

60 Lipid II in the sample. The standard curve is linear in a wide range of concentration. The method estimates that about 400 µM of Lipid II could be obtained with our extraction method. This method should be the main method used in quantifying Lipid II since it does not rely on comparing two different species of Lipid II. To improve this method further, we should use D-

Ala isotope as an internal standard.

The yield of E. coli Lipid II was a respectable 100 µg/L but the yield of B. subtilis Lipid

II was noticably lower at 20 µg/L. Nevertheless, these amounts are sufficient for use in a large number of biochemical assays that I will cover in the next chapter.

2.3 Vancomycin bead for Park nucleotide extraction

During the development of our Lipid II extraction method, we noticed that a large amount of Park nucleotide was also detected. This extremely valuable precursor, which is a by- product in our method, could be used in biochemical studies of MraY and could also be converted into Lipid II with a known enzymatic route.8

Multiple routes to obtain Park nucleotide have previously been reported. In 1998,

Hitchcock et al. published the first total chemical synthesis.22 Chi-Huey Wong and coworkers reported a chemoenzymatic strategy using MurA and MurB enzymes to help prepare UDP-

MurNAc.23 On ther other hand, Dean Crick and coworkers developed a different chemoenzymatic route that chemically synthesizes UDP-MurNAc and uses MurCDEF enzymes to install the amino acids on the stem peptide.24 As is the case with Lipid II syntheses, these approaches are difficult to follow. Few labs are capable of handling the long and difficult syntheses required or purifying multiple necessary enzymes.

A direct isolation of Park Nucleotide from bacterial culture is a great way to easily obtain a large quantity of this PG precursor. In 2004, Stachyra et al. reported a large scale isolation

61 method.25 A large culture of Bacillus cereus was grown to stationary phase and then supplemented with the building block of Park nucleotide: uracil, D-glutamic acid, L-lysine, and m-DAP. Chloramphenicol was added to stop protein synthesis. Lastly the culture was treated with vancomycin to ensure Park nucleotide build-up. Cells were disrupted and Park nucleotide purified after four chromatographic steps, yielding about 100 mg of the nucleotide per 1 L of cell culture.

I want to develop a simpler method to purify Park nucleotide and potentially Lipid II extracted from bacterial culture. One thing both molecules have in common is the D-Ala-D-Ala moiety, which has a natural binder vancomycin with low micromolar binding constant.26 I envisioned beads functionalized with vancomycin for pulling down Park nucleotide or Lipid II.

Although vancomycin is widely utilized as an enantioselective separator in chiral chromatography27, the use of vancomycin to pull down PG precursors have not been reported.

2.3.1 Vancomycin bead synthesis

My first design of vancomycin bead is to covalently attached vancomycin to N- hydroxysuccinimide functionalized agarose resin via amide bond formation (Figure 2.10). A solution of vancomycin (2 mg/mL) in coupling buffer was added to dry agarose resin and incubated at room temperature for 1-2 hours. I developed an HPLC method for detecting vancomycin. Using the HPLC method, I measured the amount of vancomycin before and after coupling. I found that over 85% of the loaded vancomycin could be bound to the agarose bead with the loading of approximately 2 mg of vancomycin per 100 mg of the dry resin (about 1 mL volume of wet resin).

Figure 2.10. Vancomycin bead for Park nucleotide purification. (a) Schematic depicting a synthesis of vancomycin and its use to purify Park nucleotide. (b) Chemical structure of vancomycin. The labeled free amine is used to couple to agarose bead. I tested whether the vancomycin column could retain S. aureus Lipid II and separate

Lipid II from unwanted phospholipids. I loaded a solution of Lipid II in a mixture of chloroform/methanol (4:1) on to an agarose column functionalized with vancomycin

(vancomycin column) and another column that did not have vancomycin (control column). I then washed the column with chloroform and increasing concentration of methanol. Lastly, I attempted to flush all the Lipid II by eluting with a basic solution of ammonium hydroxide

(NH4OH) that could disrupt the hydrogen bond of vancomycin and Lipid II. I collected all the fractions, dried them, and visualized their Lipid II content by the PBP4-biotinylation method.

I found that the vancomycin column could indeed retain Lipid II. While the control column allows Lipid II to leak out in the flow-through and the first wash step, the vancomycin column holds on to Lipid II until the elute steps (Figure 2.11). However, when I compared to

63 amount of Lipid II coming out of the vancomycin column to the loading control (Lipid II that did not pass through the column), I found that a large amount of Lipid II was stuck on the column. I believe the main problem is that I did not have a solution that could solubilize Lipid II and also disrupt the binding of Lipid II to vancomycin. I then looked at the phospholipid content of the fractions. I found that most of the phosphatidylglycerol was removed in the chloroform wash.

Figure 2.11. Lipid II purification by vancomycin beads. Western blot analysis of Lipid II purification by vancomycin column and control column. Lipid II was loaded onto a vancomycin column or a control column. The columns were washed and eluted with different solvent system. The elutes were dried and its Lipid II level was quantified by PBP4 biotinylation method. Western blot was visualized by HRP-streptavidin conjugate. Agarose resin constantly requires water to remain hydrated and swelled. Lipid II however is extremely hydrophobic and not soluble in water. Most common solvent we used with Lipid II are methanol, a mixture of chloroform and methanol, and butanol pyridinium acetate. The last is difficult to work with due to its low vapor pressure and the salt in the solution making it difficult

64 to remove by evaporation. I found that the resin tolerate methanol fine but the addition of chloroform to the resin cause the resin to clump together and clash out of solution. In addition, methanol and chloroform are not compatible with the plastic spin column commonly used with agarose resin. They cause the column to lose integrity becoming noticeably softer. The organic solvent also leech out some chemicals from the plastic column as detected by TLC.

In order to avoid the plastic spin column, I tried to functionalize vancomycin to a different base. I conjugated vancomycin to amine-functionalized glass bead via glutaraldehyde crosslinking. The coupling efficiency was close to zero. By LC/MS I detected a mass that corresponds to glutaraldehyde polymerized to vancomycin. I attempted to attach vancomycin to carboyxylic acid functionalized magnetic bead but the coupling efficiency was also low.

Coupling to NHS-activated magnetic bead has higher efficiency although only microgram of vancomycin could be loaded on 1 milligram of the magnetic bead which is quite expensive.

The purity of Lipid II had not been a problem for us. The two-step extraction method results in Lipid II that is quite clean of phospholipids and other impurities already. So instead of developing a vancomycin bead capable of purifying Lipid II, I focused on using vancomycin agarose bead to purify Park nucleotide. The high water solubility of the nucleotide precursor allows us to avoid using organic solvents that are problematic in our system.

2.3.2 Park nucleotide enrichment

To use the vancomycin bead to purify Park nucleotide, I first tested whether the vancomycin column could retain Lipid II. I prepared a fresh column of vancomycin bead from

150 mg of NHS-activated agarose resin and obtained 2 mL of wet resin loaded with 3.4 mg of vancomycin. Counting on the 1:1 binding of vancomycin to Park nucleotide, I expected the column to be able to retain over 2 µmol of Park nucleotide. I loaded 2 mL of 100 µM Park

65 nucleotide to the column and incubated for 30 minutes at room temperature. I then washed with milliQ water and attempted to elute with methanol and acetonitrile (2 mL fraction). I collected the fraction and quantified the amount of Park nucleotide in each one by HPLC. In the control column, almost all Park nucleotide came out in the flow through and the first wash 1.

Meanwhile, the vancomycin column clearly retain Park nucleotide. In fact, it held to Park nucleotide so tightly that over 34% remain on the column (Figure 2.12). We encountered two problems. First, Park nucleotide still came out too early in the wash fraction. Second, we could not get all the Park nucleotide out.

Figure 2.12. Park nucleotide elution with organic solvent. HPLC analysis of Park nucleotide content from each elute/wash fraction of vancomycin bead column loaded with Park nucleotide. I solved the first problem by realizing that milliQ water is not appropriate. It would not keep the pH constant and maintain the charge state necessary for vancomycin to bind to Park nucleotide. By using PBS buffer (pH 7.2), I was able to limit the amount of Park nucleotide coming out in the flow-through and wash fractions.

66 The second problem is more challenging. I tried acidic buffers to disrupt the hydrogen bond of vancomycin and Park nucleotide. 0.1% formic acid (pH 2) and 0.1 M glycine (pH 3) was failed to release Park nucleotide. Next, I tried to compete for the vancomycin binding motif by using Acetyl-L-Lys-D-Ala-D-Ala (5 mM), a structural analogue of the stem peptide that is known to form a complex with vancomycin.26 This was a success. The substrate mimic clearly kicks out Park nucleotide, leaving less than 20% of Park nucleotide remained on the column

(Figure 2.13). This method however is not practical. Acetyl-L-Lys-D-Ala-D-Ala is expensive and we are left with purifying Park’s nucleotide from this peptide anyway. I need an eluent that I could simply remove by evaporation to yield Park nucleotide.

Figure 2.13. Park nucleotide elution with competitive binder. HPLC analysis of Park nucleotide content from each elute/wash fraction of vancomycin bead column loaded with Park nucleotide. Ac-Lys-D-Ala-D-Ala is a structural analogue of the Park nucleotide stem peptide.

I was inspired by how vancomycin was purified from bacterial extract. The stem amino acids such as D-Ala-D-Ala was immobilized or solid support in a column. The column was then loaded with vancomycin, washed and then eluted. Many elution conditions were studied. Most

67 failed to elute vancomycin from the stem peptide.28 By using both high pH and acetonitrile (0.1

M ammonium hydroxide in 50% acetonitrile), however, full recovery of vancomycin could be obtained. In my work, I am trying to disrupt the exact same binding but in reverse. I want to elute the stem peptide out from immobilized vancomycin. I found that this elution condition worked well. While approximately 20% of Park nucleotide came out early in the flow-through and wash, over 70% of Park nucleotide could be eluted from the column (Figure 2.14). One major benefit of this eluent is that all of its components are volatile. The elute fractions could be dried by rotatory evaporation to yield pure Park nucleotide.

Figure 2.14. Optimized Park nucleotide elution. HPLC analysis of Park nucleotide content from each elute/wash fraction of vancomycin bead column loaded with Park nucleotide. Elute #1-6 is 0.1 M NH4OH in 50% acetonitrile/water.

To obtain milligram quantity of Park nucleotide, I first synthesized a larger amount of vancomycin bead. 450 mg of NHS-activated dry agarose resin was coupled to vancomycin to yield 10 mL of wet resin loaded with 25 mg of vancomycin. This vancomycin loading level should be enough to retain over 10 mg of Park nucleotide.

68 Next, the aqueous layer of the first extraction in our two-step extraction method was then dried. This crude extract contains a large amount of the nucleotide (over 7 mg / 1 L of S. aureus culture) but also other water-soluble impurities in abundance. This aqueous layer is very thick and its solvent can be challenging to remove by rotatory evaporation. To obtain just Park’s nucleotide, we could also treat S. aureus culture with bacitracin, cheap and widely available antibiotic that prevent the recycling of lipid phosphate. Bacitrcin-treated S. aureus culture has depleted Lipid II but accumulated Park’s nucleotide. This drug-treated culture could then be boiled to easily release Park nucleotide.

The large-scale column was then loaded with approximately 4 mg of crude Park nucleotide extract. It was washed and eluted. impurities was then loaded on to the column. The wash and elute fractions were subjected HPLC analysis. I found that while the wash fraction contained some impurities, the elute fraction is over 95% pure (Figure 2.15). The substrate was purified further by HPLC but its purity is high enough to use as it was. I obtained 2.6 mg of pure

Park nucleotide, a significant amount for enzymatic studies.

Figure 2.15 Vancomycin bead purifies Park nucleotide in one simple step. HPLC analysis of the wash and elute fractions of Park nucleotide purification by vancomycin bead. The elute fraction is almost free of contaminants.

69 2.3.3 Enzymatic Lipid II synthesis

With a large amount of Park nucleotide in hand, Fred Rubino and I set out to synthesize

Lipid II via an enzymatic route. We focused on obtaining soluble Lipid II with truncated lipid tail that would be useful in crystallography studies.

Nerol was activatd with trichloroacetonitrile and phosphorylated with tetrabutylammoniumphosphate. The product is a mixture of nerol phosphate and nerol diphosphate which could hydrolyzed during reflux to yield primarily nerolphosphate.

The two enzymes needed for the synthesis are MraY to attach Park nucleotide to the lipid carrier and MurG to install the GlcNAc. Our lab previously purified E. coli MurG and in the past two decades has been using the enzyme to convert Lipid I into Lipid II in our chemoenzymatic route.1 MurG has a broad substrate scope and could handle different length of lipid tail. Our lab has not purified MraY previously. MraY is an integral membrane protein that is challenging to purify. It was not until 2004, that Bouhss et al reported the first successful purification N-terminal His-tagged MraY requiring multiple rounds of solubilization by n- dodecyl-β-D-maltoside (DDM) detergent.29 With this construct however it is still difficult to obtain an active MraY. When Chi-Huey Wong and coworkers attempted to purify this His- tagged MraY in 2014, they failed and had to fuse MraY to maltose-binding protein to improve expression level and solubility.8 The first crystal structure of MraY was also obtained with MBP

30 fusion. We decided to use His6-SUMO-Flag-MraY construct, the same fusion Andrew Kruse lab successfully used to purify RodA, another integral membrane protein.31 SUMO protease

Ulp1 was coexpressed resulting in in vivo cleavage of the SUMO tag generating an amino- terminal aspartic acid in the Flag tag that could be pulled down by the M1 anti-Flag monoclonal antibody. The immunoaffinity purified protein was then further purified by size exclusion

70 chromatography yielding active MraY. The yield however is still low. Most MraY aggregates. In addition, M1-antibody preparation is time-consuming. We decided to purify MraY without the

Ulp1 expression resulting in the intact His6-SUMO-tag that we could purify with nickel resin.

We also decided not to purify it further with size exclusion chromatography. We used this His-

SUMO-Flag-MraY to convert nerol phosphate and Park nucleotide to Lipid I and found the enzyme to be active. Although the new construct is slightly less than Flag-MraY, the yield was higher and the enzyme was much easier to purify.

To synthesize Lipid II, we carried out a coupled enzymatic reaction containing nerol phosphate (as a limiting reagent), Park nucleotide, UDP-GlcNAc, MraY and MurG. We also added alkaline phosphatase to cleave UDP the by-product of MurG reaction into UMP in order to drive the reaction to completion. The reaction went to completion. No nerol phosphate and

Lipid I was detected by LC/MS. We obtained over 0.5 mg of C10-Lys-Lipid II.

2.4 Conclusions

We developed a simple strategy to obtain native Lipid II from bacteria. Our strategy employs different methods to inhibit the consumption of Lipid II in three bacteria causing accumulation of the peptidoglycan precursor. Our two-step extraction method yield relatively pure Lipid II, mostly free of phospholipids and water soluble metabolites. Our S. aureus Lipid II contains both the pentaglycine branch and the carboxamide on the stem peptide. This work was the first practical preparation of native S. aureus Lipid II. We estimated over 500 µg of Lipid II could be obtained from 1 litre of moenomycin-treated culture. All in less than a day of work. The method could also be adapted to obtain S. aureus Lipid II with a truncated peptide branch or E. coli and B. subtilis Lipid II. Our Lipid II preparation method led to biochemical studies of important cell wall enzymes such as RodA, FtsW, MurJ, and LCPs.

71 We also developed vancomycin bead for a straight-forward purification of Park nucleotide, a by-product in our Lipid II isolation method. We demonstrated the usefulness of the bead by purifying milligrams of Park nucleotide in one simple chromatographic step and used

Park nucleotide to prepare soluble Lipid II. This purification method and our easy to purify

MraY construct will make Lipid II and its analogues accessible to more researchers, leading to more breakthroughs in the study of cell wall synthesis.

2.5 Materials and methods

Materials and general methods

Biotin-D-lysine (BDL) and synthetic Lipid II analog were prepared as previously described.32

Moenomycin A was isolated from Flavomycin stock. Vancomycin hydrochloride was purchased from Sigma-Aldrich. 2-sulfonatoethyl methanethiosulfonate (MTSES) was purchased from

Toronto Research Chemicals. Beta-lactam drugs were purchased from the indicated vendors: piperacillin sodium salt (VWR), imipenem monohydrate (Toronto Research Chemicals), methicillin sodium salt, mecillinam vetranal, cefaclor, oxacillin sodium salt and cephradine

(Sigma-Aldrich). Teflaro (ceftaroline fosamil) was provided from David Hooper (Massachusetts

General Hospital). Zevtera (ceftobiprole medocaril) was provided by Rolf Müller (Helmholtz

Centre for Infection Research). S. aureus lipids, 16:0 phosphatidylglycerol (abbreviated as PG),

14:0 cardiolipin (abbreviated as CL), and 16:0 lysyl-phosphatidylglycerol (abbreviated as LPG), were purchased from Avanti Polar Lipids. Nalgene Oak Ridge High-Speed Centrifuge Tubes used for lipid extractions were purchased from Thermo Scientific. Streptavidin-HRP antibody was purchased from Pierce (Catalog #21130). Amersham ECL Prime Western Blotting Detection

Reagent was purchased from GE Healthcare. Primers were purchased from Integrated DNA

72 Technologies. Restriction endonucleases were purchased from New England Biolabs. Vectors and expression hosts were obtained from Novagen. Non-stick conical vials and pipette tips used for enzymatic reactions were from VWR. Tryptic Soy Broth and Luria Broth were purchased from Becton Dickinson. FemB deletion S. aureus strain and femA deletion S. aureus strain, which were described previously, were obtained from the Kim lab (Baylor University). Alkaline phosphatase (from calf intestine) was purchased from Roche.

