Danielbarrosortegatese2019.Pdf

Figure 1 Average biomass of Paenibacillus elgii strain AC13 cultures at 37°C expressing swarming or non-swarming behavior.Pró- CulturesReitoria were scraped Acadêmica from agar and weighted (n=147)...... 58

Escola de Saúde e Medicina

Programa de Pós-Graduação Stricto Sensu em Ciências Genômicas e

Biotecnologia

Production of Polypeptin A analogues by Paenibacillus elgii strain

AC13 and quorum sensing dependent bacterial processes

Autor: Daniel Barros Ortega

Orientadora: Cristine Chaves Barreto

Brasília - DF 1

2019 Daniel Barros Ortega

Production of Polypeptin A analogues by Paenibacillus elgii strain AC13 and quorum sensing dependent bacterial processes

Tese apresentada ao Programa de Pós- graduação Stricto Sensu em Ciências Genômicas e Biotecnologia da Universidade Católica de Brasília como requisito parcial para obtenção do Título de Doutor em Ciências Genômicas e Biotecnologia

Orientadora: Prof.ª Dr.ª Cristine Chaves Barreto

Coordenadora: Prof.ª Dr.ª Maria Sueli Soares Felipe

Brasília 2019 2

Para o Léo, João, Raissa, Nez e Neusa. Amo vocês.

AGRADECIMENTOS

À Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, CAPES, pelo financiamento prestado; À Universidade Católica de Brasília e aos professores, por acreditarem em novas Ideias e no desenvolvimento da Pesquisa; À professora Dra. Beatriz Simas Magalhães, por ter me apresentado ao Projeto; À professora Dra. Cristine Chaves Barreto, por, apesar de tantas divergências, conseguiu aparar arestas e me ajudou a produzir este documento; À coordenação do programa de pós-graduação da UCB, pela compreensão e apoio; À Universidade da Califórnia San Diego e aos membros do Laboratório do Professor Dorrestein, pelas palestras, disponibilização do espaço e equipamentos; Aos amigos desde a infância até agora e colegas de laboratório Tamires, Clarissa, Rosiane, Flávio, Michel, Marise, Thiago entre tantos que compartilharam momentos de planejamento, organização, frustrações, conquistas e também de bem estar e insights, dos mais coerentes aos mais absurdos! Fundamentais. À minha mãe, Neusa; irmã, Raissa; e à Nez, pelo Todo. Aos meus sobrinhos, por serem a luz que ilumina tanto espaço e tempo deste Todo.

“Awareness is the world's mystical experience.” Thus Spoke Zarathustra; F. Nietzsche

Resumo

Quorum sensing (QS) são sofisticadas redes de comunicação baseadas em trocas metabólicas que influenciam processos bacterianos como a bioluminescência, reprodução, esporulação, formação de biofilme, expressão de virulência e produção de antimicrobianos. Peptídeos não ribossomais (NRP) utilizam aminoácidos não proteinogênicos como blocos de construção, aumentando o número de possíveis arcabouços quando comparados aos obtidos pelo processo de síntese ribossômica. Naturalmente produzida por Paenibacillus elgii, a família das Pelgipeptinas (PGP) é constituída por cinco NRP cíclicos e catiônicos compostos por nove aminoácidos e uma cadeia de ácido graxo. Após a linearização de NRP semelhantes às PGP, concentrações mais baixas foram suficientes para inibição do crescimento de Candida albicans patogênica, ampliando a polivalência deste grupo de moléculas. A cepa de P. elgii AC13 foi isolada de amostras de solo do Cerrado brasileiro. PGP produzidas por AC13 tiveram atividade contra Escherichia coli, Staphylococcus aureus e Klebsiella pneumoniae multirresistente. Isoformas lineares de PGP também foram encontradas em culturas de AC13 em caldo nutriente. Parâmetros de crescimento, quantificação de PGP e caracterização parcial de lipopeptídeos produzidos sugerem que a cepa AC13 é capaz de modular a produção PGP em função de variações de seu ambiente ou de sua população inicial. Associações por Redes Moleculares indicam que o íon de massa 1.101,7052 Da (PGP- B) relaciona-se ao gene leu B, responsável por codificar 3-isopropilmalato desidrogenase em P. dendritiformis C454, P. selenitireducens, P. yonginensis e P. amylolyticus. A enzima é parte da via biossintética da leucina e é responsável por catalisar a oxidação do 3- isopropilmalato. Analises semelhantes também sugerem que a enzima responsável por sintetizar O-acetil-L-homoserina em Lysinibacillus sp. cepa FJAT-14222 está relacionada ao íon 416.7368 Da, ambos encontrados em cultura de AC13 por LC-MS. Entre as vias de sinalização de QS, N-acil-homoserina lactonas são moléculas de sinalização para mais de 70 espécies diferentes de bactérias Gram-negativas. Como forma de se aumentar as estratégias no combate aos microrganismos multirresistentes, sugere-se a interferência nos sistemas de comunicação QS, que podem ser interrompidos pela redução da atividade ou bloqueio da síntese de proteínas receptoras de sinalização ou ainda por meio de moléculas sintéticas análogas aos compostos sinalizadores de QS. Contudo, uma melhor compreensão da interface entre a biologia de sistemas e técnicas avançadas de redes moleculares se faz necessária.

