Naturally Occurring Antimicrobial Peptides for Enhancing Food Safety and Protecting the Public Against Emerging Antibiotic-Resistant Pathogens

Home , Daptomycin, Intralytix, Paenibacillus polymyxa

Naturally Occurring Antimicrobial Peptides for Enhancing Food Safety and Protecting the Public against Emerging Antibiotic-resistant Pathogens

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the Ohio State University

En Huang

Graduate Program in Food Science and Nutrition

The Ohio State University

2013

Dissertation Committee:

Dr. Ahmed E. Yousef, Advisor

Dr. V.M. Balasubramaniam

Dr. Jiyoung Lee

Dr. Luis Rodriguez-Saona

Copyrighted by

En Huang

2013

Abstract

Food spoilage is a complex process in which foods become unsafe or undesirable for human consumption. Microbial spoilage causes food deterioration that may lead to slime and/or off-odors and off-flavors. In addition, human pathogens can be transmitted by food vehicles and the consumption of contaminated foods may lead to foodborne illnesses. The

Centers for Disease Control and Prevention (CDC) estimated that foodborne pathogens caused 9.4 million illnesses, 55,961 hospitalizations and 1,351 deaths annually in the United

States. Many food preservation and processing technologies have been developed to minimize the microbiological hazards in foods. Conventional chemical preservatives such as nitrites, and natural antimicrobial agents, such as nisin and essential oils, are used to preserve foods by preventing or retarding the growth of foodborne pathogens and spoilage microorganisms. The increasing use of nisin by food processors has promoted many investigators to search for other natural antimicrobial peptides with improved properties.

The widespread of antibiotic-resistant pathogens has become a global public health concern. Most of current antibiotics in the market are derived from the chemical scaffolds discovered between mid-1930s and early 1960s. After the “golden era” of discovery, only four new classes of antibiotics were introduced to the market in the past 40 years. The emergence of drug resistant pathogenic strains has increased mortality rate due to infections

with these particular pathogens. For example, the methicillin-resistant Staphylococcus aureus

(MRSA) caused ~19,000 deaths each year in the United States. Therefore, new and potent antibiotics are urgently needed against emerging antibiotic-resistant pathogens. Peptide antibiotics, including polymyxins and daptomycin, have received great attention in recent years for treating infections caused by drug-resistant pathogens.

The overall objective of this study was to search for safe and effective natural antimicrobial agents from microbial sources. To better use the newly found antimicrobials, we aimed to elucidate the chemical structure of the new compounds, examine the mechennisms of action and identify the biosynthetic pathway of the new antimicrobial agents.

Additionally, we investigated the efficacy of of new discovered compounds using different model food systems.

In this current study, we described three antimicrobial peptides, including a nonribosomal peptide, paenibacterin, and two ribosomal peptides, paenibacillin and enterocin

RM6. Paenibacterin is a lipopeptide antibiotic that is active against both Gram-negative and

Gram-positive bacteria, including antibiotic-resistant pathogens. Paenibacterin is a promising antibiotic scaffold for developing new antibiotics targeting drug-resistant pathogens. On the other hand, paenibacillin and enterocin RM6 are anti-Gram positive agents that belong to the bacteriocin family, which are ribosomally synthesized peptides. Paenibacillin is lanthionine- containing lantibiotic (class I bacteriocin) whereas enterocin RM6 is a cyclic unmodified class II bacteriocin. Paenibacillin and enterocin RM6 are naturally occurring antimicrobial peptides with great potential for food preservation.

Paenibacterin is produced by a soil isolate, Paenibacillus thiaminolyticus OSY-SE.

iii

The compound was extracted from cells with acetonitrile and purified to homogeneity by high performance liquid chromatography (HPLC). The chemical structure of paenibacterin was elucidated using mass spectrometry (MS) and nuclear magnetic resonance (NMR).

Paenibacterin is a cyclic lipopeptide consisting of 13 amino acids and an N-terminal C15 fatty acyl (FA) chain. The deduced sequence is: FA-Orn-Val-Thr-Orn-Ser-Val-Lys-Ser-Ile-

Pro-Val-Lys-Ile. The carboxyl terminal Ile is connected with Thr by an ester linkage forming a macrolactone ring. The mechanisms of action of paenibacterin involve direct cell membrane damage and indirect oxidative cellular damage. Paenibacterin is a cationic lipopeptide with four positive charges. The electrostatic interaction between paenibacterin and LPS can displace the divalent cations on the LPS network and promote the uptake of paenibacterin. The cytoplasmic membrane is the direct target of paenibacterin. Paenibacterin depolarized cell membrane, triggered K+ release and increased the uptake of hydrophobic nucleic acid stain, propidium iodide. In addition, paenibacterin led to production of hydroxyl radicals in bacterial cells. The radical scavenger, thiourea, and the iron chelator, 2, 2’- dipyridyl, reduced the killing effect of paenibacterin against Escherichia coli and

Staphylococcus aureus.

The presence of non-proteinogeneic amino acids (Orn, ornithine) in the peptide sequence suggested that paenibacterin is synthesized by nonribosomal synthetases. In order to determine the genes for paenibacterin biosynthesis, we sequenced the whole genome of the producer strain using the next-generation sequencing technology. The gene cluster was identified within 52-kb region, encoding three non-ribosomal peptide synthetases (PbtA,

PbtB and PbtC) and two ABC-transporters (PbtD and PbtE). As deduced from the sequence

data, each PbtA and PbtB enzyme consists of five modules, whereas PbtC is composed of three modules. Each of the 13 modules assembles one amino acid into the paenibacterin peptide. Sequence analysis revealed that Orn1, Orn4, Lys7 and Ser8 in paenibacterin may have

D-configuration. The absolute configuration of two ornithine residues was confirmed by chirality analysis using Marfey's reagents. In addition, the substrate specificities of selected adenylation domains were confirmed by overexpression in Escherichia coli and in vitro protein function analyses.

The lantibiotic, paenibacillin, was discovered by our group and its chemical structure has been elucidated in 2007-2008. In the current study, we investigated the mechanism of action and biosynthesis of this lantibiotic. Paenibacillin depolarized cell membrane and triggered potassium ions efflux, leading to cell death. The gene cluster of paenibacillin in

Paenibacillus polymyxa OSY-DF was identified using a PCR-based method combined with whole genome sequencing. The paenibacillin gene cluster (11.7 kb) consisted of 11 open reading frames (ORFs) encoding proteins for production, modification, regulation, immunity and transportation of the lantibiotic. Disruption of the lantibiotic dehydratase (PaenB) by targeted mutagenesis completely eliminated the production of paenibacillin.

Additionally, we tested the efficacy of paenibacillin against Listeria in meat products.

The crude extract of paenibacillin inhibited or delayed the growth of Listeria monocytogenes in sausage and irradiated ground beef products.

Enterocin RM6 is produced by Enterococcus faecalis isolated from raw milk. In this study, we purified enterocin RM6 by HPLC and identified its chemical structure by mass spectrometry. This compound is a 70-residue cyclic peptide with a head-to-tail linkage

between the N-terminal methionine and C-termianl tryptophan. The peptide sequence was further confirmed by sequencing the structural gene of the peptide. Enterocin RM6 (final concentration, 80 AU/ml) caused a rapid 4-log reduction of L. monocytogenes in cottage cheese within 30 min, and no viable cells were detected after 26 hrs of treatment. Therefore, enterocin RM6 is potentially useful in inhibiting L. monocytogenes in foods.

Acknowledgment

I would like to offer my sincerest thanks to my advisor Dr. Ahmed E. Yousef for his constant support and guidance throughout my study at the Ohio State University. His valuable instruction and encouragement have allowed me to grow to a better researcher in the field of food microbial safety. His passion for research and teaching will have a life-long influence on my future career. To my committee members, Dr. V.M. Balasubramaniam, Dr.

Jiyoung Lee and Dr. Luis Rodriguez-Saona, thank you all for your suggestions, support and patience.

I would also like to thank all current and past members in the Yousef’s group for their friendship and kind help: Dr. Jennifer Perry, Dr. Yaoqi Guo, Dr. Yan Yuan, Dr. Yoon Chung,

Dr. Jin-Gab Kim, Dr. Joy White, Dr. Zengguo He, David Kasler, Mustafa Yesil, Jiang He,

Greg Culbertson, Xu Yang, Amrish Chawla, Seth Costello, Goksel Tirpanci, Marilia Pena,

Joe Jones, and Joshua Kim. Special thanks to Jennifer for making the birthday’s cake for every group member. To my colleague, Yaoqi Guo: we have been working together in the lab for two years, sharing the happiness of new discovery and encouraging each other when things went wrong. I am also grateful to Dr. Liwen Zhang for mass spectrometry analysis and to Dr. Chunhua Yuan for NMR analysis. I would like to thank Dr. Charles Brooks for his suggestions and allowing me to use the fluorometer in his lab.

I am especially grateful to my father, Qigao Huang, my mother Xiaozhen Song, my

vii

brother, Chunhua Huang, and my sister Yaqin Huang for their love and support. Lastly, I thank my wife Ruilan for her love, support and motivation. Thank you for making home the best place to live.

viii

Vita

September 7, 1982……………………………………….Born, Yangjiang, Guangdong, China

2004………..……….…….………….……………………B.S., Food Science and Technology, Southwest University, China

2007………..…………………….…………………………...M.S., Fermentation Engineering, South China University of Technology

2009………..….……………………………………………….....M.S., Plant and Soil Science, Southern Illinois University

2009-present………………………………………...Graduate research and teaching associate, The Ohio State University

Publications

Huang E. and Yousef A.E. 2012. Draft Genome Sequence of Paenibacillus polymyxa OSY- DF That Coproduces a Lantibiotic, Paenibacillin, and Polymyxin E1. J. Bacteriol. 194: 4739- 4740.

Guo, Y., Huang E., Yuan C., Zhang L., and Yousef A.E. 2012. Isolation of a Paenibacillus sp. strain and structural elucidation of its broad-spectrum lipopeptide antibiotic. Appl. Environ. Microbiol. 78: 3156-3165.

Huang E., Guo, Y., and Yousef, A.E. 2012. Draft Genome Sequence of Paenibacillus sp. OSY-SE, a Bacterium Produces a Novel Broad-spectrum Lipopeptide Antibiotic, Paenibacterin. J. Bacteriol. 194: 6306

Fields of Study

Major Field: Food Science and Nutrition

Table of Contents Abstract ...... ii Acknowledgment ...... vii Vita ...... ix List of Tables ...... xv List of Figures ...... xvi Chapter 1: Literature Review ...... 1 1.1. Protection of food safety ...... 1 1.1.1. Food spoilage and foodborne illnesses ...... 1 1.1.2. Physical preservation technologies ...... 2 1.1.3. Chemical preservatives and naturally occurring antimicrobial agents ...... 7 1.1.4. Biopreservation using bacteriophages ...... 9 1.2. Antibiotics: the miracle drugs ...... 10 1.2.1. Action of antibiotics: from targets to network ...... 11 1.2.2. Antibiotic resistance ...... 12 1.2.3. New antibiotics are urgently needed ...... 13 1.3. Natural antimicrobial peptides ...... 14 1.3.1. Lantibiotics: ribosomally-synthesized peptides ...... 15 1.3.2. Cyclic lipopeptides: nonribosomal peptides ...... 23 1.4. Application of next-generation sequencing in food microbiology ...... 29 1.4.1. Community profiling during fermentation ...... 29 1.4.2. NGS-assisted process optimization ...... 31 1.4.3. RNA-sequencing (RNA-seq) Transcriptomics ...... 32 1.4.4. Whole genome sequencing typing for outbreak investigation ...... 33 References ...... 35 Chapter 2 Isolation of a Paenibacillus thiaminolyticus Strain and Structural Elucidation of Its Broad- spectrum Antibacterial Agent, Paenibacterin ...... 46 Abstract ...... 46 2.1. Introduction ...... 47 2.2. Materials and methods ...... 48 x

2.2.1. Strain screening ...... 48 2.2.2. Strain identification ...... 49 2.2.3. Isolation and purification of antimicrobial agents ...... 51 2.2.4. Antimicrobial activity determination ...... 52 2.2.5. Sensitivity to heat, pH and enzymes ...... 52 2.2.6. Alkaline hydrolysis ...... 53 2.2.7. MALDI-TOF MS analysis ...... 53 2.2.8. Quadrupole-time of flight MS/MS ...... 54 2.2.9. NMR analysis ...... 54 2.2.10. GC/MS analysis for confirmation of acyl moiety ...... 55 2.2.11. LC/MS/MS analysis ...... 56 2.3. Results ...... 57 2.3.1. Isolation and identification of an antimicrobial-producing strain ...... 57 2.3.2. Purification of antimicrobial agents produced by Paenibacillus OSY-SE ...... 59 2.3.3. Antimicrobial spectrum and stability ...... 61 2.3.4. Amino acid sequence of paenibacterin ...... 63 2.3.5. Linkage elucidation ...... 68 2.3.6. Determination of the acyl moiety ...... 69 2.3.7. Structure confirmation by GC/MS and LC/MS/MS ...... 72 2.4. Discussion ...... 81 References ...... 84 Chapter 3 Paenibacterin Biosynthesis in Paenibacillus thiaminolyticus: Identification of the Synthetase Gene Cluster by Whole Genome Sequencing ...... 88 Abstract ...... 88 3.1. Introduction ...... 89 3.2. Materials and Methods ...... 90 3.2.1. Strains and media ...... 90 3.2.2. Whole genome sequencing of producer strain ...... 91 3.2.3. Mining the paenibacterin gene cluster in the whole genome ...... 92 3.2.4. Determination of the absolute configuration of amino acids in paenibacterin ...... 93 3.2.5. Determination of adenylation domain substrate specificity ...... 94 3.2.6. Amino acid specificities of purified A-domains ...... 98 3.3. Results ...... 99 3.3.1. Whole genome sequencing and bioinformatic analyses ...... 99 xi

3.3.2. Organization of the paenibacterin gene cluster (pbt)...... 100 3.3.3. Absolute configuration of amino acids in paenibacterin ...... 107 3.3.4. Cloning, expression, purification and functional analysis of A-domains ...... 109 3.4. Discussion ...... 111 References ...... 113 Chapter 4 Paenibacterin Exerts Its Antimicrobial Activity through Membrane Damage and Radical Formation ...... 118 Abstract ...... 118 4.1. Introduction ...... 119 4.2. Materials and Methods ...... 123 4.2.1. Effect of lipopolysaccharides (LPS) on paenibacterin activity ...... 123 4.2.2. Binding of paenibacterin to LPS and Gram-negative bacteria ...... 123 4.2.3. Membrane potential depolarization ...... 124 4.2.4. Potassium release assay ...... 125 4.2.5. Membrane permeability assay ...... 125 4.2.6. Radical production measurement ...... 126 4.2.7. Effect of thiourea and 2, 2'-dipyridyl on paenibacterin activity ...... 126 4.3. Results ...... 127 4.3.1. Lipopolysaccharides antagonize paenibacterin activity ...... 127 4.3.2. Paenibacterin binds to LPS components of outer membrane ...... 128 4.3.3. Paenibacterin depolarizes cytoplasmic membrane ...... 130 4.3.4. Paenibacterin induces intracellular potassium release...... 132 4.3.5. Paenibacterin increases cell membrane permeability ...... 133 4.3.6. Radical production measurement ...... 134 4.3.7. Paenibacterin inactivates cells via radical production ...... 135 4.4. Discussion ...... 137 References ...... 139 Chapter 5 Biosynthesis, Mode of Action and Application of Paenibacillin ...... 142 Abstract ...... 142 5.1. Introduction ...... 143 5.2. Materials and Methods ...... 144 5.2.1. Bacterial strains and growth conditions ...... 144 5.2.2. Minimum inhibitory concentration of paenibacillin ...... 145 5.2.3. Identification of the paenibacillin structural gene, paenA ...... 145 xii

5.2.4. Identification of paenB and paenP by genome walking technique ...... 147 5.2.5. Disruption of lantibiotic dehydratase gene, paenB ...... 147 5.2.6. Identification of the paen gene cluster by whole genome sequencing ...... 150 5.2.7. Mode of action of paenibacillin...... 151 5.2.8. Applications of paenibacillin in foods ...... 153 5.3. Results ...... 155 5.3.1. Minimum inhibitory concentration of paenibacillin ...... 155 5.3.2. Identification of paenibacillin genes paenA, paenP and paenB ...... 156 5.3.3. Disruption of the lantibiotic dehydratase gene, paenB ...... 158 5.3.4. Whole genome sequencing of P. polymyxa OSY-DF ...... 159 5.3.5. Organization of the paenibacillin gene cluster ...... 161 5.3.6. Paenibacillin targets cell membrane ...... 169 5.3.7. Purification of paenibacillin and its application in meat products ...... 172 5.4. Discussion ...... 175 References ...... 178 Chapter 6 Characterization of Enterocin RM6, A Bacteriocin from Enterococcus faecalis Inhibitory to Listeria monocytogenes in Cottage Cheese ...... 183 Abstract ...... 183 6.1. Introduction ...... 185 6.2. Materials and Methods ...... 186 6.2.1. Bacterial strains and media ...... 186 6.2.2. Strain identification by 16S rDNA sequencing ...... 186 6.2.3. Purification of enterocin RM6 from cultured broth ...... 187 6.2.4. Antimicrobial activity determination and inhibition spectrum ...... 189 6.2.5. ESI-MS and MS/MS analyses ...... 190 6.2.6. Structural gene of enterocin RM6 ...... 191 6.2.7. Efficacy of enterocin RM6 on Listeria in inoculated cottage cheese ...... 191 6.3. Results ...... 192 6.3.1. Strain identification ...... 192 6.3.2. Purification of enterocin RM6 ...... 192 6.3.3. Molecular weight and peptide sequence determination by MS and MS/MS ...... 194 6.3.4. Antimicrobial spectrum of enterocin ...... 198 6.3.5. Bactericidal effect of enterocin RM6 on Listeria in cottage cheese ...... 198 6.4. Discussions ...... 200 xiii

References ...... 207 Bibliography ...... 209

xiv

List of Tables

Table 1. Representative regulatory-approved (US-FDA) food antimicrobialsa ...... 7 Table 2. Regulatory-approved bacteriophage preparations ...... 10 Table 3. Classification and mechanism of action of antibioticsa ...... 12 Table 4. Classification of lanthionine-containing peptides ...... 17 Table 5. Selected lipopeptide-derived antibiotics ...... 25 Table 6. Biochemical properties of Paenibacillus OSY-SE...... 59 Table 7. Relative antimicrobial activity of purified paenibacterin against selected bacteria ...... 62 Table 8. Chemical shift assignments of peptidyl fragment of paenibacterin in aqueous solution at pH 4.5, 298.0 K...... 77 Table 9. Chemical shift assignments of fatty acyl chain of paenibacterin in aqueous solution at pH 4.5, 298.0 K...... 78

Table 10. Chemical shift assignments of peptidyl fragment of paenibacterin in methanol-d4, 298.0 K...... 79

Table 11. Chemical shift assignments of fatty acyl chain of paenibacterin in methanol-d4, 298.0 K...... 80 Table 12. Primers used in the study...... 93 Table 13. The biosynthetic cluster in P. thiaminolyticus OSY-SE responsible for paenibacterin synthesis ...... 100 Table 14. Conserved Motifs in adenylation (A), condensation (C), thiolation (T) and epimerization (E) domains from paenibacterin NRPS ...... 104 Table 15. List of primers used in the current study ...... 146 Table 16. Minimum inhibitory concentration (MICa) of antimicrobial agents...... 156 Table 17 Antimicrobial activity of the OSY-DF wild type and its mutant against two indicatorsa ...... 158 Table 18. Open reading frames (ORFs) of the putative paenibacillin gene cluster ...... 162 Table 19. Preparation of crude extract of enterocin RM6...... 193 Table 20. Antimicrobial spectrum of enterocin RM6...... 199

List of Figures

Figure 1. Structure of nisin and its unusual residues ...... 16 Figure 2. Chemical structure of daptomycin (A) and polymyxin B (B)...... 26 Figure 3. Schematic module of nonribosomal peptide synthetase ...... 28 Figure 4. Scanning electron microscope (SEM) examination of OSY-SE cells ...... 58 Figure 5. High performance liquid chromatography profile of the crude extract of OSY-SE cells ...... 60 Figure 6. MALDI-TOF MS analysis of paenibacterin and its linear form produced by alkaline hydrolysis...... 61 Figure 7. NMR analysis of the peptidyl fragment in the amide region...... 66 Figure 8. Elucidation of amino acid sequence and linkage of paenibacterin by HMBC...... 67 Figure 9. The tertiary structure of the peptide moiety of paenibacterin calculated from NMR constraints in aqueous solution...... 69 Figure 10. Fragmentation of b and y ion series of linearized paenibacterin, examined by MS/MS...... 71 Figure 11. 1D 13C NMR spectrum revealing iso- and anteiso- fatty acyl chain...... 71 Figure 12. Gas chromatography (GC) profile of fatty acid methyl esters from paenibacterin...... 73 Figure 13. MS/MS spectra of tryptic-digested products of paenibacterin...... 75 Figure 14. The molecular structure of paenibacterin; the dotted cycle indicated the macrolactone ring formed between Thr3 and Ile13...... 76 Figure 15. schematic presentations of cloning, expressing and purification of A- domain proteins...... 97 Figure 16. malachite green colorimetric assay of substrate specificity ...... 99 Figure 17. Schematic overview of paenibacterin biosynthesis in P. thiaminolyticus OSY-SE. The peptide core is assembled by three NRPSs PbtA, PbtB, PbtC...... 106 Figure 18. Chiral analyses of standard amino acids and constituent amino acids from paenibacterin using Marfey’s reagent...... 108 Figure 19. Cloning, expression and purification of adenylation domains...... 110 Figure 20. Determination of substrate specificity of purified A-domains by phosphate detection assay. .. 111

xvi

Figure 21. Chemical structures of paenibacterin and polymyxin B. The functional groups are colored as followed: red, hydrophobic domains; blue, positively charged residues; green, hydroxyl-containing amino acids...... 121 Figure 22. Effect of lipopolysaccharide (LPS) on the activity of paenibacterin (32 µg/ml) against E. coli ATCC 25922 ...... 128 Figure 23. BODIPY FL-labeled polymyxin B displacement assays...... 130 Figure 24. Effect of paenibacterin (paen), nisin vancomycin (van) and polymyxin E (pmx) on bacterial membrane potential...... 131 Figure 25. Effect of paenibacterin (paen), nisin, vancomycin (van) and polymyxin E (pmx) on K+ release...... 133 Figure 26/ Effect of paenibacterin on bacterial cell membrane permeability...... 134 Figure 27. Hydroxyl radical production in paenibacterin treated cells of E. coli ATCC 25922...... 135 Figure 28. Effect of dipyridyl (Dip, 450 μM) and thiourea (Thio, 100 mM) on the killing efficacy of paenibacterin (Pbt)...... 136 Figure 29. Chemical structure of paenibacillin ...... 144 Figure 30. Ion exchange chromatography and solid phase extraction for paenibacillin purification ...... 154 Figure 31. Alignment of paenibacillin with related lantibiotic prepropeptides...... 157 Figure 32. MALDI-MS analysis of paenibacillin production of Paenibacillus polymyxa OSY-DF wild type and ∆paenB mutant strains...... 159 Figure 33. Genes involved in paenibacillin biosynthesis...... 161 Figure 34. Post-translational modifications of paenibacillin...... 166 Figure 35. Alignment of putative signal precursor AgrD...... 168 Figure 36 Paenibacillin structural gene (paenA) and its preceding noncoding sequences...... 168 Figure 37. Effect of paenibacillin (Pln) and nisin (Nis) on membrane potential...... 170 Figure 38. Effect of paenibacillin (Pln) and nisin (Nis) on potassium release...... 171 Figure 39. Activity tests of elution fractions from ion exchange chromatography...... 172 Figure 40. Paenibacillin was separated from polymyxin E1 by gradient elution using the Macro-prep High S strong cationic support...... 173 Figure 41. Inhibition of Listeria monocytogenes Scott A in Vienna sausagea...... 174 Figure 42. Inhibition of Listeria monocytogenes Scott A by Paenibacillin in ground beef ...... 174 Figure 43. High performance liquid chromatography profile of the crude extract of enterocin RM6...... 193 Figure 44. ESI-MS analysis of enterocin RM6...... 196 Figure 45. Fragmentation of enterocin RM6 examined by MSMS...... 197 Figure 46 Structural gene of enterocin RM6 and the deduced peptide sequence...... 197 xvii

Figure 47. Effect of enterocin RM6 on survival of L. monocytogenes Scott A inoculated on cottage cheese using different mediaa...... 200

xviii

Chapter 1: Literature Review

1.1. Protection of food safety 1.1.1. Food spoilage and foodborne illnesses

Food spoilage is a complex process in which foods become unsafe or undesirable for human consumption. Microbial spoilage usually results in slime and/or off-odors and off- flavors (Gram et al., 2002). Food spoilage leads to considerable economic and environmental loss. It was estimated that over 10% of cereal grains and legumes, and 50% of vegetables and fruits were lost due to spoilage microorganisms in the developing countries (Sperber, 2009).

According to USDA-Economic Research Service (ERS), a large proportion of foods were lost in the United States by retailers, foodservice and consumers: fruits and vegetables

(19.6%), fluid milk (18.1%), grain products (15.2%), caloric sweeteners (12.4%), processed fruits and vegetables (8.6%), meat, poultry and fish (8.5%), and fat and oils (7.1%) (Kantor,

Lipton, Manchester, & Oliveira, 1997).

On the other hand, foodborne illnesses caused by consumption of contaminated foods have significant economic and public health impact. The Centers for Disease Control and

Prevention (CDC) estimated that 31 major pathogens caused 9.4 million foodborne illnesses,

55,961 hospitalizations and 1,351 deaths annually in the United States (Scallan et al., 2011).

Norovirus accounted for 58% of the foodborne illnesses, followed by Salmonella spp. (11%),

Clostridium perfrengens (10%) and Campylobacter spp. (9%). Salmonella spp. were the

leading causative agents of hospitalizations (35%), followed by norovirus (26%),

Campylobacter spp. (15%) and Toxoplasma gondii (8%). The leading causes of death were

Salmonella spp. (28%), T. gondii (24%), Listeria monocytogenes (19%) and norovirus (11%)

(Scallan et al., 2011). As a consequence, the health-related economic cost due to foodborne illnesses was as high as $77.7 billion annually in the U.S. (Scharff, 2012).

According to CDC-Foodborne Diseases Active Surveillance Network (FoodNet), the overall infections caused by six major bacterial pathogens (Campylobacter, Escherichia coli

O157, Listeria, Salmonella, Yersinia and Vibrio) decreased 23% in 2010 compared with

1996-1998 in the United States. Remarkably, infections due to E. coli O157 declined 44% and reached the 2010 national health objective target (≤ 1 case per 100, 000). However, the incidence of Salmonella remained at a high level in the past 15 years. Furthermore, Vibrio infections increased 115% from 1996 (CDC, 2011).

E. coli O157:H7 is the most well known Shiga toxin-producing E. coli (STEC). The

O157 STEC is responsible for 63,153 foodborne illnesses while non-O157 STEC causes

112,752 illnesses annually in the United States (Scallan et al., 2011). With the increasing concern regarding to non-O157 STEC, the USDA-Food Safety and Inspection Service (FSIS) intended to test the “big six” non-O157 STEC (O26, O45, O103, O111, O121, and O145) in beef manufacturing trim and ground beef products (USDA-FSIS, 2012).

1.1.2. Physical preservation technologies

Conventional and innovative thermal processing

Heat treatment is the most widely used technology to preserve foods. Heat inactivates microorganisms in foods through irreversible denaturation of proteins, nuclear acids or other

vital cellular components (Osaili, 2012). Thermal processing is safe and chemical-free, but high temperature processing usually results in texture changes and nutrition deterioration.

Traditional thermal processing technologies mainly include pasteurization, canning, extrusion, cooking and frying (Rahman, 2007a) . Recently, innovative electroheating processing technologies, including ohmic heating, microwave heating, and radio frequency (RF) heating, have received great interest. During microwave and RF prcessing, heat is generated in foods internally by molecular fractions when polar molecules trying to align themselves to the alternating electric field (Piyasena, Dussault, Koutchma, Ramaswamy, & Awuah, 2003).

Ohmic treatment is a direct electroheating process in which electric current is passed through the food product, generating heat within the foods as a result of electrical resistance of the food matrix (Osaili, 2012). Electroheating has been successfully used in cooking, roasting, drying, thawing, defrosting, pasteurization and sterilization (Osaili, 2012; Piyasena et al.,

2003)

Low temperature storage-chilling and freezing

Refrigeration and freezing is the essential part for food processing, transportation and storage. Refrigeration and freezing slow down the physical, chemical and biological changes in foods and therefore maintain the desirable properties of perishable food products. The shelf life of perishable foods can be extended for several days by refrigeration. Frozen foods stored at -18°C to -35°C have a much longer storage life.

Vacuum cooling is a rapid evaporative cooling process for moist and porous products through moisture removal under vacuum conditions (Zheng & Sun, 2004). Compared with conventional cooling, vacuum cooling provides a much more uniform cooling in a shorter

time. The major applications of vacuum cooling have extended from cooling of fresh produce to other food products, including fishery products, bakery products, cooked meats and ready meals (Zheng & Sun, 2004).

Vacuum packaging and modified-atmosphere packaging

In additional to storage temperature, gaseous composition of storage environment is an important factor that affects the storage life of food products. Vacuum packaging (VP) preserves food products by evacuating the air and packing the foods in a film of low oxygen permeability. Modified-atmosphere packaging (MAP) involves the enclosure of food products in a gas-barrier film, removal of air from the pack and replacement with a single gas or gas mixture (Blakistone, 1999). VP and MAP can extend the shelf life of perishable foods by retarding the growth of microorganisms. For example, most aerobic spoilage and pathogenic bacteria as well as molds are inhibited by 5% carbon dioxide (CO2) (Blakistone,

1999).

Control of water activity

Water activity (aw) is an intrinsic factor of food, which can be used to predict the stability of food products. The control of water activity in foods is a common preservation technique to extend the shelf life. Reduction of water activity can be achieved by drying, freezing, or by addition of solutes such as salt or sugars (Adams & Moss, 2008). Pathogenic bacteria, including the salt-tolerant Staphylococcus aureus, are inhibited when the water activity of the food product falls below 0.85 (Russell & Gould, 2003). The growth of yeasts and molds in foods is also inhibited at aw between 0.88 and 0.80; however, some osmophilic yeasts and xerophilic molds can grow at water activity as low as 0.65 (De Man, 1995).

High pressure processing

High pressure processing (HPP) involves using elevated pressure (50 to 1000 M Pa), with or without additional heating, to inactivate pathogenic and spoilage microorganisms in foods (Balasubramaniam & Farkas, 2008; Hogan, Kelly, & Sun, 2005). High-pressure pasteurization of foods can be achieved at 600 M Pa range for several minutes at ambient temperature. High-pressure pasteurized products, such as smoothies, guacamole, ready-meal components, oysters, ham, fruit juices and salsa, are commercially available in the United

States (Balasubramaniam & Farkas, 2008). Food quality related enzymes and bacterial spores can be inactivated by a pressure-thermal combination at elevated pressure (up to 800 M Pa) and temperature (90-120°C) (Balasubramaniam & Farkas, 2008).

Pulsed electric field

Pulsed electric field processing is an emerging nonthermal technology that involves exposing the liquid foods to high intensity short electric pulses (ms or μs) at low or moderate temperature (<60°C) . PEF treatment maintains the freshness of food products and preserves the heat-sensitive components in foods. PEF ruptures the cytoplasmic membrane of microorganisms, leading to the leakage of intracellular contents and cell death (Jeyamkondan,

Jayas, & Holley, 1999). Liquid foods, such as juices, milk products, liquid eggs, starch solutions, yogurts, syrups and soups are suitable for treatment with PEF (Sobrino-López &

Martín-Belloso, 2010).

Ultrasound

High-intensity ultrasound is one of the new preservation technologies. Ultrasonic waves kill microorganisms by localized heating, thinning of cell membrane, and production

of radicals (Chemat & Khan, 2011). The combination of ultrasound with heat or/and pressure is a promising method for rapid inactivation of vegetative cells, bacterial spores, and enzymes, such as peroxidase, endopolygalacturonases (PGs) and pectinmethylesterase (PME)

(Chemat & Khan, 2011; Rastogi, 2011).

Irradiation

Food irradiation is a physical process that involves exposing pre-packaged or bulk foods to ionizing radiation sources such as gamma rays from radioisotopes ( 60Co and 137Cs),

X-rays, or energetic electrons from particle accelerators (Farkas, 2006). Ionizing radiations inactivate microorganisms by damaging nucleic acids and by generating oxidative radicals which induce direct or indirect cellular damages (Farkas, 2006). Currently, more than 50 countries have approved the use of irradiation for food processing. The applications of food irradiation include disinfection of of foods, and inhibition of sprouting of garlic and potato.

In 2005, the total irradiated food in the world was 405,000 tons, comprising 46% spices and dry vegetables, 22% garlic and potato, 20% grains and fruits, and 8% meat and seafood

(Kume, Furuta, Todoriki, Uenoyama, & Kobayashi, 2009).

Ultraviolet and pulsed light

Ultraviolet (UV) light is a powerful nonionizing bactericidal agent that induces cross- linking of neighboring pyrimidine residues in the DNA strand. The formation of pyrimidine dimer blocks DNA replication and transcription, and thus compromises cellular functions and leads to cell death. UV light has been traditionally used for disinfection of air, surfaces and water (Guerrero-Beltrán & Barbosa-Cánovas, 2004). Recently, there has been an increased interest in using UV light to process liquid food (e.g. juices), and to reduce microbial load in

fruits and vegetables, meat products and shell eggs (Guerrero-Beltrán & Barbosa-Cánovas,

2004; Rahman, 2007b). The US-FDA has approved the use of UV light for processing juice products to reduce human pathogens and other microorganisms (Codes of Federal

Regulations, 21CFR179.39).

The classical UV treatment as discussed above works in a continuous mode. In contrast, pulsed light (PL) involves using pulses of high intense broad spectrum light (100 nm to 1100 nm) for rapid inactivation of microorganisms on food surfaces, equipments and packaging materials (Elmnasser et al., 2007; Gómez-López, Ragaert, Debevere, &

Devlieghere, 2007; Oms-Oliu, Martín-Belloso, & Soliva-Fortuny, 2010).

1.1.3. Chemical preservatives and naturally occurring antimicrobial agents

Some food additives are used to preserve foods by preventing or retarding chemical or biological deterioration (Davidson & Branen, 2005). Food antimicrobials are “substances used to preserve food by preventing growth of microorganisms and subsequent spoilage, including fungistats, mold and rope inhibitors” (Codes of Federal Regulations, 21CFR170.3).

Table 1. Representative regulatory-approved (US-FDA) food antimicrobialsa Antimicrobials Spectrum Examples of use CFR designation Weak lipophilic organic acids and esters Sorbic acid/ sorbates Yeasts, molds, Dressings, wines, 182.3089; 182.3795; bacteria beverages, cheeses 182.3640; 182.3225 Parabens Yeasts, molds, G+ Bakery product, 184.1490; 184.1670; bacteria beverages, pickles, 172.145 salad dressings Benzoic acid/ benzoates Yeasts, molds Beverages, dressings, 184.1021; 184.1733 margarine Other organic acids Acetic acid/ acetates Yeasts, bacteria Baked goods, sources, 184.1005; 184.1721; dairy products Diacetates Yeasts, bacteria Baked goods, soup, 184.1754 fats/oils, sauces, meat products Dehydroacetic acid Yeasts, bacteria Cut or peeled squash 172.130 7

Table 1 continued. Propionic acid/ Molds Bakery products, dairy 184.1081; 184.1221 propionates products Lactic acid/ lactates Bacteria Meats, fermented foods 184.1061; 184.1639; 184.1768; 184.1207 Inorganic anions Sulfites Yeasts, molds Dried fruits, wines 182.3739; 182.3766 182.3616; 182.3637 Nitrite/ nitrate Clostridium spp. Cured meats 172.160;172.170; 172.175;172.177 Natural occurring antimicrobials Lysozyme Clostridium spp. and Cheeses, cooked meats GRAS Notice No. other bacteria GRN 000064 Lactoperoxidase bacteria Yogurts GRAS Notice No. 000196 Lactoferrin bacteria Meats GRAS Notice No. 000067 Nisin Clostridium spp., Cheeses, canned meats GRAS Notice No. other G+ bacteria GRN 000065 Natamycin Molds Cheeses, fermented 172.155 meats a Modified from Davidson & Branen, 2005; Jay, Loessner, & Golden, 2005; Kussendrager & Van Hooijdonk, 2000; Russell & Gould, 2003.

Food antimicrobial agents can be classified into traditional chemical preservatives and naturally occurring antimicrobials. Table 1 shows a summary of the representative regulatory-approved (US-FDA) food antimicrobial agents. The selection of proper antimicrobial agents depends on many factors such as antimicrobial spectrum, properties of antimicrobial agents (e.g. solubility and polarity), food-related factors (e.g., pH, lipid content), and food processing and storage conditions. Sensory effect, cost of use and product labeling should be considered when choosing proper antimicrobials (Davidson & Branen,

2005).

In addition to FDA-approved preservatives, many other natural antimicrobial agents from animals, plants or microorganisms are proposed for use in foods. Examples of some promising natural antimicrobial agents include ovotransferrin from egg-white, chitosan from

crustaceans and arthropods, essential oils of spices and herbs, pediocin-like bacteriocin, and microbial fermentates of starter cultures (Juneja, Dwivedi, & Yan, 2012).

1.1.4. Biopreservation using bacteriophages

Bacteriophages (or phages) are viruses that specifically infect and multiply in bacteria.

They are ubiquitously distributed in nature and have been isolated from a variety of foods, such as meat products, seafood, fermented dairy products, fresh produce and mushrooms

(Fieseler, Loessner, & Hagens, 2011). Bacteriophages rely on their obligate bacterial hosts to replicate; therefore they are harmless to human, animals and plants (García, Rodríguez,

Rodríguez, & Martínez, 2010). Phages can be classified as lytic or lysogenic depending on their life cycles. Lytic phages strictly follow the lytic life cycle in which they reproduce themselves in the bacterial host and immediately release the phage progeny by destroying the host. Lysogenic phages, on the other hand, integrate the viral DNA into the bacterial chromosome and become the prophages. Prophages replicate as part of the bacterial host genome without immediately transcribing and making new phages. The prophages may be activated under adverse environmental conditions, multiplying in the bacteria and destroying the host cells (García, Martínez, & Rodríguez, 2011).

Lytic bacteriophages can be used as biocontrol agents against bacterial pathogens.

The use of bacteriophages in food safety and agriculture has received great attention in the past five years. The advantages of using bacteriophage over other antimicrobial agents are its strict host specificity and the self-replicating capacity. A single phage or mixed phage cocktail only kills specific pathogens without interfering with other desired bacteria (e.g. starter culture) in foods. Therefore, the consumption of bacteriophages in food products will

not harm the commensal microflora in human gastrointestinal tract (Fieseler et al., 2011;

García et al., 2010). The US-Environmental Protection Agency (EPA) has approved a bacteriophage cocktail (AgriPhage) for biological control of plant bacterial diseases. In addition, the US-FDA has approved the use of bacteriophage preparations as processing aid or food additives for human consumption (Table 2).

Table 2. Regulatory-approved bacteriophage preparations

Commercial Target microorganisms Uses Regulatory Bacteriophages designation AgriPhageTM , Xanthomonas campestris Biological control of EPA Reg. No. Omnilytics, Inc. pv. vesicatoria and diseases on tomatoes 67986-1 Pseudomonas syringae pv. and peppers tomato LISTEXTM, Listeria monocytogenes Processing aid in all FDA(GRAS EBI Food food products No. 000198) Safety ListShield™, Listeria monocytogenes Direct application onto 21CFR172.785 Intralytix, Inc. foods; surfaces in food EPA Reg. No. facilities 74234-1 EcoShield™, Escherichia coli O157:H7 Red meat parts and FCN No. 1018 Intralytix, Inc. trim prior to grinding

1.2. Antibiotics: the miracle drugs The discovery of antibiotics in the 20th century represents one of the key achievements in modern medicine. The use of antibiotics to treat infections has saved numerous lives and has greatly improved human health. Antibiotics are molecules that inhibit the growth of microorganisms (bacteriostatic) or cause bacterial cell death (bactericidal). Most antibiotics in therapeutic use are natural products from microorganisms, or their semi-synthetic derivatives with improved properties. Actinomycetes represent the major group of antibiotics-producing bacteria. On the other hand, there are three classes of synthetic antibiotics in clinical use: the sulfa drugs, quinolones and oxazolidinone (Walsh, 2003). The

global sales of antibiotics reached $42 billion in 2009. Three classes of antibiotics accounted for 65% of the global market: cephalosporins (29%), penicillins (19%) and fluoroquinolones

(17%) (Hamad, 2010).

1.2.1. Action of antibiotics: from targets to network

Antibiotics exert the antimicrobial activity by inhibiting essential cell functions. As shown in Table 3, antibiotics are categorized into different classes by their chemical structure and mechanisms of action. β-lactams and glycopeptides are cell wall biosynthesis inhibitors. β-lactam antibiotics inhibit the transpeptidases (known as penicillin-binding proteins, PBPs), whereas glycopeptides bind to the terminal dipeptide (D-Ala-D-Ala) of peptidoglycan precursor, blocking the synthesis cell wall.

The cytoplasmic metabolic activity of bacteria can be disrupted by antibiotics. Some antibiotics, including macrolides, tetracyclines, aminoglycosides and oxazolidinones, interfere with protein translation by binding to 30 or 50S subunits of the bacterial ribosome.

Other antibiotics target the synthesis of bacterial DNA or RNA. Fluoroquinolones inhibit bacterial DNA replication by the inhibiting the activity of topoisomerase II or IV, while rifamycins inhibit RNA polymerase and thus arrest RNA transcription. Synthetic sulfa drugs inhibit the synthesis of folic acid, which is essential for bacterial DNA synthesis. Lipopeptide antibiotics act not by inhibiting a specific enzyme, but rather by disruption the cell membrane

(Walsh, 2003).