S. aureus strain was grown at 37 °C in Tryptic Soy Broth (TSB) media under aeration with shaking. B. subtilis and E. coli MurJA29C strain were grown at 37 °C in Luria Broth (LB) media under aeration with shaking. LC/MS chromatograms were obtained on an Agilent

Technologies 1100 series LC-MSD instrument using electrospray ionization (ESI). HRMS data was obtained on a Bruker Maxis Impact LC-q-TOF Mass Spectrometer using ESI. Western blots were developed using Biomax Light Film (Kodak) or imaged using an Azure C400 imaging system. ImageJ was used for densitometric analysis of western blots.

E. coli PBP5 was purified as previously described.33 E. coli MurG was also purified as previously described.34

Small-scale Lipid II accumulation in bacteria

The small-scale lipid extraction from bacteria was modified from a previously published protocol.32 For S. aureus sample, an overnight culture of S. aureus RN4220 was diluted to

OD600= 0.1, and allowed to grow to mid-exponential phase (OD600= 0.4–0.5) in TSB at 37 ˚C.

The culture was divided into two (10 mL each). One culture was treated with moenomycin (0.6

µg/mL, 2x minimum inhibitory concentration (MIC)), whereas the other was not. At the indicated time, a 1-mL aliquot was taken out from each culture to measure O.D. and then

73 transferred to a glass tube containing 3.5 mL of CHCl3: MeOH (1: 2). The mixture was vortexed for 10 min at 25 ˚C, following which any cell debris was removed with centrifugation for 10 min at 4,000 x g. The supernatant was collected and transferred to a new glass tube with 2 mL CHCl3 and 1.5 mL PBS (pH 7.4). The mixture was vortexed for 10 min, and centrifuged for 10 min at

4,000 x g to achieve phase separation. The bottom organic layer was collected and dried under nitrogen stream. The dried fractions were resuspended in 20 µL DMSO.

For B. subtilis sample, a similar protocol was performed with the following modifications: a culture of B. subtilis py79 (10 mL) in LB at OD600 = 0.4–0.5 was treated with vancomycin (8 µg/mL, 8x MIC) for 20 min. The culture was centrifuged to collect cell pellet, which was then resuspended in 1 mL PBS (pH 7.4) for lipid extraction as described above.

During phase separation, the top aqueous phase appeared cloudy, and was acidified with 20%

H3PO4 (25 µL) to pH 3 to limit Lipid II solubility in water. The bottom organic layer was collected and dissolved in DMSO (20 µL).

For E. coli Lipid II isolation, a 6-mL culture of E. coli MurJA29C strain (NR2186)35 grown in LB at OD600 = 0.3 was treated with MTSES (1 mM, 8x MIC) for 10 min. The culture was pelleted at 4,000 x g and resuspended in 1 mL of PBS (pH 7.4) and added with 2.7 mL CHCl3 and 1.3 mL MeOH. The mixture was pelleted 4000 x g for 2 min. The supernatant was collected and transferred to a new glass tube containing 2 mL of CHCl3 and 1.3 mL of PBS (pH 7.4). 3 mL of 100 mM HCl was added to adjust the aqueous phase to pH 1. The mixture was vortexed for 5 min and centrifuged at 4,000 x g for two minutes. The aqueous layer was removed. The organic layer was washed with 3 mL of H2O. The organic layer was collected, dried and resuspended in

320 µL of DMSO.

74 Western blot analysis of biotinylated Lipid II

The protocol for Lipid II biotinylation and detection was previously reported.32 Briefly, the lipid extract dissolved in DMSO (2 µL) was added to a mixture containing S. aureus PBP4 (4

μM), BDL (3 mM) in a reaction buffer (12.5 mM HEPES (pH 7.5), 2 mM MnCl2, and 250 μM

Tween-80) to reach a total volume of 10 μL with a final DMSO concentration of 20%. The reaction was incubated at room temperature for 1 h, and quenched with 10 μL of 2x SDS loading buffer. 3 μL of the final mixture was loaded onto a 4-20% gradient polyacrylamide gel and let run at 200 V for 40 min. The products were transferred onto Immuno-Blot PVDF membrane

(BioRad). BDL-Lipid II was detected by blotting with streptavidin-HRP (1:10000 dilution). In contrast, BDL-labeled Park nucleotide did not transfer onto the membrane and gave no signals after blotting.

Large-scale Lipid II accumulation and its two-step extraction from bacteria

Large-scale lipid extraction was performed on 6 L (4 x 1.5L) of S. aureus RN4220 cultures. An overnight culture of S. aureus RN4220 in TSB media (15 mL) was used to inoculate each 1.5-L culture. The cultures were grown at 37 ˚C with shaking. Moenomycin was added at a final concentration of 0.6 μg/mL to each 1.5 L culture when OD600= 0.5–0.6 to accumulate of

Lipid II. The moenomycin-treated cultures were grown for another 20 min before harvesting cell pellets. The pellets from 6 L cultures were resuspended in 60 mL PBS (pH 7.4), and divided equally into 4 x 125-mL Erlenmeyer flaks, each of which contains a mixture of 52.5 mL CHCl3:

MeOH (1:2). The mixture was stirred for 1 hour at room temperature to ensure cell lysis. The mixture (about 70 mL) from each Erlenmeyer flask was poured into two Teflon tubes, which were centrifuged at 4,000 x g for 10 min at 4 ˚C. The cell debris was visible as a pellet at the

75 bottom of the tube, while the supernatant that contains the solubilized cellular contents was collected. For each two tubes, the supernatants were combined and poured into a clean 125-mL

Erlenmeyer flask containing 30 mL CHCl3 and 22.5 mL PBS. The mixture was stirred at high speed for 1 h for thorough mixing of the layers. The homogenized mixture was quickly poured into three clean Teflon tubes and centrifuged at 4,000 x g for 10 min at 4 ˚C. It is important to quickly transfer the heterogeneous mixture into three tubes so that the composition of the mixture in each tube is roughly the same. In each Teflon tube, white materials were observed in between the top aqueous and bottom organic layer, resulting in an interface fraction. The top aqueous layer may appear hazy at first, but the haziness settled into the interface fraction upon the Teflon tube warmed up to room temperature, giving a clear aqueous layer. The interface fraction was collected by first removing the aqueous layer slowly using a Pasteur pipette. It was unavoidable to take up some volume of the aqueous layer while transferring the interface. The combined interface was dried in vacuo. We note that the interface fraction was not observed in the small-scale lipid extraction from S. aureus, since the volume of organic solvent used per cell mass was greater in small-scale extraction. Extraction of Gly1 and Gly3-Lipid II was performed with the same protocol but on ∆femA and ∆femB S. aureus strains.

For the second extraction to remove the Park nucleotide in the interface, the combined dried interface was dissolved in a 15 mL organic mixture of 6M pyridinium acetate: n-butanol

(1:2) (note: 6M pyridinium acetate was prepared by mixing 51.5 mL glacial acetic acid with 48.5 mL pyridine51,52), and washed with 15 mL of aqueous solvent (n-butanol saturated water) in a separatory funnel. The aqueous layer was extracted again with 10 mL organic solvent (1: 2/ 6M pyridinium acetate: n-butanol) to maximize Lipid II extraction. The organic layers were combined and washed with aqueous solvent (n-butanol saturated water) for three times (10 mL 

76 3) to remove the water-soluble Park nucleotides. The clean organic layer was concentrated in vacuo, and re-dissolved in MeOH.

In large-scale Lipid II extraction, a 1.5-L culture of B. subtilis py79 grown in LB at 37 ˚C was treated with vancomycin (8 µg/mL, 8x MIC) at OD600 = 0.5–0.6 for 20 min, and a 1.5-L culture of E. coli MurJA29C (NR2186) in LB at 37 ˚C was treated with MTSES (1 mM, 8x MIC) at OD600 = 0.5–0.6 for 20 min. The identical protocol of the two-step extraction of Lipid II described above was performed for both cultures.

Thin-layer chromatography (TLC) analyses of phospholipids and Lipid II

For phospholipid analysis, the dried interface and organic layers from the first extraction were each dissolved in 200 µL MeOH, and 10 µL was spotted for TLC. Authentic phospholipids dissolved in MeOH were used as standards. The TLC plate was eluted in solvent CHCl3: MeOH:

CH3COOH of 60: 30: 10 (by volume), and detected by spraying with a solution of CuSO4 (100 mg/mL) in 8% phosphoric acid, followed by heating at 180 ˚C for 10 min until brown spots appeared.36

For Lipid II analysis, isolated Lipid II or synthetic Lipid II analog in MeOH was spotted on a TLC plate. The plate was eluted in solvent CH2Cl2: MeOH: 1% NH4OH of 6: 3.5: 1, stained with cerium ammonium molybdate reagent, and heated on a heating block until blue spots appeared.37

LC/MS analysis of delipidated Lipid II

For removal of the lipid tail, a Lipid II sample in DMSO (10 μL) was incubated with H2O

(80 μL) and 100 mM ammonium acetate at pH 4.2 (10 μL). The mixture was boiled at 100 ˚C on

77 a heating block for 30 min. Under this condition, Lipid II is cleaved at the phosphodiester linkage. The reaction was then lyophilized.

For LC/MS analysis, the lyophilized sample was resuspended in H2O, centrifuged at

16,000 x g for 10 min to remove precipitates. The supernatant was subjected to LC/MS analysis conducted with ESI-MS operating in negative mode. The instrument was equipped with a Waters

Symmetry Shield RP18 column (5 μM, 3.9 x 150 mm) with a matching column guard. The fragments were separated using the following method: 0.5 mL/min H2O (0.1% formic acid) for 5 min followed by a gradient of 0% acetonitrile (ACN) (0.1% formic acid)/H2O (0.1% formic acid) to 20% ACN (0.1% formic acid)/H2O (0.1% formic acid) over 40 min. The molecular ion corresponding to the hydrolyzed Lipid II was extracted. MS/MS fragmentations of the species were also obtained.

Estimations of Lipid II quantities isolated from bacteria

A sample of native bacterial Lipid II in DMSO (4 μL) was mixed with varying known concentrations of synthetic Lipid II analog in DMSO (4 μL of 200 μM, 100 μM or 50 μM). The mixture was subjected to hydrolysis of the lipid tail prior to LC/MS analysis as described above.

In the extracted ion chromatograms (EICs), the integrated area corresponding to each species was tabulated and plotted. A linear standard curve was obtained for the synthetic analog, and was used to estimate the concentration of native Lipid II (Supplementary Figure 1).

In addition, an orthogonal quantification approach was used to estimate the amount of S. aureus Lipid II. A serial dilution of S. aureus Lipid II or the synthetic analog was prepared separately, and subjected to chemoenzymatic biotinylation by PBP4. BDL-Lipid II signals on western blot were quantified using ImageJ, and a standard curve was obtained for the synthetic

78 analog. The linear region of the curve was used to estimate the amount of S. aureus Lipid II

(Supplementary Figure 1).

The third quantification method utilizes E. coli PBP5 to cleave the terminal D-alanine.

Unknown concentration of Lipid II (previously estimated to be around 100 µM) was incubated with 10 µM of E. coli PBP5 in 50 µL of 1 x TGase buffer (50 mM HEPES (pH 7.5), 10 mM

CaCl2) for 1 hour at room temp. The reaction was heat quenched, frozen, and lyophilized over night. The dried sample was dissolved with 100 µL of coupling agent (acetonitrile: pyridine: triethyamine, water 30: 15: 6: 9), dried, dissolved with another 100 µL of coupling buffer and added with 5 µL of fresh phenylisothiocyanate (Thermo Fisher). The reaction was incubated at room temp for 10 minutes, dried, dissolved in 50 µL of 70: 20 water: acetonitrile. The sample was spun down at 15000 x g for 10 minutes to remove precipitates, filtered, and analyzed by

LC/MS. The functionalized D-Ala product (m/z = 223.1) was quantified and compared to a standard curve created from known concentrations of D-Ala.

Vancomycin Bead synthesis

445 mg of Pierce NHS-activated agarose dry resin (Thermo Fischer) was added to a spin column (Thermo Fischer, part no. 89896), dissolved with 10 mL of 3 mg/mL vancomycin solution in coupling buffer (0.1 M sodium phosphate, 0.15 M NaCl, pH 7.2), and incubated for 3 hours at room temp. Flow-through was collected. The resin was washed with 10 mL of coupling buffer twice (W1, W2) and quenched with 10 mL of 1 M Tris pH 7.4. The vancomycin content of flow-through, W1 and W2 were analyzed by HPLC to determine the amount of vancomycin bound to the resin. The HPLC instrument Agilent Technologies 1260 Infinity was equipped with

Phenomenex Luna C18 (5 µm, 250 x 10 mm). Vancomycin was separated with the following

79 method: 3 mL/min, buffer A = water (0.1% formic acid), buffer B = acetonitrile (0.1% formic acid); 0-5 min, 10% buffer B; 5-15 min, 10-40% gradient buffer B, 15-21 min, 100% buffer B,

22-27 min, 10% buffer B. Absorbance at 254 nm was used to compared vancomycin content of each fraction to the 3 mg/mL solution.

Enrichment of Park Nucleotide

The aqueous layer from the first part of the two-step Lipid II extraction method was dried with rotatory evaporation and dissolved in 10 mL of coupling buffer (0.1 M sodium phosphate,

0.15 M NaCl, pH 7.2). The solution is then added to 5 mL of prewashed vancomycin bead resin

(with 25 mg of vancomycin bound). After flow-through was collected, the resin was washed with

10 mL of coupling buffer once and eluted with 10 mL of 0.1 M ammonium hydroxide in 50% acetonitrile for up to seven times. The fractions were analyzed by HPLC to determine the Park nucleotide content. The HPLC instrument Agilent Technologies 1260 Infinity was equipped with

Phenomenex Luna C18 (5 µm, 250 x 10 mm). Park nucleotide was separated with the following method: 3 mL/min, buffer A = water (0.1% formic acid), buffer B = acetonitrile (0.1% formic acid); 0-3 min, 0% buffer B; 3-20 min, 0-25% gradient buffer B, 21-26 min, 100% buffer B, 27-

32 min, 0% buffer B. Absorbance at 254 nm was used to compared the Park nucleotide content of each fraction to the synthetic standard.

Purification of MraY

The plasmid encoding His-SUMO-FLAG-Clostridium bolteae MraY was transformed into E.coli C43(DE3). A 15 mL overnight culture of this strain grown in LB supplemented with

50 µg/mL carbenicillin was used to inoculate 2 x 1.5 L of LB supplemented with 50 µg/mL

80 carbenicillin. The culture was grown at 37˚c for 3.5 hours until OD600 is around 0.5 and then cooled down to 18˚c for 30 minutes before induction with 1 mM IPTG for 18 hours.

Cells were harvested by centrifugation (5,250 x g, 20 min, 4 ˚c). Cell pellets were resuspended in 100 mL of lysis buffer (50 mM HEPES (pH 7.5), 150 mM NaCl, 20 mM MgCl2,

0.5 mM PMSF, 100 µg/mL lysozyme and DNase), lysed via 3 passes through cell disruptor

(10,000 psi, 4 ˚c). Membranes were pelleted (150,000 x g, 30 min, 4 ˚c) and solubilized in 100 mL of buffer (20 mM HEPES (pH7.5), 500 mM NaCl, 20% glycerol, 1% DDM) overnight.

Insoluble fractions were pelleted (150,000 x g, 30 min, 4 ˚c). The solubilized fraction was added with imidazole to 30 mM and applied to 2 mL of pre-washed Ni-NTA resin. After collecting the flow-through, the resin was washed with 10 mL of wash buffer (20 mM HEPES (pH 7.5), 500 mM NaCl, 20% glycerol, 0.1% DDM, 30 mM imidazole) three times and eluted with 10 mL of elution buffer (20 mM HEPES (pH 7.5), 500 mM NaCl, 20% glycerol, 0.1% DDM, 300 mM imidazole). The protein was concentrated with 30 kDa MWCO filter to ~80 µM, flash frozen in liquid nitrogen and stored at -80 ˚c.

Enzymatic nerol-Lipid II synthesis.

Nerol (Sigma Aldrich) was phosphorylated with tetrabutylammonium phosphate and trichloroacetonitrile as previously described.38 In 200 µL of reaction buffer (30 mM Tris pH 8

10 mM MgCl2, 10 mM NaCl, 0.1 mM Tween-20), His-SUMO-FLAG-CbMraY (1 µM) was added to 375 µM of neryl-phosphate, 450 µM of S. aureus Park nucleotide, 450 µM of UDP-

GlcNAc, 1 µM of E. coli MurG, and 0.04 u/µL of alkaline phosphatase. The reaction was incubated at 25 ˚c for 3 hours with 600 rpm shaking, quenched by freezing in liquid nitrogen and lyophilized. The C10-Lipid II product (m/z = 1261.5) was quantified by LC/MS.

81 2.6 References

(1) Men, H.; Park, P.; Ge, M.; Walker, S. Substrate Synthesis and Activity Assay for MurG. Journal of the American Chemical Society 1998, 120 (10), 2484-2485.