Palavras-chave: Paenibacillus; Pelgipeptinas; microbiologia de solos; quorum-sensing

Abstract

Quorum sensing (QS) are sophisticated communication networks based on metabolic exchanges that influence bacteria processes such as reproduction, sporulation, biofilm formation, virulence and antimicrobials expression. Non-ribosomal peptides (NRP) use non- proteinogenic amino acids as building blocks, which increase the number of possible templates when compared to standard ribosomal synthesis process. Naturally produced by Paenibacillus elgii, the NRP family of Pelgipeptins (PGP) consists of five cyclic and cationic lipopeptides comprising nine amino acids and a fatty acid chain. Linearization of NRP similar to PGPs have improved antifungal activity against pathogenic Candida albicans. Paenibacillus elgii strain AC13 was isolated from soil samples of the Brazilian Cerrado. PGPs produced by AC13 were active against Escherichia coli, Staphylococcus aureus and multidrug resistant Klebsiella pneumoniae. Strain AC13 also synthesizes linear counterparts or PGPs in nutrient broth. Growth parameters, PGPs quantification, and partial characterization of produced lipopeptides suggest that strain AC13 is able to modulate its lipopeptides production based on variations of its environment or initial population. Molecular networking analysis indicates that the precursor mass of 1,101.7052 Da (PGP-B) is connected to the gene leuB pathway, responsible to code 3-isopropylmalate dehydrogenase described for P. dendritiformis C454, P. selenitireducens, P. yonginensis and P. amylolyticus. The enzyme is part of leucine biosynthetic pathway and is responsible to catalyze the oxidation of 3-isopropylmalate. Molecular networking analysis also indicates that the enzyme responsible to synthesize O-acetyl-L-homoserine (gene metAA) expressed by Lysinibacillus sp. strain FJAT-14222 is connected to precursor mass 416.7368 Da. Among QS signaling pathways, N-acyl-homoserine lactones occur as signaling molecules for more than 70 different species of diderm bacteria. QS communication can be interrupted by reducing the activity of signaling receptor proteins, their synthesis or by synthetic molecules analogous to QS signaling compounds, increasing the strategies on fighting against multidrug resistant microorganisms. Notwithstanding, a better understanding of the overlap between system biology and advanced molecular network techniques is needed.

Key-words: Paenibacillus; Pelgipeptin; soil microbiology; quorum-sensing.

List of Abbreviations

QS – Quorum Sensing; NRP – Non ribosomal peptides; PGP – Pelgipeptin; AHL – Acyl-homoserine lactones; AIP – Auto-inducing peptides; RNA – Ribonucleic acid; DNA – Deoxyribonucleic acid; EPS – Extracellular polymeric substances; GMP – Guanosine 3’, 5’-monophosphate; NRPS – Nonribosomal Peptide Synthases; LC – Liquid Chromatography; MS – Mass Spectrometry; rRNA – ribosomal RNA; SDS – Sodium dodecyl sulfate; PGAP – Prokaryotic Genome Annotation Process Pipeline; NCBI – National Center for Biotechnology Information; PGAAP – Prokaryotic Genome Automatic Annotation Pipeline; ORF – Open reading frame; HMM – Hidden Markov model; CDS – Coding sequences; tRNA – transporter RNA; ANI – Average Nucleotide Identity; MALDI-TOF – Matrix assisted laser desorption ionization time of flight; RP-HPLC – Reversed-Phase High Performance Liquid Chromatography; UPLC – Ultra Performance Liquid Chromatography; SG – Schaeffer’s Glucose; ACN – Acetonitrile; MMP – Mineral Medium for Paenibacillus; NB – Nutrient Broth; MHB – Mueller-Hinton Broth; PMHB – Peptone enriched Mueller-Hinton broth;