Table 3. Classification and mechanism of action of antibioticsa

Antibiotic classes Mechanism of action Representative drugs β-lactams Block cell wall biosynthesis Cephalosporins, penicillins Glycopeptides Block cell wall biosynthesis Vancomycin, Teicoplanin Macrolides Inhibit protein synthesis Erythromycin, clarithromycin, azithromycin Tetracyclines Inhibit protein synthesis Tetracycline, tigicycline Aminoglycosides Inhibit protein synthesis Gentamicin, streptomycin, kanamycin, tobramycin Oxazolidinone Inhibit protein synthesis Linezolid Fluoroquinolones Block DNA replication Ciprofloxacin, ofloxacin Rifamycins Inhibit RNA polymerase Rifampin Sulfa drugs Block folic acid biosynthesis Sulfamethoxazole, trimethoprim Lipopeptides Cell membrane disruption Polymyxin, daptomycin a Modified form Walsh, 2003.

Recently, a common mechanism was proposed for antibiotic-mediated cell death following the interaction between drug and cell target. All major classes of bactericidal antibiotics kill both Gram-positive and Gram-negative bacteria through the production of the lethal hydroxyl radicals (Kohanski, Dwyer, Hayete, Lawrence, & Collins, 2007). The primary antibiotic-target interaction results in the depletion of NADH and stimulates the formation of superoxide. Superoxide can destabilize the iron-sulfur cluster and trigger the release of ferrous iron, which reduces hydrogen peroxide and generates hydroxyl radicals via

Fenton reaction. The formation of hydroxyl radicals can damage DNA, proteins and lipids, leading to cell death (Kohanski et al., 2007; Kohanski, Dwyer, & Collins, 2010).

1.2.2. Antibiotic resistance

The three main molecular mechanisms of antibiotic resistance are (i) efflux of antibiotics, (ii) inactivation of antibiotics, and (iii) modification of the drug target in bacteria

(Walsh, 2003). The active efflux pumps are transmembrane proteins, which exports antibiotics such that the drugs do not accumulate in the bacterial cytoplasm. Some bacterial

can directly inactivate antibiotics. For example, the most notorious enzymes β-lactamases can destroy penicillins and cephalosporins. Lastly, some resistant bacterial strains replace or modify the drug target to decrease the antibiotic sensitivity. For example, methylation of adenine in the 23S RNA in the 50S ribosome subunit contributes to the resistance to macrolides (Walsh, 2003).

The intensive use of antibiotics has dramatically increased the prevalence of resistant human pathogens. The selection pressure of antibiotics in the environments has promoted the wide spread of antibiotic resistant bacteria. Bacteria can evolve rapidly and acquire resistance via de novo mutations. The hypothesis was proved in an artificial device in laboratory. In a microfluidic device mimicking the natural bacterial niches, after exposure to ciprofloxacin for 10 hrs, Escherichia coli cells acquired resistance to the antibiotic tested. Genome analysis indicated that single-nucleotide-polymorphisms (SNPs) in the mutant strains contributed to ciprofloxacin resistance. One of the mutations occurred in gyrase A, which is the drug target of fluoroquinolones (Zhang et al., 2011). In addition to mutations, bacteria acquire resistance by horizontal gene transfer (HGT) through mobile genetic elements such as plasmids, transposons, naked DNA or bacteriophages. For example, the tet (M) genes for tetracycline resistance in bacteria are typically carried by the transposon Tn916 (Levy & Marshall, 2004).

1.2.3. New antibiotics are urgently needed

The emerging and rapid dissemination of antibiotic-resistant pathogens has become an ever-increasing global threat to public health. In 2009, the Infectious Diseases Society of

America (IDSA) identified a group of problematic antibiotic-resistant pathogens that escape the therapeutic effect of antimicrobial agents. The ‘ESKAPE’ pathogens are Enterococcus

faecium, Staphylococcus aureus, Klebsiella pneumonia, Acinetobacter baumannii,

Pseudomonas aereuginasa, and Enterobacter species (Boucher et al., 2009). The drug resistant strains usually lead to a higher mortality rate. For example, the methicillin-resistant

Staphylococcus aureus (MRSA) caused ~19,000 deaths each year in the United States

(Fischbach & Walsh, 2009). The carbapenems, which are stable to most β-lactamases, are considered as the last line of antibiotics for treating the drug resistant bacteria. However, the emergence of carbapenem-resistant Enterobacteriaceae compromised the therapeutic options

(Bonomo, 2011). Facing the shortage of effective antibiotic, physicians are forced to use some old antibiotics. Polymyxins, which were discovered 50 years ago and were discarded due to concerns of toxicity, are used to treat infections caused by multidrug-resistant Gram- negative pathogens (Landman, Georgescu, Martin, & Quale, 2008).

Most today’s antibiotics in the market are derived from the chemical scaffolds discovered between mid-1930s and early 1960s. After the “golden era” of discovery, there was a 40-year innovation gap in introducing new classes of antibiotics to the market. In the last decade, 20 antibiotics have been launched worldwide but only four of them belong to novel classes: oxazolidinones (linezolid), lipopeptides (daptomycin), mutilins (Retapamulin) and macrocyclics (Fidaxomicin) (Artsimovitch, Seddon, & Sears, 2012; Butler & Cooper,

2011). Currently, there are a total of ~40 compounds in the clinical trial stage, of which five are in phase-III and 22 are in phase-II (Butler & Cooper, 2011). However, there are very few antibiotics undergoing clinical trials targeting Gran-negative bacteria (Coates & Halls, 2012).

1.3. Natural antimicrobial peptides Antimicrobial peptide can be categorized into two classes: ribosomally synthesized

peptides and non-ribosomally synthesized peptides. Bacteriocins are ribosomally synthesized antimicrobial peptides produced by Gram-positive bacteria. The majority of bacteriocins fall into two categories: lanthionine-containing lantibiotics (class I) and unmodified bacteriocins

(class II). For example, nisin and lacticin 418 are lantibiotics while pediocin and enterocin are class II bacteriocins (Molly 2011). On the other hand, many peptide antibiotics, such as vancomycin and penicillin, are synthesized by the nonribosomal peptide synthetases (NRPSs)

(Walsh, 2003).

1.3.1. Lantibiotics: ribosomally-synthesized peptides

Structure and classification of lantibiotics

The term lantibiotics is the abbreviation of lanthionine-containing antibiotics (Schnell et al. 1988). Lantibiotics are ribosomally-synthesized peptides with formation of lanthionines during post-translational modifications (Willey et al. 2007). The unique residues in lantibiotics include dehydroalanine (Dha), didehydrobutyrine (Dhb), lanthionine and methyllanthionine. Dha and Dhb residues are dehydroamino acids derived from serine and threonine residues respectively after loss of one water molecule. The unsaturated Dha or Dhb residue can form a thioether bond with cysteine residue in the peptide, resulting in the unusual residues, lanthionine (Lan) or methyllanthionine (MeLan), respectively (Willey &

Van der Donk, 2007). The chemical structure of nisin and the unusual residues are shown in

Figure 1.

Figure 1. Structure of nisin and its unusual residues

Currently, over 60 lantibiotics with different size, structure or modes of action have been discovered (Kuipers et al. 2011). Lanthionine-containing peptides are divided into four classes (

Table 4) based on their biosynthetic pathway and biological activity (Rea, Ross,

Cotter, & Hill, 2011). Class I lantibiotics, such as nisin, have an elongated structure while class II lantibiotic have a more compact and globular structure. For class I lantibiotics, two distinct enzymes, LanB and LanC, are responsible for dehydration and formation of thioether linkages, respectively. For other classes, lantibiotics are modified by bifunctional enzymes

(LanM in class II, RamC/ LabKC in class III, and LanL in class IV) that exhibit dehydratase and cyclase activities (Rea et al., 2011; Willey & Van der Donk, 2007).

Biosynthesis of lantibiotics

The genes for lantibiotic synthesis, export, and self-immunity are generally found in a gene cluster and are designated by the generic locus symbol lan. The gene cluster are either

located on the chromosome (e.g. sublilin), plasmid (e.g. epidermin) or transposon (e.g. nisin)

(Kuipers, Rink, & Moll, 2011). The biosynthesis of lantibiotics involves several steps: (i) prepeptide (lanA) formation, (ii) posttranslational modifications including dehydration and cyclization (lanB and lanC for class I, lanM for class II), (iii) cleavage of the leader peptide

(lanP), and (iv) secretion (lanT) of the mature peptides. In addition, Genes involving in regulation (lanRK) and self-immunity (lanEFG and lanI) are often found in the same gene cluster (Chatterjee, Paul, Xie, & van der Donk, 2005; McAuliffe, Ross, & Hill, 2001).

Table 4. Classification of lanthionine-containing peptides

Lantibiotics Modification Export and Conserved motif in Other features and enzymes leader cleavage leader peptide examples Class I LanB and LanC LanT and LanP “FDLD” linear structure, e.g., nisin Class II Bifunctional Bifunctional “GC” or “GA” globular structure, LanM LanT(P) e.g., mersacidin Class III RamC/ LabKC - - no antimicrobial activity, e.g., SapB Class IV LanL - - no antimicrobial activity, e.g., venezuelin

Lantibiotic precursor peptide (LanA)

The structural gene (lanA) encodes the precursor peptide comprising a propeptide and an N-terminal leader sequence. The propeptide is the region corresponding to the mature lantibiotic. The propeptide region is rich in cysteine residues while the leader sequence lacks this amino acid (McAuliffe et al., 2001). The cysteine residues couple with Dha or Dhb residues, forming lanthionine or methyllanthionine, respectively. The leader peptide provides a recognition motif for posttranslational modifications, and serves as signals for the secretion

of the mature lantibiotics (Chatterjee et al., 2005).

Dehydratase (LanB) and cyclase (LanC)

In class I lantibiotics, the dehydratase, LanB, catalyzes the selective dehydration of serine and threonine in the propeptide, forming Dha and Dhb residues, respectively. LanB comprises approximately 1000 amino acids and is likely associated with the cytoplasmic membrane (Sahl & Bierbaum, 1998). Following dehydration, the cyclase LanC is responsible for the formation of the thioether bond between cysteine and Dha or Dhb. Koponen et al.

(2002) provided the first direct evidence that NisB and NisC are required for dehydration and cyclization during nisin maturation. The deletion of nisB resulted in totally unmodified nisin precursor, whereas nisC-deficient mutant produced dehydrated precursor without lanthionine formation (Koponen et al., 2002). NisC is a zinc protein where Zn2+ is ligated by two cysteine residues (Cys284, Cys330), one histidine (His331) and a water molecule. Zinc ions activate cysteine residues for the nucleophilic attack on Dha or Dhb in the dehydrated nisin precursor

(B. Li & van der Donk, 2007).

Bifunctional enzyme (LanM)

In class II lantibiotics, the bifunctional enzyme LanM is responsible for dehydration and cyclization reactions. The LanM enzyme possesses 900-1000 amino acids and share 20-

27% similarity with the C-terminus of LanC proteins (Chatterjee et al., 2005). LanM requires

ATP and Mg2+ for dehydration and correct cyclization (Xie et al., 2004). In a two-step dehydration model, the phosphoryl group of ATP is transferred to serine or threonine residue; then the resulting phosphor-Ser (pSer) or phosphor-Thr (pThr) undergoes β-elimination,

forming the dehydrated Dha or Dhb (Chatterjee et al., 2005; You & van der Donk, 2007).

Cyclization of the peptide is catalyzed by the C-terminal domain of LanM; the zinc binding ligand is essential for the catalytic activity of LanM (Paul, Patton, & van der Donk, 2007).

Proteases (LanP) and transporters (LanT)

All lantibiotic precursors consist of an N-terminal leader peptide and a C-terminal propeptide that undergoes dehydration and cyclization by LanB/C or LanM enzyme(s). The modified precursor with a leader peptide is devoid of antimicrobial activity. The removal of the leader peptide gives rise to mature lantibiotics with activity. In class I lantibiotics, the cleavage of leader peptide is catalyzed by a serine protease, LanP, before or after the peptide is exported by the dedicated ABC-transporter, LanT. In class II lantibiotics, a multifunctional transporter, LanT (P), removes the leader peptide by its N-terminal peptidase domain during the export of the peptide (McAuliffe et al., 2001; Sahl & Bierbaum, 1998).

Regulation and self-immunity

The production of lantibiotics is coordinately regulated by cellular events and the signal transduction pathway. Nisin biosynthesis is controlled by a typical two-component regulatory system (Chatterjee et al., 2005). The system comprises a histidine kinase (NisK) and a transcriptional response regulator (NisR). Nisin molecule is the signal for inducing the expression of the nis gene cluster. In the presence of nisin molecules, the membrane protein

NisK passes the signal to NisR. Then the activated NisR binds to nisA and nisF operators and triggers the transcription of nisin gene cluster (Chatterjee et al., 2005). The production of other lantibiotics may be regulated by different mechanisms. For example, lacticin 481

production from Lactococcus lactis is induced by acidification due to lactic acid production

(Hindré, Pennec, Haras, & Dufour, 2006).

Lantibiotic producing strains protect themselves from killing by their own products by specific immunity proteins. The self-immunity mechanisms involve two distant systems: the membrane-associated lipopeptide (LanI), and a dedicated ABC-transporter, LanFEG

(Sahl & Bierbaum, 1998). LanI lipopeptide works as an intercepting molecule by binding to the lantibiotic peptides (Kuipers et al., 2011). On the other hand, the EpiFEG transporter in

Staphylococcus epidermidis Tü3298 protects the producer by expelling the epidermin molecules from cell membrane to the surrounding medium (Otto, Peschel, & Götz, 1998).

Lantibiotic bioengineering

Lantibiotic variants can be produced by manipulation of the structural gene lanA. In the whole-cell system, lantibiotics variants are generated in vivo in the natural host organism or in a genetically well-characterized host by heterologous expression (Cortés, Appleyard, &

Dawson, 2009). To avoid production of the mixture of wild-type and mutated lantibiotics, a lanA- host is usually created by inactivation of the lanA gene. Inactivation can be achieved by deletion of lanA, gene replacement through recombination, or generating a frame-shift mutation (Cortés et al., 2009). The lanA- gene can be replaced by a mutated lanA gene in the same locus, or can be complemented in trans with a separate mutated lanA on a plasmid. The cis complementation system has been used to produce mutants such as subtilin, mutacin II, pep5, mersacidin, cinnamycin and lacticin 3147 (Cortés et al., 2009). Some example of lantibiotic variants generated by trans complementation system included nisin, mersacidin gallidermin/epidermin, actagardine, lacticin 3417 and nukacin ISK-1 (Cortés et al., 2009).

The in vitro reconstitution of the bifunctional LanM enzyme provides the foundation for cell-free enzymatic synthesis of lantibiotics. Two class II lantibiotics, lacticin 481 and haloduracin have been generated by the in vitro systems (McClerren et al., 2006; Xie et al.,

2004). For synthesis of lacticin 481, lctA and lctM were first cloned and overexpressed in E. coli. The His6-tagged lanticin 481prepeptide and LctM enzyme were isolated and purified by affinity chromatography, followed by RP-HPLC and cationic exchange chromatography, respectively. In the presence of Mg2+ and ATP, the purified LctM catalyzed the formation of fully modified lanticin 481. Proteolytic removal of the leader peptide resulted in the active lacticin 481. Several lacticin 481 mutants or analogues containing nonproteinogenic amino acids were generated by the cell-free system (Levengood, Knerr, Oman, & van der Donk,

2009; Xie et al., 2004).

Chemical synthesis of lantibiotics and analogues

The first total solution-phase synthesis of nisin was achieved successfully through desulfurization reaction of disulfide peptides (Fukase et al. 1988). In recent years, a great progress has been made in solid-phase synthesis of the lantibiotics (Tabor, 2011). The advancement of solid-phase peptide synthesis (SPPS) and the use of orthogonally protected building blocks allows for the on-resin synthesis of lantibiotics (Tabor, 2011). In 2009, Ross et al. reported the first solid-supported synthesis of the lantibiotic, lactocin S, using peptide cyclization on solid phase. In addition, lactocin S analogues with improved oxidative stability were chemically synthesized by replacing the sulfur in lanthionine with a methylene unit

(Ross, McKinnie, & Vederas, 2012). Many lantibiotics, such as nisin, epilancin 15X and lacticin 3147A2, possess the interlocking lanthionine or methyllanthionine rings, which

represents a challenge of lantibiotic synthesis. The successful synthesis of both components of lacticin 3147 (Liu, Chan, Liu, Cochrane, & Vederas, 2011) and epilancin 15X (Knerr & van der Donk, 2012) provided a general methodology for the total synthesis of the lantibiotic family.

Mode of action

Class I lantibiotics, such as nisin, disrupt the membrane function by pore formation and inhibit bacterial cell wall biosynthesis. Nisin molecules utilize lipid II, the precursor of cell wall synthesis, as a docking molecule for pore formation. When nisin molecule reaches cell membrane, the N-terminus of nisin (rings A and B) binds to pyrophosphate moiety in lipid II via intermolecular hydrogen bonds. The positive charged C-terminus of nisin inserts into anion cell membrane, resulting in pores on cell membrane (Breukink & de Kruijff, 2006).

On the other hand, class II lantibiotics (e.g. mersacidin) inhibit cell wall synthesis by binding to lipid II; but the interaction does not form pores on cell membrane (Brötz, Bierbaum,

Leopold, Reynolds, & Sahl, 1998).

Application of lantibiotics in food and medical fields

The application of lantibiotics in foods has great potential to extend the shelf-life and increase food safety. Lantibiotics can be applied to foods through three different routes, including the direct adding of the purified or semipurified peptide, incorporating of powdered fermentate, or introduction of the lantibiotic-producing strains to fermented foods (Mills,

Stanton, Hill, & Ross, 2011). Nisin is the most studied lantibiotic and it has been approved as a food preservative by US-FDA in 1988. Nisin has no known toxicity to human and currently is allowed to be used in over 40 countries (Healy, O’Mahony, Hill, Cotter, & Ross, 2011).

In addition to food application, some lantibiotics are promising candidates for treating bacterial infections. For example, nisin is active against drug-resistant pathogens, including methicillin-resistant S. aureus (MRSA), vancomycin-resistant Enterococcus (VRE) and

Clostridium difficile (Piper, Cotter, Ross, & Hill, 2009). Actagardine is a 19-amino acid class

II lantibiotic which is produced by Actinoplanes garbadinensis (Boakes, Cortés, Appleyard,

Rudd, & Dawson, 2009). The semisynthetic derivative of deoxyl-actagardine (NVB302) with selective activity against Clostridium difficile has completed the phase I clinical trial (Shi,

Bueno, & van der Donk, 2012). The lantibiotic salivaricin A is produced by a probiotic strain

Streptococcus salivarius K12. This lantibiotic-producing strain has been used in lozenges to suppress the oral bacteria implicated in halitosis (Burton, Wescombe, Moore, Chilcott, &

Tagg, 2006).

1.3.2. Cyclic lipopeptides: nonribosomal peptides

Cyclic lipopeptide antibiotics are a diverse group of nonribosomally synthesized peptides with a fatty acid size chain. Lipopeptide antibiotics include acidic anti-Gram- positive peptides (e.g. daptomycin), cationic anti-Gram-negative peptides (e.g. polymyxins), and antifungal agents (e.g. echinocandins and surfactin) (Table 5). Daptomycin is a prototype of the acidic lipoeptide family (Figure 2A). In the presence of Ca2+, daptomycin kills the susceptible bacteria by inducing membrane depolarization or disrupting some membrane- associated processes such as cell wall biosynthesis, energetic, and cell division (Robbel &

Marahiel, 2010; Straus & Hancock, 2006). The marked daptomycin for injection (Cubicin) is indicated for treatment of complicated skin and skin structure infections (cSSSI), and for S. aureus bloodstream infections (bacteremia) (http://www.cubicin.com/).

Polymyxins (Figure 2B) are cationic lipopeptides that were discovered over 50 years ago. The group of antibiotics is active against Gram-negative bacteria, but polymyxins were once discarded for decades due to the concerns on nephrotoxicity and neurotoxicity. However, the lack of effective antibiotics for emgering multidrug-resistant Gram-negative pathogens has stimulated a renewed interest in polymyxins (Landman et al., 2008). Polymyxins have high affinity toward lipopolysaccharides (LPS) on outer membrane of Gram-negative bacteria. Polymyxins interact with LPS and replace the divalent cations that stabilize the LPS network on outer membrane. This is followed by the uptake of polymyxins and disruption of cytoplasmic membrane, leading to cell death (Velkov, Thompson, Nation, & Li, 2010).

The echinocandins are antifungal lipopeptides comprising a hexapeptide ring with an

N-terminal acyl chain. These compounds have rapid fungicidal activity against most Candida spp. and fungistatic effect against Aspergillus species (Denning, 2003). Three compounds derived from the echinocandins, including caspofungin, micafungin, and anidulafungin, have been approved by the US-FDA for the treatment of candidiasis and aspergillosis (Cappelletty

& Eiselstein‐McKitrick, 2007). The echinocandins block the fungal cell wall biosynthesis and lead to cell death by inhibiting the β-(1, 3)-glucan synthetase (Denning, 2003).

Table 5. Selected lipopeptide-derived antibiotics

Compound Producer Notes Anti-Gram-positive lipopeptidesa Acidic peptides, calcium dependent activity Daptomycin Streptomyces roseosporus NRRL11379 Form pores in membrane, Approved by FDA in 2003 CDA Streptomyces coelicolor A3(2) / Streptomyces violaceoruber Kutner 673 Streptomyces lividans A54145 Streptomyces fradiae NRRL18158 / Streptomyces refuineus sp. thermotolerans Friulimicins Actinoplanes friuliensis DSM 7358 Inhibit cell wall synthesis Amphomycin Streptomyces canus ATCC 12237 / Glycinocin Actinomycete AW998 / Laspartomycin Streptomyces viridochromogenes ATCC / 29814 Anti-Gram-negative lipopeptidesb Cationic peptides, high affinity for LPS Polymyxin Paenibacillus polymyxa Renewed interest in their Colistin Paenibacillus polymyxa use against MDR-resistant Gram-negative strains Antifungal lipopeptidesc Caspofungin Glarea lozoyensis Echinocandin-derived Micafungin Coleophoma empetri drugs approved by FDA, Anidulafungin Aspergillus nidulans Inhibit glucan synthetase Iturin family Bacillus subtilis Use as biocontrol of plant Bacillus amyloliquefaciens pathogens Surfactin Bacillus coagulans family Bacillus pumilus Bacillus licheniformis Fengincin Bacillus cereus family Bacillus thuringiensis Bacillus subtilis Bacillus amyloliquefaciens Fusaricidin Paenibacillus polymyxa a Modified from Baltz, Miao, & Wrigley, 2005 b Referred to Landman et al., 2008 c Referred to Beatty & Jensen, 2002; Ongena & Jacques, 2008; Pirri, Giuliani, Nicoletto, Pizzuto, & Rinaldi, 2009

Figure 2. Chemical structure of daptomycin (A) and polymyxin B (B). Modified from Heinis, Rutherford, Freund, & Winter, 2009

Biosynthesis of nonribosomal peptides

Lipopeptide antibiotics are produced by nonribosomal peptide synthetases (NRPSs).

The NRPS catalysts are organized in modules, each of which is responsible for incorporation of one amino acid. There are three core domains in each module: condensation (C) domain, adenylation (A) domain, and peptidyl carrier protein domain (PCP or thiolation domain)

(Walsh, 2003).

A-domain in the C-A-PCP module recognizes the amino acid to be added in the growing peptide chain. The selection of substrate is determined by 10 amino acids

(nonribosomal code) in the binding pocket of A-domain (Stachelhaus, Mootz, & Marahiel,

1999). In addition to the 20 proteinogenic amino acids, A-domains can recognize unusual, nonproteinogeneic amino acids, which contribute to the structural diversity of nonribosomal peptide natural products (Konz & Marahiel, 1999). A-domain activates the selected amino acid at the expense of ATP consumption, forming pyrophosphate (PPi) and aminoacyl-AMP.

The activated aminoacyl group is then transferred to the PCP domain, generating aminoacyl-

S-PCP intermediate. The C domain in the C-A-PCP module catalyzes the peptide bond formation between two neighboring PCP-bound amino acids (Walsh, 2003). A schematic model of NRPS module is presented in Figure 3.

The starter C domain in lipopeptides NRPSs was proposed for the coupling the acyl side chain to the N-terminal amino acid (Kraas, Helmetag, Wittmann, Strieker, & Marahiel,

2010; Miao et al., 2005; Miao et al., 2006). The C-terminal thioesterase (TE) domain in

NRPSs is responsible for the macrocyclization of the newly formed linear peptide (Kohli &

Walsh, 2003).

Figure 3. Schematic module of nonribosomal peptide synthetase

Chemical synthesis of lipopeptides

Isolation and purification of a large amount of lipopeptide antibiotics sometimes are difficult particularly when the producer strain produces a mixture of lipopeptides with single or multiple amino acid substitution, or different length of lipid chain. As an alternative, the total peptide synthesis can provide easy access to the synthetic analogues of lipopeptides for the studies of structure-activity relationship (Stawikowski & Cudic, 2007). The challenges facing the total synthesis of cyclic lipopeptides include the synthesis and incorporation of unusual amino acids, and the closure of the macrolactone or macrolactam ring (Stawikowski

& Cudic, 2007). The antifungal lipopeptide fusaricidin A was successfully synthesized using the standard Fmoc solid-phase peptide synthesis method (Cochrane et al., 2010; Cochrane,

McErlean, & Jolliffe, 2010; Stawikowski & Cudic, 2009). The key macrolactone ring closure step was accomplished either in solution using Yamaguchi reagents after cleavage from the resin (Cochrane et al., 2010), or when the peptide was still attached on the resin

(Stawikowski & Cudic, 2009). In addition, the solid-phase synthesis of surfactin was also reported (Pagadoy, Peypoux, & Wallach, 2005). However, the total synthesis of daptomycin was hampered by the lack of commercially available 3-methyl-glutamic acid (Baltz et al.,

2005; Nguyen et al., 2006).

1.4. Application of next-generation sequencing in food microbiology The revolutionary advancement in non-Sanger-based sequencing technologies has profound impacts on today’s biological science (Shendure & Ji, 2008). The advantage of the next-generation sequencing (NGS) technology lies in producing a large volume of sequence data in a very short time (Metzker, 2009). For example, a typical bacterial genome can be sequenced within a week. The new technologies are currently used in many research areas including applied and food microbiology. The NGS platforms provide important advantages over some traditional techniques, such as hybridization-based microarray for transcription analysis, and PCR-denaturing gradient gel electrophoresis (PCR-DGGE) for profiling microbial community. The RNA-Seq technique and pyrosequencing using the NGS platforms are expected to replace microarrays and PCR-DGGE in the future (Wang, Gerstein, & Snyder,

2009; Roh et al., 2009). In addition, NGS technologies can also be used in subtyping of foodborne pathogens and tracking the source of outbreak strains. The applications of NGS in food microbiology will be discussed in the following texts.

1.4.1. Community profiling during fermentation

NGS-based community profiling allows for comparison of microbial diversity and dynamics under different environmental or processing conditions. NGS technologies for studying microbial diversity have advantages over other culture-independent approaches

such as PCR-DGGE. NGS is a high throughput method and usually provides a complete microbial profile for a certain ecosystem.

Humblot & Guyot (2009) investigated the feasibility of pyrosequencing-based method for profiling the microbial community in fermented pearl millet slurries. In this study, the variable V3 region of the 16S rRNA gene was amplified by PCR, and 10 samples were analyzed at the same time using barcoded pyrosequencing. The study suggested that pyrosequencing were promising for rapid and preliminary microbial characterization in complex food samples. Alegría, Szczesny, Mayo, Bardowski, & Kowalczyk (2012) studied the microbial composition and dynamics during the manufacture of traditional Oscypek cheese using the culturing and culture-independent methods. The studied showed that pyrosequencing analysis revealed a greater diversity than PCR-DGGE in the cheese ecosystem.

Fungal, viral and archaeal diversity have been studied using NGS technologies. The internal transcribed spacer (ITS1 and ITS2) regions of the nuclear ribosomal repeat show a high rate of evolution and are typically species-speciﬁc (Nilsson, Ryberg, Abarenkov,

Sjökvist, & Kristiansson, 2009). X. R. Li et al. (2011) analyzed the diversity of fungi in the traditional Chinese liquor fermentation process using pyrosequencing targeting the ITS1 region. Park et al. (2011) studied the overall viral assemblages (viromes) in fermented shrimp, kimchi, and sauerkraut using viral metagenome pyrosequencing. The linker-amplified shotgun library (LASL) technique was used to amplify double-stranded viral DNA. The metagenomic analysis showed that the fermented food communities were dominated by bacteriophage belonging to the order Caudovirales. Roh et al. (2009) investigated the

bacterial and archaeal diversity in fermented seafood using barcoded pyrosequencing of V3 region of 16S rRNA gene. The results showed that the most abundant archaea in the seafood sample was the extremely halophlic archaea of family Halobacteriaceae.

Despite the tremendous potential in deciphering the diversity of microbial community,

NGS techniques have some limitations. The 16S rRNA gene does not provide enough variation to unambiguously classify reads to species level. In one study, up to 43% of the sequences amplified from universal primers were attributed to Enkarya sequences of plant origin (Humblot & Guyot, 2009). Prosdocimi et al. (2012) pointed out that universal primers flanking the V3 region (e.g., 341F and 518R) amplified 18S rRNA genes from eukaryotes.

Therefore, primer selection is critical when analyzing the bacterial community associated with eukaryotes. Prosdocimi et al. (2012) proposed to use nested PCR to avoid amplification of eukaryotic sequences.

1.4.2. NGS-assisted process optimization

Application of NGS in profiling food microbial community is beneficial to food processing optimization. Ercolini et al. (2011) used pyrosequencing to monitor changes in beef microbiota during storage under different packaging conditions over 45 days. In air storage condition, the dominant microorganisms in beef shifted from Brochothrix thermosphacta at the early stage to genera Pseudomonas at the late stage of storage.

Sakamoto, Tanaka, Sonomoto, & Nakayama (2011) studied the bacterial community in a fermented rice bran mash (nukadoko) in a laboratory model. When an aged nukadoko was inoculated to the fresh bran, Lactobacillus spp. dominated the fermentation and ripening process; however, unfavorable bacteria, including Staphylococcus and Bacillus species, were

observed in the inoculant-free fermentation process (Sakamoto et al., 2011). Lopez‐Velasco,

Welbaum, Boyer, Mane, & Ponder (2011) compared the composition of spinach bacterial community under typical retail storage conditions (4°C) or temperature abuse conditions

(10°C). At both temperatures, Pseudomonas spp. and members of the Enterobacteriaceae were the most abundant population in fresh-packed spinach after 15 days of storage. The results also revealed that the growth of Escherichia spp. was inhibited at 4°C but not at 10°C storage. The study highlighted the need for optimizing the processing parameters to suppress the growth of pathogenic bacteria.

1.4.3. RNA-sequencing (RNA-seq) Transcriptomics

Transcriptome represents the complete collection of transcribed sequences in a cell, including the coding RNA (mRNA) and noncoding RNA (Van Vliet, 2009). Microarrays have been the method of choice for high-throughput analysis of the relative abundence of transcripts for more than a decade. However, microarrays have some pitfalls. For example, oligonucleotides probes based on a certain strain may not be optimal for other strains due to genomic variation (Van Vliet, 2009). In addition, microarrays do not capture novel sequences that have not been annotated in the genome. Currently, RNA sequencing (RNA-seq) technique is competing with microarrays for transcription analysis. Leimena et al. (2012) compared the RNA-sequencing method with DNA microarray for transcription analysis of

Lactobacillus plantarum grown in two different media. Both methods detected similar patterns of transcript abundance and fold-change levels of differentially expressed genes.

Deng, Li, & Zhang (2011) used RNA sequencing technology to study the global gene expression of Salmonella Enteritidis in a peanut oil (water activity= 0.30). Under starvation

and desiccation conditions, only 5% or less of the Salmonella genome was transcribed, which indicated that the bacteria were in the metabolically dormant state. Among the very few expressed genes, transcripts for cold and heat shock proteins, DNA protection and regulatory functions were detected (Deng et al., 2011). Change of transcription levels under different conditions can help understand the mechanism of adaptive process to harsh environments.

Transcriptional factors (e.g., alternative σ factors) are critical regulatory components contributing to the survival of the pathogen to harsh environmental conditions

(Chaturongakul, Raengpradub, Wiedmann, & Boor, 2008). Listeria monocytogenes can persist on food processing equipment for extended periods of time (Carpentier and Cerf,

2011). Oliver et al. (2009) studied the transcriptome of L. monocytogenes and its isogenic

ΔsigB mutant, which does not express the alternative σ factor, using Illumina RNA- sequencing technology. A total of 96 σB -dependent transcripts as well as 67 non-coding RNA molecules were identified in the stationary phase.

Temperature plays a very critical role in cellular activity. Yu et al. (2011) studied the transcription levels of aflatoxin gene cluster in Aspergillus flavus at two temperatures using

RNA-sequencing technology. They found that aflatoxin biosynthesis was tightly controlled by temperature. At lower temperature 30°C, the level of aflatoxin biosynthetic transcript was

3300 times greater as compared with 37°C (Yu et al., 2011). The research may shed some light on controlling storage conditions for crops to reduce the incidence of aflatoxin.

1.4.4. Whole genome sequencing typing for outbreak investigation

The NGS technologies are useful in subtyping pathogenic strains and tracking the source of outbreak. In 2009-2010, the multistate outbreak of Salmonella enterica subsp.

enterica serovar Montevideo sickened 272 people across the US (den Bakker et al., 2011).

The currently used subtyping method, the pulsed field gel electrophoresis (PFGE), was unable to differentiate the implicated spiced-meat food and other clinical isolates that were associated with previous contaminated pistachio nuts (Lienau et al., 2011). The whole genome single nucleotide polymorphism (SNP)-based typing confirmed that the outbreak strain was from an implicated food processing facility (Lienau et al., 2011).

The NGS technologies can facilitate the whole genome characterization of outbreak strains at the early stage. This strategy will reduce the time for identifying the unique feature of outbreak strains. In 2011, a deadly foodborne outbreak in Germany caused 3816 illnesses and 54 deaths (Frank et al., 2011). Scientists rapidly determined the whole genome sequence of outbreak isolates, Escherichia coli O104:H4, within a week after receiving the bacterial genomic DNA (Kupferschmidt, 2011). The genome information helped researchers across the world to understand the unique features of the deadly pathogen. Comparative genomics analyses revealed that the pathogen shares 93% of its genome sequences with an enteroaggregative E. coli (EAEC) strain 55989; however, the outbreak strain carries the

Shiga-toxin gene in a prophage (Denamur, 2011; Mellmann et al., 2011).

In the current study, the NGS technology was used to sequence two bacterial strains that produce antimicrobial agents of interest. The whole genome sequencing is critical for identifying the genes responsible for the peptide biosynthesis.

References

Adams, M., & Moss, M. (2008). Food microbiology (pp. 112-115). Cambridge, UK: The Royal Society of Chemistry.

Alegría, Á., Szczesny, P., Mayo, B., Bardowski, J., & Kowalczyk, M. (2012). Biodiversity in Oscypek, a Traditional Polish Cheese, Determined by Culture-Dependent and- Independent Approaches. Applied and Environmental Microbiology, 78(6), 1890-1898.

Artsimovitch, I., Seddon, J., & Sears, P. (2012). Fidaxomicin is an inhibitor of the initiation of bacterial RNA synthesis. Clinical Infectious Diseases, 55(suppl 2), S127-S131.

Balasubramaniam, V., & Farkas, D. (2008). High-pressure food processing. Food Science and Technology International, 14(5), 413-418.

Baltz, R. H., Miao, V., & Wrigley, S. K. (2005). Natural products to drugs: daptomycin and related lipopeptide antibiotics. Natural Product Reports, 22(6), 717-741.

Beatty, P. H., & Jensen, S. E. (2002). Paenibacillus polymyxa produces fusaricidin-type antifungal antibiotics active against Leptosphaeria maculans, the causative agent of blackleg disease of canola. Canadian Journal of Microbiology, 48(2), 159-169.

Blakistone, B. A. (1999). Introduction. In B. A. Blakistone (Ed.), Principles and applications of modified atmosphere packaging of foods (pp. 1-13). Gaithersburg, MD: Aspen Publishers.

Boakes, S., Cortés, J., Appleyard, A. N., Rudd, B. A. M., & Dawson, M. J. (2009). Organization of the genes encoding the biosynthesis of actagardine and engineering of a variant generation system. Molecular Microbiology, 72(5), 1126-1136.

Bonomo, R. A. (2011). New Delhi metallo-β-lactamase and multidrug resistance: a global SOS? Clinical Infectious Diseases, 52(4), 485-487.

Boucher, H. W., Talbot, G. H., Bradley, J. S., Edwards, J. E., Gilbert, D., Rice, L. B., Scheld, M., Spellberg, B., & Bartlett, J. (2009). Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clinical Infectious Diseases, 48(1), 1- 12.

Breukink, E., & de Kruijff, B. (2006). Lipid II as a target for antibiotics. Nature Reviews Drug Discovery, 5(4), 321-323.

Brötz, H., Bierbaum, G., Leopold, K., Reynolds, P. E., & Sahl, H. G. (1998). The lantibiotic mersacidin inhibits peptidoglycan synthesis by targeting lipid II. Antimicrobial Agents and Chemotherapy, 42(1), 154-160.

Burton, J. P., Wescombe, P. A., Moore, C. J., Chilcott, C. N., & Tagg, J. R. (2006). Safety assessment of the oral cavity probiotic Streptococcus salivarius K12. Applied and Environmental Microbiology, 72(4), 3050-3053.

Butler, M. S., & Cooper, M. A. (2011). Antibiotics in the clinical pipeline in 2011. The Journal of Antibiotics, 64(6), 413-425.

Cappelletty, D., & Eiselstein‐McKitrick, K. (2007). The echinocandins. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy, 27(3), 369-388.

CDC. (2011). Trends in foodborne illness, 1996–2010. Retrieved 1/17, 2013, from http://www.cdc.gov/foodborneburden/PDFs/FACTSHEET_B_TRENDS.pdf

Chatterjee, C., Paul, M., Xie, L., & van der Donk, W. A. (2005). Biosynthesis and mode of action of lantibiotics. Chemical Reviews, 105(2), 633-684.

Chaturongakul, S., Raengpradub, S., Wiedmann, M., & Boor, K. J. (2008). Modulation of stress and virulence in Listeria monocytogenes. Trends in Microbiology, 16(8), 388-396.

Chemat, F., & Khan, M. K. (2011). Applications of ultrasound in food technology: Processing, preservation and extraction. Ultrasonics Sonochemistry, 18(4), 813-835.

Coates, A. R. M., & Halls, G. (2012). Antibiotics in phase II and III clinical trials. In A. R. M. Coates (Ed.), Antibiotic resistance (pp. 167-183). New York, NY: Springer.

Cochrane, J. R., McErlean, C. S. P., & Jolliffe, K. A. (2010). Total synthesis and assignment of the side chain stereochemistry of LI-F04A: an antimicrobial cyclic depsipeptide. Organic Letters, 12(15), 3394-3397.

Cortés, J., Appleyard, A. N., & Dawson, M. J. (2009). Whole‐Cell generation of lantibiotic variants. In D. A. Hopwood (Ed.), Complex enzymes in microbial natural product biosynthesis, part A: Overview articles and peptides (pp. 559-574). San Diego: Elsevier.

Davidson, P. M., & Branen, A. L. (2005). Food antimicrobials-an introduction. In P. M. Davidson, J. N. Sofos & A. L. Branen (Eds.), Antimicrobials in food (pp. 1-10). Boca Raton, FL: CRC Press.

De Man, J. M. (1995). Principles of food chemistry (pp. 23). Gaithersburg, MD: Aspen Publishers.

den Bakker, H. C., Switt, A. I. M., Cummings, C. A., Hoelzer, K., Degoricija, L., Rodriguez- Rivera, L. D., Wright, E. M., Fang, R., Davis, M., & Root, T. (2011). A whole-genome single nucleotide polymorphism-based approach to trace and identify outbreaks linked to a common Salmonella enterica subsp. enterica serovar Montevideo pulsed-field gel electrophoresis type. Applied and Environmental Microbiology, 77(24), 8648-8655.

Denamur, E. (2011). The 2011 Shiga toxin‐producing Escherichia coli O104: H4 German outbreak: a lesson in genomic plasticity. Clinical Microbiology and Infection, 17(8), 1124-1125.

Deng, X., Li, Z., & Zhang, W. (2011). Transcriptome Sequencing of Salmonella enterica serovar Enteritidis under Desiccation and Starvation Stress in Peanut Oil. Food Microbiology, 30(1), 311-315.

Denning, D. W. (2003). Echinocandin antifungal drugs. Lancet, 362(9390), 1142-1151.

Elmnasser, N. E. N., Guillou, S. G. S., Leroi, F. L. F., Orange, N. O. N., Bakhrouf, A. B. A., & Federighi, M. F. M. (2007). Pulsed-light system as a novel food decontamination technology: a review. Canadian Journal of Microbiology, 53(7), 813-821.

Ercolini, D., Ferrocino, I., Nasi, A., Ndagijimana, M., Vernocchi, P., La Storia, A., Laghi, L., Mauriello, G., Guerzoni, M. E., & Villani, F. (2011). Monitoring of microbial metabolites and bacterial diversity in beef stored under different packaging conditions. Applied and Environmental Microbiology, 77(20), 7372-7381.

Farkas, J. (2006). Irradiation for better foods. Trends in Food Science & Technology, 17(4), 148-152.

Fieseler, L., Loessner, M. J., & Hagens, S. (2011). Bacteriophages and food safety. In C. Lacroix (Ed.), Protective cultures, antimicrobial metabolites and bacteriophages for food and beverage biopreservation (pp. 161-178). Philadelphia, PA: Woodhead Publishing.

Fischbach, M. A., & Walsh, C. T. (2009). Antibiotics for emerging pathogens. Science, 325(5944), 1089-1093.

Frank, C., Werber, D., Cramer, J. P., Askar, M., Faber, M., an der Heiden, M., Bernard, H., Fruth, A., Prager, R., & Spode, A. (2011). Epidemic profile of Shiga-toxin–producing Escherichia coli O104: H4 outbreak in Germany. New England Journal of Medicine, 365(19), 1771-1780.

García, P., Martínez, B., & Rodríguez, A. (2011). Bacteriophages and phage-encoded proteins: Prospects in food quality and safety. In M. Rai, & M. Chikindas (Eds.),

Natural antimicrobials in food safety and quality (pp. 10-26). Cambridge, MA: CABI Publishing.

García, P., Rodríguez, L., Rodríguez, A., & Martínez, B. (2010). Food biopreservation: promising strategies using bacteriocins, bacteriophages and endolysins. Trends in Food Science & Technology, 21(8), 373-382.