(2) Ye, X.-Y.; Lo, M.-C.; Brunner, L.; Walker, D.; Kahne, D.; Walker, S. Better Substrates for Bacterial Transglycosylases. Journal of the American Chemical Society 2001, 123 (13), 3155-3156.

(3) Lebar, M. D.; Lupoli, T. J.; Tsukamoto, H.; May, J. M.; Walker, S.; Kahne, D. Forming Cross-Linked Peptidoglycan from Synthetic Gram-Negative Lipid II. Journal of the American Chemical Society 2013, 135 (12), 4632-4635.

(4) Chatterjee, A. N.; Park, J. T. Biosynthesis of Cell Wall Mucopeptide by a Particulate Fraction from Staphylococcus aureus. Proceedings of the National Academy of Sciences 1964, 51 (1), 9-16.

(5) Schneider, T.; Senn, M. M.; Berger-Bächi, B.; Tossi, A.; Sahl, H.-G.; Wiedemann, I. In vitro assembly of a complete, pentaglycine interpeptide bridge containing cell wall precursor (lipid II-Gly5) of Staphylococcus aureus. Molecular Microbiology 2004, 53 (2), 675-685.

(6) Münch, D.; Roemer, T.; Lee, S. H.; Engeser, M.; Sahl, H. G.; Schneider, T. Identification and in vitro Analysis of the GatD/MurT Enzyme-Complex Catalyzing Lipid II Amidation in Staphylococcus aureus. PLOS Pathogens 2012, 8 (1), e1002509.

(7) Lloyd, A. J.; Gilbey, A. M.; Blewett, A. M.; De Pascale, G.; El Zoeiby, A.; Levesque, R. C.; Catherwood, A. C.; Tomasz, A.; Bugg, T. D. H.; Roper, D. I.; Dowson, C. G. Characterization of tRNA-dependent Peptide Bond Formation by MurM in the Synthesis of Streptococcus pneumoniae Peptidoglycan. Journal of Biological Chemistry 2008, 283 (10), 6402-6417.

(8) Huang, L.-Y.; Huang, S.-H.; Chang, Y.-C.; Cheng, W.-C.; Cheng, T.-J. R.; Wong, C.-H. Enzymatic Synthesis of Lipid II and Analogues. Angewandte Chemie International Edition 2014, 53 (31), 8060-8065.

(9) Somner, E. A.; Reynolds, P. E. Inhibition of peptidoglycan biosynthesis by ramoplanin. Antimicrobial agents and chemotherapy 1990, 34 (3), 413-419.

82 (10) van Heijenoort, J. Lipid intermediates in the biosynthesis of bacterial peptidoglycan. Microbiology and molecular biology reviews : MMBR 2007, 71 (4), 620-635.

(11) Guan, Z.; Breazeale, S. D.; Raetz, C. R. H. Extraction and identification by mass spectrometry of undecaprenyl diphosphate-MurNAc-pentapeptide-GlcNAc from Escherichia coli. Analytical Biochemistry 2005, 345 (2), 336-339.

(12) Schwartz, B.; Markwalder, J. A.; Wang, Y. Lipid II: Total Synthesis of the Bacterial Cell Wall Precursor and Utilization as a Substrate for Glycosyltransfer and Transpeptidation by Penicillin Binding Protein (PBP) 1b of Eschericia coli. Journal of the American Chemical Society 2001, 123 (47), 11638-11643.

(13) VanNieuwenhze, M. S.; Mauldin, S. C.; Zia-Ebrahimi, M.; Winger, B. E.; Hornback, W. J.; Saha, S. L.; Aikins, J. A.; Blaszczak, L. C. The First Total Synthesis of Lipid II: The Final Monomeric Intermediate in Bacterial Cell Wall Biosynthesis. Journal of the American Chemical Society 2002, 124 (14), 3656-3660.

(14) Breukink, E.; van Heusden, H. E.; Vollmerhaus, P. J.; Swiezewska, E.; Brunner, L.; Walker, S.; Heck, A. J. R.; de Kruijff, B. Lipid II Is an Intrinsic Component of the Pore Induced by Nisin in Bacterial Membranes. Journal of Biological Chemistry 2003, 278 (22), 19898-19903.

(15) Baizman, E. R.; Branstrom, A. A.; Longley, C. B.; Allanson, N.; Sofia, M. J.; Gange, D.; Goldman, R. C. Antibacterial activity of synthetic analogues based on the disaccharide structure of moenomycin, an inhibitor of bacterial transglycosylase. Microbiology (Reading, England) 2000, 146 Pt 12, 3129-3140.

(16) McPherson, D. C.; Popham, D. L. Peptidoglycan synthesis in the absence of class A penicillin-binding proteins in Bacillus subtilis. Journal of bacteriology 2003, 185 (4), 1423-1431.

(17) Meeske, A. J.; Riley, E. P.; Robins, W. P.; Uehara, T.; Mekalanos, J. J.; Kahne, D.; Walker, S.; Kruse, A. C.; Bernhardt, T. G.; Rudner, D. Z. SEDS proteins are a widespread family of bacterial cell wall polymerases. Nature 2016, 537 (7622), 634-638.

(18) Silhavy, T. J.; Kahne, D.; Walker, S. The bacterial cell envelope. Cold Spring Harbor perspectives in biology 2010, 2 (5), a000414-a000414.

83 (19) Sampson, B. A.; Misra, R.; Benson, S. A. Identification and characterization of a new gene of Escherichia coli K-12 involved in outer membrane permeability. Genetics 1989, 122 (3), 491.

(20) Anderson, J. S.; Matsuhashi, M.; Haskin, M. A.; Strominger, J. L. Biosythesis of the peptidoglycan of bacterial cell walls. II. Phospholipid carriers in the reaction sequence. The Journal of biological chemistry 1967, 242 (13), 3180-3190.

(21) Hoyland, C. N.; Aldridge, C.; Cleverley, R. M.; Duchene, M. C.; Minasov, G.; Onopriyenko, O.; Sidiq, K.; Stogios, P. J.; Anderson, W. F.; Daniel, R. A.; Savchenko, A.; Vollmer, W.; Lewis, R. J. Structure of the LdcB LD-carboxypeptidase reveals the molecular basis of peptidoglycan recognition. Structure (London, England : 1993) 2014, 22 (7), 949-960.

(22) Hitchcock, S. A.; Eid, C. N.; Aikins, J. A.; Zia-Ebrahimi, M.; Blaszczak, L. C. The First Total Synthesis of Bacterial Cell Wall Precursor UDP−N-Acetylmuramyl-Pentapeptide (Park Nucleotide). Journal of the American Chemical Society 1998, 120 (8), 1916-1917.

(23) Liu, H.; Sadamoto, R.; Sears, P. S.; Wong, C.-H. An Efficient Chemoenzymatic Strategy for the Synthesis of Wild-Type and Vancomycin-Resistant Bacterial Cell-Wall Precursors: UDP-N-acetylmuramyl-peptides. Journal of the American Chemical Society 2001, 123 (40), 9916-9917.

(24) Kurosu, M.; Mahapatra, S.; Narayanasamy, P.; Crick, D. C. Chemoenzymatic synthesis of Park’s nucleotide: toward the development of high-throughput screening for MraY inhibitors. Tetrahedron Letters 2007, 48 (5), 799-803.

(25) Stachyra, T.; Dini, C.; Ferrari, P.; Bouhss, A.; van Heijenoort, J.; Mengin-Lecreulx, D.; Blanot, D.; Biton, J.; Le Beller, D. Fluorescence Detection-Based Functional Assay for High-Throughput Screening for MraY. Antimicrobial Agents and Chemotherapy 2004, 48 (3), 897.

(26) Bugg, T. D.; Wright, G. D.; Dutka-Malen, S.; Arthur, M.; Courvalin, P.; Walsh, C. T. Molecular basis for vancomycin resistance in Enterococcus faecium BM4147: biosynthesis of a depsipeptide peptidoglycan precursor by vancomycin resistance proteins VanH and VanA. Biochemistry 1991, 30 (43), 10408-10415.

(27) Armstrong, D. W.; Tang, Y.; Chen, S.; Zhou, Y.; Bagwill, C.; Chen, J.-R. Macrocyclic Antibiotics as a New Class of Chiral Selectors for Liquid Chromatography. Analytical Chemistry 1994, 66 (9), 1473-1484.

84 (28) Nagarajan, R. Glycopeptide antibiotics; Marcel Dekker: New York, 1994; Vol. 63.

(29) Bouhss, A.; Crouvoisier, M.; Blanot, D.; Mengin-Lecreulx, D. Purification and Characterization of the Bacterial MraY Translocase Catalyzing the First Membrane Step of Peptidoglycan Biosynthesis. Journal of Biological Chemistry 2004, 279 (29), 29974- 29980.

(30) Chung, B. C.; Zhao, J.; Gillespie, R. A.; Kwon, D. Y.; Guan, Z.; Hong, J.; Zhou, P.; Lee, S. Y. Crystal structure of MraY, an essential membrane enzyme for bacterial cell wall synthesis. Science (New York, N.Y.) 2013, 341 (6149), 1012-1016.

(31) Meeske, A. J.; Riley, E. P.; Robins, W. P.; Uehara, T.; Mekalanos, J. J.; Kahne, D.; Walker, S.; Kruse, A. C.; Bernhardt, T. G.; Rudner, D. Z. SEDS proteins are a widespread family of bacterial cell wall polymerases. Nature 2016, 537, 634.

(32) Qiao, Y.; Lebar, M. D.; Schirner, K.; Schaefer, K.; Tsukamoto, H.; Kahne, D.; Walker, S. Detection of lipid-linked peptidoglycan precursors by exploiting an unexpected transpeptidase reaction. J Am Chem Soc 2014, 136 (42), 14678-14681.

(33) Lupoli, T. J.; Tsukamoto, H.; Doud, E. H.; Wang, T.-S. A.; Walker, S.; Kahne, D. Transpeptidase-Mediated Incorporation of d-Amino Acids into Bacterial Peptidoglycan. Journal of the American Chemical Society 2011, 133 (28), 10748-10751.

(34) van den Brink-van der Laan, E.; Boots, J.-W. P.; Spelbrink, R. E. J.; Kool, G. M.; Breukink, E.; Killian, J. A.; de Kruijff, B. Membrane Interaction of the Glycosyltransferase MurG: a Special Role for Cardiolipin. Journal of Bacteriology 2003, 185 (13), 3773.

(35) Sham, L. T.; Butler, E. K.; Lebar, M. D.; Kahne, D.; Bernhardt, T. G.; Ruiz, N. Bacterial cell wall. MurJ is the flippase of lipid-linked precursors for peptidoglycan biogenesis. Science (New York, N.Y.) 2014, 345 (6193), 220-222.

(36) Oku, Y.; Kurokawa, K.; Ichihashi, N.; Sekimizu, K. Characterization of the Staphylococcus aureus mprF gene, involved in lysinylation of phosphatidylglycerol. Microbiology (Reading, England) 2004, 150 (Pt 1), 45-51.

(37) Tsukamoto, H.; Kahne, D. N-methylimidazolium chloride-catalyzed pyrophosphate formation: application to the synthesis of Lipid I and NDP-sugar donors. Bioorg Med Chem Lett 2011, 21 (17), 5050-5053.

85 (38) Danilov, L. L.; Druzhinina, T. N.; Kalinchuk, N. A.; Maltsev, S. D.; Shibaev, V. N. Polyprenyl phosphates: synthesis and structure-activity relationship for a biosynthetic system of Salmonella anatum O-specific polysaccharide. Chemistry and Physics of Lipids 1989, 51 (3), 191-203.

CHAPTER THREE: Reconstitution of S. aureus Peptidoglycan Biosynthesis

Work present in this chapter was adapted from

Contributions: VS and YQ cloned PBP2. YQ cloned PBP2 mutants and PBP2a. VS and YQ purified the enzymes and developed the transpeptidase assay. YQ did the time course analysis of PBP2 and tested PBP2 and the mutants against beta-lactams. VS reconstituted PBP2a activity, tested it against beta-lactams and studied the substrate preference of the PBPs.

87 3.1 Introduction

Peptidoglycan transpeptidases constitute an important class of enzymes. They are responsible for peptidoglycan crosslinking which is necessary for maintaining cell wall integrity against high osmotic pressure from inside the cell. A transpeptidase contains an active serine residue that forms an acyl-enzyme intermediate with the stem peptide of peptidoglycan and release the terminal D-Ala. The acyl-enzyme intermediate is attack by a nucleophile from a different peptidoglycan chain resulting in a crosslinked product. Transpeptidases are inhibited by clinical important drugs such as penicillin, bacteria cannot crosslink the cell wall and their cells burst open by osmolysis. Despite many decades that transpeptidases have been the targets of drug development and despite over twenty beta-lactams that have been developed against this class of enzymes, very little is known about how they function, how they corporate, and their substrate preference. One reason for that lack of knowledge is due to the inaccessibility of their substrate Lipid II. In this chapter, I will describe how I used Lipid II obtained by the two-step extraction method to study S. aureus transpeptidases PBP2 and PBP2a which are essential for beta-lactam resistance.

3.1.1 Existing knowledge of the peptidoglycan transpeptidases

Around 1980, Michio Matsuhashi and coworkers purified E. coli bifunctional PBP,

PBP1b from the membrane, the first peptidoglycan transpeptidase to be reconstituted.1 The purified enzyme when mixed with membrane particulate containing MraY and MurG, UDP-

GlcNAc, and Park nucleotide labeled with D-[14C]-Ala-D-[14C]-Ala produced crosslinked peptidoglycan that was separated by paper chromatography and detected by radioautograms. The degree of crosslinking was determined by comparing the radioautograms of the product before and after lysozyme cleavage and also by quantifying the amount of released D-[14C]-Ala. This

88 D-Ala release could be inhibited by penicillin G. E. coli PBP1b was shown to have PGT and TP activities. Matsuhashi next reconstituted the activity of PBP1a in vitro by incubating PBP1a with lipid II extracted from a membrane particulate synthetic route.2 PBP1a was also shown to be bifunctional. Matsuhashi also compared the activity of three different beta-lactams against this transpeptidase by measuring the antibiotics’ effect on degree of crosslinking. PBP1a and PBP1B were shown to be inhibited differently by the beta-lactams. Over the past four decades, these two

E. coli PBPs were studied extensively and much of what we know about the transpeptidases came from these two class A PBPs.

In term of acyl-donor, the substrate scope of the transpeptidases is narrow. PBP1b would not react directly with D-Ala-D-Ala terminated peptide.3 PBP1a does not form an acyl-enzyme intermediate with Park nucleotide or Lipid II directly but only react with uncrosslinked peptidoglycan chain.4 Streptomyces coelicolor transpeptidases were found to be capable of accepting the vancomycin-resistant stem peptide with D-Ala-D-Lac moiety as a substrate.5

The scope of transpeptidase acyl-acceptor substrate however is complex. Vollmer and coworkers found that PBP1a acyl-enzyme intermediate could be attacked by the m-DAP of tri-, tetra-, or pentapeptide Park nucleotide in addition to the stem peptide of another glycan strand.6

Matt Lebar from our lab synthesized both Lys-Lipid II and m-DAP Lipid II and found that while

PBP1a can polymersize both substrate, it can crosslink only glycan strand containing m-DAP.7

The interaction between the transpeptidase domain and the glycosyltransferase domain in bifunctional PBPs are complicated. Terrak et al. showed the PGT activity of PBP1b is not dependent on the TP activity.3 Inhibition of the TP domain by penicillin does not alter the rate of glycan chain elongation. Vollmer and coworkers incubated purified PBP1a with already oligomerized glycan strands and found very low level of crosslinking compared to when Lipid II

89 is the substrate.4 They also demonstrated that PBP1a mutant that has no PGT activity was also revealed to contain no TP activity.4,8 In vitro peptidoglycan synthesis time course experiments showed that PBP1a has a lack phase for transpeptidase activity while PBP1b does not.4,6,9 These evidences suggest that certain bifunctional PBPs requires ongoing polymerization to carry out transpeptidation while some do not.

The activity of PBPs is regulated by multiple enzymes. Early genetic studies showed that

FtsN stimulates the overall activity of E. coli PBP1b perhaps by promoting PBP1b dimerization.10 Later mechanistic study revealed that PBP1b glycosyltransferase activity could be inhibited by cell division subcomplex FtsBLQ.11 This PGT inhibition, which also cascades into TP inhibition, is antagonized by FtsN. When the cell is ready to divide, FtsN release PBP1b from the inhibition FtsBLQ and peptidoglycan synthesis begins. Two outer membrane lipoproteins LpoA and LpoB are found to be activators of PBP1a and PBP1b transpeptidase activity, respectively.12 Our lab later showed that the activation of two pairs are mechanistically different. LpoA directly increases the rate of PBP1a transpeptidation while LpoB influences the rate of PBP1b PG polymerization.9 Genetic study shows that S. aureus PBP2, PBP2a and PBP4 work cooperatively to synthesize highly crosslinked peptidoglycan.13

Another class of transpeptidase that crosslink peptidoglycan is the L,D-transpeptidases.