OD – Optical density; SASPs – Small acid-soluble proteins; NPAAs – Non-proteinogenic amino acids; Dab – 2, 4-diaminobutyric acid; Val – Valine Ile – Isoleucine. Phe – Phenylalanine; Leu – Leucine; Ser – Serine; AFB – American Foulbrood disease; GNPS – Global Natural Products Social Molecular Networking; LB – Luria-Bertani; CFU – Colony forming units; RT – Retention time; MDR – Multidrug resistant; TFA – Trifluoroacetic acid; MIC – Minimal Inhibitory concentration; FIC – Fractional Inhibitory Concentration; FICI – Fractional Inhibitory Concentration Index; UV – Ultraviolet; MTT – 3-[4,5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide; TSA – Trypticase soy agar; ATCC – American Type Culture Collection;

List of Figures Introduction………………………………………………………………………………...... 19

Figure 1: LuxIR bioluminescence expression system for Vibrio fischeri represented as quorum sensing mechanism model for diderm bacteria. Protein luxI is responsible to synthesizes auto-inducers (AHL). At high cell density, a critical threshold of auto-inducers triggers a chemical change in protein luxR. LuxR bounds to AHL and activates the transcription of the operon luxICDABE……………………………………………………...21

Figure 2: Agr virulence expression system for Staphylococcus aureus represented as quorum sensing mechanism model for monoderm bacteria. The two-stage signaling system starts when protein agrD synthesizes auto-inducer peptides (AIP) precursors. Precursors AIP are activated by lactone ring modification provided by agrB enzyme. Activated, AIP bind to agrC, which phosphorylates agrA. AgrA promotes the transcription of operon RNAIII, which reduces the synthesis of adhesion factors and increases the synthesis of toxins; and it also activates the operon agrBDCA, that will keep the whole communication system running………………………………………………………………………………………..22

Figure 3: Project’s hypothesis representation. P. elgii AC13 shows a complex life cycle; sporulation, germination, biofilm formation, and motility expression are represented here as Primary Metabolism (normal growth, development, and reproduction). Quorum-sensing signal molecules degradation is utilized on Second Metabolism to activate Non-ribosomal Peptide Synthetases genes and also utilized for building non ribosomal lipopeptides (NRPs)………………………………………………………………………………………...29

Chapter II……………………………………………………………………………………..…………36

Figure 1. MALDI-ToF spectra of strain AC13 after 24 hours of incubation at 37 °C in 2×SG broth. Peptides were extracted from pellet with 10 µL of formic acid 70 % and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method (A). Phase- contrast Microscopy of strain AC13 after 24 hours of incubation in 2×SG broth at 37 °C. Elongated rods are vegetative cells. Amplification of 100× (B)……………………………...41

Figure 2. MALDI-ToF spectra of strain AC13 after 72 hours of incubation at 37 °C in 2×SG broth. Peptides were extracted with from pellet with 10 µL of formic acid 70% and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method (A). Phase- Contrast Microscopy of strain AC13 after 72 hours of incubation in 2×SG broth at 37 °C. Almost all the population had already become spores. Remaining vegetative cells in the center. Refracting ellipses are the spores. Amplification of 100 × (B). ……………………...41

Figure 3. MALDI-ToF spectra of strain AC13 spore stock solution. Peptides were extracted from pellet with 10 µL of formic acid 70 % and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method. At least six groups of molecules are ionized properly after spores’ purification (A). Phase-Contrast Microscopy of strain AC13 spore stock solution. No vegetative cells were found. Elongated circular bright structures are spores. Amplification of 100× (B)……………………………………………………………………42

Figure 4. MALDI-ToF spectra of P. elgii strain AC13. Monoisotopic masses (1,660-1,720 m/z) shows similar ionization pattern for cultures composed in their majority by spores (B) and (C) in comparison to a vegetative cell majority population in (A). Peptides were extracted from pellet with 10 µL of formic acid 70 % and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method……………………………………………………..42

Figure 5. MALDI-ToF spectra of Paenibacillus elgii strain AC13. Protein profile comparison of a population compressed of free vegetative cells (A); free spores and a few vegetative cells (B) and purified spores (C). Protein content was extracted from pellet using 10 µL of formic acid 70% and 10 µL of acetonitrile (ACN) 100 %. Spectra were acquired using MBT (2,000- 20,000 m/z) method…………………………………………………………………………..43