Gómez-López, V. M., Ragaert, P., Debevere, J., & Devlieghere, F. (2007). Pulsed light for food decontamination: a review. Trends in Food Science & Technology, 18(9), 464-473.

Gram, L., Ravn, L., Rasch, M., Bruhn, J. B., Christensen, A. B., & Givskov, M. (2002). Food spoilage—interactions between food spoilage bacteria. International Journal of Food Microbiology, 78(1), 79-97.

Guerrero-Beltrán, J. A., & Barbosa-Cánovas, G. V. (2004). Advantages and limitations on processing foods by UV light. Food Science and Technology International, 10(3), 137- 147.

Hamad, B. (2010). The antibiotics market. Nature Reviews Drug Discovery, 9(9), 675-676.

Healy, B., O’Mahony, J., Hill, C., Cotter, P. D., & Ross, R. P. (2011). Lantibiotic-related research and the application thereof. In G. Wang (Ed.), Antimicrobial peptides: Discovery, design, and novel therapeutic strategies (pp. 22-26). Oxfordshire, UK: CAB International.

Heinis, C., Rutherford, T., Freund, S., & Winter, G. (2009). Phage-encoded combinatorial chemical libraries based on bicyclic peptides. Nature Chemical Biology, 5(7), 502-507.

Hindré, T., Pennec, J. P., Haras, D., & Dufour, A. (2006). Regulation of lantibiotic lacticin 481 production at the transcriptional level by acid pH. FEMS Microbiology Letters, 231(2), 291-298.

Hogan, E., Kelly, A. L., & Sun, D. W. (2005). High pressure processing. In D. W. Sun (Ed.), Emerging technologies for food processing (pp. 3-32). San Diego, CA: Elsrvier Academic Press.

Humblot, C., & Guyot, J. P. (2009). Pyrosequencing of tagged 16S rRNA gene amplicons for rapid deciphering of the microbiomes of fermented foods such as pearl millet slurries. Applied and Environmental Microbiology, 75(13), 4354-4361.

Jay, J. M., Loessner, M. J., & Golden, D. A. (2005). Modern food microbiology (pp. 301- 350). New York, NY: Springer.

Jeyamkondan, S., Jayas, D., & Holley, R. (1999). Pulsed electric field processing of foods: a review. Journal of Food Protection, 62(9), 1088-1096.

Juneja, V. K., Dwivedi, H. P., & Yan, X. (2012). Novel Natural Food Antimicrobials. Annual Review of Food Science and Technology, 3, 381-403.

Kantor, L. S., Lipton, K., Manchester, A., & Oliveira, V. (1997). Estimating and addressing America’s food losses. Food Review, 20(1), 2-12.

Knerr, P. J., & van der Donk, W. A. (2012). Chemical Synthesis and Biological Activity of Analogues of the Lantibiotic Epilancin 15X. Journal of the American Chemical Society, 134(18), 7648-7651.

Kohanski, M. A., Dwyer, D. J., & Collins, J. J. (2010). How antibiotics kill bacteria: from targets to networks. Nature Reviews Microbiology, 8(6), 423-435.

Kohanski, M. A., Dwyer, D. J., Hayete, B., Lawrence, C. A., & Collins, J. J. (2007). A common mechanism of cellular death induced by bactericidal antibiotics. Cell, 130(5), 797-810.

Kohli, R. M., & Walsh, C. T. (2003). Enzymology of acyl chain macrocyclization in natural product biosynthesis. Chemical Communications, 7(3), 297-307.

Konz, D., & Marahiel, M. A. (1999). How do peptide synthetases generate structural diversity? Chemistry & Biology, 6(2), R39-48.

Koponen, O., Tolonen, M., Qiao, M., Wahlstrom, G., Helin, J., & Saris, P. E. J. (2002). NisB is required for the dehydration and NisC for the lanthionine formation in the post- translational modification of nisin. Microbiology, 148(11), 3561.

Kraas, F. I., Helmetag, V., Wittmann, M., Strieker, M., & Marahiel, M. A. (2010). Functional dissection of surfactin synthetase initiation module reveals insights into the mechanism of lipoinitiation. Chemistry & Biology, 17(8), 872-880.

Kuipers, A., Rink, R., & Moll, G. N. (2011). Genetics, biosynthesis, structure, and mode of action of lantibiotics. In D. Drider, & S. Rebuffat (Eds.), Prokaryotic antimicrobial peptides: From genes to applications (pp. 147-169). New York, NY: Springer.

Kume, T., Furuta, M., Todoriki, S., Uenoyama, N., & Kobayashi, Y. (2009). Status of food irradiation in the world. Radiation Physics and Chemistry, 78(3), 222-226.

Kupferschmidt, K. (2011). Scientists rush to study genome of lethal E. coli. Science, 332(6035), 1249-1250.

Kussendrager, K. D., & Van Hooijdonk, A. (2000). Lactoperoxidase: physico-chemical properties, occurrence, mechanism of action and applications. British Journal of Nutrition, 84, 19-25.

Landman, D., Georgescu, C., Martin, D. A., & Quale, J. (2008). Polymyxins revisited. Clinical Microbiology Reviews, 21(3), 449-465.

Leimena, M. M., Wels, M., Bongers, R. S., Smid, E. J., Zoetendal, E. G., & Kleerebezem, M. (2012). Comparative Analysis of Lactobacillus plantarum WCFS1 Transcriptomes by Using DNA Microarray and Next-Generation Sequencing Technologies. Applied and Environmental Microbiology, 78(12), 4141-4148.

Levengood, M. R., Knerr, P. J., Oman, T. J., & van der Donk, W. A. (2009). In vitro mutasynthesis of lantibiotic analogues containing nonproteinogenic amino acids. Journal of the American Chemical Society, 131(34), 12024-12025.

Levy, S. B., & Marshall, B. (2004). Antibacterial resistance worldwide: causes, challenges and responses. Nature Medicine, 10, S122-S129.

Li, B., & van der Donk, W. A. (2007). Identification of essential catalytic residues of the cyclase NisC involved in the biosynthesis of nisin. Journal of Biological Chemistry, 282(29), 21169-21175.

Li, X. R., Ma, E. B., Yan, L. Z., Meng, H., Du, X. W., Zhang, S. W., & Quan, Z. X. (2011). Bacterial and fungal diversity in the traditional Chinese liquor fermentation process. International Journal of Food Microbiology, 146(1), 31-37.

Lienau, E. K., Strain, E., Wang, C., Zheng, J., Ottesen, A. R., Keys, C. E., Hammack, T. S., Musser, S. M., Brown, E. W., & Allard, M. W. (2011). Identification of a salmonellosis outbreak by means of molecular sequencing. New England Journal of Medicine, 364(10), 981-982.

Liu, W., Chan, A. S. H., Liu, H., Cochrane, S. A., & Vederas, J. C. (2011). Solid supported chemical syntheses of both components of the lantibiotic lacticin 3147. Journal of the American Chemical Society, 133(36), 14216-14219.

Lopez‐Velasco, G., Welbaum, G., Boyer, R., Mane, S., & Ponder, M. (2011). Changes in spinach phylloepiphytic bacteria communities following minimal processing and refrigerated storage described using pyrosequencing of 16S rRNA amplicons. Journal of Applied Microbiology, 110(5), 1203-1214.

McAuliffe, O., Ross, R. P., & Hill, C. (2001). Lantibiotics: structure, biosynthesis and mode of action. FEMS Microbiology Reviews, 25(3), 285-308.

McClerren, A. L., Cooper, L. E., Quan, C., Thomas, P. M., Kelleher, N. L., & Van Der Donk, W. A. (2006). Discovery and in vitro biosynthesis of haloduracin, a two-component lantibiotic. Proceedings of the National Academy of Sciences, 103(46), 17243-17248.

Mellmann, A., Harmsen, D., Cummings, C. A., Zentz, E. B., Leopold, S. R., Rico, A., Prior, K., Szczepanowski, R., Ji, Y., & Zhang, W. (2011). Prospective genomic characterization of the German enterohemorrhagic Escherichia coli O104: H4 outbreak by rapid next generation sequencing technology. PloS One, 6(7), e22751.

Metzker, M. L. (2009). Sequencing technologies—the next generation. Nature Reviews Genetics, 11(1), 31-46.

Miao, V., Brost, R., Chapple, J., She, K., Gal, M. F. C. L., & Baltz, R. H. (2006). The lipopeptide antibiotic A54145 biosynthetic gene cluster from Streptomyces fradiae. Journal of Industrial Microbiology & Biotechnology, 33(2), 129-140.

Miao, V., Coëffet-LeGal, M. F., Brian, P., Brost, R., Penn, J., Whiting, A., Martin, S., Ford, R., Parr, I., & Bouchard, M. (2005). Daptomycin biosynthesis in Streptomyces roseosporus: cloning and analysis of the gene cluster and revision of peptide stereochemistry. Microbiology, 151(5), 1507-1523.

Mills, S., Stanton, C., Hill, C., & Ross, R. (2011). New developments and applications of bacteriocins and peptides in foods. Annual Review of Food Science and Technology, 2, 299-329.

Nguyen, K. T., Ritz, D., Gu, J. Q., Alexander, D., Chu, M., Miao, V., Brian, P., & Baltz, R. H. (2006). Combinatorial biosynthesis of novel antibiotics related to daptomycin. Proceedings of the National Academy of Sciences, 103(46), 17462-17467.

Nilsson, R. H., Ryberg, M., Abarenkov, K., Sjökvist, E., & Kristiansson, E. (2009). The ITS region as a target for characterization of fungal communities using emerging sequencing technologies. FEMS Microbiology Letters, 296(1), 97-101.

Oliver, H., Orsi, R., Ponnala, L., Keich, U., Wang, W., Sun, Q., Cartinhour, S., Filiatrault, M., Wiedmann, M., & Boor, K. (2009). Deep RNA sequencing of L. monocytogenes reveals overlapping and extensive stationary phase and sigma B-dependent transcriptomes, including multiple highly transcribed noncoding RNAs. BMC Genomics, 10(1), 641.

Oms-Oliu, G., Martín-Belloso, O., & Soliva-Fortuny, R. (2010). Pulsed light treatments for food preservation. A review. Food and Bioprocess Technology, 3(1), 13-23.

Ongena, M., & Jacques, P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends in Microbiology, 16(3), 115-125.

Osaili, T. M. (2012). Developments in the thermal processing of food. In R. Bhat, A. K. Alias & G. Paliyath (Eds.), Progress in food preservation (pp. 211-230). West sussex, UK: Wiley-Blackwell.

Otto, M., Peschel, A., & Götz, F. (1998). Producer self‐protection against the lantibiotic epidermin by the ABC transporter EpiFEG of Staphylococcus epidermidis Tü3298. FEMS Microbiology Letters, 166(2), 203-211.

Pagadoy, M., Peypoux, F., & Wallach, J. (2005). Solid-phase synthesis of surfactin, a powerful biosurfactant produced by Bacillus subtilis, and of four analogues. International Journal of Peptide Research and Therapeutics, 11(3), 195-202.

Park, E. J., Kim, K. H., Abell, G. C. J., Kim, M. S., Roh, S. W., & Bae, J. W. (2011). Metagenomic analysis of the viral communities in fermented foods. Applied and Environmental Microbiology, 77(4), 1284-1291.

Paul, M., Patton, G. C., & van der Donk, W. A. (2007). Mutants of the zinc ligands of lacticin 481 synthetase retain dehydration activity but have impaired cyclization activity. Biochemistry, 46(21), 6268-6276.

Piper, C., Cotter, P. D., Ross, R. P., & Hill, C. (2009). Discovery of medically significant lantibiotics. Current Drug Discovery Technologies, 6(1), 1-18.

Pirri, G., Giuliani, A., Nicoletto, S. F., Pizzuto, L., & Rinaldi, A. C. (2009). Lipopeptides as anti-infectives: a practical perspective. Central European Journal of Biology, 4(3), 258- 273.

Piyasena, P., Dussault, C., Koutchma, T., Ramaswamy, H., & Awuah, G. (2003). Radio frequency heating of foods: principles, applications and related properties—a review. Critical Reviews in Food Science and Nutrition, 43(6), 587-606.

Prosdocimi, E. M., Novati, S., Bruno, R., Bandi, C., Mulatto, P., Giannico, R., Casiraghi, M., & Ferri, E. (2012). Errors in ribosomal sequence datasets generated using PCR-coupled ‘panbacterial’pyrosequencing, and the establishment of an improved approach. Molecular and Cellular Probes, 27(1), 65-67.

Rahman, M. S. (2007a). Food preservation: Overview. In M. S. Rahman (Ed.), Handbook of food preservation (pp. 3-17). Boca Raton, FL: CRC Press.

Rahman, M. S. (2007b). Light energy in food preservation. In M. S. Rahman (Ed.), Handbook of food preservation (pp. 751-757). Boca Raton, FL: CRC Press.

Rastogi, N. K. (2011). Opportunities and Challenges in Application of Ultrasound in Food Processing. Critical Reviews in Food Science and Nutrition, 51(8), 705-722. 42

Rea, M. C., Ross, R. P., Cotter, P. D., & Hill, C. (2011). Classification of bacteriocins from gram-positive bacteria. In D. Drider, & S. Rebuffat (Eds.), Prokaryotic antimicrobial peptides: From genes to applications (pp. 29-53). New York, NY: Springer.

Robbel, L., & Marahiel, M. A. (2010). Daptomycin, a bacterial lipopeptide synthesized by a nonribosomal machinery. Journal of Biological Chemistry, 285(36), 27501-27508.

Roh, S. W., Kim, K. H., Nam, Y. D., Chang, H. W., Park, E. J., & Bae, J. W. (2009). Investigation of archaeal and bacterial diversity in fermented seafood using barcoded pyrosequencing. The ISME Journal, 4(1), 1-16.

Ross, A. C., McKinnie, S. M. K., & Vederas, J. C. (2012). The Synthesis of Active and Stable Diaminopimelate Analogues of the Lantibiotic Peptide Lactocin S. Journal of the American Chemical Society, 134(4), 2008-2011.

Russell, N. J., & Gould, G. W. (2003). Major preservative technologies. In N. J. Russell, & G. W. Gould (Eds.), Food preservatives (pp. 14-24). New York, NY: Kluwer Acdemic/Plenum Pblishers,.

Sahl, H. G., & Bierbaum, G. (1998). Lantibiotics: biosynthesis and biological activities of uniquely modified peptides from gram-positive bacteria. Annual Reviews in Microbiology, 52(1), 41-79.

Sakamoto, N., Tanaka, S., Sonomoto, K., & Nakayama, J. (2011). 16S rRNA pyrosequencing-based investigation of the bacterial community in nukadoko, a pickling bed of fermented rice bran. International Journal of Food Microbiology, 144(3), 352- 359.

Scallan, E., Hoekstra, R. M., Angulo, F. J., Tauxe, R. V., Widdowson, M. A., Roy, S. L., Jones, J. L., & Griffin, P. M. (2011). Foodborne illness acquired in the United States— major pathogens. Emerging Infectious Diseases, 17(1), 7-15.

Scharff, R. L. (2012). Economic burden from health losses due to foodborne illness in the United States. Journal of Food Protection, 75(1), 123-131.

Shendure, J., & Ji, H. (2008). Next-generation DNA sequencing. Nature Biotechnology, 26(10), 1135-1145.

Shi, Y., Bueno, A., & van der Donk, W. A. (2012). Heterologous production of the lantibiotic Ala (0) actagardine in Escherichia coli. Chemical Communications, 48(89), 10966-10968.

Sobrino-López, A., & Martín-Belloso, O. (2010). Review: potential of high-intensity pulsed electric field technology for milk processing. Food Engineering Reviews, 2(1), 17-27. 43

Sperber, W. H. (2009). Introduction to the microbiological spoilage of foods and beverages. In W. H. Sperber, & M. P. Doyle (Eds.), Compendium of the microbiological spoilage of foods and beverages (pp. 1-40). New York, NY: Springer.

Stachelhaus, T., Mootz, H. D., & Marahiel, M. A. (1999). The specificity-conferring code of adenylation domains in nonribosomal peptide synthetases. Chemistry & Biology, 6(8), 493-505.

Stawikowski, M., & Cudic, P. (2007). Depsipeptide synthesis. In G. B. Fields (Ed.), Peptide characterization and application protocols (pp. 321-339). New York, NY: Springer.

Stawikowski, M., & Cudic, P. (2009). Lipodepsipeptide antibiotic fusaricidin and its analogues. total solid-phase and biological activity. In S. D. Valle, E. Escher & W. D. Lubell (Eds.), Peptides for youth: The proceedings of the 20th american peptide symposium (pp. 383-384). New York, NY: Springer.

Straus, S. K., & Hancock, R. E. W. (2006). Mode of action of the new antibiotic for Gram- positive pathogens daptomycin: comparison with cationic antimicrobial peptides and lipopeptides. Biochimica Et Biophysica Acta (BBA)-Biomembranes, 1758(9), 1215-1223.

Tabor, A. B. (2011). The challenge of the lantibiotics: synthetic approaches to thioether- bridged peptides. Organic & Biomolecular Chemistry, 9(22), 7606-7628.

USDA-FSIS. (2012). Verification activities for non-o157 shiga toxin-producing escherichia coli (non-o157 STEC) under MT60, MT52, AND MT53 sampling programs. Retrieved 1/17, 2013, from http://www.fsis.usda.gov/OPPDE/rdad/FSISNotices/63-12.pdf

Van Vliet, A. H. M. (2009). Next generation sequencing of microbial transcriptomes: challenges and opportunities. FEMS Microbiology Letters, 302(1), 1-7.

Velkov, T., Thompson, P. E., Nation, R. L., & Li, J. (2010). Structure—Activity Relationships of Polymyxin Antibiotics. Journal of Medicinal Chemistry, 53(5), 1898.

Walsh, C. (2003). Antibiotics: Actions, origins, resistance. Washington, DC: American Society for Microbiology (ASM) Press.

Wang, Z., Gerstein, M., & Snyder, M. (2009). RNA-Seq: a revolutionary tool for transcriptomics. Nature Reviews Genetics, 10(1), 57-63.

Willey, J. M., & Van der Donk, W. A. (2007). Lantibiotics: peptides of diverse structure and function. Annual Review of Microbiology, 61, 477-501.

Xie, L., Miller, L. M., Chatterjee, C., Averin, O., Kelleher, N. L., & van der Donk, W. A. (2004). Lacticin 481: in vitro reconstitution of lantibiotic synthetase activity. Science, 303(5658), 679-681.

You, Y. O., & van der Donk, W. A. (2007). Mechanistic investigations of the dehydration reaction of lacticin 481 synthetase using site-directed mutagenesis. Biochemistry, 46(20), 5991-6000.

Yu, J., Fedorova, N. D., Montalbano, B. G., Bhatnagar, D., Cleveland, T. E., Bennett, J. W., & Nierman, W. C. (2011). Tight control of mycotoxin biosynthesis gene expression in Aspergillus flavus by temperature as revealed by RNA‐Seq. FEMS Microbiology Letters, 322(2), 145-149.

Zhang, Q., Lambert, G., Liao, D., Kim, H., Robin, K., Tung, C., Pourmand, N., & Austin, R. H. (2011). Acceleration of emergence of bacterial antibiotic resistance in connected microenvironments. Science, 333(6050), 1764-1767.

Zheng, L., & Sun, D. W. (2004). Vacuum cooling for the food industry—a review of recent research advances. Trends in Food Science & Technology, 15(12), 555-568.

Chapter 2 Isolation of a Paenibacillus thiaminolyticus Strain and Structural Elucidation

of Its Broad-spectrum Antibacterial Agent, Paenibacterin

Abstract

A new bacterial strain, designated as OSY-SE, which produces a unique and potent antimicrobial agent was isolated from a soil sample. The isolate was identified as

Paenibacillus thiaminolyticus through cultural, biochemical, and genetic analyses. The antimicrobial agent was extracted from bacterial cells with acetonitrile, and purified using liquid chromatography. After analyses by mass spectrometry (MS) and nuclear magnetic resonance (NMR), the antimicrobial compound was determined to be a cyclic lipopeptide consisting of a C15 fatty acyl (FA) chain and thirteen amino acids. The deduced sequence is:

FA-Orn-Val-Thr-Orn-Ser-Val-Lys-Ser-Ile-Pro-Val-Lys-Ile. The carboxyl terminal Ile is connected with Thr by ester linkage. The new compound, designated as paenibacterin, showed activity against Gram-positive and Gram-negative bacteria, including Listeria monocytogenes, methicillin-resistant Staphylococcus aureus, Escherichia coli O157: H7, and

Salmonella enterica serovar Typhimurium. Paenibacterin is resistant to trypsin, lipase, α- glucosidase and lysozyme. But the compound is sensitive to pronase and polymyxin acylase.

Paenibacterin is readily soluble in water, and fairly stable to exposure to heat and a wide 46

range of pH values.

2.1. Introduction

There is a constant need for novel, safe and effective antimicrobial agents because bacterial can evolve rapidly and acquire resistance to antibiotics in use today. Bacterial natural products remain the important source of novel antibiotics with new scaffold and mechanism of action (Clardy, Fischbach, & Walsh, 2006). Since thousands of antibiotic compounds have been discovered from soil microorganisms (Clardy et al., 2006), new strategies should be used to combat the rediscovery of old antibiotics (Clardy et al., 2006;

Fischbach & Walsh, 2009). In this study, we used two non-conventional media, soil-extract agar (Hamaki et al., 2005) and dilute nutrient agar (Janssen, Yates, Grinton, Taylor, & Sait,

2002), to isolate new bacteria strains producing antibacterial agents. Using this new culturing media, a new isolate, designated as Paenibacillus thiaminolyticus OSY-SE, was found from a soil sample producing novel antimicrobial agent. The new compound, paenibacterin, exhibits antimicrobial activities against both Gram-positive and Gram-negative bacteria.

Paenibacillus was originally classified under Bacillus and was reclassified as a new genus (Ash, Priest, & Collins, 1993). Paenibacillus spp. are spore-forming bacteria that are widely distributed in environments such as soil and plant root. Some species promote plant growth and nitrogen fixation (Khianngam, Akaracharanya, Tanasupawat, Lee, & Lee, 2009;

McSpadden Gardener, 2004; Timmusk & Wagner, 1999; von der Weid, Duarte, van Elsas, &

Seldin, 2002). A wide variety of antimicrobial agents, including lantibiotics (He et al., 2007), 47

lipopeptides (Martin et al., 2003), and macrolide antibiotic (Wu et al., 2010) are produced by strains of Paenibacillus species.

2.2. Materials and methods

2.1.1. Strain screening

Soil samples collected in Columbus, OH in 2011 were screened for bacteria that produce antimicrobial agents. Samples were suspended in 0.1% peptone water and homogenized in a stomacher. The soil suspensions were decimally diluted and aliquots (100

μl) of dilutions were spread plated on soil-extract agar (Hamaki et al., 2005) and dilute nutrient agar (Janssen et al., 2002). The inoculated agar plates were sealed in a plastic bag and incubated at room temperature for two to eight weeks. Each new isolate was transferred to three tryptose agar plate (Becton Dickinson, Sparks, MD) and incubated at 30°C until new colonies were formed. Among the three new plates, two agar plates were used for testing the new isolates’ ability to produce antimicrobial agents, and the third plate was kept as a culture stock. Production of antimicrobials was examined using two bacterial indicators, Listeria innocua ATCC 33090 and Escherichia coli K12. The indicators were seeded in molten soft tryptic soy agar (Becton Dickinson, 0.75% agar) and overlaid on the plates with new isolates.

After incubation overnight at 37°C, an inhibition zone was generated on the lawn of indicator if a soil isolate produced an antimicrobial agent which diffused around the colony. A new isolate, designated as OSY-SE, produced antimicrobial agents and was used in the following study. 48

2.1.2. Strain identification

The cellular morphology of the new isolate, OSY-SE, was examined using Gram staining, spore staining with malachite green, and scanning electron microscopy (SEM). The following procedure was used to prepare the bacterial cells for SEM examination. The overnight culture of OSY-SE was washed three times using phosphate buffer (0.05 M, pH 7.0) and kept in a fixative (2.5% glutaraldehyde in 0.1M phosphate buffer with 0.1 M sucrose, pH

7.4) at 4°C overnight. Bacterial cells were harvested by centrifugation and resuspended in phosphate buffer. Then the cells were collected on a membrane filter (0.22 µm, Millipore

Corp., Bedford, MA) and received a post-fixation in 1% osmium tetroxide for 1 hour. After fixation, cells were dehydrated through ascending ethanol series, washed with ascending series of hexamethyldisilazane (HMDS) in ethanol, and air-dried. At last, bacterial cells were coated with a thin layer of gold-palladium using a Cressington 108 Sputter Coater (Ted Pella

Inc., Redding, CA), and visualized using a scanning electron microscope (NOVA NanoSEM

400, FEI, Hillsboro, OR). The accelerating voltage was 5 kV and images were collected from the emitted secondary electron signal.

The identity of OSY-SE was determined by sequencing its 16S ribosomal DNA

(rDNA). Genomic DNA of OSY-SE was extracted using a DNA extraction kit (DNeasy

Blood & Tissue kit; QIAGEN, Valencia, CA). The 16S rDNA gene was amplified by PCR using two universal primers specific for bacterial 16S rDNA (Weisburg, Barns, Pelletier, &

Lane, 1991). Amplification was performed using a Taq DNA polymerase kit (Qiagen) under the following conditions: the reaction mixture was subjected to an initial denaturation at

94°C for 3 min, followed by 30 cycles, including 1 min at 94°C, 1 min at 52°C and 2 min at

72°C. A final extension was carried out at 72°C for 10 min. The resultant PCR product was purified using a commercial DNA extraction kit (Qiaquick gel extraction kit; Qiagen), ligated to a vector (pGEM-T Easy; Promega, Madison, WI) and introduced into competent E. coli

DH5α cells by electroporation. The recombinant plasmid, carrying the 16S rDNA gene, was isolated from the overnight culture using spin columns (QIAprep Spin Miniprep kit; Qiagen).

The 16S rDNA gene in the plasmid was sequenced using a 3730 DNA analyzer (Applied

Biosystems, Foster City, CA) at the Plant-Microbe Genomics Facility, The Ohio State

University (Columbus, OH). The resultant 16S rDNA gene was compared to known bacterial sequences using the Basic Local Alignment Search Tool (BLAST) algorithm against the

Genbank database in National Center for Biotechnology Information (NCBI).

In addition, biochemical tests were conducted to confirm the identity of the new isolate. The tests included catalase test, oxidase test, nitrate reduction, production of acetylmethylcarbinol, dihydroxyacetone and indole, deamination of phenylalanine, and hydrolysis of starch and casein (Gordon, Haynes, Pang, & Smith, 1973). Two commercial biochemical test kits (API 50CH and API 20E strips; BioMerieux, Inc., Durham, NC) were also used to characterize the new isolate. The readings of the strips were recorded after incubation at 30°C for 24 and 48 h. The identification of the new isolate was done by 50

referring to the database provided by the kit manufacturer.

2.1.3. Isolation and purification of antimicrobial agents

The overnight culture of OSY-SE was spread-plated onto tryptose agar plates and incubated at 37°C for 4 days. Bacterial cells were collected from the surface of the agar plates. Antimicrobial agents produced by OSY-SE were extracted by acetonitrile with agitation at 200 rpm for 30 minutes. After extraction, bacterial cells were removed by centrifugation at 7710 × g for 15 minutes. The resultant supernatant with antimicrobial agents was evaporated in a chemical hood. Then the dry crude extract was dissolved in distilled water and was passed through a 0.22µm membrane filter (Millipore). The crude extract was purified using a high-performance liquid chromatography (HPLC) system (Hewlett Packard

1050, Agilent Technologies, Palo Alto, CA). The purification was achieved using a reverse- phase column with 5μm particle size (250×4.6 mm, Biobasic C18, Thermo Electron Corp.,

Bellefonte, PA). The mobile phase consisted of (A) acetonitrile with 0.1% trifluoroacetic acid

(TFA), and (B) HPLC-grade water with 0.1% TFA. For each run, aliquots (40 μl ) of crude extract were injected and separated on the column by a linear gradient of solvent A from 0 to

70% over 20 min at a flow rate of 1.0 ml/min. Elution was monitored using a UV-detector at a wavelength of 220 nm. Fractions were collected automatically using Waters Fraction

Collector II (Waters Cooperation, Milford, MA). Fractions with the same retention time from multiple runs were combined and air-dried in a chemical hood; the resulting powder was dissolved in sterile water for antimicrobial activity assay. 51

2.1.4. Antimicrobial activity determination

Spot-on-lawn method (He et al., 2007) was used to test the antimicrobial activity of purified antimicrobial agents from OSY-SE. Briefly, aliquots (10 µl) of indicator bacteria

(Table 7) were transferred into 10 ml of molten soft tryptic soy agar (0.75% agar), which was then poured onto a tryptose agar plate. After the agar solidified, aliquots (10 µl) of two-fold diluted antimicrobial agents from OSY-SE were spotted on the top layer of agar seeded with indicator bacterium. The agar plate was incubated overnight and was examined for the presence/absence of the inhibitory zone. Antimicrobial activity was expressed in arbitrary unit (AU/ml); one arbitrary unit was defined as the reciprocal of the highest dilution demonstrating antimicrobial activity.

2.1.5. Sensitivity to heat, pH and enzymes

The crude extract of OSY-SE was tested for the resistance to heat and pH change. For heat stability test, crude extract were exposed to heat treatment at 37, 55 or 80°C for 24 hrs, or at 121°C in an autoclave for 5 min. For pH stability test, the diluted crude extracts in 25 mM phosphate buffer (pH 7.0) were adjusted to pH 3.0, 5.0 or 9.0 and incubated for 12 hrs.

After exposure to different pH, the crude extracts were neutralized to pH 7.0 before testing for remaining antimicrobial activity.

Purified antimicrobial agents from OSY-SE (~1mg/ml) were tested for resistance to different enzymes, including trypsin (type I, 12705 U/mg), lipase (type I, 9 U/mg), pronase

(6.31 U/mg), α-glucosidase (type I, 50 U/mg), lysozyme (46400 U/mg) and polymyxin acylase (16 U/mg). Polymyxin acylase was obtained from (Wako Chemicals USA, Inc.,

Richmond, VA) and all other enzymes were purchased from Sigma (St. Louis, MO).

Polymyxin acylase was prepared at a concentration of 0.2 mg/ml in 50 mM phosphate buffer

(pH 8.0), whereas other enzymes were prepared at 1.0 mg/ml in 50 mM phosphate buffer (pH

7.0). Aliquots (20µl) of antimicrobials were mixed with an equal amount of one enzyme solution and incubated at 37°C for 10 hrs. The spot-on-lawn method was used to measure the remaining antimicrobial activity after these treatments.

2.1.6. Alkaline hydrolysis

A mild alkaline hydrolysis was used to open the macrolactone ring within the purified cyclic peptide. The peptide was dissolved in 1 M NaOH and incubated at 25°C for 12 hrs.

After acidification, the solution was desalted by a peptide desalting trap (Michrom

BioResources Inc., Auburn, CA) and the ring-opened compound was analyzed by MALDI-

TOF MS and MS/MS as described later.

2.1.7. MALDI-TOF MS analysis

MALDI-TOF MS analysis was performed on a mass spectrometer (Bruker Reflex III time-of-flight, Bruker Daltonics Inc., Billerica, MA). Briefly, a sample of the purified antimicrobials was mixed with a matrix at a ratio of 1:5. The matrix is α-cyano-4-hydroxy cinnamic acid, prepared as a saturated solution in 50% acetonitrile with 0.1% TFA in water.

The mixture was then spotted (1µl) on the target plate and allowed to air dry. The instrument 53

was operated in reflection-positive ion mode at an accelerating voltage of 28 kV. The N2 laser was operated at the minimum threshold level required to generate signal and minimize dissociation.

2.1.8. Quadrupole-time of flight MS/MS

The MS/MS analysis was performed on a Micromass Q-Tof II apparatus (Micromass,

Wythenshawe, UK) equipped with an orthogonal electrospray source (Z-spray) and operated in positive ion mode. The instrument was calibrated with Angiotensin fragment prior to use.

A sample of purified antimicrobial agent, diluted in the mixture of H2O-ACN-HAc

(50:50:2.5), was infused into the electrospray source at a flow rate of 2µl/min. To achieve the optimal electrospray, capillary voltage was set at 3 kV, source temperature was 100°C, and cone voltage was 40 V. The first quadrupole, Q1, was set to pass ions between 200 and 2500 m/z. The target ion was isolated and fragmented within the second quadrupole. A voltage of

20 to 40 V was adjusted for the best quality of tandem MS spectra. The fragment ions were then analyzed in the time-of-flight tube (100-2000 m/z). Data were acquired in continuum mode until well-averaged data were obtained.

2.1.9. NMR analysis

The compound was subjected to 1D and 2D NMR analysis using a standard protocol

(Wuthrich, 1986) to determine the constituents and their sequential arrangement. The first

NMR sample was prepared by dissolving ~1 mg of the purified antimicrobial agent into 500

µl 90% H2O/10% D2O (referred to as H2O, hereafter). This sample was then lyophilized and 54

reconstituted into 500 µl 100% D2O for a parallel NMR data set. The second NMR sample contained ~5 mg of the pure compound dissolved into 500 µl 99.8% CD3OD (Cambridge

Isotope Inc., Andover, MA). Unless stated otherwise, NMR experiments were performed at room temperature on a Bruker DMX-600 spectrometer (Bruker, Karlsruhe, Germany) equipped with a 5-mm (1H, 13C, 15N) triple-resonance probe and three-axis gradients. NMR experiments included 2D 1H-homonulcear COSY, TOCSY (60 ms DIPSI2 mixing time), and

NOESY (200 ms mixing rime), 2D heteronuclear 1H-13C HSQC, multiplicity-edited 1H-13C

HSQC, 1H-13C HSQC-TOCSY (60 ms DIPSI2 mixing time), 1H-13C HSQC-NOESY (200 ms mixing time), 1H-15C HMBC, and 1H-15N HSQC, all using standard Bruker pulse sequences.

Water suppression was typically achieved using 3-9-19 WATERGATE technique (Sklenar,

Piotto, Leppik, & Saudek, 1993) for the sample dissolved in H2O, or presaturation to suppress residual HDO signal for the sample in D2O or CD3OD. NMR Data were processed with NMRPipe (Delaglio et al., 1995) and visualized using NMRView (Johnson & Blevins,

1994). Data were typically zero-filled prior to application of window functions followed by

Fourier transform. Chemical shifts were referenced externally to sodium 2,2-dimethyl-2- silapentane-5-sulfonate (DSS) at 0.00 ppm.

2.1.10. GC/MS analysis for confirmation of acyl moiety

The fatty acid component of the new compound was determined using GC/MS analysis. The fatty acid chain of the compound was removed by polymyxin acylase in phosphate buffer (pH 8.0) at 37° for 24 hrs. After acidification to pH 3.0, the free fatty acids 55

from the antimicrobial compound were extracted from the reaction mixture with chloroform

(Khianngam et al., 2009). The chloroform phase was washed with saturated sodium chloride solution and distilled water to remove impurities. Free fatty acids were obtained when the chloroform was blow dried with nitrogen gas. The resultant fatty acids were dissolved in a methylating reagent (MethElute, Thermo Scientific, Bellefonte, PA) and separated by gas chromatography (TRACE2000 GC, Thermo-Finnigan, West Palm Beach, FL). The separation was achieved using a capillary column (DB-23: 30m × 0.25mm i.d. ×0.25 µm film thickness;

Agilent Technologies). The GC equipment was coupled with a mass-spectrometer (TRACE

MS, Thermo, West Palm Beach, FL) for identifying the fatty acids. Pentadecanoic acid

(Acros organics, New Jersey) was dissolved in the methylating reagent and analyzed as a reference compound.

2.1.11. LC/MS/MS analysis

The antimicrobial compound was digested by sequencing-grade trypsin (Promega,

Madison, WI) in 100 mM NH4HCO3 buffer (pH 8.0) at 37°C overnight. The reaction was quenched by adding 0.1% TFA. The digested compound was analyzed by LC/MS/MS for amino acid sequence determination. Capillary-liquid chromatography-nanospray tandem mass spectrometry was performed on a mass spectrometer (LTQ orbitrap, Thermo -Finnigan) equipped with a nanospray source, operated in positive ion mode (Michrom Bioresources Inc,

Auburn, CA). Samples were separated on a capillary column (0.2 ×150 mm Magic C18AQ,

3µ, 200Å, Michrom Bioresources Inc., Auburn, CA) using a HPLC system (UltiMate™ 3000, 56

LC-Packings-A Dionex Co., Sunnyvale, CA). Each sample was injected into the trapping column (LC-Packings), and desalted with 50 mM acetic acid for 10 minutes. The injector port was then switched to inject, and the peptides were eluted off the trap onto the column.

Mobile phase A was 0.1% formic acid in water and mobile phase B was 0.1% formic acid in acetonitrile. Flow rate was set at 2 µl/min. Typically, mobile phase B was increased from 2% to 50% in 30 min before increased again from 50% to 90% in 5 min and then kept at 90% for another 5 min before being decreased quickly to 2% in 1 min. The column was equilibrated at 2% of mobile phase B (or 98% A) for 30 min before the next sample injection. The

MS/MS was acquired with a nanospray source operated with a spray voltage of 2 kV and a capillary temperature of 175 °C is used. The scan sequence of the mass spectrometer was based on the data dependant TopTen (10 most intense peaks) method. Briefly, the analysis was programmed for a full scan recorded between 300 and 2000 Da and a MS/MS scan to generate product ion spectra to determine amino acid sequence in consecutive scans of the ten most abundant peaks in the spectrum. The resolution of full scan was set at 3×104 to achieve high mass accuracy MS determination. The collision-induced dissociation (CID) fragmentation energy was set at 35%.

2.3. Results

2.3.1. Isolation and identification of an antimicrobial-producing strain

A large number of isolates from soil samples were screened for activity against L. innocua ATCC 33090 and E. coli K-12. A new isolate, designated as OSY-SE, was found 57

showing activity against both Gram-negative and Gram-positive indicators (Table 7). The new isolate formed irregular and shiny colonies on tryptose agar and exhibited a facultative anaerobic behavior in broth. The strain OSY-SE is a rod-shaped (0.6 × 4.2 µm), Gram+, spore-forming bacterium (Figure 4). The bacterium formed ellipsoidal spores in swollen sporangia. Motile cells can be directly observed under light microscope.

Figure 4. Scanning electron microscope (SEM) examination of OSY-SE cells

The isolate is positive for catalase test, oxidase test, and hydrolysis of starch and casein but is negative for nitrate reduction, deamination of phenylalanine, and production of acetylmethylcarbinol, dihydroxyacetone and indole. Genetic analysis indicated this strain belongs to the genus Paenibacillus. Its 16S rDNA sequence shares high similarity with that of Paenibacillus apiarius (99%), Paenibacillus alvei (96%) and Paenibacillus thiaminolyticus (95%). Carbohydrates fermentation analysis (API 50CH strips) showed that

OSY-SE has 96.2% similarity with P. thiaminolyticus. In addition, biochemical tests from

API 20E strips indicated that the isolate was positive for β-galactosidase, H2S production and urea hydrolysis, but was negative for other reactions. When the results from API 50CH and

API 20E were combined, the similarity of OSY-SE with P. thiaminolyticus increased to

99.9%. Therefore, the new isolated was identified as P. thiaminolyticus OSY-SE. However, the new isolate OSY-SE is not identical to the typical P. thiaminolyticus strains in some biochemical tests. For example, OSY-SE was negative for nitrate reduction, indole production and citrate utilization, but was positive for urease and H2S production (Table 6). Table 6. Biochemical properties of Paenibacillus OSY-SE.

Tests Results Tests Results Catalasea + β-galactosidaseb + Oxidasea + Arginine dihydrolaseb - Hydrolysis of starcha + Lysine decarboxylaseb - Hydrolysis of caseina + Ornithine decarboxylaseb - Dihydroxyacetonea - Citrate utilizationb - a b Deamination of phenylalanine - H2S production + Nitrate reductionab - Ureaseb + Acetylmethylcarbinolab - Tryptophan deaminaseb - Indoleab - Gelatinaseb - aThese biochemical tests were done according to Gordon et al. (1973). bThese biochemical tests were done using API 20E strips.

2.3.2. Purification of antimicrobial agents produced by Paenibacillus OSY-SE

Antimicrobial agents produced by OSY-SE were extracted with acetonitrile from the bacterial cells grown on tryptose agar. The crude extract in water was separated by HPLC. As shown in Figure 5, a single-peak fraction with retention time of 17.02 min showed antimicrobial activity against L. innocua and E. coli. MALDI-TOF MS analysis indicated

that the fraction contained a major compound with molecular mass of 1604 Da, which was designated as paenibacterin, and three minor compounds with molecular masses of 1590,

1618 and 1632 Da (Figure 6A). MS/MS analysis indicated that the fragmentation patterns of the major and minor compounds were very similar, which suggested that paenibacterin and the three minor components were homologues.

Figure 5. High performance liquid chromatography profile of the crude extract of OSY- SE cells Peak with retention time of 17.02 min (indicated by the arrow) showed antimicrobial activity against Listeria innocua and Escherichia coli.

Figure 6. MALDI-TOF MS analysis of paenibacterin and its linear form produced by alkaline hydrolysis. (A) Spectrogram showing paenibacterin (m/z 1605.14) and its three homologues (m/z 1591.12, 1619.17 and 1633.21); (B) Linearized paenibacterin (m/z 1622.97); ion at m/z 1644.96 was corresponded to sodium adduct.

2.3.3. Antimicrobial spectrum and stability

The HPLC purified paenibacterin was used for the antimicrobial spectrum test. The selected microorganisms included Gram-positive and Gram-negative bacteria, including some foodborne and other pathogens (Table 7). Paenibacterin showed activity against the most of tested microorganisms. Paenibacterin exhibited antimicrobial activity against Gram-

negative bacteria including E. coli O157:H7, Salmonella, Pseudomonas and Yersinia. The antimicrobial agent was active against several Gram-positive pathogens including L. monocytogenes and a methicillin-resistant S. aureus strain. Paenibacterin was active against one of two Clostridium difficile strains, but showed no activity against Enterococcus faecalis

ATCC 29212 using the spot-on-lawn method.