These enzymes have cysteine instead of serine as the active residue.14 They form an acyl-enzyme intermediate with PG stem peptide that already lost the terminal D-Ala via carboxypeptidase and contain only four amino acids on the stem.14 The intermediate is then attacked by the canonical nucleophile such as m-DAP. These enzymes can be efficiently inhibited by a single class of beta- lactams, the carbapenems.15-16

90 Unfortunately, little is known about how Gram-positive PBP transpeptidases crosslink peptidoglycan. This is especially true for bacteria that have a branch peptide on their Lipid II, due to the difficulty of synthesizing those complex Lipid II molecules. So far what we know about transpeptidases that crosslink via branching peptide come from L,D-transpeptidases in E. faecium and E. faecalis. The stem peptide of E. faecium peptidoglycan has L-Lys at the third amino acid position. The amine side chain of this lysine is attached with D-iso-asparagine. E. faecalis on the other hand has L-Ala-L-Ala attached at the same lysine position. To obtain the branching substrates, Michel Arthur and coworkers isolated the cell wall of bacteria that has these branches, digested the sacculi with muramidases, reduced the anomeric sugar with sodium borohydride and purified the muropeptide by HPLC.17 These muropeptides were then used as substrate in the reconstitution of L,D-transpeptidases. While some L,D-transpeptidases tolerate substitutions at the third position of the donor, they all require a specific nucleophilic side chain for the acyl acceptor.17 Alas, this technique would not work well for studying the substrate specificity of PBP transpeptidases since the TP activity was previously shown to require ongoing

PGT activity or oligomerized glycan strand as a substrate.4,6,9

S. aureus Lipid II has pentaglycine branch on the lysine of stem peptide. As mentioned in chapter 1, this branch is synthesized by auxiliary factors FemXAB (Figure 3.1). Deletion of

FemX, FemA, or FemB in S. aureus yield Lys-Lipid II, Gly1-Lipid II or Gly3-Lipid II respectively. FemX is essential, implying that none of the PBPs could crosslink Lys-Lipid II in vivo. Since FemA and FemB activities could be removed, one or more of the S. aureus PBPs must be able to crosslink these truncated branch peptide. The substrate specificity of S. aureus

PBPs is not currently known as Lipid II with branch peptide has never been synthesized.

Figure 3.1. Schematic of peptidoglycan synthesis in methicillin-resistant S. aureus (MRSA). PBP2a can crosslink peptidoglycan when native transpeptidases (e.g. PBP2) are inhibited by beta-lactams. PBP2a requires the PGT domain of PBP2 to polymerize Lipid II into linear glycan strands. PBP2a was also proposed to only form Gly5- crosslinking and thus requiring FemXAB enzymes in order to confer beta-lactam resistance. This hypothesis has not been tested. 3.1.2 Transpeptidase activity assays

Few methods are available to study transpeptidase activity. Traditionally, researchers reconstitute E. coli transpeptidase activity by incubating a purified HMW class A PBPs like

PBP1b with synthetic or isolated Lipid II. The crosslinked and polymerized product is then digested by muramidase into disaccharide units. Some of the cleaved muropeptide products are monomeric while other are crosslinked to one or more muropeptide. If the substrate is radiolabeled, the product could be separated paper chromatography.1-2 The degree of crosslinking and thus the TP activity could be ascertained by comparing radioactivity counting of monomeric product and the crosslinked product. The products could also be separated and analyzed by liquid chromatography (LC) method. Non-radiolabeled product can be separated LC and detected by mass spectrometer (MS). Comparing the ion count of the monomeric muropeptide and the crosslinked muropeptide yield the activity of transpeptidase crosslinking.

Beta-lactams could be used to inhibit transpeptidase. The decrease in measured TP activity

92 reflect the efficiency of the beta-lactam used.2 These activity assays however only work when transpeptidase substrate is available.

When the Lipid II substrate cannot be obtained, activity of PBP transpeptidases could still be investigated but only in terms of their interaction with the beta-lactams inhibitors. In these competition binding assays, purified or crude PBPs are incubated with a beta-lactam at a gradient of concentrations. The PBPs are then labeled by a radioactive or fluorescent penicillin probe. If the first beta-lactam could covalently inhibit the transpeptidases, then PBPs would not be label by the probe. The assay is useful in determining beta-lactam affinities to PBPs. For example, a competition binding assay reveals that mecillinam binds to E. coli PBP2 and S. aureus PBP3 efficiently but not other PBPs in S. aureus. This result translates to the clinical setting. Because

PBP2 is essential in E. coli but PBP3 is not in S. aureus, mecillinam is a good antibiotic against

E. coli but has poor activity against S. aureus. Competition binding assays however measure just binding and beta-lactam acylation. It does not truly detect the inhibition of transpeptidase activity. The transpeptidase active site may have different conformation when the enzyme is actively crosslinking peptidoglycan. In addition, the competition binding assays are not well- suited for studying non-covalent TP inhibitors or beta-lactams with a fast rate of hydrolysis since the use of excess of labeled penicillin for labeling PBPs in the experiment may result in an underestimation of the binding affinity of the compounds being investigated. For example, cefaclor was previously reported to has a low affinity for S. aureus PBP2.18 Later work with different reaction conditions, however, reveals that cefaclor in fact binds to PBP2 with high affinity.19 The apparent low-affinity binding result was due to the fast hydrolysis rate of cefaclor-

PBP2 acyl-enzyme complex that hampers cefaclor from keeping the labeled probe out of PBP2

93 active site. A direct activity assay for investigating beta-lactam inhibition of PBPs is needed for the development of new antibiotics.

3.1.3 D-amino acid incorporation by transpeptidases

Using synthetic acyl-Lys-Lipid II that cannot as a transpeptidation acceptor, Tania Lupoli also from our lab demonstrated E.coli PBP1a could incorporate a wide range of D-amino acids into the terminal position of the peptidoglycan stem peptide.8 It cannot do the same for L-amino acids. This D-amino acid exchange could be seen as a reverse reaction of the acyl-enzyme intermediate formation and the release of D-Ala (Figure 3.2). The preference for the D-isomers is not surprising but the tolerance of a wide variety of side chains was unexpected. It was also interesting that E. coli PBP could incorporate D-amino acid into the fifth amino acid position in vitro. Exogeneous D-amino acids are exclusively incorporated by L,D-transpeptidases in vivo to the fourth amino acid position on the stem peptide.20 Other bacteria incorporate D-amino acids differently. In B. subtilis, HMW PBPs mediate the incorporation.20-21 Certain intracellular pathogens like Chlamydia trachomatis and Listeria monocytogenes incorporate D-amino acid into the Park nucleotide synthesis directly via the D-Ala-D-Ala ligase (Ddl) and MurF ligase.22-24

D-amino acid incorporation has become an extremely useful tool for installing fluorescent probe on to cell wall to study peptidoglycan synthesis in vivo.21-25 Yuan Qiao from our lab found that in S. aureus PBP4 is the enzyme that is responsible for incorporating exogenous D-amino acid.26

She exploited the activity of PBP4 transpeptidase to exchange D-Ala of Lipid II with biotin-D- lysine, allowing specific detection of Lipid II.26

94 H O penicillin penicilloyl-enzyme N OH H O R L-Lys N N OH H O R L-Lys N H H O H H R N H R N S Gly5 S Gly O O 5 O N HN NH H O 2 N O OH R L-Lys N OH H O Enz O O

Gly5 Crosslinked PG Enz OH NH2 transpeptidase

O O O O H H H N H H N N OH R O Enz R L-Lys OH R L-Lys N L-Lys H O Gly Gly Gly 5 5 5 Nu Hydrolyzed PG NH NH NH OH 2 2 2 PG H2N O

Acyl-enzyme intermediate Biotin NH Biotin NH

H2N COOH H O N R L-Lys N COOH H

Gly5 Biotinylated PG NH2

Figure 3.2. Schematic of transpeptidase (TP) reactions in S. aureus. The hydroxyl group of the active serine residue of TPs attack the amide bond of D-Ala-D-Al on peptidoglycan stem peptide to form an acyl-enzyme intermediate. This reactive intermediate could then be attacked by nucleophiles. Typically, it is attacked by the amine terminus of the pentaglycine branch of another glycan strand to form a new amide bond and produce crosslinked peptidoglycan. If no suitable nucleophile is present, it could also be hydrolyzed by water, resulting in hydrolyzed peptidoglycan that lost the terminal D-Ala. In the presence of exogenous D-amino acid, the acyl-enzyme intermediate can go through a semi-reverse reaction to incorporate D-amino acid. In this chapter, I will show the first reconstitution of S. aureus PBP2 and PBP2a transpeptidase activity. I will explain the new transpeptidase activity assay that we developed and how we use it to investigate beta-lactam inhibition of PBP2 and PBP2a. I will also report the substrate specificity of S. aureus PBPs that are important for beta-lactam resistance. I will show preliminary evidence of PBP2 and PBP2a cooperative activities in vitro. The tools reported should be useful for studying PBPs of other bacteria and developing new classes of TP inhibitors.

95 3.2 Purification of PBP2

Our lab previously purified full-length S. aureus PBP2 and reconstituted its PGT activity with synthetic Lys-Lipid II.27 The full-length structure however contains a hydrophobic transmembrane helix for anchoring the protein to the cytoplasmic membrane. Our His-tagged full-length protein had to be extracted by high level of detergent (sodium lauryoyl sarcosinate) from membrane pellet during purification. It was further purified by Ni2+ resin affinity column and used as it was. We later found that the purified full-length protein is prone to aggregation as observed during a size-exclusion chromatography step. The activity of the protein therefore is questionable since aggregated enzymes exhibit poor activity. The observed PGT activity of

PBP2 may come from a small amount of properly folded protein.

We chose to purify a truncated PBP2 construct that lacks the transmembrane helix. We used the PBP2[M59-S716] construct used to obtain crystal structure of PBP2 by Lovering et al.28

We cloned this construct and purified its following the reported protocol with a few changes.

PBP2 is solubilized out of cell pellets during cell disruption by 40 mM CHAPS. The His8-tagged protein is then enriched on Ni-NTA column, during which detergent was exchanged to 0.28 mM

LDAO (below critical micelle concentration). The protein is then further purified by size exclusion chromatography (SEC) and salt is removed at the same time. The FPLC analysis reveals that the majority of the PBP2 protein is properly folded (Figure 3.3). We obtained around

5-10 mg of S. aureus PBP2[M59-S716] from 1 litre of E. coli culture. We however also further purified this PBP2 construct with ion exchange chromatography to yield even purer protein that we could crystalize with conditions modified from the published report.28 The activity of the two-FPLC-purified PBP2 is not significantly improved, so we use PBP2 purified up to the SEC step for transpeptidase activity reconstitution and other assays.

Figure 3.3. Purification of S. aureus PBP2 [59-716]. (a) Size exclusion chromatography (SEC) trace of PBP2 [59-716] purification. The first peak is aggregated protein. The small shoulder is dimeric PBP2 while the tall peak is monomeric PBP2 which was collected. (b) Coomasssie gels showing PBP2 before and after SEC purification and ion- exchange purification. 3.3 Reconstitution of PBP2 PGT and TP activities

With access to both PBP2 and native Gly5-Lipid II from S. aureus, we next established the first reconstitution of the PBP2 PGT and TP activities. We incubated purified PBP2 with the

Gly5-Lipid II isolated from with our two-step extraction at room temperature for one hour. To analyze the product, we adopted the LC/MS that Yuan Qiao used to study PBP4 transpeptidase activity.26 The reaction product was digested with a muramidase, mutanolysin, to yield disaccharide muropeptide units (some of which are crosslinked through the glycine branch) The muropeptides were then reduced by NaBH4 to remove the anomeric center so that we did not have to account for the alpha and beta anomers separately later. Finally, the reduced muropeptides were separated and detected by LC/MS analysis (Figure 3.4a).

Figure 3.4. Reconstitution of peptidoglycan biosynthesis by PBP2. (a) Schematic for analysis of peptidoglycan formed by S. aureus PBP2. Crosslinked products were digested by mutanolysin and reduced by sodium borohydride before LC/MS analysis. (b) LC/MS extracted ion chromatogram of PBP2 and Lipid II produces peak A, the monomeric muropeptide, and peak B and C, the crosslinked dimeric and trimeric muropeptides; composition of in vitro peptidoglycan formed by PBP2 closely resembles that of the enzyme-digested S. aureus sacculus. The following ions were extracted from each chromatogram: A: 1253.5856 ([M+H]); B: 1029.0617 ([M+2H]2+); and C: 1194.2204 ([M+3H]3+). (c) Time-course analysis of PBP2 reaction shows that its transpeptidase activity follows glycan polymerization. The y-axis (% maximum) is calculated from the integrated EIC intensity of the monomeric muropeptide (polymerization) or the sum of dimeric and trimeric muropeptides (crosslinking), with each normalized to the highest intensity detected in the 90-minute period. Data represent averaged values of two experimental results.

We identified three major muropeptides: monomeric (peak A), crosslinked dimeric (peak

B) and doubly-crosslinked trimeric muropeptide species (Figure 3.4b and Supplementary Figure

3). The presence of peak A that the PGT domain is active as the monomeric muropeptide species could only be produced from mutanolysin digestion of oligomerized glycan strands; the muramidase does not act on Lipid II directly. The presence of crosslinked products in peak B and

98 peak C show that the transpeptidase is also active. This result demonstrates that our reconstitution was successful. PBP2 exhibited both PGT and TP activities. The composition of the peptidoglycan produced by PBP2 in vitro closely resembled that of the isolated S. aureus sacculus.

A time course of the PBP2 reaction revealed that monomeric muropeptide (resulting from

PGT activity) stopped accumulation with in 10 minutes, a time point that coincided with the appearance of substantial amount of crosslinked muropeptides (resulting from TP activity)

(Figure 3.4c). After 1 hour, both species decrease as the trimeric species increased. PBP2 transpeptidase activity does not exhibits lack phase as in E. coli PBP1a. This implies that PBP2 may not require long glycan strands before the TP domain starts to crosslink. Reaction with a

PBP2 variant lacking the catalytic serine residue (PBP2S398G) resulted in monomeric muropeptide

(peak A), identical to what was observed when monofunctional PGT SgtB was used to polymerize Lipid II (Figure 3.5).

Figure 3.5. PBP2 active serine residue in the transpeptidase domain is necessary for peptidoglycan crosslinking. LC/MS traces showing the muropeptide products from reactions with PBP2 S398G or with monofunctional PGT, SgtB. Both enzymes are not capable of crosslinking and we did not detect crosslinked muropeptide product (peak B or peak C).

99 3.4 Purification of PBP2a

Our lab has not previously purified PBP2a. We cloned PBP2a[M24-E668] construct without the N-terminal transmembrane helix as Mobashery and coworkers reported.29 PBP2a was purified following the published protocol with a few modifications.29 After protein overexpression, cells were pelleted. PBP2a was solubilized by 40 mM CHAPs and applied to Ni-

NTA resin. During this affinity chromatographic step, detergent is removed and PBP2a is eluted with in 25 mM HEPES buffer pH 7 containing 1 M sodium chloride. Unlike PBP2[MM59-S716] which has a large hydrophobic PGT domain that normally resides in the membrane,

PBP2a[M24-E668] surface is mostly solvent exposed and thus soluble without needing detergent. The protein was further purified using size-exclusion chromatography and monomeric

PBP2a was collected (Figure 3.6). We obtained approximately 6 mg per 1 liter of E. coli culture.

We also purified PBP2a with Tris buffer with detergent (20 mM Tris pH 7.5, 500 mM NaCl,

0.1% CHAPS) but later found that this enzyme from this purification is not as active as the enzyme from the first preparation.

Figure 3.6. Purification of S. aureus PBP2a [24-668]. (a) Size exclusion chromatography (SEC) trace of PBP2a [24- 668] purification. The first peak is aggregated protein. The small shoulder is dimeric PBP2a while the tall peak is monomeric PBP2a which was collected. (b) Coomasssie gel showing PBP2a after SEC purification.

100 3.5 Reconstitution of PBP2a TP activity

PBP2a activity has never been reconstituted before. So I first use the activity assay that I will describe in the next section to screen for an optimized reaction condition for PBP2a. I first looked at HEPES and Tris buffer. I found that PBP2a is more active in HEPES buffer. At the same time I looked at the pH effect. PNP2a is most active at pH 6.5. Next we look at the cation.

Many PGT require certain cation to function. While TP does not necessary need cation for catalytic activity, we found an addition of cation to be helpful. In this case, PBP2a is most active with Mg2+. 50 mM zinc chloride however abolished PBP2a activity completely. Detergents did not significantly improve the activity. Our optimized reaction buffer (10x) is 500 mM HEPES, pH 6.5, 25 mM MgCl2.

To reconstitute PBP2a crosslinking activity, I first need to polymerize Lipid II into oligomeric glycan strands; unlike PBP4, I found that PBP2a cannot act directly on Lipid. Two enzymes could be used to polymerize Lipid II without crosslinking. First, I used SgtB, the monofunctional PGT, with a mutation Y181D that cause SgtBY181D to produce glycan strands that are significantly shorter, allowing more crosslinking substrate. Alternatively I could also use

S398G PBP2 that has an inactive TP domain. I incubated extracted Gly5-Lipid with PBP2a and

PBP2S398G for three hours at room temperature (Figure 3.7). The products were analyzed like we did in the PBP2 reconstitution. Crosslinked peptidoglycan was digested by mutanolysin and reduced by NaBH4. Muropeptides were analyzed by LC/MS.