Figure 6. Standard growth curve of strain AC13 during exponential phase. Average dry weight (g·L-1) in axis y and average optical density in axis x. Growth was monitored for 5 hours in nutrient broth at 37 °C……………………………………………………………….44

Figure 7. Growth Curve of P. elgii strain AC13 over 24 hours in nutrient broth at 37 °C. 3 -1 Starter population was settled as ~10 spores·mL . Log phase from t0 to t9 hours. Exponential

-1 growth from t9 to t20 hours. Stationary phase from t20 to t24 hours. Biomass (g·L ) in axis y and time (hours) in axis x…………………………………………………….……………….45

Figure 8. Growth Curve of strain AC13 over 20 hours in nutrient broth (NB); Mueller-Hinton Broth (MHB) or Mueller-Hinton Peptone Broth (MHPB) at 37 °C. Starter population was settled as ~103 spores·mL-1. Biomass (g·L-1) in axis y and time (hours) in axis x……………45

Figure 9. Comparison of average biomass of P. elgii strain AC13 at stationary phase from two different starter populations. No meaningful differences were found (n=18)……………………………………………………………………………………...….46

Figure 10. Average biomass during the stationary phase of strain. At both conditions ended with an average biomass of ~1 g·L-1 (n=18)………………………………………………….47

Figure 11. Comparison of strain AC13 PGP production at stationary phase from two different inoculum methods (n=18). …………………………………………………………………...47

Figure 12. Comparison of average Pelgipeptin production at stationary phase from two different starter population of P. elgii strain AC13. Approximately 100 µg·mL-1 (n=18) produced by vegetative cell starter population and ~160 µg·mL-1 produced by a spore starter population. …………………………………………………………………………….48

Figure 13. Biomass (g·L-1) and PGPs production (µg·mL-1) throughout stationary phase. Strain AC13 was grown from two different starter populations. Biomass reached similar values whereas PGPs production from spores was higher. …………………………………..48

Chapter III.……………………………………………………………………………………………...53

Figure 1. Average biomass of P. elgii strain AC13 cultures at 37°C expressing motility (swarming) or forming biofilm (non-swarming). Cultures were scraped from agar and weighted (n=147)…………………………………………………………………………...... 58

Figure 2. Cultures’ phenotype of Paenibacillus elgii strain AC13 expressing motility (swarming) in nutrient agar (superior) and only forming biofilms in Luria-Bertani 1.5 % agar (inferior, right) after 24 hours of incubation at 37 °C………………………………………59

Figure 3. Average extracted natural products from P. elgii strain AC13 cultures at 37°C expressing motility (swarming) or forming biofilm (non-swarming). Cultures were scraped from agar and weighted (n=147). Overall extracted bioproducts ranged between 0.4 and 10 milligrams.……………………………………………………………………………………60

Figure 4. P. elgii AC13 proteomic collection obtained from LC-MS/MS (A). Pelgipeptin cluster and Molecular networking analysis of ion 1,101.7052 Da is connected to an enzyme from leucine biosynthetic pathway (B). Ion 416.7368 Da connects to a sub pathway of Amino-acid biosynthesis: homoserine O-acetyltransferase transfers an acetyl group from acetyl-CoA to L-homoserine, forming acetyl-L-homoserine (C)…………………………….62

Supplementary Figure 1. Leucine biosynthetic pathway is responsible to catalyze the oxidation of 3-carboxy-2-hydroxy-4-methylpentanoate (3-isopropylmalate) to 3-carboxy-4- methyl-2-oxopentanoate. The product formed decarboxylates to 4-methyl-2 oxopentanoate………………………………………………………………………………...64

Conclusions……………………………………………………………………………………………..76

Figure 4. Thesis’ interpretation. Based on results found, one can say that Pelgipeptin/Permeatin A analogues may be needed to triggers a social response, which will produce a myriad of compounds and signaling molecules that can be recycled to the first metabolism, like the synthesis of main enzymes involved in amino acid synthesis (e.g. 3- isopropylmalate dehydrogenase). Those enzymes are needed to catalyze reactions needed for the agr B analogues production. An acetylation step might be responsible to add a lactone ring modification to the agr B protein analogues, which can then be involved in the quorum sensing signaling process for the surfactants production involved in social response like swarming behavior……………………………………………………………………………78