The crude extract of OSY-SE was resistant to heat and changes in pH. The crude extract retained its antimicrobial activity after exposure to heat at 80°C for 24 hrs or even more aggressive treatment at 121°C for 5min in an autoclave. The antimicrobial agents were also resistant to acidic (pH 3.0) or alkaline (pH 9.0) conditions. The purified paenibacterin was resistant to trypsin, lipase, α-glucosidase and lysozyme. However, the compound was sensitive to pronase (a mixture of proteinases), and polymyxin acylase which deacylates lipopeptides (Misumi, Tsuruta, Furuishi, & Shoji, 1995). Therefore, inactivation by pronase and polymyxin acylase suggested that paenibacillin is a lipopeptide antibiotic. Table 7. Relative antimicrobial activity of purified paenibacterin against selected bacteria

Antimicrobial activity Straina Broth mediumd (AU/ml)e Gram-negative bacteria Escherichia coli K-12 LB 3200 E. coli O157:H7 EDL 933 LB 1600 E. coli O157:H7 ATCC 43889 LB 1600 Pseudomonas putida ATCC 45491 TSBYE 400 Salmonella enterica ser. Typhimurium TSBYE 400 S. enterica ser. Typhimurium DT 109 TSBYE 400 S. enterica ser. Enteritidis TSBYE 800 Yersinia enterocolitica TSBYE 1600

Table 7. continued. Gram-positive bacteria Bacillus cereus ATCC 14579 TSBYE 800 B. cereus ATCC 11178 TSBYE 200 Clostridium difficile A515b BHIYE 200 C. difficile CL148c BHIYE 0 Enterococcus faecalis ATCC 29212 MRS 0 Listeria monocytogenes Scott A TSBYE 800 L. monocytogenes OSY-8578 TSBYE 1600 L. innocua ATCC 33090 TSBYE 1600 Lactobacillus plantarum ATCC 8014 MRS 400 L. lactis ATCC 11454 MRS 800 Staphylococcus aureus ATCC 6538 TSBYE 100 S. aureus (methicillin-resistant) TSBYE 100 aStrains obtained from the culture collection of The Ohio State University food safety laboratory. bStrain obtained from J. T. Lejeune, College of Veterinary Medicine, The Ohio State University. cStrain obtained from W. A. Gebreyes, Department of Veterinary Preventive Medicine, The Ohio State University. dLB, Luria–Bertani medium; TSBYE, Tryptic soy broth supplemented with 0.6% yeast extract; MRS, Lactobacillus MRS broth; BHIYE, Brain heart infusion supplemented with 5% yeast extract (Rodriguez-Palacios & LeJeune, 2011). eArbitrary units/ml: reciprocal of the highest dilution displaying a zone of inhibition corresponding to 1ml of the non-diluted antimicrobial preparation.

2.3.4. Amino acid sequence of paenibacterin

Initially, MS/MS analysis has failed to sequence the antimicrobial agent due to the lack of fragmentation information, which suggested that the paenibacterin could be a cyclic compound. After the open-ring reaction under mild alkaline conditions, a peak with m/z at

1622.97 was observed (Figure 6B).The mass difference between the compound exposed to the open-ring reaction and the intact peptide was 18 Da, suggesting the compound has a ring structure that can be opened by mild alkaline hydrolysis. Further MS/MS experiment was performed on the open-ring compound using the Q-tof. Although more fragmentation

information was obtained, no conclusive result about the amino acid composition could be achieved. Therefore, we used NMR to elucidate the structure of paenibacterin.

Preliminary analysis of NH amide cross-peaks in 2D 1H-15N HSQC (Figure 7A) and

C protons in 2D 1H-13C HSQC (Figure 8A) indicated the presence of 13 amino acids for the

1/2 peptidyl fragment, including one proline residue, evidenced by the observation of CH2 .

The complete spin system of each amino acid was subsequently established from the COSY and TOCSY spectra. The results taken together with 2D 1H-13C HSQC and HMBC analyses led to identification of 3 Val, 2 Ile, 2 Ser, 1 Thr, 1 Pro, 2 Lys, and 2 Orn – the unnatural amino acid that has been reported previously(1). The sequence of these residues was first deduced by analyzing sequential NOEs such as HN(i)-HN(i+1), H(i)-HN(i+1) and H(i)-

HN(i+1). The observation of strong NOEs between Pro10 H and Val11 Hled to their sequential assignment as well as the identification of the trans-conformation adopted by

Pro10. However, the NOE-based sequential assignment could be equivocal, particularly when considering the cyclic nature of this peptide moiety as described later. For example, long-range NOEs such as the one between Thr3 HN and Ile13 HN could complicate the analysis without a prior knowledge of the linkage location (Figure 7B). Therefore a 2D 1H-

13C HMBC of very high quality was required for unambiguous sequence-specific assignments on the basis of 1H(i)-13C'(i+1) multiple-bond J-coupling correlations. A relatively large sample (~5 mg) of the purified antimicrobial agent was prepared for this insensitive 2D 1H-13C HMBC analysis. However, severe line broadening was observed when 64

the sample was dissolved in H2O. CD3OD was then used as the alternative NMR solvent, and the experiment was conducted on a Bruker DRX-800 spectrometer equipped with a cryoprobe. Some 2D experiments were also repeated to assist the NMR assignments. As shown in Figure 8B, almost all of the intra-residue 1H(i)-13C´(i) as well as sequential 13C´(i-

1)-1H(i) multiple-bond correlations have been observed, enabling the unequivocal determination of the peptide sequence as follows: Orn1-Val2-Thr3-Orn4-Ser5-Val6-Lys7-

Ser8-Ile9-Pro10-Val11-Lys12-Ile13.

Figure 7. NMR analysis of the peptidyl fragment in the amide region. 1 15 (A) 2D H- N HSQC recorded on the sample in H2O showing the 12 main-chain NH amide cross-peaks and a cluster of folded peaks (labeled as “f”) attributable to Arg, Lys or Orn + 1 sidechain NH3 group; (B) 2D H NOESY recorded on the same sample showing the amide region cross-peaks with assignment.

Figure 8. Elucidation of amino acid sequence and linkage of paenibacterin by HMBC. 1 13 13 (A) 2D H- C HSQC recorded on the sample dissolved in CD3OD showing the C resonances in the region between 45 and 75 ppm. The CH assignments of the thirteen amino     acids, together with Thr3 CH2 , Ser5 CH2 , Ser8 CH2 , and Pro10 CH2 assignments are labeled to assist the analysis of cross-peaks in (B). The unlabeled methyl group at 3.30/49.1 ppm (1H/13C) is attributed to the residual solvent methanol; (B) 2D 1H-13C HMBC acquired on the same sample showing the connectives associated with H protons. Sequential assignment is traced according to intra-residue Hα(i)-C'(i) and C'(i-1)-H(i) (marked by asterisk) multiple-bond J-coupling connectivities. The stretch starts from the fatty acid carbonyl carbon (“fat”) to Orn1 H, and ended withLys12 C´ to Ile13 H. Also noted by the broken lines are the long range J-couplings of Thr3 H-Thr3 C´ and Thr3 H-Ile13 C´. The latter is the strong evidence for a cyclic peptide with an ester bond formed between the Thr3 hydroxyl group and the Ile13 C-terminal carboxylic group. It is important to note that the tilted and spit HMBC cross peaks are due to 1H-1H coupling (J-modulation).

2.3.5. Linkage elucidation

The above HMBC spectrum also revealed multiple-bond correlations of Thr3 H proton (5.49 ppm) with two carbonyl atoms: Thr3 C´ at 170.8 ppm and Ile13 C´ at 171.8 ppm

(Figure 8B). The latter suggests that Thr3 forms an ester linkage through its hydroxyl group to the C-terminal carboxylic group of Ile13. Consistently, both of Thr3 H and C chemical shifts experience unusual downfield shift similar to those “Threonine Shifts” reported for other lipopeptides in which cyclization involves a Thr side chain (Gerard et al., 1997;

Kajimura & Kaneda, 1996). Furthermore, this cyclic nature was supported by the long-range

NOEs that have been observed, such as the one between Thr3 H and Ile13 H. Finally, it appears that the peptide moiety possesses some rigid conformation, most likely adopting a - hairpin conformation. A tertiary structure of the peptidyl fragment was calculated using CNS software(4) with a total of 162 NMR constraints, including 156 NOE-derived distance constraints (84 intra-residue, 40 sequential, and 32 non-sequential ones) and six 1 constraints (V2, T3, V8, I9, V11 and K12) extracted from COSY and NOESY data sets, all from the NMR data recorded in aqueous solution. As shown in Figure 9, the residues of

Orn4-Val6 and Ile9-Lys12 form an anti-parallel -sheet, stabilized by hydrogen-bonds between Orn4 and Val11 as well as between Val6 and Ile9. It was also noticed that four of the five bulky aliphatic side chains (Val6, Ile9, Val11, and Ile13) group are on one side of the - sheet and interact with each other. This structural feature may contribute to the amphipathic nature of paenibacterin.

Figure 9. The tertiary structure of the peptide moiety of paenibacterin calculated from NMR constraints in aqueous solution. The five bulky aliphatic side chains (V2, V6, I9, V11 and I13) are highlighted.

2.3.6. Determination of the acyl moiety

Based on the peptide sequence derived from NMR, there was a discrepancy between the molecular mass of the thirteen amino acids and that of the whole compound. This indicated that paenibacterin contains other component, R (Figure 10). MS was then performed on b2 ion at m/z 339, which confirmed that it comprises R and Orn (data not shown). Therefore, the molecular mass of R was calculated as 225 Da, either from the molecular mass difference between the thirteen amino acids and linearized paenibacterin, or from b2 ion, and the formula of R was established as C15H29O. This suggested that

paenibacterin is a lipopeptide containing a saturated C15 fatty acid and thirteen amino acids.

The analysis of the 1D 13C NMR together with 2D 1H-13C HSQC and HMBC suggested that the fatty acid is a mixture of anteiso- and iso-branched forms, as evidenced by the presence of 13C peaks at 11.9 and 23.2 ppm, respectively (Figure 11) (Lin, Minton, Sharma, &

Georgiou, 1994). In 1D 13C NMR, the furthest downfield carbonyl carbon resonating at 180.1 ppm was assigned to the first atom of the fatty acid moiety. This C´ atom shows HMBC correlations to the first CH2 group at 2.30/38.2 ppm as well as the second CH2 group at 1.59,

1.56/28.1 ppm. More importantly, it also has a HMBC correlation to Orn1 H, indicating that the lipid chain is amidated to the N-terminal amine of Orn1. NOE was also observed between

Orn1 HN and the first methylene protons (2.30 ppm) of the fatty acid side chain. A thorough analysis of the NMR data sets, particularly 2D 1H-13C HSQC-TOCSY, HSQC-NOESY,

HMBC, and multiplicity-edited HSQC, led to the complete assignments of fatty acid side chains.

Figure 10. Fragmentation of b and y ion series of linearized paenibacterin, examined by MS/MS.

Figure 11. 1D 13C NMR spectrum revealing iso- and anteiso- fatty acyl chain.

2.3.7. Structure confirmation by GC/MS and LC/MS/MS

GC/MS and LC/MS/MS were performed to verify the acyl moiety and the peptide sequence of paenibacterin, respectively. The fatty acids were released from paenibacterin by polymyxin acylase treatment, and were analyzed by GC/MS after methylation. Three peaks with retention time of 4.87, 5.06 and 5.42 min were identified as methyl esters of iso-, anteiso- and normal chain C15 fatty acid, respectively. The identity of the compounds was determined by comparing with the chromatogram of pentadecanoic acid and referring their mass spectra to Wiley database (Figure 12). Although the normal chain fatty acid was not evident in the NMR analysis, it was detected by GC/MS in low abundance. The dominant fatty acid in the sample was anteiso-chain form, but iso- and normal branched forms were also detected. Therefore, the C15 fatty acyl chain of paenibacterin could be normal, iso or anteiso forms.

C:\Xcalibur\data\6-28-11\Blank7 7/5/2011 10:06:12 AM

RT: 0.00 - 19.48 16.31 NL: 100 (A) 4.43E7 TIC F: 90 MS Blank7 80

Relative Relative Abundance 18.79 30 17.00 18.36

20 C:\Xcalibur\data\6-28-11\SE2 6/28/20119.48 4:42:41 PM 16.14 6.77 10 4.29 5.40 5.78 15.80 15.22 6.87 7.81 8.47 9.58 11.80 12.02 13.98 RT: 0.000 - 19.48 C:\Xcalibur\data\6-28-11\SE2 6/28/2011 4:42:41 PM 0 2 4 6 8 10 12 14 16 16.26 18 NL: 100 3.87E7 Time (min) TIC F: RT: 900.00 - 19.48 16.26 NL:MS SE2 Blank7 #723 RT: 16.31 AV: 1 NL: 8.05E6 100 3.87E7 T: {0,0}80 + c EI det=200.00 Full ms [ 50.00-500.00]5.06 TIC F: 90 178.17 100 MS SE2 (B) 70 4.87 80 90 5.06 60 8070 4.87 50 7060 163.15 40 5.42

6050 Relative Relative Abundance 30 5040 18.73 5.425.73 6.70 9.41 20 Relative Relative Abundance 4.65 4030 4.26 16.64 5.86 17.98 18.73

Relative Relative Abundance 10 5.73 6.70 7.36 7.50 9.41 11.95 3020 8.15 10.52 11.73 14.59 15.34 16.18 4.26 161.154.65 16.64 0 119.13 5.86 17.98 2010 135.14 0 2 4 6 7.36 7.508 10 11.9512 354.2614 16 18 57.11 8.15 10.52 11.73 313.10 14.59 15.34 16.18 105.12 148.20 Time (min) 100 91.11 191.16 216.05 265.17 366.12 440.13 0 2 4 6 8 241.14 10 298.22 12 339.24 14 387.1116419.13 18 453.16 493.13 SE2 #063 RT: 5.06 AV: 1 NL: 4.79E6 Time (min) T: {0,0} 50 + c EI det=200.00100 Full ms [ 50.00-500.00]150 200 250 300 350 400 450 500 74.11 m/z SE2100 #63 RT: 5.06 AV: 1 NL: 4.79E6 T: {0,0} + c EI det=200.00 Full ms [ 50.00-500.00] 90 74.11 (C) 100 80 90 70 87.12 80 60 70 87.12 50 60 40 50 199.20 Relative Relative Abundance 30 40 55.13 120.15 143.15 199.20

Relative Relative Abundance 20 30 97.17 55.13 120.15 213.22 143.15 10 256.24 20 97.17 157.17 171.19 227.23 213.22 288.33 313.23 326.97 341.24 368.08 405.20 426.53 454.91 479.70 100 256.24 50 100 150157.17 171.19 200 227.23 250 300 350 400 450 500 288.33 313.23 326.97 341.24 368.08 405.20 426.53 454.91 479.70 0 m/z 50 100 150 200 250 300 350 400 450 500 Figure 12. Gas chromatography (GC) profilem/z of fatty acid methyl esters from paenibacterin. (A) GC profile of solvent used for derivatization, (B) GC profile of fatty acid methyl esters from paenibacterin, (C) mass spectrometry (MS) spectrum of anteiso- chain fatty acid methyl ester, corresponding to the peak with retention time of 5.06 min.MS spectrum.

The peptide sequence was confirmed by analyzing tryptic-digested paenibacterin

using LC/MS/MS. Paenibacterin was found to be resistant to trypsin based on antimicrobial

activity test in phosphate buffer (pH 7.0). However, digested products were detected,

including fragments VTOSVKSIPVKI, SVKSIPVKI and SIPVKI (Figure 13). It was noticed

that the linkage between Thr and C-terminal Ile was probably broken during incubation in the

NH4HCO3 buffer (pH 8.0) during enzyme digestion, as evidenced by the presence of

linearized paenibacterin in the same buffer without trypsin.

(A)

Figure 13. MS/MS spectra of tryptic-digested products of paenibacterin. (continued on next page)

Figure 13. continued (B)

(C)

Figure 13. MS/MS spectra of tryptic-digested products of paenibacterin. (A) VTOSVKSIPVKI, (B) SVKSIPVKI, (C) SIPVKI.

In summary, paenibacterin was identified as a lipopeptide consisting of a C15 fatty acyl chain (normal, iso or anteiso forms) and thirteen amino acids (Figure 14). The chemical shift assignments of the peptidyl fragment and the fatty acyl chain in aqueous solution are summarized in Table 8 and Table 9, respectively, while the corresponding ones in methanol- d4 are provided in Table 10 and Table 11.

Figure 14. The molecular structure of paenibacterin; the dotted cycle indicated the macrolactone ring formed between Thr3 and Ile13.

Table 8. Chemical shift assignments of peptidyl fragment of paenibacterin in aqueous solution at pH 4.5, 298.0 K.

Residue 1HN/15N 1H/13C (ppm) 1H/13C (ppm) Others 1H/13C (15N) and C´ (ppm)   Orn1 8.26/124.6 4.24/56.5 1.80,1.75/30.7 CH2 1.74, 1.67/26.1; CH2 3.00/41.6; C´ 177.0  Val2 7.92/118.3 4.09/62.8 2.07/32.8 CH3 1.19/22.1, 0.95/21.7; C´ 178.1  Thr3 8.67/114.5 4.93/58.9 5.50/74.4 CH3 1.14/17.6; C´ 172.0   Orn4 7.82/116.9 4.62/54.5 2.03,1.76/32.7 CH2 1.58, 1.53/24.6; CH2 2.96/41.6; C´ 173.9 Ser5 8.49/113.9 5.31/57.4 3.57, 3.41/65.8 C´ 172.9  Val6 8.73/121.6 4.27/61.9 1.88/34.9 CH3 0.93/21.6, 0.90/20.7; C´ 176.9    Lys7 9.22/130.7 4.07/58.9 1.86/32.1 CH2 1.52, 1.49/25.0; CH2 1.70/29.0; CH2 2.99/41.7; C´ 178.0 Ser8 8.32/109.0 4.43/58.5 3.90, 3.86/63.1 C´ 173.7    Ile9 7.74/123.6 4.68/57.5 2.10/39.5 CH3 0.98/16.6; CH2 1.48,1.23/28.9; CH3 0.83/11.7; C´ 173.8   Pro10 4.70/63.0 2.36, 1.96/32.8 CH2 2.16, 1.97/27.5; CH2 3.95, 3.77/51.4 

77 Val11 8.28/114.0 4.86/59.6 2.33/36.0 CH3 1.01/22.0, 0.73/19.2; C´ 176.8    Lys12 8.45/118.8 4.59/56.8 2.05, 1.76/31.7 CH2 1.45, 1.42/24.9; CH2 1.67/29.1; CH2 2.99/41.7; C´ 177.3    Ile13 6.65/115.8 4.14/61.3 1.81/38.0 CH3 0.79/17.4; CH2 1.26,1.09/27.5; CH3 0.80/13.3; C´ 174.0

Table 9. Chemical shift assignments of fatty acyl chain of paenibacterin in aqueous solution at pH 4.5, 298.0 K.

Position Iso- 1H/13C (ppm) Anteiso- 1H/13C (ppm) 1 C´ 180.1 C´ 180.1 2 CH2 2.30/38.2 CH2 2.30/38.2 3 CH2 1.59,1.56/28.1 CH2 1.59, 1.56/28.1 4 CH2 1.26/31.5 CH2 1.26/31.5 5 CH2 ~1.25/31.5 CH2 ~1.25/31.5 6 CH2 ~1.25/31.5 CH2 ~1.25/31.5 7 CH2 ~1.25/31.5 CH2 ~1.25/31.5 8 CH2 ~1.25/31.5 CH2 ~1.25/31.5 9 CH2 ~1.25/31.5 CH2 ~1.25/31.5 10 CH2 1.26/29.2 CH2 ~1.25/31.5 11 CH2 1.27, 1.08/38.6 CH2 1.26/29.2 12 CH2 1.29/36.5 CH 1.14/41.2 13 CH 1.30, 1.10/31.7 CH2 1.521/29.04 14 CH3 0.81/13.4 CH3 0.82/24.8 15 CH3 0.81/21.5 CH3 0.82/24.8

Table 10. Chemical shift assignments of peptidyl fragment of paenibacterin in methanol-d4, 298.0 K.

Residue 1H/13C (ppm) 1H/13C (ppm) Others 1H/13C (15N) and C´ (ppm)   Orn1 4.267/54.61 1.768, 1.728/29.91 CH2 1.722/24.99; CH2 2.943/39.99; C´ 173.77  Val2 4.190/61.03 2.084/31.42 CH3 1.198/20.65, 0.954/19.94; C´ 175.25  Thr3 4.836/57.12 5.492/72.10 CH3 1.137/16.02; C´ 169.43   Orn4 4.636/52.62 2.050, 1.743/31.61 CH2 1.583, 1.554/23.37; CH2 2.915/40.12; C´ 172.10 Ser5 5.315/56.13 3.549, 3.419/64.32 C´ 170.52  Val6 4.370/59.75 1.927/33.44 CH3 0.969/19.64, 0.946/19.09; C´ 174.99    Lys7 4.001/57.59 1.842/30.90 CH2 1.582, 1.535/23.80; CH2 1.708/27.86; CH2 2.938/40.21; C´ 174.95 Ser8 4.440/56.75 3.938, 3.839/61.88 C´ 171.44    Ile9 4.572/56.22 2.295/37.49 CH3 1.001/14.87; CH2 1.627,1.245/25.77; CH3 0.877/10.24; C´ 172.67   Pro10 4.717/61.36 2.281, 1.948/31.43 CH2 2.197, 1.974/25.93; CH2 4.092, 3.784/49.33; C´ 174.27  Val11 4.887/57.85 2.232/34.81 CH3 1.055/20.62, 0.725/17.55; C´ 175.13

79    Lys12 4.896/55.05 2.110, 1.703/30.58 CH2 1.514/23.70; CH2 1.723/28.02; CH2 2.975/40.40; C´ 174.40    Ile13 4.114/59.42 1.732/37.18 CH3 0.835/15.88; CH2 1.396,1.158/26.68; CH3 0.869/11.40; C´ 170.37

Table 11. Chemical shift assignments of fatty acyl chain of paenibacterin in methanol-d4, 298.0 K. Position Iso- 1H/13C (ppm) Anteiso- 1H/13C (ppm) 1 C´ 176.43 C´ 176.43 2 CH2 2.256/36.73 CH2 2.256/36.73 3 CH2 1.598/26.84 CH2 1.598/26.84 4 CH2 1.321/30.39 CH2 1.321/30.39 5 CH2 ~1.294/30.74 CH2 ~1.294/30.74 6 CH2 ~1.294/30.74 CH2 ~1.294/30.74 7 CH2 ~1.294/30.74 CH2 ~1.294/30.74 8 CH2 ~1.294/30.74 CH2 ~1.294/30.74 9 CH2 ~1.294/30.74 CH2 ~1.294/30.74 10 CH2 1.266/28.07 CH2 ~1.294/30.74 11 CH2 1.312, 1.101/37.70 CH2 1.287/28.34 12 CH2 1.305/35.57 CH 1.174/40.10 13 CH 1.343, 1.142/30.47 CH2 1.521/29.04 14 CH3 0.880/11.57 CH3 0.875/22.92 15 CH3 0.861/19.48 CH3 0.875/22.92

2.4. Discussion

A new bacterial strain OSY-SE producing antimicrobial agents was isolated from a soil sample. The new isolate was identified as Paenibacillus thiaminolyticus based on its biochemical and genetic characteristics. The bacterium produces a new antimicrobial agent, paenibacterin, which is active against Gram-positive and Gram-negative pathogens.

Three homologues of paenibacterin are also produced by the bacterium and co-eluted with paenibacterin in a single HPLC peak. However, low abundance of these homologues made it difficult to elucidate their chemical structure clearly. One homologue with molecular mass of 1618 Da differed from paenibacterin either in the acyl chain or in the fourth amino acid. Similar observations were reported previously in other lipopeptides such as fusaricidins, surfactins, fengycins and iturins (Beatty & Jensen, 2002; Ongena &

Jacques, 2008).

Paenibacterin is readily soluble in water and probably forms micelles at high concentration in aqueous solution as observed by NMR. Initially, the cyclic nature of paenibacterin was assumed due to lack of MS/MS fragmentation, and then confirmed by mild alkaline hydrolysis. The acyl moiety is critical to the antimicrobial activity in lipopeptide. Deacylation of lipopeptide led to inactivation of the antibiotics while the activity can be restored by reacylation (Boeck, Fukuda, Abbott, & Debono, 1989; Debono et al., 1988). Malina & Shai (2005) demonstrated that attachment of aliphatic acid to biologically inactive cationic peptide can produce lipopeptide with antimicrobial activity. 81

Additionally, Majerle, Kidrič, & Jerala (2003) reported that development of acyl analogues of the peptide fragment of human lactoferrin enhanced its antibacterial activity.

The importance of fatty acid residue for antimicrobial activity was also confirmed in this study, in which paenibacterin lost activity after deacylation by polymyxin acylase.

Paenibacterin is a cationic lipopeptide containing four basic amino acids (two Orn and two Lys). Positive charges are helpful for the interaction of antimicrobials with bacterial cell membrane and for promotion of uptake of the antimicrobial agent (Hancock

& Chapple, 1999). It is inferred that the number of basic amino acid correlates with antimicrobial activity up to a limit that depends on the antimicrobial peptide (Findlay,

Zhanel, & Schweizer, 2010). Beside positive residues, amphipathic structure contributes to the antimicrobial activity of cationic peptides (Jenssen, Hamill, & Hancock, 2006). In spite of the fatty acyl residue, hydrophobic amino acids of paenibacterin are evenly distributed in the peptidyl fragment, and the compound adopted anti-parallel β sheet with hydrophobic chains on one side. This structure may help paenibacterin to disrupt the cytoplasmic membrane of target cell. Disruption of cytoplasmic membrane and leakage of cell components are considered as the mechanism of action of many cationic antimicrobial peptides, such as polymyxin, octapeptin and magainin (Dixon & Chopra,

1986; Duwe, Rupar, Horsman, & Vas, 1986; Matsuzaki, Murase, Fujii, & Miyajima, 1996;

Rosenthal, Swanson, & Storm, 1976). Polymyxin contains six diaminobutyric acids and has five net positive charges (Evans, Feola, & Rapp, 1999). Polymyxin is regarded as the 82

last line therapy against multi-drug resistant Gram-negative bacteria such as P. aeruginosa, Acinetobacter baumannii and Klebsiella pneumoniae (Velkov, Thompson,

Nation, & Li, 2010). However, polymyxin B has no activity against Gram-positive bacteria and anaerobes, whereas polymyxin E is not active against Gram-positive and

Gram-negative aerobic cocci, Gram-positive aerobic bacilli and all anaerobes (Falagas &

Kasiakou, 2005; Zavascki, Goldani, Li, & Nation, 2007). Compared to polymyxin, paenibacterin showed broader antimicrobial activity, including Gram-positive and anaerobic bacteria. In conclusion, the new isolate and associated antimicrobials are potentially useful in food or medical applications.

Abbreviations

MS, Mass spectrometry; MALDI-TOF MS, Matrix assisted laser desorption ionization time of flight mass spectrometry; MS/MS, Tandem mass spectrometry; 2D, two- dimensional; HMBC, Heteronuclear multiple-bond correlations; HSQC, Heteronuclear single-quantum correlations; NMR, Nuclear magnetic resonance; NOE, Nuclear overhauser effect; NOESY, NOE spectroscopy; TOCSY, Total correlation spectroscopy;

GC/MS, Gas chromatography/mass spectrometry; LC/MS/MS, Liquid chromatography/tandem mass spectrometry

References

Ash, C., Priest, F. G., & Collins, M. D. (1993). Molecular identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks and Collins) using a PCR probe test. Antonie Van Leeuwenhoek, 64(3), 253-260.

Boeck, L., Fukuda, D. S., Abbott, B. J., & Debono, M. (1989). Deacylation of echinocandin B by Actinoplanes utahensis. Journal of Antibiotics, 42(3), 382-388.

Clardy, J., Fischbach, M. A., & Walsh, C. T. (2006). New antibiotics from bacterial natural products. Nature Biotechnology, 24(12), 1541-1550.

Debono, M., Abbott, B. J., Molloy, R. M., Fukuda, D. S., Hunt, A. H., Daupert, V. M., Counter, F. T., Ott, J. L., Carrell, C. B., & Howard, L. C. (1988). Enzymatic and chemical modifications of lipopeptide antibiotic A21978C: the synthesis and evaluation of daptomycin (LY146032). The Journal of Antibiotics, 41(8), 1093-1105.

Dixon, R. A., & Chopra, I. (1986). Leakage of periplasmic proteins from Escherichia coli mediated by polymyxin B nonapeptide. Antimicrobial Agents and Chemotherapy, 29(5), 781-788.

Duwe, A. K., Rupar, C. A., Horsman, G. B., & Vas, S. I. (1986). In vitro cytotoxicity and antibiotic activity of polymyxin B nonapeptide. Antimicrobial Agents and Chemotherapy, 30(2), 340-341.

Evans, M. E., Feola, D. J., & Rapp, R. P. (1999). Polymyxin B sulfate and colistin: old antibiotics for emerging multiresistant gram-negative bacteria. The Annals of Pharmacotherapy, 33(9), 960-967.

Falagas, M. E., & Kasiakou, S. K. (2005). Colistin: the revival of polymyxins for the management of multidrug-resistant gram-negative bacterial infections. Clinical Infectious Diseases, 40(9), 1333-1341. 84

Findlay, B., Zhanel, G. G., & Schweizer, F. (2010). Cationic amphiphiles, a new generation of antimicrobials inspired by the natural antimicrobial peptide scaffold. Antimicrobial Agents and Chemotherapy, 54(10), 4049-4058.

Fischbach, M. A., & Walsh, C. T. (2009). Antibiotics for emerging pathogens. Science, 325(5944), 1089-1093.

Gerard, J., Lloyd, R., Barsby, T., Haden, P., Kelly, M. T., & Andersen, R. J. (1997). Massetolides A-H, antimycobacterial cyclic depsipeptides produced by two pseudomonads isolated from marine habitats. Journal of Natural Products, 60(3), 223-229.

Gordon, R. E., Haynes, W. C., Pang, C. H. N., & Smith, N. R. (1973). The genus bacillus. Washington, DC: Agricultural Research Service, US Department of Agriculture.

Hamaki, T., Suzuki, M., Fudou, R., Jojima, Y., Kajiura, T., Tabuchi, A., Sen, K., & Shibai, H. (2005). Isolation of novel bacteria and actinomycetes using soil-extract agar medium. Journal of Bioscience and Bioengineering, 99(5), 485-492.

Hancock, R. E. W., & Chapple, D. S. (1999). Peptide antibiotics. Antimicrobial Agents and Chemotherapy, 43(6), 1317-1323.

He, Z., Kisla, D., Zhang, L., Yuan, C., Green-Church, K. B., & Yousef, A. E. (2007). Isolation and identification of a Paenibacillus polymyxa strain that coproduces a novel lantibiotic and polymyxin. Applied and Environmental Microbiology, 73(1), 168-178.

Janssen, P. H., Yates, P. S., Grinton, B. E., Taylor, P. M., & Sait, M. (2002). Improved culturability of soil bacteria and isolation in pure culture of novel members of the divisions Acidobacteria, Actinobacteria, Proteobacteria, and Verrucomicrobia. Applied and Environmental Microbiology, 68(5), 2391-2396.

Jenssen, H., Hamill, P., & Hancock, R. E. W. (2006). Peptide antimicrobial agents. Clinical Microbiology Reviews, 19(3), 491-511.

Johnson, B. A., & Blevins, R. A. (1994). NMR View: A computer program for the visualization and analysis of NMR data. Journal of Biomolecular NMR, 4(5), 603- 614.

Kajimura, Y., & Kaneda, M. (1996). Fusaricidin A, a New Depsipeptide Antibiotic Produced by Bacillus polymyxa KT‐8. Taxonomy, Fermentation, Isolation, Structure Elucidation, and Biological Activity. Journal of Antibiotics, 49(2), 129-135.

Khianngam, S., Akaracharanya, A., Tanasupawat, S., Lee, K. C., & Lee, J. S. (2009). Paenibacillus thailandensis sp. nov. and Paenibacillus nanensis sp. nov., xylanase- producing bacteria isolated from soil. International Journal of Systematic and Evolutionary Microbiology, 59(3), 564-568.

Lin, S. C., Minton, M. A., Sharma, M. M., & Georgiou, G. (1994). Structural and immunological characterization of a biosurfactant produced by Bacillus licheniformis JF-2. Applied and Environmental Microbiology, 60(1), 31-38.

Majerle, A., Kidrič, J., & Jerala, R. (2003). Enhancement of antibacterial and lipopolysaccharide binding activities of a human lactoferrin peptide fragment by the addition of acyl chain. Journal of Antimicrobial Chemotherapy, 51(5), 1159-1165.

Malina, A., & Shai, Y. (2005). Conjugation of fatty acids with different lengths modulates the antibacterial and antifungal activity of a cationic biologically inactive peptide. Biochemical Journal, 390, 695-702.

Martin, N. I., Hu, H., Moake, M. M., Churey, J. J., Whittal, R., Worobo, R. W., & Vederas, J. C. (2003). Isolation, Structural Characterization, and Properties of Mattacin (Polymyxin M), a Cyclic Peptide Antibiotic Produced by Paenibacillus kobensis M. Journal of Biological Chemistry, 278(15), 13124-13132.

Matsuzaki, K., Murase, O., Fujii, N., & Miyajima, K. (1996). An antimicrobial peptide, magainin 2, induced rapid flip-flop of phospholipids coupled with pore formation and peptide translocation. Biochemistry, 35(35), 11361-11368.

McSpadden Gardener, B. B. (2004). Ecology of Bacillus and Paenibacillus spp. in agricultural systems. Phytopathology, 94(11), 1252-1258.

Misumi, S., Tsuruta, M., Furuishi, K., & Shoji, S. (1995). Determination of N-myristoyl peptide sequence both by MALDI TOF MASS and with an N-myristoyl cleaving enzyme (polymyxin acylase). Biochemical and Biophysical Research Communications, 217(2), 632-639.

Ongena, M., & Jacques, P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends in Microbiology, 16(3), 115-125. 86

Rodriguez-Palacios, A., & LeJeune, J. T. (2011). Moist-heat resistance, spore aging, and superdormancy in Clostridium difficile. Applied and Environmental Microbiology, 77(9), 3085-3091.

Rosenthal, K. S., Swanson, P. E., & Storm, D. R. (1976). Disruption of Escherichia coli outer membranes by EM 49. A new membrane active peptide antibiotic. Biochemistry, 15(26), 5783-5792.

Sklenar, V., Piotto, M., Leppik, R., & Saudek, V. (1993). Gradient-tailored water suppression for 1H-15N HSQC experiments optimized to retain full sensitivity. Journal of Magnetic Resonance.Series A, 102(2), 241-245.

Timmusk, S., & Wagner, E. G. H. (1999). The plant-growth-promoting rhizobacterium Paenibacillus polymyxa induces changes in Arabidopsis thaliana gene expression: a possible connection between biotic and abiotic stress responses. Molecular Plant- Microbe Interactions, 12(11), 951-959.

Velkov, T., Thompson, P. E., Nation, R. L., & Li, J. (2010). Structure—Activity Relationships of Polymyxin Antibiotics. Journal of Medicinal Chemistry, 53(5), 1898-1916. von der Weid, I., Duarte, G. F., van Elsas, J. D., & Seldin, L. (2002). Paenibacillus brasilensis sp. nov., a novel nitrogen-fixing species isolated from the maize rhizosphere in Brazil. International Journal of Systematic and Evolutionary Microbiology, 52(6), 2147-2153.

Weisburg, W. G., Barns, S. M., Pelletier, D. A., & Lane, D. J. (1991). 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology, 173(2), 697-703.

Wu, X. C., Qian, C. D., Fang, H. H., Wen, Y. P., Zhou, J. Y., Zhan, Z. J., Ding, R., Li, O., & Gao, H. (2010). Paenimacrolidin, a novel macrolide antibiotic from Paenibacillus sp. F6‐B70 active against methicillin‐resistant Staphylococcus aureus. Microbial Biotechnology, 4(4), 491-502.

Wuthrich, K. (1986). NMR of proteins and nucleic acids. New York, NY: Wiley Interscience.

Zavascki, A. P., Goldani, L. Z., Li, J., & Nation, R. L. (2007). Polymyxin B for the treatment of multidrug-resistant pathogens: a critical review. Journal of Antimicrobial Chemotherapy, 60(6), 1206-1215. 87

Chapter 3 Paenibacterin Biosynthesis in Paenibacillus thiaminolyticus: Identification

of the Synthetase Gene Cluster by Whole Genome Sequencing

Abstract

Paenibacterin is a novel cyclic lipopeptide antibiotic with activity against Gram- negative and Gram-positive pathogens. Paenibacterin consists of 13 amino acids with an

N-terminal C15 fatty acyl side chain. To elucidate the biosynthesis of paenibacterin in

Paenibacillus thiaminolyticus OSY-SE, we determined the whole genome sequence of the producer strain. Here we reported the DNA sequence of the paenibacterin biosynthetic gene cluster. The gene cluster was identified within 52-kb region, encoding three non-ribosomal peptide synthetases (PbtA, PbtB and PbtC) and two ABC- transporters (PbtD and PbtE). As deduced from the sequence data, each PbtA and PbtB enzyme consists of five modules, whereas PbtC is composed of three modules. Each of the 13 modules assembles one amino acid into the paenibacterin peptide. Sequence analysis revealed that Orn1, Orn4, Lys7 and Ser8 in paenibacterin may have D- configuration. The absolute configuration of two ornithine residues was confirmed by chirality analysis using Marfey's reagents. In addition, the substrate specificities of selected adenylation domains were confirmed by overexpression in Escherichia coli and 88

in vitro protein function analyses. Taking together, the findings allowed us to propose the biosynthetic pathway of paenibacterin.

3.1. Introduction

Bacterial natural products remain an important source of antibiotics with new scaffolds and mechanisms of action (Clardy, Fischbach, & Walsh, 2006). Paenibacterin is a natural antimicrobial peptide produced by a soil isolate, Paenibacillus thiaminolyticus

OSY-SE. The compound consists of 13 amino acids and a C15 fatty acyl moiety (chapter

2). However, lack of knowledge of the biosynthesis of paenibacterin has limited the study of structure-activity relationships of the new compound. This current study aimed to identify the gene cluster responsible for paenibacterin biosynthesis.

The presence of non-proteinogenic amino acids (ornithine) in paenibacterin

(chapter 2) suggests that this compound is synthesized by a non-ribosomal peptide synthetase (NRPS) pathway. Many important peptide antibiotics from bacteria and fungi are nonribosomal peptides. Prominent examples of such peptides include penicillin and cephalosporin families, vancomycin family (Fischbach & Walsh, 2006), polymyxin (Choi et al., 2009), daptomycin (Baltz, Miao, & Wrigley, 2005) and fusaricidin (Choi et al.,

2008; Li, Beatty, Shah, & Jensen, 2007). The NRPS machinery works as an assembly line where one module within the synthetases incorporates one amino acid into the peptide. A typical NRPS module comprises adenylation (A), thiolation (T) and condensation (C)

domains (Fischbach & Walsh, 2006). In the termination module, the Te-domain is responsible for peptide release and cyclization. In addition, some modules may possess epimerization (E) domain which converts the configuration of amino acids from L- to D- form during peptide synthesis (Fischbach & Walsh, 2006).

With the advance of DNA sequencing technology, the massively parallel DNA sequencing platforms (next-generation sequencing) have dramatically accelerated the biological research (Schuster, 2007). In the current study, Illumina sequencing system

(http://www.illumina.com) was used to determine the whole genome sequence of the paenibacterin producer, P. thiaminolyticus OSY-SE. The goal is to determine the gene cluster for paenibacterin biosynthesis using bioinformatic analyses. The function of the synthetases will be confirmed by heterologous expression of selected domains in E. coli for functional analyses.

3.2. Materials and Methods

3.2.1. Strains and media

The producer strain P. thiaminolyticus OSY-SE was obtained from the culture collection of food safety laboratory, The Ohio State University. The strain was grown in tryptic soy broth (Becton Dickinson, Sparks, MD) supplemented with 0.6% yeast extract

(TSBYE) at 37°C with agitation at 200 rpm. Escherichia coli DH5α or E. coli BL21

(DE3) was cultivated in Luria-Bertani broth (Becton Dickinson) or on the Luria-Bertani agar plate at 37°C. When appropriate, Luria-Bertani media for E. coli were supplemented 90

with 100 μg/ml ampicillin.

3.2.2. Whole genome sequencing of producer strain

The whole genome sequence of P. thiaminolyticus OSY-SE was determined using the next-generation sequencing technology. Briefly, genomic DNA of the bacterial strain was isolated using a DNA extraction kit (DNeasy Blood & Tissue kit; Qiagen,

Valencia, CA). The RNase-treated genomic DNA in Tris-Cl buffer (pH 8.5) was used for construction of a paired-end library with a Truseq DNA sample preparation kit (Illumina,

San Diego, CA) according to the manufacturer’s instruction. The constructed library was sequenced (76 cycle paired-end runs) in a flow cell lane using an Illumina Genome

Analyzer II. The de novo assembly of the short reads into longer contigs was performed using a commercial software program (CLC Genomics Workbench 4.7.2; CLCBio,

Cambridge, MA).

Automatic genome annotation was performed using the Rapid Annotation using

Subsystem Technology (RAST) (Aziz et al., 2008). The overall GC contend of the genome was calculated by the software Artemis (Rutherford et al., 2000). The rRNA and tRNA genes were predicted using RNAmmer (Lagesen et al., 2007) and tRNAscan-SE

(Lowe & Eddy, 1997), respectively. The average nucleotide identity (ANI) between P. thiaminolyticus OSY-SE and 20 Paenibacillus genomes in the Genbank was determined using in silico DNA-DNA hybridization (DDH) method implemented in the software

JSpecies (Richter & Rosselló-Móra, 2009). 91

3.2.3. Mining the paenibacterin gene cluster in the whole genome

It was assumed that paenibacterin is likely synthesized by the NRPS pathway because the peptide contains non-proteinogenic ornithine residues. Since NRPSs in different microorganisms are conserved, a local BLASTX algorithm was employed to search for the NRPSs gene cluster within the genome of P. thiaminolyticus OSY-SE. The

BLASTX program used for NRPS mining is implemented in a commercial software

(CLC Genomics Workbench 4.7.2). The NRPS for fusaricidin biosynthesis (7,908 amino acids, accession number: ABQ96384) in Paenibacillus polymyxa was used as a driver sequence for the BLASTX analysis.