101

Figure 3.7. Reconstitution of PBP2a crosslinking and inhibition. Schematic of peptidoglycan synthesis by PBP2a and samples preparation for LC/MS analysis (top). PBP2 with inactive TP domain was used to produce uncrosslinked PG as a substrate for PBP2a. Extraction ion chromatogram (EIC) traces for wild-type PBP2a show both hydrolysis and cross-linked products, but the catalytically inactive mutant (PBP2a S403A) or ceftaroline-treated PBP2a do not. We detected three major product monomeric muropeptide (pink peak), hydrolyzed monomeric muropeptide missing the terminal D-Ala (blue peak), and crosslinked dimeric muropeptide (orange peak) (Figure 3.7). The hydrolyzed monomer and dimer peaks disappeared when PBP2a was genetically inactivated by mutating the active site serine to alanine (Figure

3.7). The presence of the crosslinked dimeric muropeptide shows that PBP2a transpeptidase is active, establishing the first reconstitution of PBP2a activity. The hydrolyzed monomer lacking

D-Ala however is also the result of PBP2a transpeptidase activity. If the acyl-enzyme intermediate is not promptly attacked by the pentaglycine branch of another glycan strand, the intermediate could instead be attacked by water resulting in the release of D-Ala and the hydrolyzed tetrapeptide product (Figure 3.2). The TP activity of PBP2a is significantly slower than PBP2.

102 I showed that PBP2a transpeptidase could be inhibited by ceftaroline (50 µM to ensure complete inhibition), a fifth generation cephalosporin antibiotic designed to combat MRSA

(Figure 3.7). The transpeptidase activity could also be impeded by K406A mutation which was reported to have slower acylation rate (Supplementary Figure 4).30 This result demonstrated for the first time that the ceftaroline directly inhibits PBP2a activity. Previous evidences that this drug targets PBP2a are from competition binding assays and from the structure of the drug- enzyme complex.31-32

3.6 Direct assay of transpeptidase inhibition by beta-lactams

We next sought to develop a simpler transpeptidase assay with higher throughput. Our

LC/MS method relies on difficult separation of the muropeptide products on a reverse phase column and a high sensitivity detector of high resolution mass spectrometer to detect and quantify the low abundant crosslinked products. Each LC/MS run takes over 50 minutes.

We were inspired by the ability of PBP4 transpeptidase to incorporate functionalized D- amino acid on to Lipid II that led us to develop a simple and sensitive method to detect the cellular pools of Lipid II in bacteria. We wondered whether we could use the transpeptidase domain of HMW PBPs such as PBP2 or PBP2a to incorporate labeled D-amino acid probe during in vitro peptidoglycan synthesis.26 Our lab previously show that HMW PBPs from E. coli and B. subtilis could incorporate D-amino acid into peptidoglycan in vitro as detected by the

LC/MS method.7,33 We wanted to use S. aureus HMW PBPs to incorporate biotin-D-lysine

(BDL) into peptidoglycan so that we could directly visualized the PG product.

To label PG during its synthesis, we incubated purified PBP2 with native Gly5-Lipid II and also BDL at room temperature for one hour (Figure 3.8a). The reaction was quenched and the products were separated by molecular weight with SDS-PAGE and transferred to PVDF

103 membrane for Western blot analysis. The biotinylated PG was visualized with streptavidin-HRP conjugate. We were surprised to detect only one band at the very top of the gel, sitting in the loading well. We reasoned that the product must be too highly crosslinked and unable to enter the gel matrix; such situation was previously observed for highly crosslinked S. pneumoniae peptidoglycan.34-35 The band at the top of well does not transfer properly and could not be reliably detected or quantified (Figure 3.8b). To better analyze the product, we added lysostaphin, an endopeptidase that specifically cleaves S. aureus pentaglycine crosslinks to the reaction mixture before quenching.36 The lysostaphin-treated sample showed strong chemiluminescence signals across a wide range of molecule weights, corresponding to BDL- labeled linear peptidoglycan polymers of different lengths (Figure 3.8b). When PBP2S398G mutant that lacks TP activity is used in the reaction, we observed no signal in the presence of absence of lysostaphin. We conclude that PBP2 PGT domain polymerizes Gly5-Lipid II into PG oligomer while the TP domain crosslinks the glycan strands via Gly5-bridge and mediate BDL incorporation at the same time. This TP-mediated labeling provides a simple assay for monitoring PBP2 transpeptidase activity in vitro.

104

Figure 3.8. Direct transpeptidase activity assay enables comparison of beta-lactam inhibition of S. aureus PBP2. (a) Schematic depicting lysostaphin digestion of polymerized, crosslinked and biotinylated PG.(b) Western blot of crosslinked peptidoglycan produced by PBP2 with (right) and without (left) post-reaction lysostaphin treatment. Product detection was enabled by BDL incorporation during PBP2 reaction. (c) Structures of beta-lactams examined in c-d. (d) Cefaclor inhibits PBP2 activity but mecillinam does not up to the highest concentration tested. (e) Ceftizoxime inhibits wild-type PBP2 at lower concentrations than the PBP2P458L variant identified in a cefitzoxime-resistant S. aureus mutant. For all experiments, 1 µM of enzyme was used. We next investigated inhibition of transpeptidase activity by beta-lactams using this assay

(Figure 3.8c). We briefly incubated PBP2 with various concentrations of beta-lactams before adding native Gly5-Lipid II and BDL. The product is processed as described previously. Without any inhibitor, PBP2 would make labeled peptidoglycan as shown before. If PBP2 TP domain is

105 inhibited however, the linear PG polymers would still be made but the products would not be crosslinked or BDL-labeled revealing no signal by Western blot analysis. Cefaclor is a second- generation cephalosporin whose PBP2 binding affinity was underestimated in competition binding assays due to its rapid deacylation rate.19 With our method we found that 5 µM of cefaclor completely inhibited PBP2 (1 µM) transpeptidase activity (Figure 3.8d). We found that the concentration of cefaclor needed to completely inhibit PBP2 TP activity does not change with reaction time but with the amount of the PBP2 enzyme to be inhibited. By contrast, mecillinam showed no detectable inhibition of PBP2 up to the highest concentration used (Figure

3.8d). This is consistent with mecillinam poor antibacterial activity against S. aureus and its reported low affinity for S. aureus PBP2.18 We also characterized other beta-lactams such as penicillin G, methicillin, and imipenem (Supplementary Figure 5). We found their inhibitory activities to be consistent with the result of competition binding assays.18

We wondered whether we could also characterize beta-lactam resistance mutations with our assay. Many laboratory and clinical isolates of S. aureus have resistance to certain beta- lactams that arose from point mutation in the pbp2 gene.13,37 A single amino acid substitution in

PBP2 (PBP2P458L) was previously identified in a ceftizoxime-resistant S. aureus strain; the mutant enzyme was reported to display decreased binding affinity to ceftizoxime.13 This mutant has not been characterized previously. It is unknown whether the mutation directly confers resistance. We purified PBP2P458L and used it in our BDL transpeptidase assay. In comparison to the wild-type PBP2, we found that PBP2P458L is indeed more resistant to ceftizoxime (Figure

3.8e). On the other hand, we found less oxacillin was needed to inhibit PBP2P458L compared to the PBP2 wild-type (Figure 3.9). This is consistent with a previous report that the mutant isolate was more susceptible to oxacillin killing that wild-type bacteria.13

106

Figure 3.9. Direct transpeptidase activity assay demonstrated the effect of point mutation on PBP2 susceptibility to a beta-lactam. The cefitzoxime-resistant mutant protein, PBP2P458L, shows no notable resistance to oxacillin compared to wild-type PBP2. 1 µM of enzyme was used. We next sought to modify the BDL transpeptidase assay to work with PBP2a, but two issues need to be addressed. First, PBP2a lacks PGT activity and it cannot label Lipid II directly.

To resolve this problem, I incubated PBP2a with Gly5-Lipid II. The PG synthesis was initiated by an addition of SgtB, a monofunctional PGT that cannot crosslink PG. The second issue is that

PBP2a TP activity is significantly lower than PBP2. I addressed this problem by incubating the reaction for 3 times longer (3 hours) and increasing the film exposure time to the blot during visualization.

I found that SgtB wild-type and PBP2a produced products that did not enter the gel matrix (Figure 3.10). Treating the reaction product with lysostaphin could not completely resolve the band at the top of the well into streak as we did with PBP2 reaction. I reasoned that SgtB must produces linear glycan strands that is much longer than the product of PBP2, these extra long PG cannot enter the gel even when it is not crosslinked. To observe PBP2a product on the gel, I instead used moenomycin-resistant SgtB Y181D protein. This point mutation confers weak moenomycin resistant, but also cause defective PGT activity, resulting in significantly shorter oligomer PG products (Figure 3.10).

107

Figure 3.10. SgtB wild type produces nascent glycan strands that are too long. Western blot analysis direct transpeptidase activity using PBP2a and SgtB wild-type (lane 1 and 2) or SgtB Y181D mutant (lane 3 and 4). The product of SgtB polymerization is too long to enter the SDS-PAGE gel matrix. SgtB Y181D mutation cause PGT activity defect. The mutant enzyme produces shorter linear peptidoglycan. Lane 1 and 3 have CaCl2 in the reaction buffer while lane 2 and 4 have MnCl2 and tween-80. I then used this BDL transpeptidase assay to examine the beta-lactam susceptibility profile PBP2a. Previous studies showed that PBP2a can be inactivated by fifth generation cephalosporins, ceftaroline and ceftobiprole, and is weakly inhibited by meropenem, but not other beta-lactams.38-40 I incubated PBP2a with these beta-lactams and Lipid II. The addition of

SgtB Y181D initiates PG synthesis allowing PBP2a, if not already inhibited by the beta-lactam, to crosslink PG and label PG with BDL (Figure 3.11a). I found that PBP2a could be inhibited by ceftaroline and ceftobiprole at concentration close to the enzyme concentration (Figure 3.11b-c).

Meropenem could also inhibit PBP2a TP activity but at much higher concentration and PBP2a inhibition by cefoxitin (PBP4 inhibitor) requires even higher concentration of the drug (Figure

3.11b-c). The result is consistent with the competition binding assays results.39,41-42 The transpeptidase inhibition assay here allows us to observe inhibition of PBP2a activity for the first time. This tool should be useful for developing the next generation of antibiotics against MRSA.

108

Figure 3.11. Direct transpeptidase activity assay enables comparison of beta-lactam inhibition of S. aureus PBP2a. (a) Schematic depicting PBP2a crosslinking peptidoglycan and incorporating biotin-D-lysine probe. (b-c) Western blot showing inhibition of PBP2a transpeptidase by different beta-lactams. In all experiments, 1 µM of PBP2a was used.

109

3.7 Substrate specificity of S. aureus PBPs

We next looked at the substrate preferences of PBP2a, PBP2, and PBP4. These three

PBPs are important for beta-lactam resistance. PBP2a could carry out PG crosslinking when other transpeptidases are inhibited. PBP2 PGT domain is required for PBP2a to function. PBP4 can confer high resistance to beta-lactams. S. aureus Lipid II contains a pentaglycine branch that is synthesized by FemX, FemA, and FemB. Deletion of FemA yields Gly1-Lipid II while

43-44 deletion of FemB yields Gly3-Lipid II. Since both femA and femB mutants are viable, one of the PBPs must be able of crosslinking truncated glycine branch to the stem peptide.43-44 It is currently unknown which PBP does the job. The deletion of these genes increase susceptibility of beta-lactams in MRSA even when PBP2a is expressed. Therefore, it has been proposed that

PBP2a transpeptidase can only crosslink Gly5-peptidoglycan, but that hypothesis has not been

45-46 proven. We used our ability to obtain Gly1, Gly3, Gly5-Lipid II substrates and our in vitro PG synthesis assay to tackle these questions.

To investigate crosslinking preference of these transpeptidases, we incubated each of the three substrates with either (1) PBP2 S398G and PBP2a, (2) PBP2 wt, or (3) PBP2 S398G and PBP4.

For each condition, we quantified the ratio of of crosslinked muropeptide and hydrolyzed monomer lacking the terminal D-Ala using LC/MS (Figure 3.12a). We found that the crosslinking of activity of PBP2 on Gly5-substate does not change over time (Supplementary

Figure 6). The combined crosslinking and hydrolysis activities were approximately constant across different substrates for each enzyme, showing that the formation of acyl-enzyme intermediate was not sensitive to the length of the glycine branch on the acyl-donor; as I previously mentioned, the hydrolyzed product came from acyl-enzyme intermediate that was

110 cleaved by water instead of the glycine branch. The proportion of the crosslinking to hydrolysis however varied greatly (Figure 3.12b). We found that while PBP2a prefers Gly5-branch as a acyl-acceptor, it also crosslink Gly3-branch to a significant extent. PBP2a however only hydrolyze Gly1-peptidoglycan, suggesting that it cannot take Gly1-branch as a nucleophile at all.

PBP2 crosslinked all three substrates very well. This result answers how femA and femB mutants are viable. PBP2 must be able to process Gly1- and Gly3-Lipid II into crosslinked peptidoglycan in vivo. PBP4 almost exclusively crosslink Gly5-substrate. This is consistent with previous studies showing that PBP4 produces highly crosslinked PG.13 Meanwhile, PBP4 mainly poorly crosslink Gly3-peptidoglycan and only acted as a carboxypeptidase toward the Gly1-substrate.

Figure 3.12. S. aureus PBPs exhibit different tolerance for the change of the glycine branch length in the substrate. (a) Muropeptide products of the reconstitution of PBP activity with Gly3-Lipid II after mutanolysin degradation and sodium borohydride reduction. A high amount of hydrolysis product implies that the glycine branch at that length is not a suitable nucleophile for a particular PBP. (b) A bar graph showing the raios of crosslinking and hydrolysis activities of PBP2a, PBP2, and PBP4 with Gly5, Gly3-, and Gly1-peptidoglycan. Integrated LC/MS peak areas for crosslinked products (X) and hydrolysis product (Y) were quantified and the %crosslinking was calculated as X/(X+Y); % hydrolysis = Y /(X+Y). We were surprised by how poorly PBP4 crosslink shorter glycine branch in vitro and the fact that PBP2a could crosslink Gly3-peptidoglycan in vitro, so we wanted to know whether these results have any relevance in vivo.

To find out whether PBP4 mainly works as a carboxypeptidase for the Gly3-substrate, we first quantified the percentage of hydrolyzed monomer compared to other muropeptides in

111 cell wall isolated from S. aureus wild-type and the ∆femB strains (Figure 3.13a). In wild-type the hydrolyzed product muropeptide make up about 10% of the total muropeptide and this increase to 20% in the ∆femB mutant. When pbp4 is deleted in the ∆femB background, we found that the amount of hydrolyzed muropeptide decrease to wild-type levels. Complementation of

∆pbp4∆femB with pbp4 doubles the amount of hydrolyzed muropeptide. We conclude that PBP4 has substantial carboxypeptidase activity in cells producing Gly3-peptidoglycan, consistent with our in vitro findings.

Figure 3.13. In cells, Gly3-peptidoglycan undergoes hydrolysis by PBP4 yet can be crosslinked by PBP2a. (a) S. aureus sacculi were degraded and analyzed by LC/MS to determine the extent of D-Ala hydrolysis. The bar graph shows the percentage of hydrolyzed muropeptide compared to other depicted species. Deleting PBP4 reduces hydrolysis. (b) Minimum inhibitory concentra-tions (MICs) for three different β-lactams against four S. aureus strains, with the targeted PBPs for each β-lactam shown in parentheses. The presence of mecA, encoding PBP2a, in the ∆femB background increases the MIC of piperacillin and cefoxitin, but not ceftaroline, consistent with the in vitro data that PBP2a can crosslink Gly3-peptidoglycan. In the ∆femA background, mecA does not confer resistance.

112 Determining whether PBP2a can cross-link Gly3-peptidoglycan in vivo was more challenging than probing PBP4’s role. We compared the minimum inhibitory concentrations

(MIC) of three beta-lactams, ceftaroline, piperacillin, and cefoxitin against four S. aureus strains: wild-type, ∆femB, mecA∆femB, and mecA∆femA (Figure 3.13b). These three beta-lactams have different binding affinity to each PBPs. Ceftaroline is able to inhibit PBP1,2,2a, and 3.

Piperacillin inhibits PBP1,2, and 3 but not PBP2a. Cefoxitin only inhibit PBP4 efficiently and has weak activity against PBP2. We found that deleting femB has only a modest effect on the

MICs of the three drugs. This is reasonable since PBP2 could crosslink the Gly3-substrate.

Introducing mecA cassette that express PBP2a into the ∆femB background, however, increased the MICs of piperacillin and cefoxitin by 8-fold and 16-fold respectively, but did not significantly affect the MIC of ceftaroline. The resistance conferred by PBP2a in the ∆femB background supports our in vitro data showing that PBP2a can crosslink Gly3-peptidoglycan.

Expression of PBP2a did not provide resistance against ceftaroline to the ∆femB strain because ceftaroline inhibits PBP2a. In contrast, mecA∆femA strain remained susceptible to all three drugs, confirming that PBP2a does not crosslink Gly1-peptidoglycan and therefore cannot rescue the cells when PBP2 is inhibited.

3.8 Conclusions

We demonstrated the first reconstitution of S. aureus PBP2 and PBP2a activities using the native Gly5-Lipid II substrate that has been unobtainable until now. Our ability to synthesize

S. aureus peptidoglycan in vitro show that three PBPs in MRSA have dynamic properties. The activity of their transpeptidase domain switch between working as a PG transpeptidase and a carboxypeptidase for trimming the D-Ala of the stem peptide, as the glycine branch on Lipid II is

113 shortened. We found out that PBP2 is likely responsible for cell wall crosslinking in cells when the glycine branch is truncated.