The substrate specificity of adenylation domains encoded by presumptive NRPS gene was predicted using a websever, NRPSpredictor2 (Rausch, Weber, Kohlbacher,

Wohlleben, & Huson, 2005; Röttig et al., 2011). The NRPSpredictor2 service is available at http://nrps.informatik.uni-tuebingen.de/. The putative paenibacterin NRPS gene cluster was found in four non-overlapping contigs in the draft genome. The gaps among contigs were closed by PCR and Sanger-DNA sequencing. Primers were derived from the terminal DNA sequence of contigs. The oligonucleotide sequences of primers (V2F and

V2R, V6F and V6R, V11F and V11R) are listed in Table 12. PCR was carried out under the following conditions: initial denaturation at 94°C for 3 min, followed by 30 cycles of denaturing at 94°C for 1 min, annealing at 55°C for 1 min, and extension at 72°C for 3 min. The final extension was performed at 72°C for 10 min. The resultant PCR products

were sequenced using a DNA analyzer (3730 DNA analyzer; Applied Biosyntems, Foster

City, CA) at the Plant-Microbe Genomics Facility, The Ohio State University.

In addition, the epimerization (E) domains and thiolation (T) domains within

NRPSs were predicted by another webserver, PKS/NRPS analysis, which is available at http://nrps.igs.umaryland.edu/nrps/ (Bachmann & Ravel, 2009). Predictions of the transmembrane helices of ABC-transporters encoded by the gene cluster were carried out using the TMHMM server (version 2.0) (Emanuelsson, Brunak, von Heijne, & Nielsen,

2007). Table 12. Primers used in the study.

Primers Nucleotide sequences V2F 5’-CAAACGGTTGACCTATGCGGAGCTGAAT-3’ V2R 5’-CCTGCACAAAGTGTGTCGGGATCATGTA-3’ V6F 5’-CCTGACTATCCGGAGGAACGGACTAACG-3’ V6R 5’-CCAGATCGAACGGGCGAATAAAGGAAC-3’ V11F 5’-TCATCTGCTTGCCATTCTGAACGATACG-3’ V11R 5’-TTGAACACATGCCGAATCTGCTCCTCTT-3’ PbtThr3_NdeF 5’-GGGAATTCCATATGTTGACGGCAGAAGAGAAG-3’ PbtThr3_XhoR 5’-GGGTATCCGCTCGAGTATATATTCCGTGCCGGT-3’ PbtPro10_NdeF 5’-GGGAATTCCATATGGTGACTGCCGAGGAGCAG-3’ PbtPro10_XhoR 5’-GGGTATCCGCTCGAGTACGAACTCCGCTCCGGT-3’

3.2.4. Determination of the absolute configuration of amino acids in paenibacterin

The absolute configuration of constituent amino acids in paenibacterin was determined using the Marfey's reagents (Marfey, 1984) with modifications. Briefly,

HPLC-purified paenibacterin (1 mg) was dissolved in 0.5 ml HCl (6 M) in a sealed glass tube and incubated overnight at 110°C to hydrolyze the paenibacterin peptide. The resulting free amino acids from acid hydrolysis was blow-dried with nitrogen gas, 93

followed by addition of 200 μl of 1% Marfey's reagents, namely 1-Fluoro-2,4- dinitrophenyl-5-L-alanine amide (FDAA, Sigma, St. Louis, MO), and 40 μl of 1.0 M sodium bicarbonate. The contents were mixed and incubated at 40°C in a water bath for 1 hour to form diastereomers of amino acids. After cooling to room temperature, 20 μl of 2

M HCl was added to the reaction mixture.

The L- and D-diastereomers of amino acids from FDAA derivatization were separated by HPLC system equipped with a reverse phase column (Biobasic C18, 250 ×

4.6 mm, 5 μm particle size; Thermo Electron Corp., Bellefonte, PA). The mobile phases consisted of (A) acetonitrile and (B) 50 mM triethylamine phosphate at pH 3.0.

Separation was achieved by a linear gradient of acetonitrile from 10% to 45% over 45 min at a flow rate of 1ml/min. Elution was monitored using an UV detector at a wavelength of 340 nm. Meanwhile, standard amino acids (Sigma or Acros Organics, New

Jersey, USA) with known configurations were used as control for derivatization and

HPLC separation. The absolute configurations of amino acids from paenibacterin were determined by matching the retention time with the diastereomers from standard amino acids.

3.2.5. Determination of adenylation domain substrate specificity

Amplification and cloning of A-domains

The gene encoding the third and tenth A-domain in the pbt gene cluster were amplified by PCR from genomic DNA of P. thiaminolyticus OSY-SE, using the high- 94

fidelity DNA polymerase (Phusion, NEB, Ipswich, MA). Two sets of primers

(PbtThr3_NdeIF and PbtThr3_XhoR; PbtPro10_NdeIF and PbtPro10_XhoR, Table 12) were used for amplifying the genes of the third and tenth A-domain, respectively. The forward primers carried an Nde I restriction site, whereas the reverse primer possessed an

Xho I restriction site. These restriction sites were added for cloning compatibility with the expression vector, pET15b (Novagen, Madison, WI). PCR was carried out under the following conditions: initial denaturation at 98°C for 30 seconds, followed by 35 cycles of denaturing at 98°C for 10 seconds, annealing at 65°C for 30 seconds, and extension at

72°C for 90 seconds. The final extension was performed at 72°C for 10 min. The resultant PCR products were purified using spin column (QIAquick gel extraction kit,

Qiagen, Valencia, CA), double-digested with Nde I and Xho I at 37°C for 5 hours. The digested PCR products were gel-purified, and ligated to vector pET15b (Novagen,

Madison, WI) that has been cut with the same enzymes. The ligation reaction was carried out with T4 ligase (NEB) at room temperature overnight. The ligation mixture was transformed into E. coli competent cells (DH5α, NEB) by heat shock at 42°C for 30 seconds. Subsequently, the confirmed recombinant plasmid (pET15b-Thr3 or pET15b-

Pro10) carrying the A-domain sequence was introduced into the expression host E. coli

BL21 (DE3) (NEB) by heat shock at 42°C for 30 seconds.

Overexpression and purification of A-domains

A fresh recombinant E. coli BL21 (DE3) culture (10 ml) was used to inoculate 95

500 ml Luria-Bertani broth supplemented with 100 μg/ml ampicillin. Bacterial cells were cultivated at 37°C with agitation at 200 rpm. When the cell density (OD600) reached ~0.5, the expression of A-domain was induced by adding isopropyl-b-D-thiogalactopyranoside

(IPTG) at a final concentration of 400 μM. After induction, cells were grown at a 25°C with agitation at 200 rpm overnight for protein expression. Bacterial cells were harvested by centrifugation at 3,074×g for 15 min at 4°C. The cell pellets were resuspended in 40 ml chilled equilibration buffer (50 mM sodium phosphate, 300 mM NaCl, pH 7.0). To facilitate protein extraction, lysozyme (final concentration at 0.75 mg/ml, Sigma) and

DNase I (40 μl, 2 Unit μl-1, NEB) were added to the cell suspension, followed by incubation on ice for 30 min. Subsequently, cells in a beaker held on ice were treated with an ultrasonic processor (36860 series, Cole Parmer, Chicago, IL) for three times (30 seconds each pulse with a 2-min pause between each burst, 50% power). The disrupted cells were centrifuged at 11,952×g for 20 min to pellet the insoluble materials. The supernatant was carefully transferred to a clear tube without disturbing the pellet. The recombinant A-domain proteins in the supernatant were purified using an immobilized metal affinity chromatography (IMAC) resin charged with cobalt (1ml, HisTALON gravity column, Clotech, Mountain View, CA). The column was equilibrated with 10 ml chilled equilibration buffer. After loading the supernatant, the column was washed with 8 ml equilibration buffer and 7 ml of wash buffer (i.e. equilibration buffer with 10 mM imidazole). The target proteins were eluted from the column with 5 ml elution buffer (i.e. 96

equilibration buffer with 150 mM imidazole). The purified A-domains were subjected to concentration and buffer exchange through ultrafiltration (10 kDa Ultracel-10 membrane,

Millipore, Billerica, MA). Ultrafiltration was performed by centrifugation at 5,050×g at

4°C for 30 min for 3 times. Sterile water was added to ultrafiltration unit between each centrifugation step. If not used immediately, A-domain proteins in water were mixed with glycerol (final concentration, 10%) and stored at -80°C. The protein concentration of A- domain was determined using a spectrophotometer (NanoDrop 1000, Thermo Scientific,

Franklin, MA). The cloning, expression and purification of A-domains are schematically presented in Figure 15.

Figure 15. schematic presentations of cloning, expressing and purification of A- domain proteins.

3.2.6. Amino acid specificities of purified A-domains

The substrate specificity of purified A-domain proteins was determined by malachite green colorimetric assay (McQuade et al., 2009) with minor modifications

(Figure 16). All 20 proteinogenic amino acids and one non-proteinogenic amino acid, ornithine, were tested in a 96-well plate. The reaction mixture (100 µl) included the following components: reaction buffer (50 mM NaCl, 10 mM MgCl2, 50 mM Tris-Cl, pH

7.4), A-domain protein (6.5 µM), ATP (100 µM, cat. no. A7699, Sigma), amino acid (0.3 mM for tyrosine, 6 mM for all other amino acids), and inorganic pyrophosphatase (0.2 units, cat. no. I1643, Sigma). The reaction was initiated by adding ATP as the last component, followed by incubation at 25°C for 20 min. During the reactions, the activation of substrate by A-domain proteins resulted in the release of pyrophosphate, which was then converted to phosphate by pyrophosphatase. The phosphate concentration was quantified by adding 25 µl of the malachite green reagent (cat. no. POMG-25H,

Bioassay Systems, Hayward, CA). After color development at 25°C for 20 min, absorbance at 600 nm was measured using a microtiter plate reader (Molecular Devices

Corp., Menlo Park, CA). Each enzyme assay was performed with two replicates.

Figure 16. malachite green colorimetric assay of substrate specificity

3.3. Results

3.3.1. Whole genome sequencing and bioinformatic analyses

Whole genome sequencing and assembly yielded 205 contigs (>200 bp, each) with a maximum contig size of 359, 285 bp. The draft genome of P. thiaminolyticus OSY-

SE consists of 6,931,767 bases with an overall GC contend of 48.66%. This Whole

Genome Shotgun project has been deposited at GenBank under the accession number

ALKF00000000. Among the 6,475 protein-coding sequences (CDSs), 65.79 % have been assigned putative function by RAST. The bacterium has one rRNA operon and 38 tRNAs.

In silico DNA-DNA hybridization indicated that P. thiaminolyticus OSY-SE has the closest genetic relatedness with Paenibacillus lactis strain 154 (77.21%, ANI).

3.3.2. Organization of the paenibacterin gene cluster (pbt)

The presence of two non-proteinogenic ornithine residues in paenibacterin suggests that the peptide is likely synthesized by multi-modular nonribosomal peptide synthetases. The paenibacterin biosynthetic gene cluster was identified within a 52-kb region, encoding three peptide synthetases (PbtA, PbtB and PbtC) and two ABC-like transporters (PbtD and PbtE) (Table 13). Sequence analyses revealed that PbtA, PbtB and

PbtC comprise five, five and three modules, respectively. Each module consists of a set of enzymatic C-A-T domains (Figure 17).

Table 13. The biosynthetic cluster in P. thiaminolyticus OSY-SE responsible for paenibacterin synthesis

ORFsa Bases Amino acids Calculated Predicted function number number molecular mass (Dalton) pbtA 19,818 6,605 745,827.1 Peptide synthetase pbtB 19,251 6,416 723,020.7 Peptide synthetase pbtC 9,513 3,170 356,884.2 Peptide synthetase pbtD 1,713 570 63,374.5 ABC transporter pbtE 1,749 582 64,584.7 ABC transporter a ORFs open reading frames

Adenylation domains and epimerization domains

Adenylation (A) domain in each module possesses a conserved binding pocket for amino acid recognition and activation (Challis, Ravel, & Townsend, 2000; Conti,

Stachelhaus, Marahiel, & Brick, 1997; Stachelhaus & Walsh, 2000). Based on the fingerprint residues at the substrate-binding site, the substrate specificity of A-domains in the paenibacterin synthetases was predicted using the software, NRPSpredictor2 (Rausch 100

et al., 2005). The paenibacterin NRPSs follow the collinear biosynthetic logic and the predicted substrate of each module are well consistent with the peptide sequence of paenibacterin as determined by NMR (Table 14). In addition, epimerization (E) domains were found in modules for Orn1, Orn4, Lys7 and Ser8, which indicated that these amino acids in the peptide may have D-configuration. One of the prominent characteristics of nonribosomal peptide is the presence of D-amino acids (Stachelhaus & Walsh, 2000).

Condensation domains

The condensation (C) domains catalyzed the peptide bond formation during

NRPS peptide biosynthesis. C-domains are classified into three functional subtypes based

L on the types of reaction catalyzed and the chirality of substrates: (i) CL domain catalyze

D peptide bond formation between two L-amino acids; (ii) CL domain add an L-amino acid to a growing peptide chain ending with a D-amino acid; (iii) starter C domain couple the fatty acyl moiety to the first amino acid in the peptide (Rausch, Hoof, Weber, Wohlleben,

L D & Huson, 2007). Both CL and CL domains have a conserved His-motif in the active site; the consensus residues in this motif are HHxxxDG, where x donates variable amino acids

(Table 14) (Rausch et al., 2007). This signature motif is critical for catalyzing the peptide bond formation (Konz & Marahiel, 1999).

L The paenibacterin NRPSs contain one starter C-domain, eight CL domains, three

D D CL domains and one CD domain (Table 14). The starter C-domain found in the first

101

module of PbtA may involve coupling the C15 fatty acyl side chain to the first ornithine

D residue in paenibacterin. The CL domain immediately downstream of E-domain differs from other C-domains in the variable residues in the HHxxxDG (Table 14).When D- amino acid residues are present in natural products, the corresponding L-isomers are

L selected by the A-domains in the four-domain CL-A-T-E module. The E domains will

D epimerize the selected substrates to D-amino acids. The downstream CL domains will accept the D-amino acids in the D-specific donor sites for peptide bond formation. The

D coordination of E-domains and the downstream CL domains has been demonstrated in tyrocidine synthetase (Clugston, Sieber, Marahiel, & Walsh, 2003).

Thiolation domains and thioesterase domain

Thiolation domain (the peptidyl carrier protein in NRPS), contains a consensus sequence (L/IGGH/DSL/I), in which the conserved serine residue covalently binds to amino acid substrate in the reaction center in NRPS (Schlumbohm et al., 1991). In the

NRPS of paenibacterin, LGGDS motifs, rather than the more common LGGHS, were found in the T-domains within modules which incorporate D-amino acids (Table 14). The specialized signature LGGDS motifs are important for the productive interaction with E- domains (Linne, Doekel, & Marahiel, 2001). In the last module of PbtC, the thioesterase

(Te) domain may catalyze the intramolecular cyclization of peptide, forming an ester bond between Ile13 and Thr3. The paenibacterin Te domain possesses a putative catalytic triad comprising Ser67, Asp94 and His197 residues, in which the Ser67 residue is located in 102

the signature GYSLG motif in Te domains (Table 14) (Bruner et al., 2002; Kohli &

Walsh, 2003).

Transporters

Two putative ATP-binding cassette (ABC) transporters (PbtD and PbtE) were found in the paenibacterin gene cluster. As predicted using the TMHMM server, both transporters are putative membrane-associated proteins with five membrane-spanning helices. PbtD and PbtE share 38% sequence similarity to each other. PbtD and PbtE are homologues of PmxC and PmxD transporters, which are encoded by polymyxin gene cluster (Choi et al., 2008). PmxC and PmxD are believed to mediate the transport of polymyxins and fusaricidins in Paenibacillus polymyxa (Shaheen, Li, Ross, Vederas, &

Jensen, 2011). PbtD and PbtE may involve secretion of paenibacterin or self-immunity.

103

Table 14. Conserved Motifs in adenylation (A), condensation (C), thiolation (T) and epimerization (E) domains from paenibacterin NRPS

Module Binding pocket in Predicted Residues in Conserved Subtype Conserved Motif Conserved A-domain Substrate Paenibacterin Motif in C- of in T-domain motif in domain C-domain E/Te- domain PbtA1 DVGEIGSVDK D-Orn Orn INHIIADGVT starter EHFFE LGGDSI FNYLGQa D PbtA2 DAFWLGGTFK Val Val SHHILMDGWC CL DSFFE LGGHSL L PbtA3 DFWNIGMVHK Thr Thr MHHIISDGAS CL DNFFE LGGHSL L a PbtA4 DVGEIGSVDK D-Orn Orn MHHIISDGVS CL DHFFE LGGDSI FNYLGQ D PbtA5 DVWHFSLVDK Ser Ser SHHILMDGWC CL DDFFE LGGHSL L PbtB1 DAFWLGGTFK Val Val MHHIISDGVS CL DSFFE IGGHSL 104 L a PbtB2 DVGDVGSIDK D-Orn/ Lys/ Lys MHHIISDGVS CL DHFFE LGGDSI FNYLGQ Agr D a PbtB3 DVWHFSLVDK D-Ser Ser SHHILMDGWC CD DHFFE LGGDSI FNYLGQ D PbtB4 DGFFLGVVFK Ile Ile SHHILMDGWC CL DNFFE LGGHSL L PbtB5 DVQYIAHVVK Pro Pro MHHIVSDGTS CL DNFFDLGGHSL L PbtC1 DAFWLGGTFK Val Val MHHIISDGAS CL DSFFE IGGHSL L PbtC2 DVGDVGSIDK Orn/ Lys/ Agr Lys MHHIISDGVS CL DNFFDLGGHSL L b PbtC3 DGFFLGVVFK Ile Ile MHHIISDGVT CL DNFFE LGGHSI GYSLG

a. conserved motif in E-domain. b. conserved motif in Te-domain.

104

Figure 17. Schematic overview of paenibacterin biosynthesis in P. thiaminolyticus OSY-SE. (continued on next page) 105

Figure 17. continued

Figure 17. Schematic overview of paenibacterin biosynthesis in P. thiaminolyticus OSY-SE. The peptide core is assembled by three NRPSs PbtA, PbtB, PbtC.

106

3.3.3. Absolute configuration of amino acids in paenibacterin The presence/absence of E-domains in the NRPS modules was used as criteria to predict the absolute configuration of amino acids in paenibacterin. The predicted configuration of the amino acids was verified using Marfey's reagent. This reagent reacts stoichiometrically with the primary amides in L- or D-amino acids, resulting in derivatives with different polarity (Bhushan & Brückner, 2004). Paenibacterin peptide was completely hydrolyzed with HCl as the catalyst; the released amino acids reacted with Marfey's reagent, followed by separation by HPLC with a simple linear gradient of acetonitrile /triethylamine phosphate.

As shown in Figure 18, Val2, Thr3, Val6, Ile9, Pro10, Val11 and Ile13 are L-amino acids, whereas Orn1and Orn4 residues are D-amino acids. These findings supported the predicted configurations of amino acids in paenibacterin (Table 14). According to sequence analysis,

Lys7 is likely D-amino acid while Lys10 may have L-configuration. The chiral analysis also confirmed that two Lysine residues in paenibacterin have different configurations (Figure 18).

However, the chirality of each lysine residue as a function of position cannot be finalized by this method due to the inherent limitation of the method. In addition, the peak of D-Ser in the

HPLC profile overlapped with the FDAA reagents (Figure 18); therefore, the configuration of two Ser residues has not been finalized by this method.

107

(A)

(B)

Figure 18. Chiral analyses of standard amino acids and constituent amino acids from paenibacterin using Marfey’s reagent. (A) HPLC profile of derivatives of standard amino acids. (B) HPLC profile of derivatives of the constituent amino acids from paenibacterin.

108

3.3.4. Cloning, expression, purification and functional analysis of A-domains A-domains in NRPS determine the primary structure of the peptide. The third and tenth A-domains in paenibacterin NRPS were predicted to select Thr and Pro residues, respectively (Table 14). To study the substrate specificity of the third and tenth A-domains in vitro, both A-domains were cloned into pET15b vector and expressed in E. coli BL21 (DE3)

(Figure 19A). The purified A-domain proteins with an N-terminal His-tag (Figure 19B) were used for substrate specificity assay. The 10th A-domain showed the highest activity on proline

(Figure 20B), which was consistent with the previous prediction based on the binding pocket in A-domain (Table 14). In addition, the third A-domain in paenibacterin NRPS, which is assumed to activate threonine, showed relatively relaxed specificity on hydroxyl-containing amino acids, serine and threonine (Figure 20A). These observations may explain the coproduction of paenibacterin and its analogues by P. thiaminolyticus OSY-SE. Therefore, these findings agreed well with the chemical structure of paenibacterin and thus confirmed the function of paenibacterin biosynthetic gene cluster.

109

(A) (B)

Figure 19. Cloning, expression and purification of adenylation domains. (A) agarose gel electrophoresis showing DNA bands for A-domain cloning. lane 1, DNA ladder (2-log, NEB); lane 2, intact pET15b plasmid; lane 3, pET15b digested with Nde I and Xho I; lane 4-5, PCR products encoding the 3rd and 10th A-domain ; lane 6-7, recombinant plasmids pET15b-Thr3 and pET15b-Pro10 that were digested with Nde I and Xho I. (B) commassie blue-stained 10% Tris-HCl SDS-PAGE gel showing the recombinant A-domains expressed in E. coli BL21 (DE3). lane 1-2, the 3rd and 10th A-domain proteins purified by Co2+-chelate affinity chromatography; lane 4, prestained protein standard (Precision plus, Bio-Rad, Hercules, CA).

110

(A)

(B)

Figure 20. Determination of substrate specificity of purified A-domains by phosphate detection assay. (A) Relative activity of the third A-domain in paenibacterin NRPS, showing highest activity on hydroxyl containing amino acids, serine and threonine. (B) Relative activity of the tenth A-domain in paenibacterin NRPS, showing highest activity on proline.

3.4. Discussion P. thiaminolyticus OSY-SE produces a potent antibacterial lipopeptide that shows 111

activity against Gram-negative and Gram-positive pathogens. Paenibacterin is a promising candidate for treating infections caused by antibiotic-resistant pathogens. Here we described the identification of a gene cluster encoding nonribosomal peptide synthetases responsible for paenibacterin biosynthesis. This conclusion was based on the following findings: (i) the whole genome sequence of the producer; (ii) the total numbers of modules and the predicted amino acid specificity are in agreement with the primary structure of paenibacterin; (iii) the amino acid specificity of the third and tenth A-domains were confirmed in vitro by functional analysis; (iv) The position of epimerization domains agreed well with the chiral analysis of paenibacterin amino acids. However, attempts have failed to disrupt the paenibacterin gene cluster using two different integration vectors, pMUTIN4 (Vagner, Dervyn, & Ehrlich, 1998) or pMAD (Arnaud, Chastanet, & Debarbouille, 2004). No confirmed transformants were obtained after electroporation (Murray & Aronstein, 2008) or protoplast transformation

(Chang & Cohen, 1979). It is not umcommon that wild type strains are not easily manipulated by genetic methods.

Lipopeptide features a fatty acid moiety that is attached to the N-terminus of the peptide chain. The lipid side chain of lipopeptides can be incorporated by different ways

(Chooi & Tang, 2010). During surfactin biosynthesis, the starter C domain selects the free fatty acyl-CoA from primary metabolism, and catalyzes the amide bond formation between the fatty acyl-CoA and an activated glutamate (Kraas, Helmetag, Wittmann, Strieker, &

Marahiel, 2010). For daptomycin biosynthesis, the dedicated acyl-CoA ligase (DptE) activates and transfers fatty acids to the acyl carrier protein ACP (DptF) (Miao et al., 2005;

Wittmann, Linne, Pohlmann, & Marahiel, 2008). For mycosubtilin lipointiation, the fatty

112

acyl ligase (AL) activates and loads the palmitic acid to an acyl carrier protein (ACP); additional modifications occur before lipid chain is added to peptide backbone (Duitman et al., 1999; Hansen, Bumpus, Aron, Kelleher, & Walsh, 2007). Paenibacterin biosynthesis may use the simple mechanism as surfactin for lipoinitiation. Sequence analysis indicated that no acyl-CoA ligase or ACP is encoded by the paenibacterin cluster; therefore it is likely that the starter C domain specifically selects the activated pentadecanoic acids from primary metabolism and catalyzes the amide bond formation with N-terminal ornithine.

Based on the information obtained from the study, we proposed the biosynthetic pathway for paenibacterin synthesis in P. thiaminolyticus OSY-SE (Figure 17). Paenibacterin is synthesized by three nonribosomal peptide synthetases and is secreted by two ABC- transporters. The elucidation of biosynthetic pathway is critical for producing paenibacterin derivatives with improved properties or for the study of the structure-activity relationships.

References

Arnaud, M., Chastanet, A., & Debarbouille, M. (2004). New vector for efficient allelic replacement in naturally nontransformable, low-GC-content, gram-positive bacteria. Applied and Environmental Microbiology, 70(11), 6887-6891.

Aziz, R., Bartels, D., Best, A., DeJongh, M., Disz, T., Edwards, R., Formsma, K., Gerdes, S., Glass, E., & Kubal, M. (2008). The RAST Server: rapid annotations using subsystems technology. BMC Genomics, 9(75)

Bachmann, B. O., & Ravel, J. (2009). Methods for in silico prediction of microbial polyketide and nonribosomal peptide biosynthetic pathways from DNA sequence data. Methods in Enzymology, 458, 181-217.

Baltz, R. H., Miao, V., & Wrigley, S. K. (2005). Natural products to drugs: daptomycin and related lipopeptide antibiotics. Natural Product Reports, 22(6), 717-741. 113

Bhushan, R., & Brückner, H. (2004). Marfey’s reagent for chiral amino acid analysis: a review. Amino Acids, 27(3), 231-247.

Bruner, S. D., Weber, T., Kohli, R. M., Schwarzer, D., Marahiel, M. A., Walsh, C. T., & Stubbs, M. T. (2002). Structural basis for the cyclization of the lipopeptide antibiotic surfactin by the thioesterase domain SrfTE. Structure, 10(3), 301-310.

Challis, G. L., Ravel, J., & Townsend, C. A. (2000). Predictive, structure-based model of amino acid recognition by nonribosomal peptide synthetase adenylation domains. Chemistry & Biology, 7(3), 211-224.

Chang, S., & Cohen, S. N. (1979). High frequency transformation of Bacillus subtilis protoplasts by plasmid DNA. Molecular and General Genetics MGG, 168(1), 111-115.

Choi, S. K., Park, S. Y., Kim, R., Kim, S. B., Lee, C. H., Kim, J. F., & Park, S. H. (2009). Identification of a polymyxin synthetase gene cluster of Paenibacillus polymyxa and heterologous expression of the gene in Bacillus subtilis. Journal of Bacteriology, 191(10), 3350-3358.

Choi, S. K., Park, S. Y., Kim, R., Lee, C. H., Kim, J. F., & Park, S. H. (2008). Identification and functional analysis of the fusaricidin biosynthetic gene of Paenibacillus polymyxa E681. Biochemical and Biophysical Research Communications, 365(1), 89-95.

Chooi, Y. H., & Tang, Y. (2010). Adding the lipo to lipopeptides: do more with less. Chemistry & Biology, 17(8), 791-793.

Clardy, J., Fischbach, M. A., & Walsh, C. T. (2006). New antibiotics from bacterial natural products. Nature Biotechnology, 24(12), 1541-1550.

Clugston, S. L., Sieber, S. A., Marahiel, M. A., & Walsh, C. T. (2003). Chirality of peptide bond-forming condensation domains in nonribosomal peptide synthetases: the C5 domain of tyrocidine synthetase is a DCL catalyst. Biochemistry, 42(41), 12095-12104.

Conti, E., Stachelhaus, T., Marahiel, M. A., & Brick, P. (1997). Structural basis for the activation of phenylalanine in the non-ribosomal biosynthesis of gramicidin S. The EMBO Journal, 16(14), 4174-4183.

Duitman, E. H., Hamoen, L. W., Rembold, M., Venema, G., Seitz, H., Saenger, W., Bernhard, F., Reinhardt, R., Schmidt, M., & Ullrich, C. (1999). The mycosubtilin synthetase of Bacillus subtilis ATCC6633: a multifunctional hybrid between a peptide synthetase, an amino transferase, and a fatty acid synthase. Proceedings of the National Academy of Sciences, 96(23), 13294-13299.

114

Emanuelsson, O., Brunak, S., von Heijne, G., & Nielsen, H. (2007). Locating proteins in the cell using TargetP, SignalP and related tools. Nature Protocols, 2(4), 953-971.

Fischbach, M. A., & Walsh, C. T. (2006). Assembly-line enzymology for polyketide and nonribosomal peptide antibiotics: logic, machinery, and mechanisms. Chemical Reviews, 106(8), 3468-3496.

Hansen, D. B., Bumpus, S. B., Aron, Z. D., Kelleher, N. L., & Walsh, C. T. (2007). The loading module of mycosubtilin: an adenylation domain with fatty acid selectivity. Journal of the American Chemical Society, 129(20), 6366-6367.

Kohli, R. M., & Walsh, C. T. (2003). Enzymology of acyl chain macrocyclization in natural product biosynthesis. Chemical Communications, 7(3), 297-307.

Konz, D., & Marahiel, M. A. (1999). How do peptide synthetases generate structural diversity? Chemistry & Biology, 6(2), R39-48.

Lagesen, K., Hallin, P., Rødland, E. A., Stærfeldt, H. H., Rognes, T., & Ussery, D. W. (2007). RNAmmer: consistent and rapid annotation of ribosomal RNA genes. Nucleic Acids Research, 35(9), 3100-3108.

Li, J., Beatty, P. K., Shah, S., & Jensen, S. E. (2007). Use of PCR-targeted mutagenesis to disrupt production of fusaricidin-type antifungal antibiotics in Paenibacillus polymyxa. Applied and Environmental Microbiology, 73(11), 3480-3489.

Linne, U., Doekel, S., & Marahiel, M. A. (2001). Portability of epimerization domain and role of peptidyl carrier protein on epimerization activity in nonribosomal peptide synthetases. Biochemistry, 40(51), 15824-15834.

Lowe, T. M., & Eddy, S. R. (1997). tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Research, 25(5), 955-964.

Marfey, P. (1984). Determination of D-amino acids. II. Use of a bifunctional reagent, 1, 5- difluoro-2, 4-dinitrobenzene. Carlsberg Research Communications, 49(6), 591-596.

McQuade, T. J., Shallop, A. D., Sheoran, A., Delproposto, J. E., Tsodikov, O. V., & Garneau-Tsodikova, S. (2009). A nonradioactive high-throughput assay for screening and characterization of adenylation domains for nonribosomal peptide combinatorial biosynthesis. Analytical Biochemistry, 386(2), 244-250.

115

Murray, K. D., & Aronstein, K. A. (2008). Transformation of the Gram-positive honey bee pathogen, Paenibacillus larvae, by electroporation. Journal of Microbiological Methods, 75(2), 325-328.

Rausch, C., Hoof, I., Weber, T., Wohlleben, W., & Huson, D. H. (2007). Phylogenetic analysis of condensation domains in NRPS sheds light on their functional evolution. BMC Evolutionary Biology, 7, 78.

Rausch, C., Weber, T., Kohlbacher, O., Wohlleben, W., & Huson, D. H. (2005). Specificity prediction of adenylation domains in nonribosomal peptide synthetases (NRPS) using transductive support vector machines (TSVMs). Nucleic Acids Research, 33(18), 5799- 5808.

Richter, M., & Rosselló-Móra, R. (2009). Shifting the genomic gold standard for the prokaryotic species definition. Proceedings of the National Academy of Sciences, 106(45), 19126-19131.

Röttig, M., Medema, M. H., Blin, K., Weber, T., Rausch, C., & Kohlbacher, O. (2011). NRPSpredictor2—a web server for predicting NRPS adenylation domain specificity. Nucleic Acids Research, 39(suppl 2), W362-W367.

Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M. A., & Barrell, B. (2000). Artemis: sequence visualization and annotation. Bioinformatics, 16(10), 944-945.

Schlumbohm, W., Stein, T., Ullrich, C., Vater, J., Krause, M., Marahiel, M., Kruft, V., & Wittmann-Liebold, B. (1991). An active serine is involved in covalent substrate amino acid binding at each reaction center of gramicidin S synthetase. Journal of Biological Chemistry, 266(34), 23135-23141.

Schuster, S. C. (2007). Next-generation sequencing transforms today’s biology. Nature, 200(8), 16-18.

Shaheen, M., Li, J., Ross, A. C., Vederas, J. C., & Jensen, S. E. (2011). Paenibacillus polymyxa PKB1 produces variants of polymyxin B-type antibiotics. Chemistry & Biology, 18(12), 1640-1648.

Stachelhaus, T., & Walsh, C. T. (2000). Mutational analysis of the epimerization domain in the initiation module PheATE of gramicidin S synthetase. Biochemistry, 39(19), 5775- 5787. 116

Vagner, V., Dervyn, E., & Ehrlich, S. D. (1998). A vector for systematic gene inactivation in Bacillus subtilis. Microbiology, 144(11), 3097-3104.

Wittmann, M., Linne, U., Pohlmann, V., & Marahiel, M. A. (2008). Role of DptE and DptF in the lipidation reaction of daptomycin. FEBS Journal, 275(21), 5343-5354.

117

Chapter 4 Paenibacterin Exerts Its Antimicrobial Activity through Membrane Damage and Radical Formation

Abstract

This objective of current study was to investigate the mechanisms of the action of paenibacterin against Escherichia coli and Staphylococcus aureus. Paenibacterin is a cationic lipopeptide with four positive charges. Paenibacterin molecules have high affinity to anionic lipopolysaccharide (LPS) of outer membrane in Gram-negative bacteria. The electrostatic paenibacterin-LPS interaction can displace the divalent cations on the LPS network and thus promote the entry of paenibacterin. The cytoplasmic membrane is the direct target of paenibacterin. Paenibacterin depolarized cell membrane, triggered K+ release and increased the uptake of hydrophobic nucleic acid stain, propidium iodide. In addition to damaging cell membrane, paenibacterin led to production of hydroxyl radicals in bacterial cells. The radical scavenger thiourea and the iron chelator dipyridyl reduced the killing effect of paenibacterin against E. coli and S. aureus. Therefore, the mechanisms of action of paenibacterin involve direct cell membrane damage and indirect oxidative cellular damages.

118

4.1. Introduction

Paenibacterin is a new cationic lipopeptide antibiotic that is active against Gram- negative and Gram-positive bacteria, including multi-drug resistant pathogens.

Paenibacterin comprises 13 amino acids with a C15 fatty acyl chain at the N-terminus.

Paenibacterin is produced by a bacterial strain, Paenibacillus thiaminolyticus OSY-SE, through the nonribosomal peptide synthetases (NRPSs). The mechanisms of action of paenibacterin were explored in the current study.

Cationic antimicrobial peptides have distinctive features: positively charged at neutral pH and amphiphilic (Hancock, 1997). The cationic and amphiphilic nature of peptides is important for their antimicrobial activity. Bacterial cell surfaces contain acidic components, such as lipopolysaccharides and teichoic acids in Gram-negative and Gram- positive bacteria, respectively. The positively charged peptides accumulate on the anionic bacterial cell surfaces and subsequently disrupt the membrane integrity of the cytoplasmic membrane (Hancock & Sahl, 2006).

Polymyxins are cationic lipopeptide antibiotics that exert their antimicrobial activity by permeabilizing the outer membrane and damaging the physical integrity of the phospholipid bilayer of inner membrane (Velkov, Thompson, Nation, & Li, 2010).

Polymyxins carry five positive charges due to the diaminobutyryl (Dab) residues. The amphipathicity of polymyxin molecules is attributed to the hydrophobic domains (the fatty acyl chain and the Phe-Leu segment) and the hydrophilic segments consisting of Thr 119

and Dab residues (Figure 21A). Similarly, paenibacterin is cationic lipopeptide with four

120

121

positive charges (Figure 21B). As described in chapter 2, seven residues of Orn4-Val6 and Ile9-Lys12 form an anti-parallel β-sheet with four hydrophobic side chains (Val6,

Ile9, Val11, and Ile13) on one side. The second hydrophobic domain in paenibacterin is the N-terminal lipid acyl chain, whereas four positive charged residues and two hydroxyl- containing amino acids (Ser) constitute the hydrophilic portion of the molecule (Figure

21B). The cationic and amphiphilic nature of paenibacterin is assumed to contribute to its antimicrobial activity. In general, cationic amphiphilic peptides target anionic cell membrane. The current study investigated the interaction of paenibacterin with outer membrane in Gram-negative bacteria and cytoplasmic membrane in both Gram-negative and Gram-positive bacteria.

Recently, the radical theory has been used to explain the cellular events that lead to cell death after the drug-target interaction (Kohanski, Dwyer, Hayete, Lawrence, &

Collins, 2007). Bactericidal antibiotics of different classes stimulate the production of the detrimental hydroxyl radicals in bacteria, which eventually kill the bacterial cells

(Kohanski et al., 2007; Kohanski, Dwyer, & Collins, 2010). The new radical-mediated killing proposal has been supported by studying other antibiotics, including the cationic lipopeptide polymyxins (Sampson et al., 2012). In the study, we also explored whether paenibacterin induce radical production in bacteria.

122

4.2. Materials and Methods

4.2.1. Effect of lipopolysaccharides (LPS) on paenibacterin activity

Population of Escherichia coli ATCC 25922 at exponential phase were diluted

100 times in tryptic soy broth (Becton Dickinson, Sparks, MD). LPS from E. coli O111:

B4 (Sigma, St. Louis, MI) were added to the cell suspension at a final concentration of

0.1 mg/ml. Paenibacterin was added to the suspension at a final concentration of 32

µg/ml (2×MIC). The mixture was incubated for at 37°C with agitation at 200 rpm.

Aliquots of cells were sampled for enumeration at 60 minutes after treatment.

4.2.2. Binding of paenibacterin to LPS and Gram-negative bacteria

A fluorescent-labeled polymyxin B (BODIPY FL conjugate, Invitrogen, Carlsbad,

CA) was used to measure the binding affinity of paenibacterin to purified LPS from E. coli O111: B4 and to the intact bacterial cells of E. coli ATCC 25922. Our preliminary experiments indicated that the fluorescence of the polymyxin conjugate is quenched when the probe binds to LPS. Aliquots (70 μl) of LPS-polymyxin conjugate were added to wells of black non-binding surface (NBS) microplate (Corning, Tewksbury, MA). Then

30 μl of MgCl2, polymyxin E or paenibacterin at different concentrations were added to the microplate wells. The final concentration of LPS and polymyxin conjugate were at 5

μg/ml and 0.1 mg/ml, respectively. Changes in fluorescence due to the displacement of

LPS-bound BODIPY FL-labeled polymyxin B were recorded using a luminescence spectrometer (LS55, Perkin-Elmer, Wellesley, MA) with an excitation at 490 nm and an

123

emission at 515 nm. The binding affinity of paenibacterin to live E. coli ATCC 25933 cells was measured using a similar procedure as above but the LPS component was replaced with ~106 E. coli cells in this binding assay.

4.2.3. Membrane potential depolarization

The membrane potential assays were performed using a fluorescence probe, 3,3'-

Dipropylthiadicarbocyanine Iodide (DiSC3(5)) (Invitrogen, Carlsbad, CA) according to previously reported method (Zhang, Dhillon, Yan, Farmer, & Hancock, 2000) with minor modifications. The DiSC3(5) probe accumulates in polarized cell membrane; therefore depolarization of cell membrane leads to the release of the probe. Overnight bacterial culture of Staphylococcus aureus ATCC 6538 or E. coli ATCC 25922 was inoculated

(1/100 dilution) into tryptic soy broth and incubated at 37°C with agitation at 200 rpm for

~5 hours. Cells were harvested by centrifugation at 4 °C at 3,660×g for 10 min, followed by washing twice using 5mM HEPES buffer (Sigma) supplemented with 5 mM glucose

(Buffer A). Cells of S. aureus were resuspended in buffer A. E. coli was suspended in buffer A modified with 0.2 mM EDTA; this facilitated the uptake of the DiSC3(5) probe by Gram-negative bacteria (Zhang, 2000). The fluorescent probe DiSC3(5) was added to the cell suspension to a final concentration of 0.5 μM from a 1mM stock solution in dimethyl sulfoxide (DMSO). Bacterial cells with DiSC3(5) probe were incubated at room temperature for 15 min. After incubation, KCl was added to the cell suspension of S. aureus at a final concentration of 100 mM; this step was omitted for E. coli cells. 124

Aliquots (90 μl) of the cell suspension were added to wells of the black NBS microplate.

This was followed by adding 10 μl of paenibacterin or other antimicrobial agents to the wells. Fluorescence change due to membrane depolarization was recorded using the LS55 luminescence spectrometer (Perkin-Elmer, Wellesley, MA) with an excitation at 622 nm and an emission at 670 nm.

4.2.4. Potassium release assay

Potassium ion release assays were performed using the cell impermeable K+- sensitive probe (PBFI, Invitrogen) (Silverman, Perlmutter, & Shapiro, 2003). Bacterial cells of S. aureus and E. coli were grown, harvested and washed using the same procedure in the membrane potential assay. After washing, S. aureus and E. coli were resuspended in buffer A. Aliquots (90 μl) of the cell suspension were added to wells of the black NBS microplate. The cell impermeable K+-sensitive probe was added to the cell suspension at a final concentration of 2 μM. Then 10 μl of paenibacterin or other antimicrobial agents was added to the wells. Change of fluorescence corresponding to potassium concentration was recorded using a using a luminescence spectrometer (LS55,

Perkin-Elmer, Wellesley, MA) with an excitation at 346 nm and an emission at 505 nm.

4.2.5. Membrane permeability assay

Cytoplasmic membrane permeability assays were performed using the

LIVE/DEAD Baclight bacterial viability kit (Invitrogen) according to the manufacturer’s instructions. Bacterial cells of S. aureus and E. coli were grown to exponential phase in 125

trypic soy broth and washed with saline solution (0.85% NaCl). The cell suspension in saline was treated with paenibacterin at 37 °C for 1 hr. Then two nuclear acid stains,

SYTO-9 and propidium iodine, were added to the treated cells at final concentrations of at 7.5 μM and 30 μM, respectively. The mixtures were incubated in dark at room temperature for 15 min. The stained cells (5 μl) were spotted on a microscopic slide and covered with a glass cover slip. Digital images were obtained using a fluorescence microscope (BX 61, Olympus, Melville, NY) at the following settings: excitation/emission of 480/500 nm for SYTO-9 and 490/635 nm for propidium iodine.