Our most intriguing result is that PBP2a unexpectedly crosslinks Gly3-peptidoglycan unlike previously proposed.45-46 The reason that femB deletion cause beta-lactam hypersensitivity could be due to PBP2a slightly slower crosslinking rate for the truncated glycine branch compared to the native pentaglycine branch. It could be due to the slightly weaker cell wall in general as femB deletion cause cell separation defect and septum placement defect.43 In addition, the hypersensitivity could be due to the decrease in cell surface protein. S. aureus anchors many of its cell surface proteins via sortase (SrtA and SrtB) to the pentaglycine branch of cell wall.47 It was found that SrtA does not accept stem peptide lacking glycine a substrate.

With the shortened glycine branch, SrtA activity is impaired.48

One strategy to combat MRSA infection is to find a compound that resensitize the bacteria to older-generation drugs that MRSA is normally resistant against. Such a compound could be an inhibitor of FemA or FemB as the deletion of these enzymes was shown to cause beta-lactam hypersensitivity.43-44 Our result implies that FemA would be a better target for such potentiator as it directly prevent the beta-lactam resistance determinant PBP2a from crosslinking the cell wall. This strategy is validated in nature. A natural product cyslabdan was found to inhibit FemA but not FemX or FemB.49 The non-antibiotic cyslabdan resensitizes MRSA to beta- lactams that were previously useless against the resistant bacteria.

114 3.9 Materials and methods

Materials

PBP4 deletion Staphylococcus aureus strains (∆pbp4) and complementation (∆pbp4 complemented with pbp4 gene), which were described previously, were obtained from the

Cheung Lab (Dartmouth University) (Supplementary Table 2). S. aureus was grown in tryptic soy broth (TSB) or on TSB with 1.5% agar at 30 and 37 °C. Primers were purchased from

Integrated DNA Technologies. Restriction endonucleases were purchased from New England

Biolabs. Vectors and expression hosts were obtained from Novagen. Non-stick conical vials and pipet tips used for enzymatic reactions were from VWR. Nalgene Oak Ridge High-Speed

Centrifuge Tubes used for lipid extractions were purchased from Thermo Scientific. Fmoc-D-

Lys(biotinyl)-OH was purchased from VWR. Teflaro (ceftaroline fosamil) was provided from

David Hooper (Massachusetts General Hospital). Zevtera (ceftobiprole medocaril) was provided by Rolf Müller (Helmholtz Centre for Infection Research). Alkaline phosphatase (from calf intestine) was purchased from Roche.

PBP4 and SgtB Y181D were purified as previously described.26

LC/MS chromatograms were obtained on an Agilent Technologies 1100 series LC-MSD instrument using electrospray ionization (ESI). Azure C400 was used to image Western blots.

Deprotection of ceftaroline fosamil

50 mg of ceftaroline fosamil powder from Teflaro (containing arginine as a stabilizer) was dissolved in 2.5 mL of H2O to make a 20 mg/mL solution. To a 5 mL Eppendorf tube, 2250

µL of 20 mg/mL ceftaroline fosamil solution, 440 µL of H2O, 300 µL of Roche 10x dephosphorylation buffer, and 10 µL of 20 U/µL alkaline phosphatase were added. The mixture

115 was rotated end over for 1 minute to mix and incubated in a 37°C shaker for one hour. The reaction mixture was loaded onto a pre-equilibrated Strata C18-E SPE column (Phenomenex,

Part number 8B-S001-EAK), washed with 3 mL of H2O (0.1% NH4OH), and eluted with 6 mL of MeOH (0.1% NH4OH). The collected elute fractions were dried by rotatory evaporation to afford 26.3 mg of yellow solid powder which was dissolved in 2 mL of 50:50 MeOH:H2O. The solution was purified by reverse-phase high pressure liquid chromatography (HPLC, Agilent

1260 Infinity) using a C18 stationary phase (Agilent Zorbax 300SB C18, 5 µM, 9.4 x 250 mM) with 12.5% to 13.5% acetonitrile (0.1% formic acid) in H2O gradient over 22.5 minutes. The

HPLC elutes were collected in a 50 mL round-bottom flask, and concentrated by rotatory evaporation to afford 10.4 mg of ceftaroline.

Deprotection of ceftobiprole medocaril

To a 1.5 mL Eppendorf tube (with a small hole on the cap to allow CO2 release), 10.2 mg of ceftobiprole medocaril sodium from Zevtera and 1020 µL of 1x PBS buffer (pH 7.4) were added. The solution (pH 6.5) was adjust to pH 7.5 with 20 µL of 0.1 M NaOH. The mixture was vortexed and incubated in a 37°C shaker for 5 hours, and quenched by adjusting the pH to 6 with

0.1 M HCl. The reaction mixture was purified by HPLC (Agilent 1260 Infinity) using a C18 stationary phase (Agilent Zorbax 300SB C18, 5 µM, 9.4 x 250 mM) with 0% to 30% acetonitrile

(0.1% formic acid) in H2O gradient over 22.5 minutes. The HPLC elute fraction was promptly dried by rotatory evaporation to afford 1.4 mg of ceftobiprole.

116 Cloning of S. aureus PBP2[M59-S716] and its point mutants

The S. aureus pbp2[M59-S716] construct was cloned into pET42a(+) plasmid using

Gibson assembly protocol. Briefly, the nucleotide region of S. aureus pbp2 gene encoding amino acid 59 to 716 was amplified using primer pair F’pET42a_PBP2 and R’pET42a_PBP2

(Supplementary table 1). The PCR product was analyzed and purified using agarose gel electrophoresis. The pET42a(+) plasmid was amplified using the primer pair F’pET42a and

R’pET42b (Supplementary table 1). The purified PCR product and the linearized vector were assembled using Gibson assembly protocol. The inserted pbp2[M59-S716] gene was confirmed by sequencing (Beckman Coulter Sequencing Facility). E. coli NovaBlue strain was used for cloning.

The point mutants of pbp2[M59-S716] were made with the appropriate primers

(Supplementary table 1) using QuickChange site-directed mutagenesis kit (Stratagene). The construct was confirmed by sequencing.

Overexpression and purification of S. aureus PBP2[M59-S716] and its point mutants

The plasmid encoding S. aureus PBP2[M59-S716] collected from the NovaBlue cloning strain was transformed into E. coli. BL21(DE3) culture for overexpression and purification. A 15 mL culture of E. coli BL21 (DE3) harboring the plasmid encoding for S. aureus PBP2[M59-

S716] grown in LB supplemented with 50 μg/ mL kanamycin was grown overnight, which was then used to inoculate 1.5 L LB medium (1:100) supplemented with 50 μg/mL kanamycin. The culture was grown at 37 ˚C with shaking until OD600 reached 0.4–0.5, and then cooled down to

17 ˚C before induction with 0.5 mM IPTG for 17 h with shaking. Cells were harvested by centrifugation (5250 x g, 20 min, 4 ˚C) and pellets were resuspended on ice with 30 mL of lysis

117 buffer (10 mM Tris pH 8.0, 1 M NaCl, 10 mM MgCl2, 10% v/v glycerol and 40 mM CHAPs) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) and 100 μg/mL DNase from bovine pancreas (Sigma Aldrich). The resuspended cells were passed through a cell disrupter (3 x 10,000 psi, 4 ˚C) for three times. The cell lysate was then pelleted by ultracentrifugation

(100,000 x g, 30 min, 4 ˚C). The resulting supernatant containing PBP2[M59-S716] protein was applied to 1.5 mL pre-washed Ni-NTA resin (Qiagen) at 4 ˚C (the Ni-NTA resin was washed with dH2O and equilibrated with lysis buffer). After collecting flow through (FT), the resin was washed with buffer A (10 mM Tris pH 8.0, 0.2 M NaCl, and 0.28 mM LDAO) to facilitate detergent exchange (20 mL, W1). The resin was then washed with buffer A containing 60 mM imidazole (20 mL, W2). The protein was eluted with buffer A containing increasing imidazole concentrations (10 mL of 100 mM, 200 mM and 500 mM each, E1-E3). The fractions were analyzed on using SDS-PAGE electrophoresis. The fractions containing PBP2 protein (E1- E3) were combined and concentrated to ~10 mg/mL using a 50 kD MWCO Amicon Ultra Centrifuge

Filter Device (Millipore). The concentrated PBP2 sample was further purified using size- exclusion chromatography with a Superdex S200 column equilibrated in buffer A. The fractions indicating monomeric protein were combined and concentrated using the 50 kD MWCO Amicon

Ultra Centrifuge Filter Device, while the concentration was measured on nanodrop using the calculated extinction coefficient of PBP2 [M59-S716]. The final yield was approximately 10 mg per 1.5 L culture. S. aureus PBP2 [M59-S716] is referred to PBP2 subsequently. PBP2S398G and PBP2P458L were purified using the same protocol.

118 LC/MS assays for evaluations of PBP2 activities in vitro

S. aureus Lipid II in DMSO (40 μM) was incubated with purified PBP2 (1 μM) in 1x reaction buffer (50 mM HEPES, pH 7.5, 10 mM CaCl2) in a total of 10 μL reaction volume for 1 h at 25 ˚C. The reaction was quenched at 95 ˚C for 5 min, and then treated with mutanolysin

(from Streptomyces globisporus, Sigma, 1 U) for 1.5 h at 37 ˚C followed by another 1 U aliquot for 1.5 h. The resulting disaccharides were reduced with sodium borohydride (10 µL of 10 mg/mL solution, 30 min). Phosphoric acid (20%, 1.2 μL) was then added to adjust the pH to ~4.

Then reaction mixture was the lyophilized, redissolved in 12 μL H2O and subjected to

LC/HRMS analysis, conducted with ESI-MS operating in positive mode on a Bruker qTOF mass spectrometer. The instrument was equipped with a Waters Symmetry Shield RP18 column (5

μM, 3.9 x 150 mm) with a matching column guard. The fragments were separated using the following method: 0.5 mL/min H2O (0.1% formic acid) for 5 min followed by a gradient of 0% acetonitrile (ACN) (0.1% formic acid)/H2O (0.1% formic acid) to 40% ACN (0.1% formic acid)/H2O (0.1% formic acid) over 25 min. Molecular ions corresponding to expected muropeptides were extracted. The reactions with PBP2S398G or SgtB were carried out using identical conditions.

The time-course analysis was performed by quenching the PBP2 reaction at various time points and subjected to mutanolysin digestion. The integrated areas of the peaks corresponding to monomeric muropeptides and dimeric muropeptides based on ion counts were measured and plotted; chromatograms showing the peak corresponding to the trimeric muropeptide were shown.

119 Western blot assay to study PBP2 transpeptidase activity

Briefly, Lipid II (1 μL of 100 μM stock in DMSO) was incubated with PBP2 (1 μL of 10

μM), BDL (1.5 μL of 20 mM) and 10x reaction buffer (1 μL of 500 mM HEPES, pH 7.5, 100 mM CaCl2) to reach a total volume of 10 μL. The reaction was incubated at room temperature for 15 min, and heat quenched briefly at 100 ˚C for 1 min. Lysostaphin (0.5 μL of 1 mg/mL) was added to the reaction mixture to resolve the crosslinked product. The reaction was shaken at 37

˚C for 3 h. To quench the reaction, 10 μL of 2x SDS loading buffer was added. The protocol for western blot analysis described in the earlier section was used. Reactions using PBP2S398G with or without lysostaphin treatment were performed.

Characterization of PBP2 enzymatic inhibition by beta-lactams

The aforementioned western blot assay for PBP2 transpeptidase activity was modified slightly to allow studies of beta-lactam inhibition. Briefly, PBP2 (1 μL of 10 μM) was pre- incubated for 10 minutes with varying concentrations of beta-lactams (0- 20 μM) in a reaction mixture containing BDL (1.5 μL of 20 mM) and 10x reaction buffer (1 μL of 500 mM HEPES, pH 7.5, 100 mM CaCl2) for 5 min. Lipid extract (1 μL of 100 μM stock in DMSO) was then added to the reaction (final reaction volume: 10 μL) and let to incubate at room temperature for

15 min, and heat quenched briefly at 100 ˚C for 1 min. Lysostaphin (0.5 μL of 1 mg/mL) was added to the reaction mixture to resolve the crosslinked product. The reaction was shaken at 37

˚C for 3 h. To quench the reaction, 10 μL of 2x SDS loading buffer was added. Western blot analysis was the same as above. For studies on the resistance mutation, PBP2 P458L was used in place of the wild-type protein.

120 Cloning of S. aureus PBP2a[24-688] and its point mutants

Following the previously reported method,[2] mecA was amplified from chromosomal

DNA of S. aureus ATCC706986 with the reported primers and cloned into pET38b.

The point mutants of mecA[24-688] were made with the appropriate primers

(Supplementary table 1) using QuickChange site-directed mutagenesis kit (Stratagene). The construct was confirmed by sequencing.

Overexpression and purification of S. aureus PBP2a wt and S403A, K406A mutants

The plasmid pET38b encoding S. aureus mecA[24-688] collected from the

NovaBlue(DE3) cloning strain was transformed into E. coli BL21(DE3) culture for overexpression and purification. An overnight culture of E. coli BL21 (DE3) harboring the plasmid encoding for S. aureus PBP2a was diluted 1:100 in 1.5 L of LB supplemented with kanamycin (50 µg/mL). The culture was grown at 37 °C with shaking until OD600 reached 0.4–

0.5, and then cooled down to 15 °C before induction with 0.5 mM IPTG for 18 hours with shaking. Cells were harvested by centrifugation (5250 x g, 15 min, 4 °C) and pellets were resuspended on ice with 40 mL of lysis buffer (10 mM Tris (pH 8.0), 1 M NaCl, 10 mM MgCl2,

10% v/v glycerol and 40 mM CHAPs) supplemented with 1 mM phenylmethylsulfonyl fluoride

(PMSF) and 100 μg/mL DNase from bovine pancreas (Sigma Aldrich). The resuspended cells were passed through a cell disrupter (3 x 10,000 psi, 4 °C) for three times. The cell lysate was then pelleted by ultracentrifugation (100,000 x g, 30 min, 4 °C).

The resulting supernatant containing PBP2a protein was applied to 1.5 mL pre-washed

Ni-NTA resin (Qiagen) at 4 °C (the Ni-NTA resin was washed with dH2O and equilibrated with lysis buffer). The resin was washed with 20 mL buffer A (25 mM HEPES (pH 7.0), 1 M NaCl)

121 and 20 mL of buffer A containing 40 mM imidazole. The protein was eluted with 10 mL of buffer A containing 200 mM imidazole. The fractions were analyzed using SDS-PAGE electrophoresis. The elute fraction was concentrated to 10 mg/mL using a 50 kD MWCO

Amicon Ultra Centrifuge Filter Device (Millipore). The concentrated PBP2a sample was further purified using size-exclusion chromatography with a Superdex S200 column equilibrated in buffer A. The fractions indicating monomeric protein were combined and concentrated using the

50 kD MWCO Amicon Ultra Centrifuge Filter Device, while the concentration was measured using Biorad protein assay. The final yield was approximately 18 mg per 1.5 L culture. S. aureus

PBP2a is referred to PBP2a subsequently. PBP2a S403A and PBP2 K406A were purified using the same protocol.

LC/MS assay to evaluate PBP2a activity in vitro

S. aureus Gly5-Lipid II in DMSO (12.5 µM) was incubated with purified PBP2a (2 µM) and SgtB Y181D (2 µM) in 1x reaction buffer (50 mM HEPES, pH 6.5, 2.5 mM MgCl2) in a total of 20 µL reaction volume for 3 hours at room temp. The reaction was quenched at 95 ˚C for

5 min, and then treated with mutanolysin (from Streptomyces globisporus, Sigma, 1 U) for 1.5 hours at 37˚C followed by another 1 U aliquot for 1.5 hours. The resulting disaccharides were reduced with sodium borohydride (10 µL of 10 mg/mL solution, 30 min). Phosphoric acid (20%,

1.4 µL) was then added to adjust the pH to 4. The reaction mixture was then lyophilized, redissolved in 20 µL H2O and subjected to LC/HRMS analysis, conducted with ESI-MS operating in positive mode on a Bruker qTOF mass spectrometer. The instrument was equipped with a Waters Symmetry Shield RP18 column (5 μM, 3.9 x 150 mm) with a matching column guard. The fragments were separated using the following method: 0.5 mL/min H2O (0.1% formic

122 acid) for 5 min followed by a gradient of 0% acetonitrile (ACN) (0.1% formic acid)/H2O (0.1% formic acid) to 40% ACN (0.1% formic acid)/H2O (0.1% formic acid) over 25 min. The following ions were extracted from each chromatogram: monomer (pink peak): 1253.5856

([M+H]); monomer minus D-Ala (blue peak): 1029.0617 ([M+2H]2+); and dimer (orange peak):

1194.2204 ([M+3H]3+). The reaction with PBP2a S403A and PBP2a K406A was carried out using identical conditions. The reaction with PBP2a and ceftaroline was carried out in the presence of 50 µM ceftaroline.