4.2.6. Radical production measurement

The hydroxyl radical levels were measured by the hydroxyl radical indicator, hydroxyphenyl fluorescein (HPF, Invitrogen) (Sampson et al., 2012). E. coli cells were grown, harvested and washed using the same procedures in the membrane potential assay.

The cell suspension in buffer A was mixed with paenibacterin (16 μg/ml or 64 μg/ml) and

HPF at a final concentration of 5 µM. The mixture was incubated in dark at 37°C for 1 hr.

Aliquots (90 μl) of the treated cells were added to wells of black NBS microplate.

Fluorescence corresponding to the concentration of hydroxyl radicals was recorded using a luminescence spectrometer (LS55, Perkin-Elmer, Wellesley, MA) with an excitation at

490 nm and an emission at 515 nm.

4.2.7. Effect of thiourea and 2, 2'-dipyridyl on paenibacterin activity

The effect of thiourea and 2, 2'-dipyridyl on antibiotic activity was performed 126

using the method of Kohanski et al. (2007) with minor modification. Overnight cultures of E. coli and S. aureus were inoculated 1:30 into Mueller Hinton II broth (MH II, Becton

Dickinson) and incubated at 37 °C with agitation at 200 rpm for ~6 hrs. The bacterial culture was then diluted (1:100) to ~107 CFU/ml in MH II broth. Bacterial cells were treated with paenibacterin at 37°C with shaking at 200 rpm for two hrs. At 1 and 2 hours post treatment, aliquots of cells were taken and the survivor cells were enumerated using tryptic soy agar after incubation overnight at 37°C.When 2, 2'-dipyridyl (Sigma) and/or thiourea (Sigma) was utilized, cells were pretreated with the indicated compound for 10 minutes at room temperature before the addition of paenibacterin.

4.3. Results

4.3.1. Lipopolysaccharides antagonize paenibacterin activity

As shown in Figure 22, the purified lipopolysaccharides (LPS) from E. coli O111:

B4 antagonized the antimicrobial activity of paenibacterin against E. coli. The results indicated that paenibacterin had a high affinity to LPS component from outer membrane in Gram-negative bacteria. Therefore, the interaction of paenibacterin and the LPS components is associated with the activity of paenibacterin.

127

Figure 22. Effect of lipopolysaccharide (LPS) on the activity of paenibacterin (32 µg/ml) against E. coli ATCC 25922

4.3.2. Paenibacterin binds to LPS components of outer membrane

Our preliminary results showed that the fluorescence of the fluorescent-labeled polymyxin was quenched when it binds to E. coli cells or the purified LPS component from outer membrane of Gram-negative bacteria (data not shown). Therefore, the cell/LPS-bound polymyxin probes can be used for testing the LPS binding affinity of other compounds. As shown in Figure 23A, the addition of divalent ions (Mg2+) and polymyxin E caused an increase of fluorescence from free labeled polymyxin, which indicated that the labeled polymyxin probes were displaced and released from LPS.

Similarly, paenibacterin molecules competed for the binding sites on LPS and released the labeled polymyxin probe from LPS on a concentration-dependent manner

128

(Figure 23A). In an in vivo test with live E. coli cells, paenibacterin also displaced the cell membrane-bound polymyxin probe from E. coli cells (Figure 23B). These results indicated that paenibacterin and polymyxin molecules have the same binding sites on

LPS in the outer membrane of Gram-negative bacteria. Polymyxins bind to the negatively charged Lipid A phosphates and displace the divalent cations (Mg2+ and Ca2+) that stabilize the normal barrier property of the outer membrane (Hancock, 1997; Velkov et al.,

2010). Therefore, the electrostatic interaction between paenibacterin with LPS can disrupt the outer membrane permeability and promote the uptake of the antibiotic.

(A)

Figure 23. BODIPY FL-labeled polymyxin B displacement assays. (continued on next page)

129

Figure 23. continued.

(B)

Figure 23. BODIPY FL-labeled polymyxin B displacement assays. (A) in vitro polymyxin B-LPS displacement (B) in vivo polymyxin B-cells displacement assay.

4.3.3. Paenibacterin depolarizes cytoplasmic membrane

Aerobic or facultative bacteria maintain a proton gradient (proton-motive force) across the cell membrane that can be used to make ATP and transport nutrients into the cells. The largest portion of proton-motive force is an electrical potential gradient (Δψ)

(Alberts et al., 2002). DiSC3(5) is a potential-sensitive dye that accumulates in polarized cell membrane and become self-quenched. Any compound that depolarizes the membrane electrical potential will lead to the release of DiSC3(5) with an increase in fluorescence

(Zhang et al., 2000). Paenibacterin can permeabilize the cell membrane of S. aureus at concentration of as low as 8 μg/ml (Figure 24A). Nisin, a membrane active peptide, also caused considerable membrane depolarization, whereas the cell wall inhibitor 130

vancomycin had no effect on cell membrane (Figure 24A). However, paenibacterin exhibited little membrane potential disturbance in E. coli at low concentrations (< 32

μg/ml); a higher concentration (64 μg/ml) of paenibacterin was required to depolarize the

E. coli cells (Figure 24B).

(A)

(B)

Figure 24. Effect of paenibacterin (paen), nisin vancomycin (van) and polymyxin E (pmx) on bacterial membrane potential. (A) S. aureus ATCC 6538 (B) E. coli ATCC 25922

131

4.3.4. Paenibacterin induces intracellular potassium release

The effect of paenibacterin on membrane potential promoted to examine other

membrane-associated functions. Paenibacterin caused a concentration-dependent K+

efflux from S. aureus and E. coli cells. Paenibacterin at 64 µg/ml rapidly triggered the

release of intracellular potassium (Figure 25). These results indicated that paenibacterin

can damage cell cytoplasmic membrane. On the contrast, nisin and polymyxin at their

bactericidal concentrations did not cause significant potassium release from S. aureus and

E. coli cells, respectively (Figure 25).

(A)

Add PBFI probe

Add antimicrobials

Figure 25. Effect of paenibacterin (paen), nisin, vancomycin (van) and polymyxin E (pmx) on K+ release. (continued on next page)

132

Figure 25. continued.

(B)

Add PBFI probe

Add antimicrobials

Figure 25. Effect of paenibacterin (paen), nisin, vancomycin (van) and polymyxin E (pmx) on K+ release. (A) S. aureus ATCC 6538 (B) E. coli ATCC 25922.

4.3.5. Paenibacterin increases cell membrane permeability

The permeability of cytoplasmic membrane was visualized by two fluorescent

nucleic acid stains. Viable cells with intact cytoplasmic membrane are stained green by

the cell-membrane permeable SYTO-9, whereas bacterial cells with compromised

membrane appear red when the propidium iodine (PI) enters the cells and queches the

green fluorescence from SYTO-9. The untreated S. aureus and E. coli cells were stained

by the green fluorescent dye SYTO-9 (Figure 26A and C). Paenibacterin permeabilized

cell membrane and allow the uptake of the red fluorescent dye propidium iodine (Figure

26B and D).

133

(A) (B)

Figure 26/ Effect of paenibacterin on bacterial cell membrane permeability. (A) Untreated cells of S. aureus ATCC 6538 (B) S. aureus ATCC 6538 treated with 80 μg/ml paenibacterin for 1hr; (C) untreated E. coli ATCC 25922 (D) E. coli ATCC 25922 treated with 64 μg/ml paenibacterin for 1 hr.

4.3.6. Radical production measurement

Antibiotic-mediated cell death begins with the drug-target interaction and involves disruption of multiple cellular functions (Hurdle, O'Neill, Chopra, & Lee, 2010;

Kohanski et al., 2010). Oxidative cellular damage was proposed as a common mechanism of cell death induced by bactericidal antibiotics (Kohanski et al., 2007). This oxidation- based mechanism has been supported by a growing body of evidences (Liu et al., 2009;

Liu, Annamalai, Sutherland, & Tse-Dinh, 2009; Sampson et al., 2012; Wang & Zhao,

2009). Similarly, paenibacterin treatment also induced hydroxyl radical production in E.

134

coli cells on a concentration-dependent manner (Figure 27). The membrane-targeting antibiotics, polymyxins, caused rapid killing of Acinetobacter baumannii, which was also mediated by the hydroxyl radical pathway (Choi & Lee, 2012).

b ab a

Figure 27. Hydroxyl radical production in paenibacterin treated cells of E. coli ATCC 25922. Means with different letters are significantly different (p< 0.05).

4.3.7. Paenibacterin inactivates cells via radical production

The observation of increased hydroxyl radical after paenibacterin treatment promoted us to examine whether hydroxyl radicals caused cell death. Thiourea, a hydroxyl radical scavenger, rescued cells from paenibacterin-mediated killing in S. aureus and E. coli (Figure 28). Paenibacterin caused a 2-log reduction in S. aureus in 2 hr while pretreatment with thiourea for 10 min reduced the killing by 1 log (Figure 28A).

Similarly, thiourea also protected the E. coli cells from killing by paenibacterin. In addition, the combination of thiourea and iron chelator (dipyridyl) provided a better 135

protection to E. coli cells (Figure 28B). These results indicated that hydroxyl radicals contributed to the cell death induced by paenibacterin. The protection from iron chelator, dipyridyl, suggested that irons were associated with radical production (Kohanski et al.,

2007).

(A)

(B)

Figure 28. Effect of dipyridyl (Dip, 450 μM) and thiourea (Thio, 100 mM) on the killing efficacy of paenibacterin (Pbt). (A) S. aureus ATCC 6538, Pbt at 96 μg/ml (B) E. coli ATCC 25922, Pbt at 64 μg/ml

136

4.4. Discussion

This study explored the mechanisms of action of paenibacterin against both

Gram-positive and Gram-negative bacteria. Gram-negative bacteria are resistant to many antibiotics due to the permeability barrier of outer membrane (Vaara, 1992). The molecular basis of permeability lies in the lipopolysaccharide networks that are electrostatically linked by divalent cations (Mg2+ and Ca2+). Cationic agents and ion chelators are known permeabilizers that disorganize the outer membrane (Vaara, 1992).

Cationic peptides, such as polymyxins, enter bacterial cells by a self-promoted uptake pathway (Hancock, 1997). In this study, we found that paenibacterin had high affinity to lipopolysaccharides (Figure 22 and Figure 23). The interaction with lipopolysaccharides may displace the Mg2+ and Ca2+ and promote the uptake of paenibacterin. The LPS binding ability of paenibacterin suggested that paenibacterin can neutralize LPS and prevent endotoxinamia during antibiotic treatment (Hancock, 1997). As an outer membrane permeabilizer, the combination of paenibacterin with other antibiotics may show synergistic interaction against drug-resistant pathogens.

The bacterial cytoplasmic membrane is the direct target of paenibacterin for

Gram-positive and Gram-negative bacteria. Paenibacterin depolarized cell membrane, triggered K+ release and increased the uptake of propidium iodide (Figure 24, Figure 25 and Figure 26). The cell membrane is essential to bacterial life as it contains one-third of proteins in a cell and is the place for crucial processes (Hurdle et al., 2010). Therefore, by

137

damaging the cell membrane, paenibacterin may interfere with numerous cellular functions such as function of electrical transport chain.

Reactive oxygen species (ROS), such as superoxide and hydroxyl radicals, are highly deleterious to bacterial cells by oxidizing macromolecules such as DNA (Imlay,

2002; Imlay, 2003). In addition to membrane damage, paenibacterin led to production of hydroxyl radicals in E. coli cells (Figure 27). Supporting the oxidative damage notion, we found that the radical scavenger, thiourea, partially protected cells from killing by paenibacterin. Additionally, pretreatment with the iron chelator, dipyridyl, reduced the killing effect of paenibacterin against E. coli (Figure 28). Kohanski et al. (2010) pointed out that the drug-target interaction resulted in hyperactivation of electron transport chain

- and generation of superoxides (O2 ) formation. Superoxide in turn damages the iron- sulfur cluster and thus releases ferrous irons, which stimulated the production of hydroxyl radical through Fenton reaction (Imlay, 2003; Kohanski et al., 2010).

In this study, we presented the dual mechanisms of action of paenibacterin, namely direct cell membrane damage and indirect oxidative cellular damages due to hydroxyl radical formation. Unlike other conventional antibiotics, bacteria are more difficult to acquire resistance to membrane-active peptides (Hancock & Sahl, 2006;

Hurdle et al., 2010). Facing the challenges of antibiotic resistance, paenibacterin has the potential to become a new therapeutic approach against bacterial infections.

138

In addition, we developed a rapid method for determining the binding affinity of compounds toward lipopolysaccharides (LPS) based on the BODIPY® FL Conjugate- polymyxin B/LPS displacement assay. When fluorescent-labeled polymyxins bind to LPS, the fluorescence from the mixture becomes quenched. The addition of test compound, such as paenibacterin, leads to dequenching of fluorescence. This approach provides an alternative way to study the interactions of polycation and LPS using dansyl-polymyxin

(Moore, Bates, & Hancock, 1986). This method can be extended to screen compounds with antiendotoxin activity.

References

Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., & Walter, P. (2002). Molecular biology of the cell (4th ed.). New York, NY: Garland Science.

Choi, H., & Lee, D. G. (2012). Antimicrobial peptide pleurocidin synergizes with antibiotics through hydroxyl radical formation and membrane damage, and exerts antibiofilm activity. Biochimica Et Biophysica Acta (BBA)-General Subjects, 1820(12), 1831-1838.

Hancock, R. E. W. (1997). Peptide antibiotics. The Lancet, 349(9049), 418-422.

Hancock, R. E. W., & Sahl, H. G. (2006). Antimicrobial and host-defense peptides as new anti-infective therapeutic strategies. Nature Biotechnology, 24(12), 1551-1557.

Hurdle, J. G., O'Neill, A. J., Chopra, I., & Lee, R. E. (2010). Targeting bacterial membrane function: an underexploited mechanism for treating persistent infections. Nature Reviews Microbiology, 9(1), 62-75.

Imlay, J. A. (2002). How oxygen damages microbes: oxygen tolerance and obligate anaerobiosis. Advances in Microbial Physiology, 46, 111-153.

139

Imlay, J. A. (2003). Pathways of oxidative damage. Annual Reviews in Microbiology, 57(1), 395-418.

Kohanski, M. A., Dwyer, D. J., & Collins, J. J. (2010). How antibiotics kill bacteria: from targets to networks. Nature Reviews Microbiology, 8(6), 423-435.

Kohanski, M. A., Dwyer, D. J., Hayete, B., Lawrence, C. A., & Collins, J. J. (2007). A common mechanism of cellular death induced by bactericidal antibiotics. Cell, 130(5), 797-810.

Liu, I. F., Annamalai, T., Sutherland, J. H., & Tse-Dinh, Y. C. (2009). Hydroxyl radicals are involved in cell killing by the bacterial topoisomerase I cleavage complex. Journal of Bacteriology, 191(16), 5315-5319.

Moore, R. A., Bates, N. C., & Hancock, R. (1986). Interaction of polycationic antibiotics with Pseudomonas aeruginosa lipopolysaccharide and lipid A studied by using dansyl-polymyxin. Antimicrobial Agents and Chemotherapy, 29(3), 496-500.

Sampson, T. R., Liu, X., Schroeder, M. R., Kraft, C. S., Burd, E. M., & Weiss, D. S. (2012). Rapid Killing of Acinetobacter baumannii by Polymyxins is Mediated by a Hydroxyl Radical Death Pathway. Antimicrobial Agents and Chemotherapy, 56(11), 5642-5649.

Silverman, J. A., Perlmutter, N. G., & Shapiro, H. M. (2003). Correlation of daptomycin bactericidal activity and membrane depolarization in Staphylococcus aureus. Antimicrobial Agents and Chemotherapy, 47(8), 2538-2544.

Vaara, M. (1992). Agents that increase the permeability of the outer membrane. Microbiological Reviews, 56(3), 395-411.

Velkov, T., Thompson, P. E., Nation, R. L., & Li, J. (2010). Structure—Activity Relationships of Polymyxin Antibiotics. Journal of Medicinal Chemistry, 53(5), 1898.

Wang, X., & Zhao, X. (2009). Contribution of oxidative damage to antimicrobial lethality. Antimicrobial Agents and Chemotherapy, 53(4), 1395-1402.

Zhang, L., Dhillon, P., Yan, H., Farmer, S., & Hancock, R. E. W. (2000). Interactions of Bacterial Cationic Peptide Antibiotics with Outer and Cytoplasmic Membranes of

140

Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy, 44(12), 3317- 3321.

141

Chapter 5 Biosynthesis, Mode of Action and Application of Paenibacillin

Abstract

Paenibacillin, a newly-found lantibiotic from Paenibacillus polymyxa OSY-DF, showed potency against Listeria monocytogenes, methicillin-resistant Staphylococcus aureus and other Gram-positive bacteria. This compound exerts its antimicrobial activity by damaging cell membrane. The crude extract of paenibacillin from the producer strain inhibited or delayed the growth of L. monocytogenes in meat products. This study also investigated the biosynthesis of paenibacillin. The structural gene (paenA) was identified by polymerase chain reaction. The complete biosynthetic gene cluster was revealed by whole genome sequencing of the producer strain. The paenibacillin gene cluster (11.7 kb) comprises

11 open reading frames (ORFs) encoding proteins for production, modification, regulation, immunity and transportation of the lantibiotic. Lantibiotic dehydratase (PaenB) is essential for paenibacillin biosynthesis because disruption of paenB gene eliminated the production of paenibacillin. The paenibacillin cluster includes a gene encoding a putative acetylase

(PaenN), which may catalyze the N-terminal acetylation of paenibacillin during its biosynthesis. This finding supports the results of a previous chemical analysis, which reported an acetyl moiety uniquely located at paenibacillin N-terminus. In addition, an 142

accessory gene regulator (agr) locus, consisting of agrB, agrD, agrC, agrA, was found in the paenibacillin gene cluster, which suggests that paenibacillin production is regulated by a cell density-dependent fashion via quorum-sensing.

5.1. Introduction

New antimicrobials are needed for preservation of minimally-processed foods, a category of emerging popular products. Bacteriocin-based strategies for food biopreservation can help in minimizing the use of chemical preservatives and permit the application of less severe heat treatments (Gálvez, Abriouel, López, & Omar, 2007). There are increased interest in lantibiotics because of the successful application of nisin in food industry and the potential therapeutic use of lantibiotics against infections (Breukink & De Kruijff, 2006; Cotter, Hill,

& Ross, 2005; van Heel, Montalban-Lopez, & Kuipers, 2011). Nisin has been in use as a natural food biopreservative for decades, but emergence of nisin-resistant pathogens was reported recently (Collins, Joyce, Hill, Cotter, & Ross, 2010; Gravesen, Axelsen, Da Silva,

Hansen, & Knöchel, 2002).

Paenibacillin is a lantibiotic produced by Paenibacillus polymyxa OSY-DF (He et al.,

2007). The chemical structure of paenibacillin (Figure 29) has been determined by nuclear magnetic resonance (NMR) and mass spectrometry (MS) analyses (He, Yuan, Zhang, &

Yousef, 2008). In the current study, we explored the mechanism of action of paenibacillin and proposed the biosynthetic pathway for paenibacillin synthesis in P. polymyxin OSY-DF.

143

Figure 29. Chemical structure of paenibacillin

5.2. Materials and Methods

5.2.1. Bacterial strains and growth conditions

Bacillus subtilis 1A771 was obtained from Bacillus Genetic Stock Center (The Ohio

State University, Columbus, OH). Other strains were obtained from the culture collection of food safety laboratory, The Ohio State University. Paenibacillus polymyxa OSY-DF was grown in tryptic soy broth (Becton Dickinson, Sparks, MD) supplemented with 0.6% yeast extract (TSBYE) at 30°C with agitation at 200 rpm. Escherichia coli DH5α was cultured in

Luria-Bertani broth (Becton Dickinson) or on Luria-Bertani agar at 37°C. When appropriate,

Luria-Bertani media were supplemented with ampicillin (100 µg /ml). The indicator strains,

Listeria innocua ATCC 33090 and B. subtilis 1A771, were grown in Luria-Bertani media at

37°C. Bacterial strains and media used in the minimum inhibitory concentration (MIC) assays are listed in Table 16. For studying the mechanism of action of paenibacillin,

Staphylococcus aureus ATCC 6538 and L. innocua ATCC 33090 were grown in tryptic soy broth.

144

5.2.2. Minimum inhibitory concentration of paenibacillin

The minimum inhibitory concentrations (MICs) of paenibacillin, nisin and vancomycin were determined using the broth micro-dilution method (Clinical and Laboratory

Standards Institute, 2006). Briefly, each antimicrobial agent was dissolved in water and diluted to appropriate concentration with the medium intended for each tested strain (Table

16). Aliquots (50μl) of serially-diluted antimicrobial agents were dispensed into wells of a

96-well plate; equal amounts of 1/10 diluted overnight bacterial (indicator) culture was added to wells. Plates were incubated at 35°C for 24h. Cell growth after incubation was examined spectrophotometrically to measure the optical density at 600 nm (OD600) using a microtiter plate reader (Molecular Devices Corp., Menlo Park, CA). MIC refers to the lowest concentration of an antimicrobial that resulted in no visible growth of bacterial cells after incubation at 35°C for 24h.

5.2.3. Identification of the paenibacillin structural gene, paenA

Genomic DNA of P. polymyxa OSY-DF was purified using a DNA isolation kit

(DNeasy Blood & Tissue kit; Qiagen, Valencia, CA). The structural gene encoding paenibacillin prepropeptide was amplified by PCR. The forward primer, PaenAF, was designed based on the conserved “FDLD” motif in the leader peptide of class I lantibiotics

(Chatterjee, Paul, Xie, & van der Donk, 2005). The reverse primer, PaenAR, was based on the nucleotide sequences encoding the C-terminus of a putative paenibacillin homologue,

145

BtA1 (ZP_04136593.1) in Bacillus thuringiensis genome (Figure 31). All primers used in this study are listed in Table 15. Table 15. List of primers used in the current study

Primer Name Oligonucleotide Sequences PaenAF 5’-ATGAATAAAGAATTATTTGATTTAGATATT-3’ PaenAR 5’-CTTACAATTAGAGCATGANCCAGTACA-3’ Walkdn1 5’-ACGTGATTTAGGGGTACCACTGAAATC-3’ Walkdn2 5’-ATGAAAGTAGACCAAATGTTTGACCTT-3’ AP1 5'-GTAATACGACTCACTATAGGGC-3' PMuDelBEcoF 5'-CCGGAATTCAAGTGCTGAGTTCTTCGGTGAATGC-3' PMuDelBBamR 5'-CGCGGATCCTTCGGATCAATAATCGCAGCCATAG-3' PMUTIN4_EryF 5'-AAAGGGCATTTAACGACGAA-3' PMUTIN4_EryR 5'-TTGAGTGTGCAAGAGCAACC-3' LacZR1 5'-GTGCTGCAAGGCGATTAAGTT-3' PCR amplification was performed using a Taq DNA polymerase kit (Qiagen) under the following conditions: the reaction mixture (50 µl) was subjected to an initial denaturation at 94°C for 3 min, followed by 35 cycles, including 1 min at 94°C, 1 min at 59°C and 30 seconds at 72°C. A final extension was carried out at 72°C for 10 min. The amplified PCR product was purified using a commercial extraction kit (Qiaquick; Qiagen), ligated to a vector (pGEM-T Easy; Promega, Madison, WI) and introduced into E. coli competent cells

(TOP10; Invitrogen, Carlsbad, CA) by heat shock at 42°C for 30 seconds. The recombinant plasmid, carrying the paenibacillin structural gene, was isolated from the overnight culture using spin columns (QIAprep Spin Miniprep kit; Qiagen). Resultant plasmid DNA was sequenced using a DNA analyzer (3730 DNA Analyzer; Applied Biosystems, Foster City, CA) at the Plant-Microbe Genomics Facility, The Ohio State University (Columbus, OH).

146

5.2.4. Identification of paenB and paenP by genome walking technique

The DNA sequences flanking paenA were amplified by PCR using a genome walking procedure (GenomeWalker™ universal kit; Clontech, Mountain View, CA) according to the manufacturer’s instruction with some modifications. Briefly, samples of OSY-DF genomic

DNA were digested with each of the four restriction enzymes (Dra I, EcoR V, Pvu II and Stu

I) at 37°C overnight. The resulting blunt-ended DNA fragments were purified using spin columns (Qiaquick; Qiagen) and ligated to the genome walking adaptors at 16°C overnight.

PCR-based DNA genome walking was performed using a specific primer (Walkdn1 or

Walkdn2) from paenA gene and a universal primer AP1 derived from the adaptor sequence

(Table 15).

PCR amplification was carried out in a thermalcycler using a commercial kit

(Advantage® 2 PCR Kit; Clontech) under the following two-step cycle parameters: 94°C for

25 seconds and 72°C for 3 min (7 cycles), 94°C for 25 seconds and 65°C for 3 min (32 cycles), followed by a final extension at 72°C for 7 min. Selected PCR products were purified using an extraction kit (Qiaquick; Qiagen) and sequenced at the Plant-Microbe

Genomics Facility. Proteins were deduced from the DNA sequences and their putative functions were predicted using the BLASTX program against NCBI protein database.

5.2.5. Disruption of lantibiotic dehydratase gene, paenB

For disruption of the paenB gene, a 541 bp DNA fragment internal to lantibiotic 147

dehydratase gene (paenB) was amplified from strain OSY-DF by PCR using the primers,

PMuDelBEcoF and PMuDelBBamR (Table 15). The PCR product was purified by spin column (QIAquick gel extraction kit; Qiagen) and double digested with the restriction enzymes EcoR I and BamH I. The digested PCR product was cloned into the plasmid pMUTIN4 (Vagner, Dervyn, & Ehrlich, 1998) between the EcoR I and BamH I sites. The resulting vector, pMUTIN4_DelB, was introduced into P. polymyxa OSY-DF cells by electroporation. The pMUTIN4-derived plasmid lacks the origin of replication in Gram- positive bacteria but it can integrate into bacterial genome by homologous recombination

(Vagner et al., 1998).

Competent cells of P. polymyxa OSY-DF were prepared as described by (Murray &

Aronstein, 2008) with modifications. Briefly, strain OSY-DF was grown in MYPGP medium

(Murray & Aronstein, 2008) at 30°C with shaking at 200 rpm to OD600 of 0.3-0.4.

Subsequently, the cells were washed 3 times with the ice-cold electroporation buffer (EB;

0.625M sucrose with 1mM MgCl2) and resuspended in cold EB buffer (1/200 initial culture volume). Approximate 400 ng of plasmid pMUTIN4_DelB were added to aliquots (50μl) of competent cells for transformation. Electroporation was performed in a 0.2 cm cuvette, using a pulser apparatus connected to a pulse controller (Gene Pulser; Bio-Rad, Richmond, CA) at the following conditions: 8.5 kV/cm, 200Ω and 25μFD. Immediately after the application of pulse, cells were recovered with 1ml of MYPGP medium followed by incubation at 30°C for

148

4h with shaking at 200 rpm. The presumptive mutants were selected on tryptic soy agar (TSA;

Becton Dickinson) plate supplemented with 0.5μg/ml erythromycin after incubation at 30°C for 2-3 days. PCR were performed directly on colonies using pMUTIN4 specific primers,

PMUTIN4_EryF and PMUTIN4_EryR (Table 15), to detect the plasmid sequence in the presumptive mutants. Gene disruption due to plasmid integration was further confirmed by

PCR using two primers (PMuDelBEcoF and LacZR1, Table 15): the primer PMuDelBEcoF was derived from the target gene paenB while the other primer, LacZR1, was specific to plasmid pMUTIN4 (São-José, Baptista, & Santos, 2004).

The antimicrobial activity of wild type strain OSY-DF or its mutants was determined using a microtiter plate assay. Aliquots (100 μl) of overnight cultures were mixed with 15 ml

Luria-Bertani broth in a 50 ml Falcon tube (BD Biosciences, Bedford, MA). The mixture was incubated at 30°C for 36 hrs with shaking at 200 rpm. Samples withdrawn at 24 and 36h were passed through a 0.22μm membrane filter and used for antimicrobial tests. Two susceptible indicators, L. innocua ATCC 33090 and B. subtilis 1A771, were used in the antimicrobial tests. Briefly, aliquots (100 μl) of 10-1 diluted culture (~ 107 CFU/ml) of each indicator were mixed with equal amounts of filtered culture samples in wells of a 96-well plate. The mixtures in the plate were incubated at 37°C for 9 hrs and OD600 was measured using a microtiter plate reader (Molecular Devices Corp.).

Paenibacillin production in the wild type and mutant strains was also tested using

149

matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF

MS), on a mass spectrometer (Bruker Reflex III time-of-flight; Bruker Daltonics Inc.,

Billerica, MA). Briefly, a sample of the filtered culture (36 hrs) was mixed with a matrix at a ratio of 1:5. The matrix was α-cyano-4-hydroxy cinnamic acid, prepared as a saturated solution in 50% acetonitrile with 0.1% TFA in water. The mixture was then spotted (1µl) on the target plate and allowed to air dry. The instrument was operated in reflection-positive ion mode at an accelerating voltage of 28 kV. The N2 laser was operated at the minimum threshold level required to generate signal and minimize dissociation.

5.2.6. Identification of the paen gene cluster by whole genome sequencing

The whole genome sequence of P. polymyxa OSY-DF was determined using the next- generation sequencing technology. Briefly, genomic DNA of the bacterial strain was isolated using a DNA extraction kit (DNeasy Blood & Tissue kit; Qiagen, Valencia, CA). The RNase- treated genomic DNA in Tris-Cl buffer (pH 8.5) was used for construction of a paired-end library with a Truseq DNA sample preparation kit (Illumina, San Diego, CA) according to the manufacturer’s instruction. The constructed library was sequenced (76 cycle paired-end runs) in a flow cell lane using an Illumina Genome Analyzer II. The de novo assembly of the short reads into longer contigs was performed using a commercial software program (CLC

Genomics Workbench 4.7.2; CLCBio, Cambridge, MA).

Automatic genome annotation was performed using the rapid annotation using

150

subsystem technology (RAST) (Aziz et al., 2008). The overall GC contend of the genome was calculated by the software Artemis (Rutherford et al., 2000). The rRNA and tRNA genes were predicted using RNAmmer (Lagesen et al., 2007) and tRNAscan-SE (Lowe & Eddy,

1997), respectively. The average nucleotide identity (ANI) between P. polymyxa OSY-DF and other P. polymyxa genomes available in the Genbank was determined using in silico DNA-

DNA hybridization (DDH) method implemented in the software JSpecies (Richter &

Rosselló-Móra, 2009).

The CLC Genomics Workbench (CLCBio) was used to search the paenibacillin gene cluster in the assembled genome of P. polymyxa OSY-DF. Open reading frames (ORFs) of the paenibacillin gene cluster were analyzed by Artemis; and protein function was predicted by searching for homologues using BLASTP at NCBI database.

5.2.7. Mode of action of paenibacillin

Paenibacillin on membrane potential

The membrane potential assays were performed using a fluorescence probe, 3,3'-

Dipropylthiadicarbocyanine Iodide (DiSC3(5)) (Invitrogen, Carlsbad, CA) according to previously reported method (Zhang, Dhillon, Yan, Farmer, & Hancock, 2000) with minor modifications. Overnight bacterial culture of Staphylococcus aureus ATCC 6538 or Listeria innocua ATCC 33090 was subcultured (1/100 dilution) into tryptic soy broth at 37°C with agitation at 200 rpm for ~5 hours. Cells were harvested at 4 °C at 3,660×g for 10 min and

151

washed twice using 5mM HEPES buffer (Sigma) with 5 mM glucose (Buffer A). Cells of S. aureus ATCC 6538 or L. innocua ATCC was resuspended in buffer A. Then the fluorescent probe, DiSC3(5), was added to the cell suspension to a final concentration of 0.5 μM from a

1mM stock solution in dimethyl sulfoxide (DMSO). Cells were incubated at room temperature for 15 min to equilibrate the internal and external dye concentrations. To measure the change of membrane potential, aliquots (90 μl) of the cell suspension were added to wells of white non-binding surface (NBS) microplate (Corning, Tewksbury, MA).

This is followed by the addition of 10 μl of paenibacillin or nisin at different concentrations.

Change of fluorescence due to membrane depolarization was recorded using a luminescence spectrometer (LS55, Perkin-Elmer, Wellesley, MA) with an excitation at 622 nm and an emission at 670 nm. Paenibacillin on potassium release Bacterial cells were grown, harvested and washed using the same procedures in the membrane potential assay. Aliquots (90 μl) of the cell suspension in buffer A were added to wells of white non-binding surface (NBS) microplate. The cell impermeable potassium- sensitive probe (PBFI, Invitrogen) was added to the cell suspension at a final concentration of 2 μM. Then 10 μl of paenibacillin or nisin was added to the cell suspension. Change of fluorescence corresponding to potassium concentration was recorded using a using a luminescence spectrometer (LS55, Perkin-Elmer, Wellesley, MA) with an excitation at 346 nm and an emission at 505 nm.

152

5.2.8. Applications of paenibacillin in foods

Crude extract preparation

An aliquot of 10 ml of P. polymyxa OSY-DF overnight culture was inoculated into a

2-liter flask containing 1 liter tryptic soy broth supplemented with 0.6 % yeast extract. The flask was incubated at 30 °C for 36 hrs at a rotary shaker at 200 rpm. Cells in the fermentation broth were removed by centrifuge at 15,180×g for 10 min in a SLA-1500 rotor at 4 °C (Sorvall RC-5B, DuPont, Wilmington, DE). The supernatant was passed through a cationic-exchange column (Macro-Prep High S; Bio-Rad, Hercules, CA). Paenibacillin retained on the column was eluted with at 0.3, 0.5 and 1.0 M NaCl. The eluate was subjected to a second purification step using multiple Sep-Pak C18 cartridges (Waters Corporation,

Milford, MA) connected in series to increase sample binding capacity. Paenibacillin was eluted from these cartridges by 70% acetonitrile and was dried by freeze drying. The crude extract (CE) of paenibacillin was obtained by dissolving the dry powder from evaporator with HPLC-grade water. The overall purification process is shown in Figure 30.

153

Figure 30. Ion exchange chromatography and solid phase extraction for paenibacillin purification

Application in Vienna sausage

A piece of sterile Vienna sausage (~ 15 g) was placed in a sterile stomacher bag. An aliquot of 0.5 ml paenibacillin crude extract was applied on the surface of sausage after inoculation of 0.5 ml Listeria monocytogenes Scott A. The samples were mixed well with hands and incubated at room temperature (25°C). Samples were taken at 0.5, 10 and 24 hrs after treatment for enumeration.

Application in irradiated ground beef

A representative portion (~10 g) of irradiated ground beef (Omaha Steaks, Omaha

Nebraska) was placed in a sterile stomacher bag. Aliquots (1.0 ml) of paenibacillin crude 154

extract and 0.5 ml of L. monocytogenes Scott A were added to the ground beef. The beef samples were mixed with hands and incubated at refrigerated temperature (4°C) for 7 days.

Samples were taken at 1 hour, 1 day, 3 days and 7 days after treatment for enumeration using modified oxford agar (MOX, Becton Dickinson).

5.3. Results

5.3.1. Minimum inhibitory concentration of paenibacillin

As shown in Table 16, paenibacillin exhibited potent activity against the Gram- positive bacteria, including methicillin-resistant S. aureus and L. monocytogenes. This compound is also active against the acid-fast Mycobacterium smegmatis, a bacterium often used experimentally in lieu of M. tuberculosis (Chaturvedi, Dwivedi, Tripathi, & Sinha,

2007). Compared to another lantibiotic (i.e., nisin), paenibacillin has a much lower minimum inhibitory concentration (MIC) against L. monocytogenes (16 and 1 µg/ml, respectively;

Table 16).

155

Table 16. Minimum inhibitory concentration (MICa) of antimicrobial agents

Bacterial Strains Media MIC of MIC of nisin MIC of paenibacillin (μg/ml) vancomycin (μg/ml) d (μg/ml) S. aureus (methicillin MHb broth 8.00 (0) 13.3 (4.62) 1.33 (0.57) resistant) S. aureus ATCC 29213 MH broth 5.33 (2.31) 13.3 (4.62) 2.00 (0) E. faecalis ATCC 29212 MH broth 13.3 (4.62) 13.3 (4.62) 2.00 (0) B. cereus ATCC 14579 MH broth 3.33 (1.15) 8.00 (0) 2.00 (0) B. cereus ATCC 11778 MH broth 16.0 (0) 8.00 (0) 1.00 (0) L. innocua ATCC 33090 MH broth with 5% 1.00 (0) 16.0 (0) 1.00 (0) lysed horse blood L. monocytogenes ScottA MH broth with 5% 1.00 (0) 16.0 (0) 1.00 (0) lysed horse blood S. agalactiae MH broth with 5% 26.7 (9.24) 16.0 (0) 1.00 (0) lysed horse blood M. smegmatis 7H9 brothc 8.00 (0) 8.00 (0) 8.00 (0) aMIC: lowest concentration of an antimicrobial that results in no visible growth of bacteria b Cation adjusted Mueller-Hinton broth (Becton Dickinson) c 7H9 broth from Becton Dickinson d Average of three replicates; values in parentheses were standard deviation.

5.3.2. Identification of paenibacillin genes paenA, paenP and paenB

The primary structure of paenibacillin was proposed previously (accession number:

P86013.1) (He et al., 2008), and it is important to confirm the sequence by revealing the structural gene of the lantibiotic. Comparison between lantibiotics with similarity to paenibacillin produced a likely conserved region within the leader peptide of the prepropeptides (Figure 31). Multiple sequence alignment also showed high similarity between paenibacillin and a putative lantibiotic, BtA1 in the C-terminal residues (Figure 31).

This information was useful in developing a functional primer set for amplifying the structural gene of paenibacillin. Using these primers, the structural gene of paenibacillin was

156

successfully amplified by PCR and the deduced amino acid sequence of the PCR products matched exactly with the chemically-determined paenibacillin primary structure.

Paenibacillin prepropeptide (PaenA) comprises 53 amino acid residues with a leader peptide from residues -1 to -23 (Figure 31). Alignment of PaenA with other lantibiotics (Figure 31), such as epilancin K7, epilancin 15X and epicidin 280, revealed a conserved cleavage site

( PQ) and a conserved FDLD motif in the leader peptide; leader peptides of lantibiotics are required for introducing posttranslational modification during biosynthesis (Plat, Kluskens,

Kuipers, Rink, & Moll, 2011).

DNA sequences flanking paenA were obtained by a PCR-based genome walking method. Two ORFs encoding putative lantibiotic dehydratase (PaenB) and subtilisin-like serine peptidase (PaenP) were identified in the downstream region of paenA.

Leader peptide -1 1 propeptide

PaenA MKV-DQMFDLDLRKSYEA--SELSPQ A-SIIKTTIKVSKAVCK----TLTCICTG-SCSNCK 53 BtA1 MN--KELFDLDINKKMETP-TEMTAQ T---VPTTIFVSRSVCK----TLTCICTI-SCSNCK 51 BtA2 MN--KELFDLDINKKMEAP-TEMTVQ T---WGTVVKVSKAICK----TGTCIGTI-SCTNCK 51 BsnA MEK-NNIFDLDINKKMEST-SEVSAQ TWATIGKTIVQSVKKCR----TFTCGCSLGSCSNCN 56 ElkA MN--NSLFDLNLNKGVETQKSDLSPQ SASVLKTSIKVSKKYCKGV--TLTCGCNI---TGGK 55 ElxA MK--KELFDLNLNKDIEAQKSDLNPQ SASIVKTTIKASKKLCRGF--TLTCGCHF---TGGK 55 EciA MENKKDLFDLEIKKDNMENNNELEAQ S---LGPAIKATRQVCPKATRFVTVSCKK---SDCQ 56 PaenA2 -MK--NQFDLDLQVAKNEVAPKE-VQ ---PASGLIC-TPS-CATG--TLNCQVSLSFCKTC- 50

Figure 31. Alignment of paenibacillin with related lantibiotic prepropeptides. PaenA: paenibacillin, BtA1 and BtA2: putative lantibiotic in Bacillus thuringiensis str. T01001, BsnA: putative lantibiotic in Bacillus subtilis Bsn5, ElkA: epilancin K7, ElxA: epilanicin 15X, EciA: epicidin 280, PaenA2: putative lantibiotic in Paenibacillus polymyxa OSY-DF.

157

5.3.3. Disruption of the lantibiotic dehydratase gene, paenB

The gene encoding putative lantibiotic dehydratase (PaenB) was inactivated by targeted mutagenesis. The confirmed mutant, with expected insertion in the paenB gene, was tested for paenibacillin production. The mutant strain (mutant B) lost its activity against two bacterial indicators, Listeria and Bacillus (Table 17). Moreover, paenibacillin production was not detected by MALDI-MS in the filtered sample (0.22µm, 36 hrs) from the mutant strain

(Figure 32). These findings proved that the mutation occurred within the paenibacillin gene cluster, and confirmed the importance of PaenB for the synthesis of the lantibiotic.

Table 17 Antimicrobial activity of the OSY-DF wild type and its mutant against two indicatorsa

Producer Listeria innocua Bacillus subtilis Strains 24 h 36 h 24 h 36 h Wild type 0.054±0.001 0.050±0.004 0.049±0.007 0.047±0.002 Mutant B 0.429±0.022 0.401±0.002 0.489±0.103 0.384±0.089 a The growth of indicators was determined by the microtiter plate assay, and reported as optical density measurements at 600 nm (OD600); samples from producers were taken after 24 or 36 h of incubation.

158

Figure 32. MALDI-MS analysis of paenibacillin production of Paenibacillus polymyxa OSY-DF wild type and ∆paenB mutant strains. (A) Mutant B, paenibacillin was not detected in the 36 h culture in Luria-Bertani broth. (B) Wild type strain, peak (arrow) [M+H]+=2984.54 and peak [M+Na]+=3006.57 corresponding to paenibacillin and its sodium adduct ion were detected in the 36 h culture in Luria-Bertani broth.

5.3.4. Whole genome sequencing of P. polymyxa OSY-DF

The whole genome shotgun project of P. polymyxa OSY-DF has been deposited at

GenBank under the accession AIPP00000000. The draft genome of P. polymyxa OSY-DF consists of 5,695,430 bases; the overall GC content of the genome was calculated as 45.35%.