Western blot assay to evaluate PBP2a inhibition by beta-lactams

To a non-stick reaction vial, 3.5 µL H2O was mixed with 1 µL of 10x reaction buffer

(500 mM HEPES, pH 6.5, 25 mM MgCl2) and 1.5 µL of BDL probe (20 mM in H2O). The mixture was vortexed before it was added with 1 µL of beta-lactam solution of varying concentrations (0-5120 µM in DMSO). The mixture was vortexed again before it was added with

1 µL of 50 µM SgtB Y181D and 1 µL of 20 µM PBP2a. The tube was flicked vigorously to mix and incubated for 10 minutes at room temp. Lipid extract (1 µL of 150 µM stock in DMSO) was then added to the reaction (final reaction volume: 10 µL) and let to incubate at 25 ˚C with shaking (400 rpm) for 3 hours. The reaction was quenched with 10 µL of 2x SDS loading dye. 5

µL of the quenched reaction mixture was loaded onto a 4-20% gradient polyacrylamide gel and let run at 180 V for 40 minutes. The products were transferred onto Immuno-Blot PVDF membrane (BioRad). Biotinylated peptidoglycan polymer was detected by blotting with streptavidin-HRP (1:7000).

123 LC/MS Analysis of PBP2a, PBP2, PBP4 and reactions

For the PBP2a reactions, S. aureus Gly5-Lipid II (or Gly3-Lipid II or Gly1-Lipid II) in

DMSO (25 µM) was incubated with purified PBP2a (2 µM) and PBP2 S398G (1 µM) in 1x

PBP2a buffer (50 mM HEPES (pH 6.5), 2.5 mM MgCl2) in a total of 20 µL reaction volume for

3 hours at room temp. The reaction was quenched at 95 ˚C for 5 min, and then treated with mutanolysin (from Streptomyces globisporus, Sigma, 1 U) for 1.5 h at 37˚C followed by another

1 U aliquot for 1.5 h. The resulting disaccharides were reduced with sodium borohydride (10 µL of 10 mg/mL solution, 30 min). Phosphoric acid (20%, 1.4 µL) was then added to adjust the pH to 4. The reaction mixture was then lyophilized, redissolved in 20 µL H2O and subjected to

LC/HRMS analysis, conducted with ESI-MS operating in positive mode on a Bruker qTOF mass spectrometer. The instrument was equipped with a Waters Symmetry Shield RP18 column (5

μM, 3.9 x 150 mm) with a matching column guard. The fragments were separated using the following method: 0.5 mL/min H2O (0.1% formic acid) for 5 min followed by a gradient of 0% acetonitrile (ACN) (0.1% formic acid)/H2O (0.1% formic acid) to 40% ACN (0.1% formic acid)/H2O (0.1% formic acid) over 25 min. Molecular ions corresponding to expected muropeptides were extracted. Integrated peak areas for monomer minus D-ala muropeptide (X) and for crosslinked-dimer and crosslinked-trimer muropeptides combined (Y) were quantified and the %hydrolysis was calculated as X/(X+Y); % crosslinking = Y /(X+Y).

For the PBP2 reactions, S. aureus Gly5-Lipid II (or Gly3-Lipid II or Gly1-Lipid II) in

DMSO (25 µM) was incubated with purified PBP2 (1 µM) in 1x TGase buffer (50 mM HEPES

(pH 7.5), 10 mM CaCl2) in a total of 20 µL reaction volume for 3 hours at room temp. The reactions were worked up and analyzed as described above.

124 For the PBP4 reaction, S. aureus Gly5-Lipid II (or Gly3-Lipid II or Gly1-Lipid II) in

DMSO (25 µM) was incubated with purified PBP2 S398G (1 µM) in 1x MTG buffer (12.5 mM

HEPES (pH 7.5), 2.0 mM MnCl2, 0.25 mM Tween-80) in a total of 18 µL reaction volume for 3 hours at room temp. PBP4 (4 µM) was added and the mixture was incubated for 3 more hours at room temp. The reactions were worked up and analyzed as described above.

The reactions of PBP2a with S. aureus Gly5-Lipid II (or Gly3-Lipid II or Gly1-Lipid II) were also carried out in TGase buffer and MTG buffer using the same protocol. PBP2 reactions were also performed in PBP2a buffer and MTG buffer. PBP4 reactions were also carried out in

PBP2a buffer and TGase buffer (Supplementary Table 3).

The time-course analysis [SI Fig. 9] was performed by quenching the PBP2 reaction at various time points and subjected to mutanolysin digestion.

Transduction of ∆femB

Phage 80alpha was used for transductions to move ∆femB::Tn551 from UT34-2 into

BB255 (NCTC 8325), MW2, MW2 ∆pbp4, MW2 ∆pbp4::pbp4 in order to create NCTC 8325

∆femB, MW2 ∆femB, MW2 ∆pbp4∆femB, and MW2 ∆femB ∆pbp4::pbp4, respectively. The methods for transduction were described previously.50

Isolation and enzyme digestion of S. aureus sacculus

The protocol was modified from a previous report.51 Briefly, an overnight culture of S. aureus (MW2, MW2 ∆femB, MW2 ∆femB ∆pbp4, MW2 ∆femB∆pbp4::pbp4) (2 mL) was centrifuged at 10,000 rpm for 5 min. The pellet was resuspended in 1 mL 0.25% SDS in 0.1 M

Tris/HCl, pH ~7.0, and boiled at 100 ˚C for 20 min. The suspension was centrifuged at 10,000

125 rpm for 5 min, and the pellet was washed with 1.5 mL H2O for at least three times to remove

SDS. The washed pellet was resuspended in 1 mL H2O and sonicated in a water bath for 30 min, after which, 500 μL of a solution containing 15 μg/mL DNase from bovine, 60 μg/mL RNase in

0.1 M Tris-HCl, pH 6.8 was added. After shaking at 37 ˚C for 2 h, the mixture was boiled at 100

˚C for 3 min to inactivate enzymes, centrifuged (5 min, 10,000 rpm) and washed with water once

(1 mL). To release wall teichoic acid, the pellet was suspended in 500 μL of 1M HCl and incubated with shaking at 37 ˚C for 4 h. The pellet was centrifuged (5 min, 10,000 rpm) and washed with water until pH is 5 ~6. The pellet was resuspended in a 100 μL digestion buffer of

12.5 mM NaH2PO4, pH 5.5, and treated with 10 μL of mutanolysin (5 U/mL in H2O), and was incubated with shaking at 37 ˚C for 16 h. After digestion, the sample was boiled at 100 ˚C for 3 min to inactivate enzymes. The sample was centrifuged (5 min, 10,000 rpm), and added 50 μL

NaBH4 (10 mg/mL) at room temperature for 30 min. The pH of the sample was adjusted to 4 with 20% phosphoric acid, and lyophilized. The lyophilized materials were resuspended in 500

μL H2O, and 20 μL was used for LC/MS analysis. The LC/MS condition is the same as described above for PBP2 reactions.

Measurement of minimum inhibitory concentration (MIC) of beta-lactams against S. aureus

A serial dilution of antibiotic was prepared, and 1.5 μL of each was added to a clear 96- well plate (Corning). To each well containing antibiotic, 150 μL of a culture of S. aureus

(NCTC8325, NCTC8325 ∆femB, NCTC8325 mecA∆femB, NCTC8325 mec ∆femA) at

OD600= 0.1 (diluted from a saturated overnight culture in TSB medium) was seeded. The plate

126 was shaken at 37 °C and OD600 reading was taken after 18 hr. The concentration of antibiotic that showed no bacteria growth was designated the MIC.

3.10 References

(1) Nakagawa, J.-i.; Tamaki, S.; Matsuhashi, M. Purified Penicillin Binding Proteins 1Bs from Escherichia coli Membrane Showing Activities of Both Peptidoglycan Polymerase and Peptidoglycan Crosslinking Enzyme. Agricultural and Biological Chemistry 1979, 43 (6), 1379-1380.

(2) Ishino, F.; Mitsui, K.; Tamaki, S.; Matsuhashi, M. Dual enzyme activities of cell wall peptidoglycan synthesis, peptidoglycan transglycosylase and penicillin-sensitive transpeptidase, in purified preparations of Escherichia coli penicillin-binding protein 1A. Biochemical and Biophysical Research Communications 1980, 97 (1), 287-293.

(3) Terrak, M.; Ghosh, T. K.; Van Heijenoort, J.; Van Beeumen, J.; Lampilas, M.; Aszodi, J.; Ayala, J. A.; Ghuysen, J.-M.; Nguyen-Distèche, M. The catalytic, glycosyl transferase and acyl transferase modules of the cell wall peptidoglycan-polymerizing penicillin- binding protein 1b of Escherichia coli. Molecular Microbiology 1999, 34 (2), 350-364.

(4) Born, P.; Breukink, E.; Vollmer, W. In Vitro Synthesis of Cross-linked Murein and Its Attachment to Sacculi by PBP1A from Escherichia coli. Journal of Biological Chemistry 2006, 281 (37), 26985-26993.

(5) Hugonnet, J.-E.; Haddache, N.; Veckerlé, C.; Dubost, L.; Marie, A.; Shikura, N.; Mainardi, J.-L.; Rice, L. B.; Arthur, M. Peptidoglycan cross-linking in glycopeptide- resistant Actinomycetales. Antimicrobial Agents and Chemotherapy 2014, 58 (3), 1749.

(6) Bertsche, U.; Breukink, E.; Kast, T.; Vollmer, W. In Vitro Murein (Peptidoglycan) Synthesis by Dimers of the Bifunctional Transglycosylase-Transpeptidase PBP1B from Escherichia coli. Journal of Biological Chemistry 2005, 280 (45), 38096-38101.

(7) Lebar, M. D.; Lupoli, T. J.; Tsukamoto, H.; May, J. M.; Walker, S.; Kahne, D. Forming Cross-Linked Peptidoglycan from Synthetic Gram-Negative Lipid II. Journal of the American Chemical Society 2013, 135 (12), 4632-4635.

127 (8) Lupoli, T. J.; Tsukamoto, H.; Doud, E. H.; Wang, T.-S. A.; Walker, S.; Kahne, D. Transpeptidase-Mediated Incorporation of d-Amino Acids into Bacterial Peptidoglycan. Journal of the American Chemical Society 2011, 133 (28), 10748-10751.

(9) Lupoli, T. J.; Lebar, M. D.; Markovski, M.; Bernhardt, T.; Kahne, D.; Walker, S. Lipoprotein activators stimulate Escherichia coli penicillin-binding proteins by different mechanisms. Journal of the American Chemical Society 2014, 136 (1), 52-55.

(10) Müller, P.; Ewers, C.; Bertsche, U.; Anstett, M.; Kallis, T.; Breukink, E.; Fraipont, C.; Terrak, M.; Nguyen-Distèche, M.; Vollmer, W. The Essential Cell Division Protein FtsN Interacts with the Murein (Peptidoglycan) Synthase PBP1B in Escherichia coli. Journal of Biological Chemistry 2007, 282 (50), 36394-36402.

(11) Boes, A.; Olatunji, S.; Breukink, E.; Terrak, M. Regulation of the Peptidoglycan Polymerase Activity of PBP1b by Antagonist Actions of the Core Divisome Proteins FtsBLQ and FtsN. mBio 2019, 10 (1), e01912-01918.

(12) Typas, A.; Banzhaf, M.; van den Berg van Saparoea, B.; Verheul, J.; Biboy, J.; Nichols, R. J.; Zietek, M.; Beilharz, K.; Kannenberg, K.; von Rechenberg, M.; Breukink, E.; den Blaauwen, T.; Gross, C. A.; Vollmer, W. Regulation of Peptidoglycan Synthesis by Outer-Membrane Proteins. Cell 2010, 143 (7), 1097-1109.

(13) Łęski, T. A.; Tomasz, A. Role of Penicillin-Binding Protein 2 (PBP2) in the Antibiotic Susceptibility and Cell Wall Cross-Linking of Staphylococcus aureus: Evidence for the Cooperative Functioning of PBP2, PBP4, and PBP2A. Journal of Bacteriology 2005, 187 (5), 1815.

(14) Magnet, S.; Dubost, L.; Marie, A.; Arthur, M.; Gutmann, L. Identification of the L,D- transpeptidases for peptidoglycan cross-linking in Escherichia coli. Journal of Bacteriology 2008, 190 (13), 4782.

(15) Mainardi, J. L.; Hugonnet, J. E.; Rusconi, F.; Fourgeaud, M.; Dubost, L.; Moumi, A. N.; Delfosse, V.; Mayer, C.; Gutmann, L.; Rice, L. B.; Arthur, M. Unexpected inhibition of peptidoglycan LD-transpeptidase from Enterococcus faecium by the beta-lactam imipenem. The Journal of biological chemistry 2007, 282 (42), 30414-30422.

(16) Cordillot, M.; Dubée, V.; Triboulet, S.; Dubost, L.; Marie, A.; Hugonnet, J.-E.; Arthur, M.; Mainardi, J.-L. In Vitro Cross-Linking of Mycobacterium tuberculosis Peptidoglycan by L,D-Transpeptidases and Inactivation of These Enzymes by Carbapenems. Antimicrobial Agents and Chemotherapy 2013, 57 (12), 5940.

128 (17) Magnet, S.; Arbeloa, A.; Mainardi, J.-L.; Hugonnet, J.-E.; Fourgeaud, M.; Dubost, L.; Marie, A.; Delfosse, V.; Mayer, C.; Rice, L. B.; Arthur, M. Specificity of L,D- Transpeptidases from Gram-positive Bacteria Producing Different Peptidoglycan Chemotypes. Journal of Biological Chemistry 2007, 282 (18), 13151-13159.

(18) Georgopapadakou, N. H.; Liu, F. Y. Binding of beta-lactam antibiotics to penicillin- binding proteins of Staphylococcus aureus and Streptococcus faecalis: relation to antibacterial activity. Antimicrobial agents and chemotherapy 1980, 18 (5), 834-836.

(19) Chambers, H. F.; Miick, C. Characterization of penicillin-binding protein 2 of Staphylococcus aureus: deacylation reaction and identification of two penicillin-binding peptides. Antimicrob Agents Chemother 1992, 36 (3), 656-661.

(20) Cava, F.; de Pedro, M. A.; Lam, H.; Davis, B. M.; Waldor, M. K. Distinct pathways for modification of the bacterial cell wall by non-canonical D-amino acids. The EMBO journal 2011, 30 (16), 3442-3453.

(21) Kuru, E.; Hughes, H. V.; Brown, P. J.; Hall, E.; Tekkam, S.; Cava, F.; de Pedro, M. A.; Brun, Y. V.; VanNieuwenhze, M. S. In Situ probing of newly synthesized peptidoglycan in live bacteria with fluorescent D-amino acids. Angewandte Chemie (International ed. in English) 2012, 51 (50), 12519-12523.

(22) Liechti, G. W.; Kuru, E.; Hall, E.; Kalinda, A.; Brun, Y. V.; VanNieuwenhze, M.; Maurelli, A. T. A new metabolic cell-wall labelling method reveals peptidoglycan in Chlamydia trachomatis. Nature 2014, 506 (7489), 507-510.

(23) Kuru, E.; Tekkam, S.; Hall, E.; Brun, Y. V.; Van Nieuwenhze, M. S. Synthesis of fluorescent D-amino acids and their use for probing peptidoglycan synthesis and bacterial growth in situ. Nature protocols 2015, 10 (1), 33-52.

(24) Shieh, P.; Siegrist, M. S.; Cullen, A. J.; Bertozzi, C. R. Imaging bacterial peptidoglycan with near-infrared fluorogenic azide probes. Proceedings of the National Academy of Sciences 2014, 111 (15), 5456.

(25) Gautam, S.; Kim, T.; Spiegel, D. A. Chemical Probes Reveal an Extraseptal Mode of Cross-Linking in Staphylococcus aureus. Journal of the American Chemical Society 2015, 137 (23), 7441-7447.

129 (26) Qiao, Y.; Lebar, M. D.; Schirner, K.; Schaefer, K.; Tsukamoto, H.; Kahne, D.; Walker, S. Detection of lipid-linked peptidoglycan precursors by exploiting an unexpected transpeptidase reaction. J Am Chem Soc 2014, 136 (42), 14678-14681.

(27) Barrett, D.; Leimkuhler, C.; Chen, L.; Walker, D.; Kahne, D.; Walker, S. Kinetic characterization of the glycosyltransferase module of Staphylococcus aureus PBP2. J Bacteriol 2005, 187 (6), 2215-2217.

(28) Lovering, A. L.; de Castro, L. H.; Lim, D.; Strynadka, N. C. J. Structural Insight into the Transglycosylation Step of Bacterial Cell-Wall Biosynthesis. Science (New York, N.Y.) 2007, 315 (5817), 1402.

(29) Fuda, C.; Suvorov, M.; Vakulenko, S. B.; Mobashery, S. The Basis for Resistance to β- Lactam Antibiotics by Penicillin-binding Protein 2a of Methicillin-resistant Staphylococcus aureus. Journal of Biological Chemistry 2004, 279 (39), 40802-40806.

(30) Fuda, C.; Suvorov, M.; Vakulenko, S. B.; Mobashery, S. The Basis for Resistance to β- Lactam Antibiotics by Penicillin-binding Protein 2a of Methicillin-resistant Staphylococcus aureus. J. Biol. Chem. 2004, 279 (39), 40802-40806.

(31) Moisan, H.; Pruneau, M.; Malouin, F. Binding of ceftaroline to penicillin-binding proteins of Staphylococcus aureus and Streptococcus pneumoniae. The Journal of antimicrobial chemotherapy 2010, 65 (4), 713-716.

(32) Otero, L. H.; Rojas-Altuve, A.; Llarrull, L. I.; Carrasco-López, C.; Kumarasiri, M.; Lastochkin, E.; Fishovitz, J.; Dawley, M.; Hesek, D.; Lee, M.; Johnson, J. W.; Fisher, J. F.; Chang, M.; Mobashery, S.; Hermoso, J. A. How allosteric control of Staphylococcus aureus penicillin binding protein 2a enables methicillin resistance and physiological function. Proceedings of the National Academy of Sciences 2013, 110 (42), 16808.

(33) Lebar, M. D.; May, J. M.; Meeske, A. J.; Leiman, S. A.; Lupoli, T. J.; Tsukamoto, H.; Losick, R.; Rudner, D. Z.; Walker, S.; Kahne, D. Reconstitution of Peptidoglycan Cross- Linking Leads to Improved Fluorescent Probes of Cell Wall Synthesis. Journal of the American Chemical Society 2014, 136 (31), 10874-10877.

(34) Helassa, N.; Vollmer, W.; Breukink, E.; Vernet, T.; Zapun, A. The membrane anchor of penicillin-binding protein PBP2a from Streptococcus pneumoniae influences peptidoglycan chain length. The FEBS journal 2012, 279 (11), 2071-2081.

130 (35) Zapun, A.; Philippe, J.; Abrahams, K. A.; Signor, L.; Roper, D. I.; Breukink, E.; Vernet, T. In vitro Reconstitution of Peptidoglycan Assembly from the Gram-Positive Pathogen Streptococcus pneumoniae. ACS Chemical Biology 2013, 8 (12), 2688-2696.

(36) Sabala, I.; Jagielska, E.; Bardelang, P. T.; Czapinska, H.; Dahms, S. O.; Sharpe, J. A.; James, R.; Than, M. E.; Thomas, N. R.; Bochtler, M. Crystal structure of the antimicrobial peptidase lysostaphin from Staphylococcus simulans. The FEBS journal 2014, 281 (18), 4112-4122.

(37) Nadarajah, J.; Lee, M. J. S.; Louie, L.; Jacob, L.; Simor, A. E.; Louie, M.; McGavin, M. J. Identification of different clonal complexes and diverse amino acid substitutions in penicillin-binding protein 2 (PBP2) associated with borderline oxacillin resistance in Canadian Staphylococcus aureus isolates. Journal of Medical Microbiology 2006, 55 (12), 1675-1683.

(38) Fishovitz, J.; Rojas-Altuve, A.; Otero, L. H.; Dawley, M.; Carrasco-López, C.; Chang, M.; Hermoso, J. A.; Mobashery, S. Disruption of Allosteric Response as an Unprecedented Mechanism of Resistance to Antibiotics. Journal of the American Chemical Society 2014, 136 (28), 9814-9817.

(39) Bobba, S.; Ponnaluri, V. K.; Mukherji, M.; Gutheil, W. G. Microtiter plate-based assay for inhibitors of penicillin-binding protein 2a from methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 2011, 55 (6), 2783-2787.

(40) Lovering, A. L.; Gretes, M. C.; Safadi, S. S.; Danel, F.; de Castro, L.; Page, M. G. P.; Strynadka, N. C. J. Structural Insights into the Anti-methicillin-resistant Staphylococcus aureus (MRSA) Activity of Ceftobiprole. Journal of Biological Chemistry 2012, 287 (38), 32096-32102.

(41) Davies, T. A.; Page, M. G. P.; Shang, W.; Andrew, T.; Kania, M.; Bush, K. Binding of ceftobiprole and comparators to the penicillin-binding proteins of Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and Streptococcus pneumoniae. Antimicrobial Agents and Chemotherapy 2007, 51 (7), 2621.

(42) Moisan, H.; Pruneau, M.; Malouin, F. Binding of ceftaroline to penicillin-binding proteins of Staphylococcus aureus and Streptococcus pneumoniae. Journal of Antimicrobial Chemotherapy 2010, 65 (4), 713-716.

(43) Henze, U.; Sidow, T.; Wecke, J.; Labischinski, H.; Berger-Bächi, B. Influence of femB on methicillin resistance and peptidoglycan metabolism in Staphylococcus aureus. Journal of Bacteriology 1993, 175 (6), 1612.

131 (44) Maidhof, H.; Reinicke, B.; Blumel, P.; Berger-Bachi, B.; Labischinski, H. femA, which encodes a factor essential for expression of methicillin resistance, affects glycine content of peptidoglycan in methicillin-resistant and methicillin-susceptible Staphylococcus aureus strains. J Bacteriol 1991, 173 (11), 3507-3513.

(45) Tschierske, M.; Ehlert, K.; Strandén, A. M.; Berger-Bächi, B. Lif, the lysostaphin immunity factor, complements FemB in staphylococcal peptidoglycan interpeptide bridge formation. FEMS Microbiology Letters 1997, 153 (2), 261-264.

(46) Stranden, A. M.; Ehlert, K.; Labischinski, H.; Berger-Bachi, B. Cell wall monoglycine cross-bridges and methicillin hypersusceptibility in a femAB null mutant of methicillin- resistant Staphylococcus aureus. J Bacteriol 1997, 179 (1), 9-16.

(47) Ruzin, A.; Severin, A.; Ritacco, F.; Tabei, K.; Singh, G.; Bradford, P. A.; Siegel, M. M.; Projan, S. J.; Shlaes, D. M. Further evidence that a cell wall precursor [C(55)-MurNAc- (peptide)-GlcNAc] serves as an acceptor in a sorting reaction. J Bacteriol 2002, 184 (8), 2141-2147.

(48) Ton-That, H.; Labischinski, H.; Berger-Bachi, B.; Schneewind, O. Anchor structure of staphylococcal surface proteins. III. Role of the FemA, FemB, and FemX factors in anchoring surface proteins to the bacterial cell wall. The Journal of biological chemistry 1998, 273 (44), 29143-29149.

(49) Koyama, N.; Tokura, Y.; Münch, D.; Sahl, H.-G.; Schneider, T.; Shibagaki, Y.; Ikeda, H.; Tomoda, H. The Nonantibiotic Small Molecule Cyslabdan Enhances the Potency of β-Lactams against MRSA by Inhibiting Pentaglycine Interpeptide Bridge Synthesis. PLOS ONE 2012, 7 (11), e48981.

(50) Santa Maria, J. P.; Sadaka, A.; Moussa, S. H.; Brown, S.; Zhang, Y. J.; Rubin, E. J.; Gilmore, M. S.; Walker, S. Compound-gene interaction mapping reveals distinct roles for Staphylococcus aureus teichoic acids. Proceedings of the National Academy of Sciences 2014, 111 (34), 12510.

(51) Kuhner, D.; Stahl, M.; Demircioglu, D. D.; Bertsche, U. From cells to muropeptide structures in 24 h: peptidoglycan mapping by UPLC-MS. Scientific reports 2014, 4, 7494.

132

CHAPTER FOUR: Conclusions and Future Directions

133 4.1 Conclusions

In this thesis, I reported a number of biochemical tools that will enable a better understanding of how the bacterial cell wall is synthesized and inhibited:

1. A simple method to accumulate and isolate Lipid II (Chapter 2.2)

2. A one-step technique for purifying Park nucleotide (Chapter 2.3)

3. A direct assay of transpeptidase inhibition by beta-lactams (Chapter 3.6)

The accumulation and isolation allows us to obtain a practical quantity of Lipid II, a synthetically-challenging molecule that was previously available to only a handful of labs around the world, in just about a day of work. Vancomycin bead provides an easy access to milligrams of

Park nucleotide, a by-product of Lipid II accumulation, which can be converted into Lipid II in one coupled enzymatic reaction. Our direct transpeptidase assay allows us to see an inhibition of transpeptidase activity directly.

These tools established a platform for discovering the next generation of beta-lactam antibiotics. A new cephalosporin compound being developed against MRSA can be tested directly against PBP2a. Resistance-conferring mutation in clinical isolates can be validated. The transpeptidase assay can also be used to find compound that triggers PBP2a conformation change without inhibiting it directly. Such compounds will likely render MRSA susceptible to currently available beta-lactams.

I also used these tools to reveal the biological function of S. aureus PBPs when the S. aureus could not produce its native Gly5-Lipid II. Unexpectedly, I found that Gly3-peptidoglycan can be crosslinked by PBP2a but almost exclusively hydrolyzed by PBP4. The result suggests that

FemA, the enzyme that produces Gly3-Lipid II, is an ideal target for beta-lactam potentiators.

134 The reported tools have a few limitations. The accumulation of Lipid II requires bacteria that can be cultured in a lab. The bacteria needs to have a sufficiently large pool of cellular Lipid

II. Its peptidoglycan synthesis must be susceptible to inhibition by one of the chemical probes.

Most Gram-negative positive bacteria are naturally resistant to vancomycin and moenomycin due to the outer membrane barrier and thus require a different accumulation strategy. We were fortunate to be able to stop PG synthesis in E. coli since E. coli is genetically tractable. The outer membrane of Mycobacteria might present a similar problem. Many of these bacteria however share a simple stem peptide D-Ala-D-Ala-m-DAP-D-iso-Glu-L-Ala stem peptide which we already have an access too. Gram-positive bacteria which have a more elaborate stem peptide are generally susceptible to vancomycin and moenomycin.

The direct transpeptidase assay requires a sufficient amount of the enzyme for a detectable labeling of peptidoglycan. In our assay, we use 1 µM of PBP2 or PBP2a. This means that our assay would not be able to distinguish the potency of sub-micromolar inhibitors. While we can use less of PBP2 due to its highly active TP domain, we cannot decrease the concentration of PBP2a much lower. The assay does not distinguish between PGT domain inhibitors (e.g. moenomycin), TP domain inhibitors (e.g. beta-lactams), or Lipid II binder (e.g. vancomycin).

4.2 Future directions

PBP2 and PBP2a are proposed to work cooperatively to synthesize peptidoglycan in the presence of beta-lactams. The tools reported here allow us to reconstitute the activity of the two enzymes together. In fact I already did used PBP2 S398G which has an inactive TP domain with

PBP2a. Our current reconstitution however rely on soluble enzymes lacking the transmembrane helices but the transmembrane helix of both proteins may be important for protein-protein

135 interaction. Functional membrane microdomains (or lipid rafts) may also be necessary for the interaction. More experiments are needed to show that the cooperativity in vitro.

The transpeptidase assay is good for validating an inhibitor of the PBPs. Its throughput however is still quite low. A higher throughput assay is needed for screening a new inhibitor. 96- well filter plates could perhaps be use to separate Lipid II from highly crosslinked and labeled peptidoglycan, allowing straightforward detection of transpeptidase inhibition. D-amino acid incorporation could be exploited to install two fluorescent probes for Forster resonance energy transfer (FRET). Inhibition of transpeptidase would result in no FRET signal.

Our Lipid II isolation method and the enzymatic synthesis of Lipid II from isolated Park nucleotide provide an easy access of Lipid II and its analogues. The PG substrates could be used recently discovered PG biosynthetic enzymes such RodA, FtsW, and MurJ. Little is currently known about their substrate preferences and how they interact with Lipid II.

136 APPENDIX

Supplementary Table 1. Primers used in this study. Sequence (5’-3’) Primer Name F’pET42a_PBP2 AGGAGATATACATATGAAAGCACCTGCTTTTACCGAAGC

R’pET42a_PBP2 GTGGTGCTCGAGAGATTGTTGAGATCTAGTATTGTTATTTGA TTGTGCAGT F’pET42a ATCTCAACAATCTCTCGAGCACCACCACC

R’pET42a AGGTGCTTTCATATGTATATCTCCTTCTTAAAGTTAAACAAA ATTATTTC F’PBP2_S398G CAACAGATCCTCACCCTACTGGTGGATCTTTAAAACCTTTCT TAGCGTAT R’PBP2_S398G ATACGCTAAGAAAGGTTTTAAAGATCCACCAGTAGGGTGAG GATCTGTTG F’PBP2_P458L GACAAAGTTTCAATATCCTAGCTTTAAAAG

R’PBP2_P458L CTTTTAAAGCTAGGATATTGAAACTTTGTC

PBP2a-S403A_FW GATTACAACTTCACCAGGTGCAACTCAAAAAATATTAAC

PBP2a-S403A_RC GTTAATATTTTTTGAGTTGCACCTGGTGAAGTTGTAATC

PBP2a-K406A_FW CTTCACCAGGTTCAACTCAAGCAATATTAACAGC

PBP2a-K406A_RC GCTGTTAATATTGCTTGAGTTGAACCTGGTGAAG

137 Supplementary Table 2: Bacterial strains used in this study.

Strain Characteristics Source UT34-2 NCTC 8325 mecA Ω2006 (femB::Tn551) (1) BB255 NCTC 8325 (1) UK17 NCTC 8325 mecAΔfemA (ochre) (1) MW2 CA-MRSA, wild-type strain (2) MW2 ∆pbp4 ∆pbp4 in-frame deletion mutant of parental strain MW2 (3) MW2 ∆pbp4::pbp4 ∆pbp4 complemented with pbp4 with pMAD cycling (3) JVS-100 MW2 ∆femB This study JVS-101 MW2 ∆pbp4∆femB This study JVS-102 MW2 ∆femB ∆pbp4::pbp4 This study JVS-104 NCTC 8325 ∆femB This study

Supplementary Table 3. The ratio of crosslinking to hydrolysis for PBP2a, PBP2, and PBP4 with three substrates in all three buffer conditions.

The reaction buffers highlighted here in yellow were used to report the ratios in Figure 3 of the main text. The PBP2a buffer contains 50 mM HEPES, 2.5 mM MgCl2, pH 6.5, TGase buffer contains 50 mM HEPES, 10 mM CaCl2, pH 7.5, and MTG buffer contains 12.5 mM HEPES, 2 mM MnCl2, 0.25 mM tween-80, pH 7.5. Data represent averaged values of two (white boxes) or three experimental results (yellow boxes).

138

Supplementary Figure 1. Sufficient quantities of Lipid II can be isolated from bacteria. S. aureus Lipid II was quantified by two orthogonal methods: western blot analysis of biotinylated Lipid II (b) and LC/MS analysis of delipidated Lipid II (c). The estimated yields by both methods agree. LC/MS analysis of delipidated Lipid II was used to quantify B. subtilis and E. coli Lipid II (d-e). Structures of delipidated Lipid II species are shown (a).

139

Supplementary Figure 2. Quantification of Lipid II using PBP5 degradation and Edman reagent. A standard curve of integrated extracted chromatogram of functionalized D-Ala at various concentration (top) and an estimation of unknown concentration of Lipid II using the curve (bottom).

140

Supplementary Figure 3. Reconstitution of crosslinked peptidoglycan by S. aureus PBP2 using native Lipid II. (a) Extracted ion chromatogram (EIC) of muropeptide products. Peak A is the monomeric muropeptide, peak B is the crosslinked dimer, and peak C is the crosslinked trimer. High-resolution mass spectra of peak A and B are close to the theoretical mass spectra. The following ions were extracted: A: 1253.5856 (M+1), B: 1209.0617 ((M+2)/2), C: 1194.2204 ((M+3)/3). (b-c) Reaction with S. aureus PBP2S398G (TP inactive mutant) or SgtB, a monofunctional PGT does not yield crosslinked muropeptides.

141

Supplementary Figure 4. LC/MS extracted ion chromatogram of S. aureus PBP2a K406A and Gly5-Lipid II shows no crosslinking and hydrolysis activity. PBP2a K406A was reported to undergo extremely sluggish acylation.4

142

Supplementary Figure 5. A direct transpeptidase activity assay enables characterization of inhibitory potencies of different beta-lactams. (a) Western blot of crosslinked peptidoglycan produced by PBP2 with (left) requires active transpeptidase activity (middle-right). Product detection was enabled by BDL incorporation during PBP2 reaction. (b) Structures of beta- lactams examined in c-d. (c) Cefradine does not inhibit PBP2 activity up to the highest concentration tested in the experiment; whereas the other beta-lactams show potent inhibition. (d) The cefitzoxime-resistant mutant protein, PBP2P458L, shows no notable resistance to oxacillin compared to wild-type PBP2. For all experiments, 1 µM of enzyme was used.

143

Supplementary Figure 6. Time-course analysis of PBP2 reaction with Gly5-Lipid II in TGase buffer shows that the ratio of crosslinking to hydrolysis remains relatively constant over time. Data represent averaged values of two experimental results.

References

(1) Sharif, S.; Kim, S. J.; Labischinski, H.; Chen, J.; Schaefer, J. Uniformity of glycyl bridge lengths in the mature cell walls of fem mutants of methicillin-resistant Staphylococcus aureus. J. Bacteriol. 2013, 195 (7), 1421-1427.

(2) Baba, T.; Takeuchi, F.; Kuroda, M.; Yuzawa, H.; Aoki, K.; Oguchi, A.; Nagai, Y.; Iwama, N.; Asano, K.; Naimi, T.; Kuroda, H.; Cui, L.; Yamamoto, K.; Hiramatsu, K. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 2002, 359 (9320), 1819-1827.

(3) Memmi, G.; Filipe, S. R.; Pinho, M. G.; Fu, Z.; Cheung, A. Staphylococcus aureus PBP4 Is Essential for β-Lactam Resistance in Community-Acquired Methicillin-Resistant Strains. Antimicrob. Agents Chemother. 2008, 52 (11), 3955-3966.

(4) Fuda, C.; Suvorov, M.; Vakulenko, S. B.; Mobashery, S. The Basis for Resistance to β- Lactam Antibiotics by Penicillin-binding Protein 2a of Methicillin-resistant Staphylococcus aureus. J. Biol. Chem. 2004, 279 (39), 40802-40806.

144