159

Among the 5,139 protein-coding sequences (CDSs), 72% have been assigned putative function by RSAT. The chromosome of P. polymyxa OSY-DF has 1 rRNA operon and 42 tRNAs. In silico DNA-DNA hybridization (DDH) indicated that has the closest genetic relatedness with strain ATCC842T (97.50% ANI), followed by strains SC2 (94.52% ANI), M-

1(94.52% ANI) and E681 (91.78% ANI).

Some gene clusters encoding non-ribosomal peptide synthetase (NRPS) were identified in the genome of P. polymyxa OSY-DF. For instance, genes for polymyxin E1 production were found in four non-overlapping contigs. The complete fusaricidin gene cluster was found in a large contig, whose predicted protein shows 93% identity with its homologue in P. polymyxa PKB1 or P. polymyxa E681(Choi et al., 2008; J. Li, Beatty, Shah,

& Jensen, 2007).

Most importantly, two complete lantibiotic gene clusters were identified in the genome of P. polymyxa OSY-DF. One of lantibiotic gene clusters is responsible for paenibacillin biosynthesis (Figure 33) while the other one may encode a new lantibiotic as predicted by the bacteriocin mining tool, Bagel2 (De Jong, Van Heel, Kok, & Kuipers, 2010).

160

Figure 33. Genes involved in paenibacillin biosynthesis.

5.3.5. Organization of the paenibacillin gene cluster

The paenibacillin gene cluster (11.7 kb) comprised 11 putative ORFs. The length of each ORF, nearest homolog, identity percentage and the proposed function are shown in

Table 18. The organization of the gene cluster will be described in detail in the following texts.

161

Table 18. Open reading frames (ORFs) of the putative paenibacillin gene cluster

ORF Proposed function Length Closest homolog Identity no. (aa) (In NCBI database on 2/8/2012) (across × aa) 1 PaenA, paenibacillin 53 ZP_04136593, Lantibiotic 58% (50) prepropeptide paenibacillin [Bacillus thuringiensis T01001] 2 PaenP, peptidase 324 ZP_04136674, hypothetical 34% (317) protein bthur0003_58920[Bacillus thuringiensis T01001] 3 PaenB, lantibiotic 1027 ZP_04136675, hypothetical 31% (1044) dehydratase protein bthur0003_58930 [Bacillus thuringiensis T01001]

4 PaenC, lantibiotic 423 YP_004206154.1, lanthionine 33% (396) cyclase synthetase C-like protein [Bacillus subtilis BSn5] 5 PaenI, putative 185 YP_004206155, hypothetical 27% (138) immunity protein protein BSn5_12570 [Bacillus subtilis BSn5] 6 PaenT, transporter 597 NP_763973, ATP-binding 41% (571) cassette transporter A [Staphylococcus epidermidis ATCC 12228] 7 AgrB, accessory gene 137 ZP_03108595, accessory gene 34% (136) regulator B, regulator protein B [Bacillus processing signal cereus NVH0597-99] molecule 8 AgrD, autoinducing 57 ZP_00237849.1, conserved 33% (57) peptide (AIP) hypothetical protein [Bacillus cereus G9241] 9 AgrC, histidine 442 CAA75398, histidine kinase 30% (441) kinase [Lactobacillus plantarum] 10 AgrA, response 244 CAA75399, response regulator 43% (217) regulator protein [Lactobacillus plantarum] 11 PaenN, putative 256 ZP_05346156, protein TraX 34% (252) acetylase [Bryantella formatexigens DSM 14469]

162

Modification enzymes, PaenB, PaenC and PaenN

PaenB consists of 1027 amino acids with a theoretical molecular weight of 119.0 kDa, which is a homolog of lantibiotic dehydratase (LanB). PaenB shares the highest sequence similarity to LanB-like proteins from Bacillus thuringiensis T01001 (ZP_04136675; 31% identity), B. thuringiensis IBL200 (ZP_04075567.1; 31% identify) and B. subtilis Bsn5

(YP_004206153.1; 29% identity). These three Bacillus genomes encode paenibacillin-like lantibiotics (Figure 31) that have not been purified from the strains. Homologues of PaenB from well-studied lantibiotics included PepB (pep5), EpiB (epidermin), EciB (epicidin 280),

SpaB (subtilin) and NisB (nisin). PaenB is likely responsible for formation of Dha and Dhb from serine and threonine residues, respectively, in the paenibacillin propeptide.

PaenC shows homology to lantibiotic cyclases (LanC) such as PepC (pep5), EciC

(epicidin 280), EpiC (epidermin), SpaC (subtilin) and NisC (nisin). PaenC contains 423 amino acids with a calculated molecular mass of 47.3 kDa. The well-studied nisin cyclase

(NisC) is a zinc metalloprotein with a metal ligand and an acid-base catalytic site (B. Li et al.,

2006; B. Li & van der Donk, 2007). The zinc ligands (Cys 286, Cys335 and His336) and the active site residues (Asp153 and His216) are present in paenibacillin cyclase (PaenC). The results suggested that PaenC may catalyze the formation of the thioether bonds between Cys and Dha/Dhb during paenibacillin synthesis via a catalytic mechanism similar to that of NisC.

The N-terminal of paenibacillin contains a unique acetyl group (He et al., 2008). A putative acetylase (PaenN) may catalyze the N-terminal acetylation of alanine after the leader 163

peptide is cleaved. PaenN consists of 256 amino acids with a theoretical mass of 28.7 kDa.

PaenN shows high sequence homology with the TraX protein family (pfam05857); TraX is required for N-terminal acetylation of alanine in F-pilin in E. coli (Marchler-Bauer et al.,

2011; Moore et al., 1993). The proposed post-translational modifications of paenibacillin by

PaenB, PaenC and PaenN are shown in Figure 34.

Protease (PaenP), transporter (PaenT) and self-immunity (PaenI)

PaenP, a putative peptidase consisting of 324 amino acids with a theoretical mass of

36.0 kDa, shows high sequence similarity to subtilisin-like serine protease, LanP, such as

PepP (pep5), ElkP (epilancin K7) and NisP (nisin). PaenP contains the conserved catalytic triad residues (Asp43, His102 and Ser290) as well as the oxyanion hole, Asn185, which are the characteristics of this protease family (van der Meer et al., 1993). However, PaenP lacks the N-terminal sec-signal sequence and C-terminal cell wall anchor sequence (LPXTGX) found in NisP (Sahl & Bierbaum, 1998), suggesting that PaenP is likely located in the cytoplasm. Therefore, the leader peptide of paenibacillin may be cleaved within the cytoplasm and the mature lantibiotic is produced inside the cell.

PaenT is a putative ATP-binding cassette (ABC) transporter that may export the processed paenibacillin to the extracellular space. PaenT consists of 597 amino acids with a theoretical molecular weight of 67.2 kDa. The N-terminal domain of PaenT contains six membrane- spanning helices as predicted by TMHMM server 2.0 (Emanuelsson, Brunak, von Heijne, & Nielsen, 2007), whereas the C-terminal domain contains an ATP-binding site 164

as determined by conserved domain analysis (Marchler-Bauer et al., 2011). ATP hydrolysis provides the energy for peptide transportation (Sahl & Bierbaum, 1998).

165

166

Figure 34. Post-translational modifications of paenibacillin. (A) Common post-translational modifications in lantibiotics. (B) Proposed biosynthetic pathway of paenibacillin synthesis. Dha, dehydroalanine; Dhb, dehydrobutyrine. 166

PaenI is a putative self-immunity protein that consists of 185 amino acids with a theoretical mass of 21.0 kDa. PaenI shows sequence similarity to a putative permease

(YP_004206177.1) in B. Subtilis Bsn5, and a multidrug-efflux transporter (CAK02436) in

Bartonella tribocorum. PaenI contains 5 membrane-spanning helices as predicted by

TMHMM server 2.0 (Emanuelsson et al., 2007); therefore, PaenI may be integrated into cell membranes and facilitate the export of paenibacillin.

Quorum sensing system, encoded in paenibacillin gene cluster

The paenibacillin gene cluster contains an accessory gene regulator (agr)-like locus whose gene products may assemble a quorum sensing system. The well-studied staphylococcal Agr system consists of a typical two-component signaling module (AgrC and

AgrA), as well as AgrB and AgrD that are essential for production of the activating ligand, a

7–9 amino acid peptide with a thiolactone ring (Novick & Geisinger, 2008). Putative AgrC

167

and AgrA in OSY-DF show sequence homology with histidine kinase and response regulator, respectively. Putative AgrB in OSY-DF contains 137 amino acids with a theoretical mass of

13.7 kDa. AgrB in OSY-DF is a transmenbrane protein with 4 membrane spanning helices as predicted by TMHMM server 2.0 (Emanuelsson et al., 2007). AgrB protein may be involved in processing the signal peptide, AgrD. Putative AgrD in OSY-DF is a polypeptide with 57 amino acids and shows sequence similarity to other autoinducing peptides (AIPs) in Bacillus and Staphylococcus. The activated response regulator, AgrA, may regulate the production of paenibacillin (Figure 35).

AgrD_Bacillus --MKNMKKRLMDEVKHNVSKALGYIAIKSGEAATEK-LCIGFGYEPSVPTELLKLNKEKE 57

AgrD_Stapth ------MDLLNG-IFKLFAFIFEKIGNLAKYN-PCLGFLDEPTVPKELLEEDK--- 45

AgrD_Paen MPIMNNIKNIKEHRNNHTQRLKAKLLEHMSKRGEGGGICFGLFYEPVNP-KVWKLNKK-- 57 Figure 35. Alignment of putative signal precursor AgrD. AgrD_Bacillus: putative AgrD (ZP_00237849.1) from Bacillus; AgrD_Stapth: AgrD (AAL65845.1) from Staphylococcus; AgrD_Paen: putative AgrD from Paenibacillus. Underlined amino acids may form the activating ligand.

1 ATATATTCTTTATTCATCGACAGCAAGATATTACATACGGTTACTCACAATGAACGTGAT

61 TTAGGGGTACCACTGAAATCATGTTTATTAGAATTTAATTCAGAGAGGAAGAGGAAAATG -35 -10 RBS M 121 AAAGTAGATCAAATGTTTGACCTTGATTTAAGAAAGAGCTATGAGGCCAGTGAGCTTAGC K V D Q M F D L D L R K S Y E A S E L S 181 CCACAAGCTTCAATCATCAAGACAACCATTAAGGTATCCAAAGCAGTATGTAAAACTCTT P Q A S I I K T T I K V S K A V C K T L 241 ACCTGTATCTGTACCGGTTCTTGTTCCAATTGTAAATAA T C I C T G S C S N C K - Figure 36 Paenibacillin structural gene (paenA) and its preceding noncoding sequences. Underlined nucleotides are putative promoter (-35 and -10 elements) and ribosome binding site (RBS). The deduced leader peptide sequences are shaded.

168

Additional features of the lantibiotic biosynthesis include a 117-bp of noncoding sequence preceding paenA. The sequence presumably contains the promoter of the paenibacillin operon. The putative promoter was predicted using the BPROM program available at http://softberry.com, and the predicted -10 and -35 elements in the putative promoter are shown in Figure 36.

5.3.6. Paenibacillin targets cell membrane

Paenibacterin depolarizes cytoplasmic membrane

The cytoplasmic membrane potential was determined using a potential-sensitive probe DiSC3(5). The probe distributes in the intact cytoplasmic cell membrane and becomes self-quenched. As shown in Figure 37, paenibacillin released the fluorescent probe from the cytoplasmic membrane in S. aureus ATCC 6538 or L. innocua ATCC 33090 on a concentration manner. Similarly, the pore-forming lantibiotic, nisin, caused significant membrane depolarization in two bacterial strains.

169

(A)

(B)

Figure 37. Effect of paenibacillin (Pln) and nisin (Nis) on membrane potential.

(A) S. aureus ATCC 6538; (B) L. innocua ATCC 33090

Paenibacillin on potassium release

As shown in Figure 38, paenibacillin caused K+ release in S. aureus ATCC 6538 or L.

170

innocua ATCC 33090 at very high concentrations. The concentrations required to induce significant potassium release were much higher than the MIC of paenibacillin (Table 16).

(A)

(B)

Figure 38. Effect of paenibacillin (Pln) and nisin (Nis) on potassium release. (A) S. aureus ATCC 6538; (B) L. innocua ATCC 33090

171

5.3.7. Purification of paenibacillin and its application in meat products

Preparation of paenibacillin crude extract without polymyxin E

Paenibacillus polymyxa OSY-DF coproduces two antimicrobial compounds, paenibacillin and polymyxin E (He et al., 2007). Paenibacillin is active against Gram-positive bacteria while polymyxin E has activity against Gram-negative bacteria. As antibiotics in clinical use are not allowed to use in foods as preservatives, it is critical to separate polymyxin E1 from paenibacillin. As shown in Figure 39 and Figure 40, paenibacillin was eluted at low salt concentration of 0.3 M and 0.5 M whereas polymyxin E1 was not eluted until the salt concentration reached 1.0 M. This method provided an effective way to separate the antibiotic polymyxin E1 from the lantibiotic paenibacillin.

0.3 M NaCl 0.5 M NaCl

1.0 M NaCl

Figure 39. Activity tests of elution fractions from ion exchange chromatography. (A) Lactobacillus plantarum was used as indicator; fractions from 0.3 and 0.5 M NaCl were active. (B) Escherichia coli K12 was used as indicator, fraction from 1.0 M NaCl was active.

172

1 0.9 0.8 0.7 paenibacillin 0.6 0.5 0.4 0.3 polymyxin 0.2 NaCl concentration NaCl (M) 0.1 0 0 50 100 150 200 250 300 Elution fraction (ml)

Figure 40. Paenibacillin was separated from polymyxin E1 by gradient elution using the Macro-prep High S strong cationic support.

Effect of paenibacillin against Listeria inoculated in sausage and beef products

The initial Listeria count was ~5 log for both control and treatment. For the control, the population of Listeria increased to 7 log in 10 hrs and 8.7 log within 24 hrs. For the treatments, a 0.5 log reduction was observed in 10 hrs, and the growth of Listeria was significantly inhibited by paenibacillin in 24 hrs (Figure 41).

173

10.0 9.0

8.0

7.0 6.0 5.0 Control 4.0 Paenibacillin

3.0 Listeria Listeria (CFU/g) log 2.0 1.0 0.0 0 0.5 10 24 Time after treatment (h)

Figure 41. Inhibition of Listeria monocytogenes Scott A in Vienna sausagea. a average result from two replicates on modified oxford agar

7.0

6.0

5.0

4.0 Log (CFU/g) Log 3.0 Control Paenibacillin

2.0 Listeria

1.0

0.0 0 1 3 7 Time after treatments (days)

Figure 42. Inhibition of Listeria monocytogenes Scott A by Paenibacillin in ground beef a average result from two replicates on modified oxford agar; b Samples were taken 1 hour after treatment on day 1

The total plate count representing the natural microbiota in the irradiated ground beef 174

before inoculation was 2.2 × 104 CFU/g. The initial total plate count reached ~6 log for both control and treatment after inoculation. As shown in Figure 42, the population of Listeria in the control increased to 6.5 log and the total plate count reached 8 log in 3 days. For the treatment, a 0.5 log reduction was observed in 1 hour and the population of Listeria didn’t exceed its initial count (6 log) in 7 days.

5.4. Discussion

In this current study, we used the standard approach to determine the minimum inhibitory concentrations of paenibacillin against selected bacteria (Table 16). The results allow comparing the activity of paenibacillin with other compounds tested under the same conditions. Compared to the FDA-approved lantibiotic nisin, paenibacillin has a much higher antimicrobial activity against Listeria strains (Table 16). Considering the favorable heat and pH resistant properties (He et al., 2007), paenibacillin has the potential to be used in food products as a preservative. In this study, the efficacy of paenibacillin against Listeria monocytogenes in two different meat products was examined. Our results indicated that paenibacillin can effectively inhibit the growth of Listeria in sausage (Figure 41) and delay the outgrowth of the microorganism in ground beef (Figure 42).

The gene cluster of paenibacillin in P. polymyxa OSY-DF was identified using a PCR- based method combined with the whole genome sequencing. Knowledge on biosynthesis of paenibacillin is useful for genetical manipulation of the gene cluster to create paenibacillin

175

analougues with improved. In addition, some unique features in the paenibacillin gene cluster were also found. The putative acetylase, PaenN, in P. polymyxa OSY-DF may introduce the acetyl group to the N-terminal alanine of the mature paenibacillin after cleavage of the leader peptide. The function of acetylation in paenibacillin remains unknown but N-terminal modification of a peptide may protect it from proteolytic degradation (Velasquez, Zhang, & van der Donk, 2011). N-terminal acetylation is an enzymatic reaction catalyzed by N- terminal acetyltransferase (NAT); the peptide α-amino group receives the acetyl group from acetyl-CoA (Polevoda & Sherman, 2003). In eukaryotic organisms, 80-90% of mammalian cytosolic proteins and 50% of yeast proteins are acetylated (Polevoda & Sherman, 2003).

However, protein acetylation is extremely rare in prokaryotic organisms. Only 5 acetylated proteins from prokaryotes were identified in the Swiss-PROT database (Polevoda & Sherman,

2003). These acetylated proteins include the ribosomal protein L7 from Micrococcus luteus, the E. coli EFTu elongation factor and three E. coli ribosomal proteins S15, S5 and L12

(Polevoda & Sherman, 2003).

Lantibiotic production typically occurs at late exponential phase or early stationary phase. Nisin production is regulated by a two-component system which comprises a histidine kinase (NisK) and a response regulator (NisR) (Chatterjee et al., 2005). In S. epidermidis, transcription of epidermin biosynthetic genes is regulated by a response regulator, EpiQ

(Peschel, Augustin, Kupke, Stevanovic, & Götz, 1993), while extracellular processing of the

176

leader peptide by EpiP is controlled by an Agr quorum sensing system (Kies et al., 2003).

However, Agr system has not been reported in bacteria other than staphylococci. In this study, an agr-like locus was identified in the paenibacillin gene cluster in the genome of P. polymyxa OSY-DF. This is the first complete agr system, to the best of our knowledge, found in a bacteriocin biosynthetic cluster. The results suggested that paenibacillin production may be regulated by a cell density-dependent fashion via quorum-sensing.

The lantibiotic dehydratase, PaenB, plays an important role in generating the characteristic lanthionine ring structure. The production of paenibacillin was completely eliminated by knocking out the paenB gene; the loss of paenibacillin production was confirmed by MALDI-MS analysis. Data mining of the bacterial genome revealed that P. polymyxa OSY-DF may produce a second lantibiotic (PaenA2), encoded by a putative lantibiotic gene cluster (accession number: AIPP01000133). As shown in Figure 31, the predicted prepropeptide of PaenA2 comprises 50 residues with a leader sequence of 22 amino acids. The mutant strain (mutant B) grown in broth does not produce detectable antibacterial agents as determined by activity test. When the mutant strain, which lacks the production of paenibacillin, was cultured on TSA agar plate for two days, the colonies of mutant strain inhibited the growth of L. innocua that was overlaid on the mutant colonies.

The residual antimicrobial activity of the mutant strain on agar plate may result from the putative lantibiotic PaenA2 or other unidentified antibacterial compounds that are only produced on solid media. Many antimicrobial agents, including the broad-spectrum 177

lantibiotic from Bifidobacterium longum DJO10A (Lee, Li, & O'Sullivan, 2011), are only produced on the solid media. When a strain produces multiple antimicrobial agents, the producer usually remain active against indicator strains even when the critical gene for one compound is inactivated by targeted mutagenesis. For example, the disruption of the subtilosin gene, in strain B. subtilis 22a, did not completely eliminate the antilisterial activity

(Zheng, Yan, Vederas, & Zuber, 1999).

In addition, the mechanism of action of paenibacillin was investigated using fluorescent probes. Paenibacillin depolarized the bacterial cell membrane and induced potassium release from cells of S. aureus and L. innocua (Figure 37 and Figure 38). Cell membrane damage caused by paenibacillin may lead to cell death. Similarly, nisin exerts the antimicrobial activity by forming pores on cell membrane using lipid II as a ‘docking molecule’ (Chatterjee et al., 2005).

References

Breukink, E., & De Kruijff, B. (2006). Lipid II as a target for antibiotics. Nature Reviews Drug Discovery, 5(4), 321-323.

Chatterjee, C., Paul, M., Xie, L., & van der Donk, W. A. (2005). Biosynthesis and mode of action of lantibiotics. Chemical Reviews, 105(2), 633-684.

178

Chaturvedi, V., Dwivedi, N., Tripathi, R. P., & Sinha, S. (2007). Evaluation of Mycobacterium smegmatis as a possible surrogate screen for selecting molecules active against multi-drug resistant Mycobacterium tuberculosis. The Journal of General and Applied Microbiology, 53(6), 333-337.

Clinical and Laboratory Standards Institute. (2006). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Appproved Standard M7-A7,

Collins, B., Joyce, S., Hill, C., Cotter, P. D., & Ross, R. P. (2010). TelA contributes to the innate resistance of Listeria monocytogenes to nisin and other cell wall-acting antibiotics. Antimicrobial Agents and Chemotherapy, 54(11), 4658-4663.

Cotter, P. D., Hill, C., & Ross, R. P. (2005). Bacteriocins: developing innate immunity for food. Nature Reviews Microbiology, 3(10), 777-788.

De Jong, A., Van Heel, A. J., Kok, J., & Kuipers, O. P. (2010). BAGEL2: mining for bacteriocins in genomic data. Nucleic Acids Research, 38(suppl 2), W647-W651.

Emanuelsson, O., Brunak, S., von Heijne, G., & Nielsen, H. (2007). Locating proteins in the cell using TargetP, SignalP and related tools. Nature Protocols, 2(4), 953-971.

Gálvez, A., Abriouel, H., López, R. L., & Omar, N. B. (2007). Bacteriocin-based strategies for food biopreservation. International Journal of Food Microbiology, 120(1-2), 51-70.

Gravesen, A., Axelsen, A. M. J., Da Silva, J. M., Hansen, T. B., & Knöchel, S. (2002). Frequency of bacteriocin resistance development and associated fitness costs in Listeria monocytogenes. Applied and Environmental Microbiology, 68(2), 756-764.

He, Z., Yuan, C., Zhang, L., & Yousef, A. E. (2008). N-terminal acetylation in paenibacillin, a novel lantibiotic. FEBS Letters, 582(18), 2787-2792.

179

Kies, S., Vuong, C., Hille, M., Peschel, A., Meyer, C., Götz, F., & Otto, M. (2003). Control of antimicrobial peptide synthesis by the agr quorum sensing system in Staphylococcus epidermidis: activity of the lantibiotic epidermin is regulated at the level of precursor peptide processing. Peptides, 24(3), 329-338.

Lee, J. H., Li, X., & O'Sullivan, D. J. (2011). Transcription Analysis of a Lantibiotic Gene Cluster from Bifidobacterium longum DJO10A. Applied and Environmental Microbiology, 77(17), 5879-5887.

Li, B., & van der Donk, W. A. (2007). Identification of essential catalytic residues of the cyclase NisC involved in the biosynthesis of nisin. Journal of Biological Chemistry, 282(29), 21169.

Li, B., Yu, J. P. J., Brunzelle, J. S., Moll, G. N., Van der Donk, W. A., & Nair, S. K. (2006). Structure and mechanism of the lantibiotic cyclase involved in nisin biosynthesis. Science, 311(5766), 1464.

Lowe, T. M., & Eddy, S. R. (1997). tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Research, 25(5), 955-964.

Marchler-Bauer, A., Lu, S., Anderson, J. B., Chitsaz, F., Derbyshire, M. K., DeWeese-Scott, C., Fong, J. H., Geer, L. Y., Geer, R. C., & Gonzales, N. R. (2011). CDD: a Conserved Domain Database for the functional annotation of proteins. Nucleic Acids Research, 39(suppl 1), D225-D229.

Moore, D., Hamilton, C., Maneewannakul, K., Mintz, Y., Frost, L., & Ippen-Ihler, K. (1993). The Escherichia coli K-12 F plasmid gene traX is required for acetylation of F pilin. Journal of Bacteriology, 175(5), 1375-1383.

Murray, K. D., & Aronstein, K. A. (2008). Transformation of the Gram-positive honey bee pathogen, Paenibacillus larvae, by electroporation. Journal of Microbiological Methods, 75(2), 325-328. 180

Novick, R. P., & Geisinger, E. (2008). Quorum sensing in staphylococci. Annual Review of Genetics, 42, 541-564.

Peschel, A., Augustin, J., Kupke, T., Stevanovic, S., & Götz, F. (1993). Regulation of epidermin biosynthetic genes by EpiQ. Molecular Microbiology, 9(1), 31-39.

Plat, A., Kluskens, L. D., Kuipers, A., Rink, R., & Moll, G. N. (2011). Requirements of the engineered leader peptide of nisin for inducing modification, export, and cleavage. Applied and Environmental Microbiology, 77(2), 604-611.

Polevoda, B., & Sherman, F. (2003). N-terminal acetyltransferases and sequence requirements for N-terminal acetylation of eukaryotic proteins. Journal of Molecular Biology, 325(4), 595-622.

Richter, M., & Rosselló-Móra, R. (2009). Shifting the genomic gold standard for the prokaryotic species definition. Proceedings of the National Academy of Sciences, 106(45), 19126-19131.

Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M. A., & Barrell, B. (2000). Artemis: sequence visualization and annotation. Bioinformatics, 16(10), 944-945.

Sahl, H. G., & Bierbaum, G. (1998). Lantibiotics: biosynthesis and biological activities of uniquely modified peptides from gram-positive bacteria. Annual Reviews in Microbiology, 52(1), 41-79.

São-José, C., Baptista, C., & Santos, M. A. (2004). Bacillus subtilis operon encoding a membrane receptor for bacteriophage SPP1. Journal of Bacteriology, 186(24), 8337-8346.

Vagner, V., Dervyn, E., & Ehrlich, S. D. (1998). A vector for systematic gene inactivation in Bacillus subtilis. Microbiology, 144(11), 3097-3104. van der Meer, J. R., Polman, J., Beerthuyzen, M. M., Siezen, R. J., Kuipers, O. P., & De Vos, W. (1993). Characterization of the Lactococcus lactis nisin A operon genes nisP, encoding a subtilisin-like serine protease involved in precursor processing, and nisR, encoding a regulatory protein involved in nisin biosynthesis. Journal of Bacteriology, 175(9), 2578-2588. van Heel, A. J., Montalban-Lopez, M., & Kuipers, O. P. (2011). Evaluating the feasibility of lantibiotics as an alternative therapy against bacterial infections in humans. Expert Opinion on Drug Metabolism & Toxicology, 7(6), 675-680.

181

Velasquez, J. E., Zhang, X., & van der Donk, W. A. (2011). Biosynthesis of the antimicrobial peptide epilancin 15X and its N-terminal lactate. Chemistry & Biology, 18(7), 857-867.

Zhang, L., Dhillon, P., Yan, H., Farmer, S., & Hancock, R. E. W. (2000). Interactions of Bacterial Cationic Peptide Antibiotics with Outer and Cytoplasmic Membranes of Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy, 44(12), 3317-3321.

Zheng, G., Yan, L. Z., Vederas, J. C., & Zuber, P. (1999). Genes of the sbo-alb Locus of Bacillus subtilis Are Required for Production of the Antilisterial Bacteriocin Subtilosin. Journal of Bacteriology, 181(23), 7346-7355.

182

Chapter 6 Characterization of Enterocin RM6, A Bacteriocin from Enterococcus faecalis Inhibitory to Listeria monocytogenes in Cottage Cheese

Abstract

Bacteriocin-based techniques as part of hurdle technology for food preservation have received great attention in recent years. This study aimed to characterize enterocin

RM6 from Enterococcus faecalis OSY-RM6 and investigate its efficacy against Listeria monocytogenes in cottage cheese. Enterocin RM6 was purified from OSY-RM6 using ion

exchange column, multiple Sep-Pak C18 cartridges, followed by reverse-phase HPLC.

The molecular weight of enterocin RM6 is 7145.0823 as determined by ESI-MS; MS/MS analysis revealed that enterocin RM6 is a 70-residue cyclic peptide with a linkage between methionine at the N-terminus and tryptophan at the C-terminus. The sequence was also confirmed by sequencing the structural gene of the peptide. Enterocin RM6 is active against Gram-positive bacteria, including L. monocytogenes, Bacillus cereus and methicillin-resistant Staphylococcus aureus (MRSA). Enterocin RM6 (final concentration,

80 AU/ml) caused a 4-log reduction of L. monocytogenes in cottage cheese within 30 min

183

and no viable cells were detected at 26 hrs after treatment. Therefore, enterocin RM6 has potential applications for inhibiting spoilage bacteria and foodborne pathogens in foods.

184

6.1. Introduction

Listeria monocytogenes is the causative agent of human listeriosis with 20-30% fatality (Allerberger et al. 2010). Listeria is a ubiquitous microorganism that can be found in many raw and processed foods. Soft cheese products have been associated with several listeriosis outbreaks (Linnan et al., 1988; MacDonald et al., 2005; Makino et al., 2005).

Rudolf & Scherer (2001) reported that 6.4% of European red smear cheese samples were contaminated by L. monocytogenes. Recently, a multistate listeriosis outbreak occurred in the United States due to consumption of contaminated cantaloupe, resulting in 30 deaths in 28 states (Centers for Disease Control, and Prevention, 2011). Therefore, control of listeriosis remains an important issue for the food industry.

Bacteriocin-based techniques as part of hurdle technology for food preservation have gained great attention in recent years (Gálvez, Abriouel, López, & Omar, 2007). The effectiveness of many anti-listerial bacteriocins (e.g. nisin, pediocin and enterocin AS-48) has been investigated in foods including dairy products, meats and fresh produce (Davies,

Bevis, & Delves‐Broughton, 1997; Goff, Bhunia, & Johnson, 1996; Molinos et al., 2005).

In this study, we described a new isolate of Enterococcus faecalis from raw milk, which produces an anti-listerial peptide enterocin RM6. This bacteriocin was purified by liquid

185

chromatography techniques. The peptide sequence was determined by mass spectrometry and confirmed by structural gene sequencing. In addition, the efficacy of enterocin RM6 against L. monocytogenes in cottage cheese was investigated.

6.2. Materials and Methods

6.2.1. Bacterial strains and media

Strains were obtained from the culture collection of the food safety laboratory at

The Ohio State University (Columbus, OH). The producer strain E. faecalis OSY-RM6, which was isolated from raw milk (Zhang, 2008), was grown in MRS broth (Difco,

Sparks, MD). The indicator strain Lactobacillus cellobiosus OSU 919 was cultivated in

MRS broth. Other bacterial strains used for antimicrobial spectrum test are listed in

Table 20.

6.2.2. Strain identification by 16S rDNA sequencing

Genomic DNA of strain OSY-RM6 was purified using a DNA isolation kit

(DNeasy Blood & Tissue kit; QIAGEN, Valencia, CA). The 16S rDNA sequence was amplified by PCR using two universal primers (Weisburg, Barns, Pelletier, & Lane, 1991).

PCR amplification was performed using a Taq DNA polymerase kit (QIAGEN) under the following conditions: the reaction mixture was subjected to an initial denaturation at

186

94°C for 3 min, followed by 35 cycles, including 1 min at 94°C, 1 min at 52°C and 2 min at 72°C. A final extension was carried out at 72°C for 10 min. The amplified PCR product was purified using a gel extraction kit (QIAquick, QIAGEN), ligated to the pGEM-T

Easy vector (Promega, Madison, WI) and introduced into competent E. coli DH5α cells by electroporation. The recombinant plasmid carrying the 16S rDNA fragment was isolated from overnight culture of E. coli using spin column (QIAprep Spin Miniprep kit,

QIAGEN). Resultant plasmid DNA was sequenced using a 3730 DNA Analyzer (Applied

Biosystems, Foster city, CA) at the Plant-Microbe Genomics Facility at the Ohio State

University.

6.2.3. Purification of enterocin RM6 from cultured broth

An aliquot of 0.5 ml of OSY-RM6 overnight culture was inoculated into a 1-liter flask containing 500 ml MRS broth. The flask was incubated at 30°C for 18 hours without shaking. Cells in the cultured broth were removed by centrifugation at 15,180× g for 15 min at 4 °C using a SLA-1500 rotor (Sorvall RC-5B, DuPont, Wilmington, DE).

The supernatant was adjusted to pH 6.5 using 1.0 N NaOH and passed through a cationic exchange column (Macro-Prep High S support; Bio-Rad, Hercules, CA) that was equilibrated with phosphate buffer (50 mM, pH 6.5). Enterocin RM6 retained on the

187

column was eluted with 1.0 M NaCl in phosphate buffer (50 mM, pH 6.5). The resultant

eluate was subjected to solid phase extraction using ten Sep-Pak C18 cartridges (Waters

Corporation, Milford, MA) that were head-to-tail connected to increase binding capacity.

Enterocin RM6 was eluted from the cartridges by 70% acetonitrile and the solvents were removed by lyophilization. The crude extract (CE) of enterocin was obtained by dissolving the freeze-dried powder in HPLC-grade water.

Crude extract of enterocin was further purified by reverse-phase high performance liquid chromatography (RP-HPLC) (Hewlett Packard 1050, Agilent Technologies, Palo

Alto, CA). Separation was achieved using a preparative column with 5 μm particle size

(250 mm × 10mm; Alltech Associates, Inc., Deerfield, IL). The mobile phase consisted of (A) a mixture of isopropanol and acetonitrile (2:1, v/v) with 0.4% trifluoroacetic acid

(TFA) and (B) HPLC grade water with 0.1% TFA. For each run, aliquots (300 μl) of crude extract was loaded and separated on the column by a linear gradient of 0 to100% solvent A over 30 min, followed by 100% solvent A for 5 min at a flow rate of 1.5 ml/min.

Elution was monitored using a UV-detector at a wavelength of 280 nm. Fractions from each minute were collected automatically using Waters Fraction Collector II (Waters

Cooperation, Milford, MA). Fractions with the same retention time from multiple runs

188

were pooled and lyophilized; the resulting powder was dissolved in water. Antimicrobial activity of each faction was determined by microtiter plate bioassay. Active fraction against indicator strain was stored at 4°C for further analyses.

6.2.4. Antimicrobial activity determination and inhibition spectrum

Antimicrobial activity was determined by two methods: spot-on-lawn method and microtiter plate bioassay. Strain of L. cellobiosus OSU 919 was used as the indicator for both methods. Spot-on-lawn test was performed as described by (Yousef & Carlstrom,

2003). Briefly, aliquots (10 µl) of overnight indicator bacterium were transferred into 9 ml of MRS soft agar (0.75%) that was held at ~50 °C; the mixture was then poured onto a basal MRS agar plate. The tested compound or solution with enterocin RM6 was subjected to serial two-fold dilutions. Aliquots (5 µl) of diluted solution were spotted onto the soft agar layer seeded with indicator bacterium. After incubation at 30°C overnight, inhibitory areas were observed. Antimicrobial activity is expressed in arbitrary unit (AU/ml), which is the reciprocal of the highest dilution factor resulting in a clear inhibitory zone. In addition, antimicrobial spectrum of enterocin RM6 was determined by the spot-on-lawn method using the non-diluted crude extract against selected bacterial strains.

189

Antimicrobial activity of HPLC fractions was examined by microtiter plate bioassay. Briefly, overnight indicator culture was tenfold diluted using MRS medium and

100 µl of diluted cell suspension were added a well in a 96-well plate. An equal volume

(100 µl) of HPLC fraction was added to each well and incubated at 30°C for 5-8 hours, where sterile distilled water was used as negative control. Optical density (O.D.) was measured using a spectrophotometric microplate reader (Vmax Kinetic Microplate

Reader, Molecular Devices Corp., Menlo Park, CA) at 600 nm. Active fraction showed a lower O.D. value compared to the negative control because of the inhibitory effect against the indicator bacterium.

6.2.5. ESI-MS and MS/MS analyses

Accurate molecular weight determination and further peptide sequence investigation of enterocin RM6 was performed on a Thermo Finnigan LTQ orbitrap mass spectrometer operated in positive ion mode. Briefly, the sample diluted in the mixture of

H2O-MeOH-HAc (50:50:2.5) was infused into the electrospray source at a 6-µl/min flow rate. To achieve the optimal electrospray, spray voltage was set at 2,000 V; source temperature was 175°C. The data were recorded between 400-2000 Da and the resolution was set at 30000 to achieve high mass accuracy determination. The most abundant

190

enterocin RM6 peak was isolated for further MSMS study. The isolation window was set at 10Da and the CID fragmentation energy was set to 35%. Data were acquired in continuum mode until well-averaged data were obtained.

6.2.6. Structural gene of enterocin RM6

Genomic DNA from strain OSY-RM6 was used as template for amplifying the structural gene. Two primers (Ent48SF: 5’-GAGGAGTITCATGITTAAAGA-3’ and

Ent48SR: 5’-CATATTGTTAAATTACCAAGCAA-3’) were used for PCR (Martínez-

Bueno et al., 1994). PCR amplification was performed using a Taq DNA polymerase kit

(QIAGEN) under the following conditions: the reaction mixture was subjected to an initial denaturation at 95°C for 5 min, followed by 30 cycles, including 1 min at 95°C, 1 min at 52°C and 1 min at 72°C. A final extension was carried out at 72°C for 10 min. The resultant PCR product was cloned in pGEM-T Easy vector for sequencing.

6.2.7. Efficacy of enterocin RM6 on Listeria in inoculated cottage cheese

Twenty gram of cottage cheese (2% reduced milk fact, Kraft foods) was mixed with 30 ml of peptone water (0.1%) and the mixture was homogenized in a stomacher for

3 minutes. Enterocin crude extract was added to the diluted cheese sample at a final concentration of 80 AU/ml. Peptone water (0.1%) was used as negative control. The

191

mixture was inoculated with L. monocytogenes Scott A at a final concentration of ~105

CFU/ml. Inoculated cheese samples were incubated at 35°C for 26 hours. Samples were analyzed at 0.5 h, 4 h and 26 hours after treatment to examine the inhibitory effect of enterocin. Viable cells after treatment were counted using tryptic soy agar (Becton

Dickinson, Sparks, MD), PALCAM agar (Becton Dickinson) and modified Oxford agar

(Becton Dickinson). Each treatment or control included two independent experiments.

6.3. Results

6.3.1. Strain identification

The producer strain OSY-RM6 is a Gram-positive coccus. Analysis of the 16S rDNA gene showed a 99% similarity to E. faecalis. The new isolate was designated as E. faecalis OSY-RM6. This strain is positive for acid production (pH 4.8) in MRS broth and coagulates milk protein within 24 hours.

6.3.2. Purification of enterocin RM6

Crude extract of enterocin RM6 was prepared by a rapid two-step method using ion exchange chromatography and solid phase extraction. Enterocin RM6 retained on the

cationic resin was eluted with 1.0 M NaCl; the eluate was purified by Sep-Pak C18 cartridges. The procedure recovered ca. 25% total activity in the culture broth (Table 19).

192

The crude extract of enterocin RM6 was further purified by reverse-phase HPLC using a

preparative column. The retention time of enterocin RM6 on a C18 column was 31.268 min (Figure 43).

Table 19. Preparation of crude extract of enterocin RM6.

Purification step Volume Arbitrary Unit Total Activity Recovery (ml) (AU/mla) (×103AU) rate (%) Culture supernatant 500 800 400 100 Cation exchange 75 1600 120 30% Sep-Pak cartridges 60 1600 96 24% aAU/ml: the reciprocal of the highest dilution factor showing a visible inhibitory zone.

DAD1 E, Sig=280,20 Ref =550,100 (RM01012\SIG10007.D)

mAU

15.528 16.083 16.327 17.274 17.554 17.955 18.271 18.704 19.039 21.543 14 15.202

10 20.013

23.416

14.416

20.901 31.268

8 23.578

22.205 29.671

4 25.033 2

0 5 10 15 20 25 30 35 min

Retention Time (min) Figure 43. High performance liquid chromatography profile of the crude extract of enterocin RM6. The peak with retention time of 31.27 min (indicated by the arrow) showed antimicrobial activity against Lactobacillus cellobiosus OSU 919.

193

6.3.3. Molecular weight and peptide sequence determination by MS and MS/MS

HPLC purified enterocin RM6 was subjected to electrospray ionization mass spectrometry (ESI-MS) analysis for accurate molecular weight determination. Additional

MS/MS analysis was also performed to obtain the amino acid sequence of enterocin RM6.

As shown in Figure 44, peaks carrying different charge status (m/z= 894.14118+,

1021.73347+, 1191.85336+ and 1430.02425+) were observed for intact enterocin RM6.

After deconvolution, the average mono-isotopic mass (M+H) of enterocin RM6 was calculated as 7146.0828 Da. The most abundant peak (m/z=1191.85336+) was then isolated and further fragmented by CID to obtain the sequential information of this peptide. A series of MSMS product ions were observed at m/z =1419.173+, 1438.183+,

1467.193+, 1486.193+, 1519.223+, 1542.903+, 1576.583+, 1614.283+ and 1680.983+, which led to the identification of a 12 amino acids fragment sequenced as IVSILTAVGSGG or

GGSGVATLISVI (note: L and I can be switched since they have identical masses)

(Figure 45). This 12 amino acids sequence partially matches the sequence of a 70-residue cyclic peptide: AS-48 protein, whose complete sequence is MAKEFGIPAAVA

GTVLNVVEAGGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVI

AW, where a tail-to-head linkage is formed between the N-terminal methionine and the

194

C-terminal tryptophan through the dehydration of one water molecule (Martínez-Bueno et al., 1994; Maqueda et al., 2004). The theoretical molecular weight of AS-48 protein is

7145.0718Da, which matches exactly with the observed enterocin RM6 molecular weight

(7145.0745Da). Mass accuracy between theoretical and measured mass of enterocin RM6 is 0.44 ppm. In addition, more product ions were observed in the MSMS spectrum (see

Table S 1 for details), further supported that enterocin RM6 shares the same sequence with peptide AS-48 (Maqueda et al., 2004; Martínez-Bueno et al., 1994). Fragment ions were observed at m/z=575.32, 646.36, 777.40, 963.48, 1034.51, which corresponded to sequences KEFGI, AKEFGI, MAKEFGI, WMAKEFGI and AWMAKEFGI, respectively.

This observation thus confirmed that the NH2 terminus in Met1 was linked with the

COOH-terminus in Trp70 to form the head-to-tail cyclized peptide. During CID fragmentation, the cyclized peptide can be broken at different locations to generate different fragmentation patterns. For example, when the linkage was broken between Ile8 and Pro9, internal ions identified as PAAVA (m/z=410.24), PAAVAG (m/z=467.26),

PAAVAGT (m/z=568.31), PAAVAGTV (m/z=667.38) and

PAAVAGVTLNVVEAGGWVTTIV (m/z=1053.592+) were observed. Other fragment ions that were observed included ions represent amino acids sequences KEFGI, AKEFGI,

195

MAKEFGI, WMAKEFGI, AWMAKEFGI, as well as other amino acid sequences between SLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI and

EAGGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMA

KEFGI. Observation of other product ions suggested that the cyclized peptide can also be broken between Ala11 and Val12, Val12 and Ala13, Val19 and Val20. Detailed MSMS fragment assignments were listed in Table S 1. In addition, the peptide sequence was further confirmed by determination of structural gene of enterocin RM6. The deduced peptide sequence from structural gene agreed with the sequence determined by mass spectrometry (Figure 46).

Figure 44. ESI-MS analysis of enterocin RM6. 196

Figure 45. Fragmentation of enterocin RM6 examined by MSMS.

Figure 46 Structural gene of enterocin RM6 and the deduced peptide sequence.

197

6.3.4. Antimicrobial spectrum of enterocin

Enterocin RM6 is active against all tested Gram-positive bacteria, but it has no activity against Gram-negative bacteria (Table 20). Pediococcus acidilactici PO2 is the most sensitive strain to enterocin RM6. Moreover, this enterocin has strong activity against some important pathogens such as L. monocytogenes Scott A, B. cereus ATCC

14579 and methicillin-resistant S. aureus (MRSA).

6.3.5. Bactericidal effect of enterocin RM6 on Listeria in cottage cheese

When enterocin crude extract (final concentration, 80 AU/ml) was added to diluted cottage cheese inoculated with Listeria, a 4-log reduction of Listeria cells was observed within 30 min and viable cells were not detected after 26 hrs. In contrast, the population of L. monocytogenes without treatment increased to 107 CFU/ml after 26 hrs at the same incubation conditions (Figure 47).

198

Table 20. Antimicrobial spectrum of enterocin RM6.

Strainsa Mediab Diameter of inhibitory zone (mm) Gram-positive bacteria Pediococcus acidilactici PO2 MRS 17.5 Pediococcus pentosaceus MRS 16.8 Lactobacillus plantarum ATCC 8014 MRS 11.4 Lactobacillus casei ATCC 7469 MRS 13.4 Lactobacillus acidophillus ATCC 19992 MRS 10.0 Pediococcus cerevisiae MRS 11.1 Lactobacillus cellobiosus OSU 919 MRS 14.8 Listeria innocua ATCC33090 TSBYE 7.4 Enterococcus faecalis ATCC 29212 MRS 8.4 Listeria monocytogenes Scott A TSBYE 11.3 Bacillus cereus ATCC 14579 TSBYE 11.2 Bacillus cereus ATCC 11778 TSBYE 5.8 Staphylococcus aureus OSU 6538 NB 6.1 Staphylococcus aureus (methicillin sensitive ) NB 12.4 Staphylococcus aureus (methicillin resistant) NA 10.0 Gram-negative bacteria Yesinia enterocolitica TSBYE - Salmonella Typhimurium LB - Escherichia coli O157: H7 LB - aStrains obtained from the culture collection of The Ohio State University food safety laboratory. bMRS, Lactobacillus MRS broth ; TSBYE, Tryptic soy broth supplemented with 0.6% yeast extract; NB, Nutrient broth; LB, Luria–Bertani medium.

199

9.0 8.0 7.0 6.0

5.0 TSA 4.0 MOX

Log (CFU/ml) Log 3.0 PALCAM 2.0

Listeria Listeria 1.0 0.0 0.5 h 0.5 h 4 h 4 h 26h 26h Control Treatment Control Treatment Control Treatment

Figure 47. Effect of enterocin RM6 on survival of L. monocytogenes Scott A inoculated on cottage cheese using different mediaa. a TSA, tryptic soy agar; PALCAM, a selective media for Listeria; MOX, modified Oxford agar. Data represent the average of two independent experiments.

6.4. Discussions

Bacteriocins are ribosomally-synthesized antimicrobial peptides produced by different groups of Gram-positive bacteria. Many bacteriocins from Enterococcus spp. have been purified and characterized, such as enterocin A, B, P and AS-48 (Nes, Diep, &

Holo, 2007). In this study, we describe a new bacterial isolate from raw milk exhibiting antimicrobial activity against indicator strain L. cellobiosus. The new isolate was identified as E. faecalis by 16S rDNA sequencing and designated as strain OSY-RM6.

MS/MS analyses and structural gene sequencing confirmed that the antimicrobial activity was attributed to a cyclic peptide with 70 residues, whose chemical structure is the same 200

as enterocin AS-48.

The efficacy of bacteriocins is usually affected by environmental conditions in the food ecosystems such as food composition (Gálvez et al., 2007). The anti-listerial efficacy of enterocin RM6 was investigated in cottage cheese. The results indicated that enterocin RM6 has a rapid bactericidal activity against L. monocytogenes in cheese products. In all, enterocin RM6 may have a practical application in the food industry to control Listeria contamination.

201

Table S 1. Detailed MSMS analysis of Enterocin RM6.

Theoretical m/z Measured m/z Δ Mass Internal Ion Sequence (Da) Cyclized peptide broken in between I8 and P9 410.2398 410.24 PAAVA 57.02 PAAVAG 467.2613 467.26 (Gly) 101.05 PAAVAGT 568.3089 568.31 (Thr) 667.3774 667.38 99.07 (Val) PAAVAGTV 113.08 PAAVAGTVL 780.4614 780.461 (Leu) 114.04 PAAVAGTVLN 894.5043 894.501 (Asn) 993.5728/497.29032+ 993.571/497.292+ 99.07 (Val) PAAVAGTVLNV 202 1092.6412/546.82452+ 1092.64/546.822+ 99.07 (Val) PAAVAGTVLNVV

129.04 PAAVAGTVLNVVE 1221.6838/611.34582+ 1221.68/611.342+ (Glu) 71.04 PAAVAGTVLNVVEAG 1292.7209/646.86442+ 1292.72/646.862+ (Ala) 114.04 PAAVAGTVLNVVEAGG 1406.7638/703.88582+ 1406.76/703.882+ (Gly-Gly) 186.08 PAAVAGTVLNVVEAGGW 1592.8431/796.92552+ 1592.84/796.922+ (Trp) 1691.9115/846.45972+ 1691.91/846.462+ 99.07 (Val) PAAVAGTVLNVVEAGGWV 101.04 PAAVAGTVLNVVEAGGWVT 896.98352+ 896.982+ (Thr) 313.22 PAAVAGTVLNVVEAGGWVTTIV (Thr-Ile- 1053.58402+ 1053.592+ Val)

575.3188 575.32 KEFGI 71.04 AKEFGI 646.3559 646.36 (Ala) 777.3964 777.40 131.04(Me MAKEFGI

202

t) 186.08 WMAKEFGI 963.4757 963.48 (Trp) 71.03 AWMAKEFGI 1034.5128 1034.51 (Ala) 958.0586+2 958.05+2 KKGKRAVIAWMAKEFGI 370.24 EIKKKGKRAVIAWMAKEFGI 1143.1694+2 1143.17+2 (Lys-Leu- Glu) 128.10 KEIKKKGKRAVIAWMAKEFGI 1207.2169+2 1207.22+2 (Lys) 128.04 KKEIKKKGKRAVIAWMAKEFGI 1271.2643+2 1271.24+2 (Lys) 113.08 LKKEIKKKGKRAVIAWMAKEFGI 1327.8064+2 1327.78+2 (Leu) +3 +3

203 1362.4694 1362.47 SLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI +3 +3 113.07 LSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1400.1641 1400.16 (Leu) 57.03 GLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1419.1712+3 1419.17+3 (Gly) 57.03 GGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1438.1784+3 1438.18+3 (Gly) 1467.1891+3/1100.643 1467.19+3/1100.6 SGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 6+4 4+4 87.03 (Ser) 1486.1962+3/1114.899 1486.19+3/1114.9 57.00 GSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 0+4 0+4 (Gly) 1519.2190+3/1139.666 1519.22+3/1139.6 VGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1+4 4+4 99.09 (Val) 1542.8981+3/1157.425 1542.90+3/1157.4 71.04 AVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 4+4 2+4 (Ala) 101.04 TAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1576.5806+3 1576.58+3 (Thr) 1614.2753+3/1210.958 1614.28+3/1210.9 113.1 LTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 3+4 6+4 (Leu) 1651.9700+3/1239.229 1651.96+3/1239.2 113.04 ILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 3+4 3+4 (Ile)

203

1680.9807+3/1260.987 1680.98+3/1260.9 SILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 3+4 9+4 87.06 (Ser) 1714.0035+3/1285.754 1714.00+3/1285.7 VSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 4+4 5+4 99.06 (Val) 1751.6982+3/1314.025 1751.69+3/1314.0 113.07 IVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 4+4 2+4 (Leu) 1785.3807+3/1339.287 1785.38+3/1339.28 101.07 TIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 4+4 +4 (Thr) 1819.0633+3/1364.549 1819.06+3/1364.5 101.04 TTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 3+4 5+4 (Thr) 1852.0861+3/1389.316 1852.05+3/1389.3 VTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 4+4 1+4 98.97 (Val) 186.17 WVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1435.8362+4 1435.83+4 (Trp) 57.08 GWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1450.0916+4 1450.10+4 (Gly)

204 +4 +4 57.00 GGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI

1464.3469 1464.35 (Gly) 200.08 EAGGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGI 1514.3669+4 1514.37+4 (Ala+Glu) AVAGTVLNVVEAGGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWM 1396.4038+5 1396.40+5 AKEFGI 71.05 AAVAGTVLNVVEAGGWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAW 1410.6112+5 1410.61+5 (Ala) MAKEFGI Cyclized peptide broken in between A11 and V12 1124.4120+4 1124.41+4 GLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 57.04 GGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1138.6674+4 1138.67+4 (Gly) 1160.4254+4 1160.43+4 87.04 (Ser) SGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 57.00 GSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1174.6807+4 1174.68+4 (Gly) 1199.4478+4 1199.44+4 99.04 (Val) VGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 71.08 AVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1217.2071+4 1217.21+4 (Ala) 214.12 LTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1270.7401+4 1270.74+4 (Thr-Leu)

204

113.08 ILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1299.0111+4 1299.01+4 (Ile) 1320.7691+4 1320.77+4 87.04 (Ser) SILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1345.5362+4 1345.53+4 99.04 (Val) VSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 113.08 IVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1373.8072+4 1373.80+4 (Leu) 101.08 TIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1399.0691+4 1399.07+4 (Thr) 101.04 TTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA 1424.3310+4 1424.33+4 (Thr) 1449.0981+4 1449.10+4 99.08 (Val) VTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAA Cyclized peptide broken in between V19 and V20 1327.0270+4 1327.02+4 GGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 1348.7850+4 1348.78+4 87.04 (Ser) SGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 1363.0404+4 57.04 GSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 1363.04+4 205 (Gly) +4 +4

1387.8075 1387.80 99.04 (Val) VGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 71.04 AVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 1405.5667+4 1405.56+4 (Ala) 101.04 TAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAVAGTVLNV 1430.8287+4 1430.82+4 (Thr) Cyclized peptide broken in between V12 and A13 556.3089 556.31 AGTVLN 655.3774 655.38 99.07 (Val) AGTVLNV 754.4458 754.45 99.07 (Val) AGTVLNVV 129.04 AGTVLNVVE 883.4884 883.49 (Glu) 71.04 AGTVLNVVEA 954.5255 954.53 (Ala) 57.02 AGTVLNVVEAG 1011.5469 1011.55 (Gly)

1106.6527+4 1106.65+4 SLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 113.04 LSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1134.9237+4 1134.91+4 (Ala) 1531.9030+3 1531.89+3 57.03 GLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV

205

(Gly) 1163.4345+4/1550.910 1163.43+4/1550.9 57.05 GGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 2+3 1+3 (Gly) 1199.4478+4/1598.928 1199.45+4/1598.9 144.08 GSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 0+3 3+3 (Ser-Gly) 1224.2149+4/1631.950 1224.22+4/1631.9 VGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 8+3 5+3 99.08 (Val) 1241.9742+4/1655.629 1241.97+4/1655.6 71.00 AVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 9+3 3+3 (Ala) 1267.2361+4/1689.312 1267.24+4/1689.3 101.08 TAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 4+3 1+3 (Thr) 1295.5072+4/1727.007 1295.50+4/1727.0 113.04 LTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1+3 1+3 (Leu) 113.08 ILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1323.7782+4 1323.77+4 (Ile) 1345.5362+4/1793.712 1345.53+4/1793.7 SILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV +3 +3 206 5 0 87.04 (Ser) 1370.3033+4 1370.30+4 99.08 (Val) VSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV

113.04 IVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1398.5743+4 1398.56+4 (Ile) 101.08 TIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1423.8362+4 1423.83+4 (Thr) 101.08 TTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1449.0981+4 1449.10+4 (Thr) 1473.8652+4 1473.86+4 99.04 (Val) VTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 243.12 GWVTTIVSILTAVGSGGLSLLAAAGRESIKAYLKKEIKKKGKRAVIAWMAKEFGIPAAV 1534.6404+4 1534.64+4 (Trp-Gly)

206

References Allerberger, F., Wagner, M. (2010). Listeriosis: a resurgent foodborne infection. Clinical Microbiology and Infection. 16: 16-24.

Centers for Disease Control, and Prevention. (2011). Multistate outbreak of listeriosis linked to whole cantaloupes from jensen farms, colorado. Retrieved Jan. 31, 2013, from http://www.cdc.gov/listeria/outbreaks/cantaloupes-jensen-farms/120811/index.html

Davies, E., Bevis, H., & Delves‐Broughton, J. (1997). The use of the bacteriocin, nisin, as a preservative in ricotta‐type cheeses to control the food‐borne pathogen Listeria monocytogenes. Letters in Applied Microbiology, 24(5), 343-346.

Gálvez, A., Abriouel, H., López, R. L., & Omar, N. B. (2007). Bacteriocin-based strategies for food biopreservation. International Journal of Food Microbiology, 120(1-2), 51-70.

Goff, J. H., Bhunia, A. K., & Johnson, M. G. (1996). Complete inhibition of low levels of Listeria monocytogenes on refrigerated chicken meat with pediocin AcH bound to heat- killed Pediococcus acidilactici cells. Journal of Food Protection&;, 59(11), 1187-1192.

Linnan, M. J., Mascola, L., Lou, X. D., Goulet, V., May, S., Salminen, C., Hird, D. W., Yonekura, M. L., Hayes, P., & Weaver, R. (1988). Epidemic listeriosis associated with Mexican-style cheese. New England Journal of Medicine, 319(13), 823-828.

MacDonald, P. D. M., Whitwam, R. E., Boggs, J. D., MacCormack, J. N., Anderson, K. L., Reardon, J. W., Saah, J. R., Graves, L. M., Hunter, S. B., & Sobel, J. (2005). Outbreak of listeriosis among Mexican immigrants as a result of consumption of illicitly produced Mexican-style cheese. Clinical Infectious Diseases, 40(5), 677-682.

Makino, S. I., Kawamoto, K., Takeshi, K., Okada, Y., Yamasaki, M., Yamamoto, S., & Igimi, S. (2005). An outbreak of food-borne listeriosis due to cheese in Japan, during 2001. International Journal of Food Microbiology, 104(2), 189-196.

Maqueda, M., Galvez, A., Bueno, M. M., Sanchez-Barrena, M. J., Gonzalez, C., Albert, A., Rico, M., & Valdivia, E. (2004). Peptide AS-48: prototype of a new class of cyclic bacteriocins. Current Protein and Peptide Science, 5(5), 399-416.

Martínez-Bueno, M., Maqueda, M., Gálvez, A., Samyn, B., Van Beeumen, J., Coyette, J., & Valdivia, E. (1994). Determination of the gene sequence and the molecular structure of the enterococcal peptide antibiotic AS-48. Journal of Bacteriology, 176(20), 6334-6339.

Molinos, A. C., Abriouel, H., Omar, N. B., Valdivia, E., López, R. L., Maqueda, M., Cañamero, M. M., & Gálvez, A. (2005). Effect of immersion solutions containing enterocin AS-48 on 207

Listeria monocytogenes in vegetable foods. Applied and Environmental Microbiology, 71(12), 7781-7787.

Nes, I. F., Diep, D. B., & Holo, H. (2007). Bacteriocin diversity in Streptococcus and Enterococcus. Journal of Bacteriology, 189(4), 1189-1198.

Rudolf, M., & Scherer, S. (2001). High incidence of Listeria monocytogenes in European red smear cheese. International Journal of Food Microbiology, 63(1-2), 91-98.

Weisburg, W. G., Barns, S. M., Pelletier, D. A., & Lane, D. J. (1991). 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology, 173(2), 697-703.

Yousef, A., & Carlstrom, C. (2003). Food microbiology: A laboratory manual (pp. 234-236). Hoboken, New Jersey: John Wiley and Sons, Inc.

Zhang, X. (2008). Identification, properties, and application of enterocins produced by enterococcal isolates from foods. The Ohio State University).

208

Bibliography

Adams, M., & Moss, M. (2008). Food microbiology (pp. 112-115). Cambridge, UK: The Royal Society of Chemistry.

Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., & Walter, P. (2002). Molecular biology of the cell (4th ed.). New York, NY: Garland Science.

Alegría, Á., Szczesny, P., Mayo, B., Bardowski, J., & Kowalczyk, M. (2012). Biodiversity in Oscypek, a Traditional Polish Cheese, Determined by Culture-Dependent and-Independent Approaches. Applied and Environmental Microbiology, 78(6), 1890-1898.

Allerberger, F., Wagner, M. (2010). Listeriosis: a resurgent foodborne infection. Clinical Microbiology and Infection. 16: 16-24.

Artsimovitch, I., Seddon, J., & Sears, P. (2012). Fidaxomicin is an inhibitor of the initiation of bacterial RNA synthesis. Clinical Infectious Diseases, 55(suppl 2), S127-S131.

Ash, C., Priest, F. G., & Collins, M. D. (1993). Molecular identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks and Collins) using a PCR probe test. Antonie Van Leeuwenhoek, 64(3), 253-260.

Balasubramaniam, V., & Farkas, D. (2008). High-pressure food processing. Food Science and Technology International, 14(5), 413-418.

209

Baltz, R. H., Miao, V., & Wrigley, S. K. (2005). Natural products to drugs: daptomycin and related lipopeptide antibiotics. Natural Product Reports, 22(6), 717-741.

Bhushan, R., & Brückner, H. (2004). Marfey’s reagent for chiral amino acid analysis: a review. Amino Acids, 27(3), 231-247.

Blakistone, B. A. (1999). Introduction. In B. A. Blakistone (Ed.), Principles and applications of modified atmosphere packaging of foods (pp. 1-13). Gaithersburg, MD: Aspen Publishers.

Boeck, L., Fukuda, D. S., Abbott, B. J., & Debono, M. (1989). Deacylation of echinocandin B by Actinoplanes utahensis. Journal of Antibiotics, 42(3), 382-388.

Bonomo, R. A. (2011). New Delhi metallo-β-lactamase and multidrug resistance: a global SOS? Clinical Infectious Diseases, 52(4), 485-487.

Breukink, E., & De Kruijff, B. (2006). Lipid II as a target for antibiotics. Nature Reviews Drug Discovery, 5(4), 321-323.

Butler, M. S., & Cooper, M. A. (2011). Antibiotics in the clinical pipeline in 2011. The Journal of Antibiotics, 64(6), 413-425. 210

Cappelletty, D., & Eiselstein‐McKitrick, K. (2007). The echinocandins. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy, 27(3), 369-388.

CDC. (2011). Trends in foodborne illness, 1996–2010. Retrieved 1/17, 2013, from http://www.cdc.gov/foodborneburden/PDFs/FACTSHEET_B_TRENDS.pdf

Chang, S., & Cohen, S. N. (1979). High frequency transformation of Bacillus subtilis protoplasts by plasmid DNA. Molecular and General Genetics MGG, 168(1), 111-115.

Chatterjee, C., Paul, M., Xie, L., & van der Donk, W. A. (2005). Biosynthesis and mode of action of lantibiotics. Chemical Reviews, 105(2), 633-684.

Chaturongakul, S., Raengpradub, S., Wiedmann, M., & Boor, K. J. (2008). Modulation of stress and virulence in Listeria monocytogenes. Trends in Microbiology, 16(8), 388-396.

Chemat, F., & Khan, M. K. (2011). Applications of ultrasound in food technology: Processing, preservation and extraction. Ultrasonics Sonochemistry, 18(4), 813-835.

211

Chooi, Y. H., & Tang, Y. (2010). Adding the lipo to lipopeptides: do more with less. Chemistry & Biology, 17(8), 791-793.

Clardy, J., Fischbach, M. A., & Walsh, C. T. (2006). New antibiotics from bacterial natural products. Nature Biotechnology, 24(12), 1541-1550.

Clinical and Laboratory Standards Institute. (2006). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Appproved Standard M7-A7,

Coates, A. R. M., & Halls, G. (2012). Antibiotics in phase II and III clinical trials. In A. R. M. Coates (Ed.), Antibiotic resistance (pp. 167-183). New York, NY: Springer.

Cotter, P. D., Hill, C., & Ross, R. P. (2005). Bacteriocins: developing innate immunity for food. Nature Reviews Microbiology, 3(10), 777-788.

Davidson, P. M., & Branen, A. L. (2005). Food antimicrobials-an introduction. In P. M. Davidson, J. N. Sofos & A. L. Branen (Eds.), Antimicrobials in food (pp. 1-10). Boca Raton, FL: CRC Press.

212

De Jong, A., Van Heel, A. J., Kok, J., & Kuipers, O. P. (2010). BAGEL2: mining for bacteriocins in genomic data. Nucleic Acids Research, 38(suppl 2), W647-W651.

De Man, J. M. (1995). Principles of food chemistry (pp. 23). Gaithersburg, MD: Aspen Publishers.

Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G., Pfeifer, J., & Bax, A. (1995). NMRPipe: a multidimensional spectral processing system based on UNIX pipes. Journal of Biomolecular NMR, 6(3), 277-293. den Bakker, H. C., Switt, A. I. M., Cummings, C. A., Hoelzer, K., Degoricija, L., Rodriguez- Rivera, L. D., Wright, E. M., Fang, R., Davis, M., & Root, T. (2011). A whole-genome single nucleotide polymorphism-based approach to trace and identify outbreaks linked to a common Salmonella enterica subsp. enterica serovar Montevideo pulsed-field gel electrophoresis type. Applied and Environmental Microbiology, 77(24), 8648-8655.

Denamur, E. (2011). The 2011 Shiga toxin‐producing Escherichia coli O104: H4 German outbreak: a lesson in genomic plasticity. Clinical Microbiology and Infection, 17(8), 1124- 1125.

Deng, X., Li, Z., & Zhang, W. (2011). Transcriptome Sequencing of Salmonella enterica serovar Enteritidis under Desiccation and Starvation Stress in Peanut Oil. Food Microbiology, 30(1), 311-315.

Denning, D. W. (2003). Echinocandin antifungal drugs. Lancet, 362(9390), 1142-1151.

Dixon, R. A., & Chopra, I. (1986). Leakage of periplasmic proteins from Escherichia coli mediated by polymyxin B nonapeptide. Antimicrobial Agents and Chemotherapy, 29(5), 781-788.

Duwe, A. K., Rupar, C. A., Horsman, G. B., & Vas, S. I. (1986). In vitro cytotoxicity and antibiotic activity of polymyxin B nonapeptide. Antimicrobial Agents and Chemotherapy, 30(2), 340-341. 213

Emanuelsson, O., Brunak, S., von Heijne, G., & Nielsen, H. (2007). Locating proteins in the cell using TargetP, SignalP and related tools. Nature Protocols, 2(4), 953-971.

Evans, M. E., Feola, D. J., & Rapp, R. P. (1999). Polymyxin B sulfate and colistin: old antibiotics for emerging multiresistant gram-negative bacteria. The Annals of Pharmacotherapy, 33(9), 960-967.

Farkas, J. (2006). Irradiation for better foods. Trends in Food Science & Technology, 17(4), 148- 152.

Fischbach, M. A., & Walsh, C. T. (2006). Assembly-line enzymology for polyketide and nonribosomal peptide antibiotics: logic, machinery, and mechanisms. Chemical Reviews, 106(8), 3468-3496.

Fischbach, M. A., & Walsh, C. T. (2009a). Antibiotics for emerging pathogens. Science, 325(5944), 1089-1093.

Fischbach, M. A., & Walsh, C. T. (2009b). Antibiotics for emerging pathogens. Science, 325(5944), 1089-1093.

Gálvez, A., Abriouel, H., López, R. L., & Omar, N. B. (2007). Bacteriocin-based strategies for food biopreservation. International Journal of Food Microbiology, 120(1-2), 51-70.

García, P., Martínez, B., & Rodríguez, A. (2011). Bacteriophages and phage-encoded proteins: Prospects in food quality and safety. In M. Rai, & M. Chikindas (Eds.), Natural antimicrobials in food safety and quality (pp. 10-26). Cambridge, MA: CABI Publishing.

Gómez-López, V. M., Ragaert, P., Debevere, J., & Devlieghere, F. (2007). Pulsed light for food decontamination: a review. Trends in Food Science & Technology, 18(9), 464-473.

Gordon, R. E., Haynes, W. C., Pang, C. H. N., & Smith, N. R. (1973). The genus bacillus. Washington, DC: Agricultural Research Service, US Department of Agriculture.

Guerrero-Beltrán, J. A., & Barbosa-Cánovas, G. V. (2004). Advantages and limitations on processing foods by UV light. Food Science and Technology International, 10(3), 137-147.

Hamad, B. (2010). The antibiotics market. Nature Reviews Drug Discovery, 9(9), 675-676.

Hancock, R. E. W. (1997). Peptide antibiotics. The Lancet, 349(9049), 418-422.

215

Hancock, R. E. W., & Chapple, D. S. (1999). Peptide antibiotics. Antimicrobial Agents and Chemotherapy, 43(6), 1317-1323.

Hancock, R. E. W., & Sahl, H. G. (2006). Antimicrobial and host-defense peptides as new anti- infective therapeutic strategies. Nature Biotechnology, 24(12), 1551-1557.

He, Z., Yuan, C., Zhang, L., & Yousef, A. E. (2008). N-terminal acetylation in paenibacillin, a novel lantibiotic. FEBS Letters, 582(18), 2787-2792.

Heinis, C., Rutherford, T., Freund, S., & Winter, G. (2009). Phage-encoded combinatorial chemical libraries based on bicyclic peptides. Nature Chemical Biology, 5(7), 502-507.

Hindré, T., Pennec, J. P., Haras, D., & Dufour, A. (2006). Regulation of lantibiotic lacticin 481 production at the transcriptional level by acid pH. FEMS Microbiology Letters, 231(2), 291- 298.

Hogan, E., Kelly, A. L., & Sun, D. W. (2005). High pressure processing. In D. W. Sun (Ed.), Emerging technologies for food processing (pp. 3-32). San Diego, CA: Elsrvier Academic Press.

Imlay, J. A. (2002). How oxygen damages microbes: oxygen tolerance and obligate anaerobiosis. Advances in Microbial Physiology, 46, 111-153.

216

Imlay, J. A. (2003). Pathways of oxidative damage. Annual Reviews in Microbiology, 57(1), 395- 418.

Jay, J. M., Loessner, M. J., & Golden, D. A. (2005). Modern food microbiology (pp. 301-350). New York, NY: Springer.

Jenssen, H., Hamill, P., & Hancock, R. E. W. (2006). Peptide antimicrobial agents. Clinical Microbiology Reviews, 19(3), 491-511.

Jeyamkondan, S., Jayas, D., & Holley, R. (1999). Pulsed electric field processing of foods: a review. Journal of Food Protection, 62(9), 1088-1096.

Johnson, B. A., & Blevins, R. A. (1994). NMR View: A computer program for the visualization and analysis of NMR data. Journal of Biomolecular NMR, 4(5), 603-614.

Juneja, V. K., Dwivedi, H. P., & Yan, X. (2012). Novel Natural Food Antimicrobials. Annual Review of Food Science and Technology, 3, 381-403.

Kantor, L. S., Lipton, K., Manchester, A., & Oliveira, V. (1997). Estimating and addressing America’s food losses. Food Review, 20(1), 2-12.

Knerr, P. J., & van der Donk, W. A. (2012). Chemical Synthesis and Biological Activity of Analogues of the Lantibiotic Epilancin 15X. Journal of the American Chemical Society, 134(18), 7648-7651.

217

Kohanski, M. A., Dwyer, D. J., & Collins, J. J. (2010). How antibiotics kill bacteria: from targets to networks. Nature Reviews Microbiology, 8(6), 423-435.

Kohanski, M. A., Dwyer, D. J., Hayete, B., Lawrence, C. A., & Collins, J. J. (2007). A common mechanism of cellular death induced by bactericidal antibiotics. Cell, 130(5), 797-810.

Kohli, R. M., & Walsh, C. T. (2003). Enzymology of acyl chain macrocyclization in natural product biosynthesis. Chemical Communications, 7(3), 297-307.

Konz, D., & Marahiel, M. A. (1999). How do peptide synthetases generate structural diversity? Chemistry & Biology, 6(2), R39-48.

Koponen, O., Tolonen, M., Qiao, M., Wahlstrom, G., Helin, J., & Saris, P. E. J. (2002). NisB is required for the dehydration and NisC for the lanthionine formation in the post-translational modification of nisin. Microbiology, 148(11), 3561.

Kume, T., Furuta, M., Todoriki, S., Uenoyama, N., & Kobayashi, Y. (2009). Status of food irradiation in the world. Radiation Physics and Chemistry, 78(3), 222-226.

Kupferschmidt, K. (2011). Scientists rush to study genome of lethal E. coli. Science, 332(6035), 1249-1250.

Kussendrager, K. D., & Van Hooijdonk, A. (2000). Lactoperoxidase: physico-chemical properties, occurrence, mechanism of action and applications. British Journal of Nutrition, 84, 19-25.

Landman, D., Georgescu, C., Martin, D. A., & Quale, J. (2008). Polymyxins revisited. Clinical Microbiology Reviews, 21(3), 449-465.

Lee, J. H., Li, X., & O'Sullivan, D. J. (2011). Transcription Analysis of a Lantibiotic Gene Cluster from Bifidobacterium longum DJO10A. Applied and Environmental Microbiology, 77(17), 5879-5887. 218

Levy, S. B., & Marshall, B. (2004). Antibacterial resistance worldwide: causes, challenges and responses. Nature Medicine, 10, S122-S129.

Li, B., & van der Donk, W. A. (2007). Identification of essential catalytic residues of the cyclase NisC involved in the biosynthesis of nisin. Journal of Biological Chemistry, 282(29), 21169- 21175.

Lowe, T. M., & Eddy, S. R. (1997). tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Research, 25(5), 955-964.

Marfey, P. (1984). Determination of D-amino acids. II. Use of a bifunctional reagent, 1, 5- difluoro-2, 4-dinitrobenzene. Carlsberg Research Communications, 49(6), 591-596. 220

McAuliffe, O., Ross, R. P., & Hill, C. (2001). Lantibiotics: structure, biosynthesis and mode of action. FEMS Microbiology Reviews, 25(3), 285-308.

McQuade, T. J., Shallop, A. D., Sheoran, A., Delproposto, J. E., Tsodikov, O. V., & Garneau- Tsodikova, S. (2009). A nonradioactive high-throughput assay for screening and characterization of adenylation domains for nonribosomal peptide combinatorial biosynthesis. Analytical Biochemistry, 386(2), 244-250.

McSpadden Gardener, B. B. (2004). Ecology of Bacillus and Paenibacillus spp. in agricultural systems. Phytopathology, 94(11), 1252-1258.

Metzker, M. L. (2009). Sequencing technologies—the next generation. Nature Reviews Genetics, 11(1), 31-46.

Mills, S., Stanton, C., Hill, C., & Ross, R. (2011). New developments and applications of bacteriocins and peptides in foods. Annual Review of Food Science and Technology, 2, 299- 329.

Molinos, A. C., Abriouel, H., Omar, N. B., Valdivia, E., López, R. L., Maqueda, M., Cañamero, M. M., & Gálvez, A. (2005). Effect of immersion solutions containing enterocin AS-48 on Listeria monocytogenes in vegetable foods. Applied and Environmental Microbiology, 71(12), 7781-7787.

Murray, K. D., & Aronstein, K. A. (2008). Transformation of the Gram-positive honey bee pathogen, Paenibacillus larvae, by electroporation. Journal of Microbiological Methods, 75(2), 325-328.

Nes, I. F., Diep, D. B., & Holo, H. (2007). Bacteriocin diversity in Streptococcus and Enterococcus. Journal of Bacteriology, 189(4), 1189-1198.

Novick, R. P., & Geisinger, E. (2008). Quorum sensing in staphylococci. Annual Review of Genetics, 42, 541-564.

Oms-Oliu, G., Martín-Belloso, O., & Soliva-Fortuny, R. (2010). Pulsed light treatments for food preservation. A review. Food and Bioprocess Technology, 3(1), 13-23.

Ongena, M., & Jacques, P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends in Microbiology, 16(3), 115-125.

Osaili, T. M. (2012). Developments in the thermal processing of food. In R. Bhat, A. K. Alias & G. Paliyath (Eds.), Progress in food preservation (pp. 211-230). West sussex, UK: Wiley- Blackwell.

Peschel, A., Augustin, J., Kupke, T., Stevanovic, S., & Götz, F. (1993). Regulation of epidermin biosynthetic genes by EpiQ. Molecular Microbiology, 9(1), 31-39.

Piper, C., Cotter, P. D., Ross, R. P., & Hill, C. (2009). Discovery of medically significant lantibiotics. Current Drug Discovery Technologies, 6(1), 1-18.

Pirri, G., Giuliani, A., Nicoletto, S. F., Pizzuto, L., & Rinaldi, A. C. (2009). Lipopeptides as anti- infectives: a practical perspective. Central European Journal of Biology, 4(3), 258-273.

223

Polevoda, B., & Sherman, F. (2003). N-terminal acetyltransferases and sequence requirements for N-terminal acetylation of eukaryotic proteins. Journal of Molecular Biology, 325(4), 595-622.

Rahman, M. S. (2007a). Food preservation: Overview. In M. S. Rahman (Ed.), Handbook of food preservation (pp. 3-17). Boca Raton, FL: CRC Press.

Rahman, M. S. (2007b). Light energy in food preservation. In M. S. Rahman (Ed.), Handbook of food preservation (pp. 751-757). Boca Raton, FL: CRC Press.

Rastogi, N. K. (2011). Opportunities and Challenges in Application of Ultrasound in Food Processing. Critical Reviews in Food Science and Nutrition, 51(8), 705-722.

Richter, M., & Rosselló-Móra, R. (2009). Shifting the genomic gold standard for the prokaryotic species definition. Proceedings of the National Academy of Sciences, 106(45), 19126-19131.

Robbel, L., & Marahiel, M. A. (2010). Daptomycin, a bacterial lipopeptide synthesized by a nonribosomal machinery. Journal of Biological Chemistry, 285(36), 27501-27508.

Rodriguez-Palacios, A., & LeJeune, J. T. (2011). Moist-heat resistance, spore aging, and superdormancy in Clostridium difficile. Applied and Environmental Microbiology, 77(9), 3085-3091.

224

Rosenthal, K. S., Swanson, P. E., & Storm, D. R. (1976). Disruption of Escherichia coli outer membranes by EM 49. A new membrane active peptide antibiotic. Biochemistry, 15(26), 5783-5792.

Rudolf, M., & Scherer, S. (2001). High incidence of Listeria monocytogenes in European red smear cheese. International Journal of Food Microbiology, 63(1-2), 91-98.

Russell, N. J., & Gould, G. W. (2003). Major preservative technologies. In N. J. Russell, & G. W. Gould (Eds.), Food preservatives (pp. 14-24). New York, NY: Kluwer Acdemic/Plenum Pblishers,.

Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, M. A., & Barrell, B. (2000). Artemis: sequence visualization and annotation. Bioinformatics, 16(10), 944-945.

Sahl, H. G., & Bierbaum, G. (1998). Lantibiotics: biosynthesis and biological activities of uniquely modified peptides from gram-positive bacteria. Annual Reviews in Microbiology, 52(1), 41-79.

Sakamoto, N., Tanaka, S., Sonomoto, K., & Nakayama, J. (2011). 16S rRNA pyrosequencing- based investigation of the bacterial community in nukadoko, a pickling bed of fermented rice bran. International Journal of Food Microbiology, 144(3), 352-359.

São-José, C., Baptista, C., & Santos, M. A. (2004). Bacillus subtilis operon encoding a membrane receptor for bacteriophage SPP1. Journal of Bacteriology, 186(24), 8337-8346.

Scallan, E., Hoekstra, R. M., Angulo, F. J., Tauxe, R. V., Widdowson, M. A., Roy, S. L., Jones, J. L., & Griffin, P. M. (2011). Foodborne illness acquired in the United States—major pathogens. Emerging Infectious Diseases, 17(1), 7-15.

Scharff, R. L. (2012). Economic burden from health losses due to foodborne illness in the United States. Journal of Food Protection, 75(1), 123-131.

225

Schuster, S. C. (2007). Next-generation sequencing transforms today’s biology. Nature, 200(8), 16-18.

Shaheen, M., Li, J., Ross, A. C., Vederas, J. C., & Jensen, S. E. (2011). Paenibacillus polymyxa PKB1 produces variants of polymyxin B-type antibiotics. Chemistry & Biology, 18(12), 1640-1648.

Shendure, J., & Ji, H. (2008). Next-generation DNA sequencing. Nature Biotechnology, 26(10), 1135-1145.

Shi, Y., Bueno, A., & van der Donk, W. A. (2012). Heterologous production of the lantibiotic Ala (0) actagardine in Escherichia coli. Chemical Communications, 48(89), 10966-10968.

Sobrino-López, A., & Martín-Belloso, O. (2010). Review: potential of high-intensity pulsed electric field technology for milk processing. Food Engineering Reviews, 2(1), 17-27.

Stachelhaus, T., Mootz, H. D., & Marahiel, M. A. (1999). The specificity-conferring code of adenylation domains in nonribosomal peptide synthetases. Chemistry & Biology, 6(8), 493- 505.

Stachelhaus, T., & Walsh, C. T. (2000). Mutational analysis of the epimerization domain in the initiation module PheATE of gramicidin S synthetase. Biochemistry, 39(19), 5775-5787.

Stawikowski, M., & Cudic, P. (2007). Depsipeptide synthesis. In G. B. Fields (Ed.), Peptide characterization and application protocols (pp. 321-339). New York, NY: Springer.

226

Tabor, A. B. (2011). The challenge of the lantibiotics: synthetic approaches to thioether-bridged peptides. Organic & Biomolecular Chemistry, 9(22), 7606-7628.

Vaara, M. (1992). Agents that increase the permeability of the outer membrane. Microbiological Reviews, 56(3), 395-411.

Vagner, V., Dervyn, E., & Ehrlich, S. D. (1998). A vector for systematic gene inactivation in Bacillus subtilis. Microbiology, 144(11), 3097-3104. van der Meer, J. R., Polman, J., Beerthuyzen, M. M., Siezen, R. J., Kuipers, O. P., & De Vos, W. (1993). Characterization of the Lactococcus lactis nisin A operon genes nisP, encoding a subtilisin-like serine protease involved in precursor processing, and nisR, encoding a regulatory protein involved in nisin biosynthesis. Journal of Bacteriology, 175(9), 2578- 2588. van Heel, A. J., Montalban-Lopez, M., & Kuipers, O. P. (2011). Evaluating the feasibility of lantibiotics as an alternative therapy against bacterial infections in humans. Expert Opinion on Drug Metabolism & Toxicology, 7(6), 675-680.

Van Vliet, A. H. M. (2009). Next generation sequencing of microbial transcriptomes: challenges and opportunities. FEMS Microbiology Letters, 302(1), 1-7.

Velasquez, J. E., Zhang, X., & van der Donk, W. A. (2011). Biosynthesis of the antimicrobial peptide epilancin 15X and its N-terminal lactate. Chemistry & Biology, 18(7), 857-867.

227

Velkov, T., Thompson, P. E., Nation, R. L., & Li, J. (2010). Structure—Activity Relationships of Polymyxin Antibiotics. Journal of Medicinal Chemistry, 53(5), 1898. von der Weid, I., Duarte, G. F., van Elsas, J. D., & Seldin, L. (2002). Paenibacillus brasilensis sp. nov., a novel nitrogen-fixing species isolated from the maize rhizosphere in Brazil. International Journal of Systematic and Evolutionary Microbiology, 52(6), 2147-2153.

Walsh, C. (2003). Antibiotics: Actions, origins, resistance. Washington, DC: American Society for Microbiology (ASM) Press.

Wang, X., & Zhao, X. (2009). Contribution of oxidative damage to antimicrobial lethality. Antimicrobial Agents and Chemotherapy, 53(4), 1395-1402.

Wang, Z., Gerstein, M., & Snyder, M. (2009). RNA-Seq: a revolutionary tool for transcriptomics. Nature Reviews Genetics, 10(1), 57-63.

Weisburg, W. G., Barns, S. M., Pelletier, D. A., & Lane, D. J. (1991a). 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology, 173(2), 697-703.

Weisburg, W. G., Barns, S. M., Pelletier, D. A., & Lane, D. J. (1991b). 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology, 173(2), 697-703.

Willey, J. M., & Van der Donk, W. A. (2007). Lantibiotics: peptides of diverse structure and function. Annual Review of Microbiology, 61, 477-501.

Wittmann, M., Linne, U., Pohlmann, V., & Marahiel, M. A. (2008). Role of DptE and DptF in the lipidation reaction of daptomycin. FEBS Journal, 275(21), 5343-5354.

Wuthrich, K. (1986). NMR of proteins and nucleic acids. New York, NY: Wiley Interscience.

You, Y. O., & van der Donk, W. A. (2007). Mechanistic investigations of the dehydration reaction of lacticin 481 synthetase using site-directed mutagenesis. Biochemistry, 46(20), 5991-6000.

228

Yousef, A., & Carlstrom, C. (2003). Food microbiology: A laboratory manual (pp. 234-236). Hoboken, New Jersey: John Wiley and Sons, Inc.

Zhang, X. (2008). Identification, properties, and application of enterocins produced by enterococcal isolates from foods. The Ohio State University).

Zheng, L., & Sun, D. W. (2004). Vacuum cooling for the food industry—a review of recent research advances. Trends in Food Science & Technology, 15(12), 555-568.

229