Linear counterparts of the Non-Ribosomal peptides Pelgipeptin B and C are naturally produced by Paenibacillus elgii strain AC13 in nutrient broth (Appendix2)……………………………………………………………………………………………101

Figure 1. RP-HPLC chromatogram of pelgipeptins’ isoforms showing the supernatant of P. elgii AC13 cultured in Nutrient broth, after organic extraction with butanol. Analyses were conducted using a C18 Shim-pack VP-ODS column (150×4.6mm, 4.6µm; Shimadzu). Flow rate of 1 ml·min-1, after optimized method. The elution was monitored at 216 nm for 50 minutes. Solvent A: water + 0.1% TFA. Solvent B: acetonitrile + 0.1% TFA. [M + H]+ 1105 (I); [M + H]+ 1119 (II); [M + H]+ 1073 (III); [M + H]+ 1087 (IV); [M + H]+ 1087 (V); and [M + H]+1101 (VI)...... 118

Figure 2. Quantification of pelgipeptin isoforms produced by P. elgii strain AC13 in nutrient broth was carried out using a standard curve of known concentrations of pelgipeptin. After 40 hours of incubation at 37°C the concentration of PGP-B’ was more than two times higher than the concentration of PGP A-D altogether…………………………………………………...118

Figure 3. MALDI-ToF MS/MS spectra from the precursor ion [M+H]+ 1105 (Pelgipeptin C’). Sequences found for -b and -y series…………………………………………………...... 119

Figure 4. MALDI-ToF MS/MS spectra from the precursor ion [M+H]+ 1119 (Pelgipeptin B’). Sequences found for series -b and -y…………………………………...... 119

Figure 5. MALDI-ToF/MS spectra of Pelgipeptin B before alkaline hydrolysis (A) and after alkaline hydrolysis (B). [M+H]+ 1101.8 m/z; [M+Na]+ 1123.7 m/z; [M+K]+ 1139.7 m/z. Molecular weight differences between cyclic and linear isoforms are due to lactone ring hydrolysis……………………………….…………………………………...... 120

Figure 6. MALDI-ToF/MS spectra of Pelgipeptin C before alkaline hydrolysis (A) and after alkaline hydrolysis (B). [M+H]+ 1087.9 m/z; [M+Na]+ 1109.9 m/z; [M+K]+ 1125.9……...120

Figure 7. Mapping of P. elgii AC13 contigs coding for pelgipeptin in the P. elgii B69 pelgipeptin biosynthetic gene cluster (JQ74521). Accession numbers of the sequences in GenBank are showed on the left…………………………………………………………….121

Figure 8. Domain organization in pelgipeptin synthase gene plpE of P. elgii AC13 compared to the complete gene in P. elgii B69 presented at Ding et al. (2011)………………..……...121

Figure 9. Cell viability assay performed after 24 hours incubation with different concentrations of Pelgipeptins and polymyxin b. Each bar represents the mean ± SD of cellular absorbance (abs), n=3. *p<0.05, **p<0.01, ****p<0.0001. Insufficient amount of Pelgipeptin C’ were available to perform the cell viability assay…………………………...122

Figure S1. Chromatogram of Pelgipeptin Mix A-D stock solution before (A) and after (B) treatment with butanol. Analysis were conducted using a C18 Shim-pack VP-ODS column (4.6µm, 150×4.6mm; Shimadzu). Flow rate of 0,6 ml·min-1, linear gradient 5-95% B. Solvent A: water + 0.1% TFA. Solvent B: acetonitrile + 0.1% TFA. The elution was monitored at 216nm. There are no changes in retention time or chromatography profile………………...125

Figure S2. MALDI-ToF/MS spectra of P. elgii cell free supernatant before (A) and after butanol extraction (B). Monoisotopic masses correspond to PGP B [M+H]+ 1101.8 m/z; [M+Na]+ 1123.7 m/z; [M+K]+ 1139.7; PGP C [M+H]+ 1087.9 m/z; [M+Na]+ 1109.9 m/z; [M+K]+ 1125.9; PGP B’ [M+H]+ 1119.7 m/z; [M+Na]+ 1141.7 m/z; [M+K]+ 1157.7; PGP C’ [M+H]+ 1105.7 m/z; [M+Na]+ 1127.7 m/z; [M+K]+ 1143.7; The linear isoforms are present in supernatant before extraction………………………………………………………………..126

List of Tables

Introduction………………………………………………………………………………...... 19

Table 1. Natural Lipopeptides Produced By Paenibacillus sp. strain OSY-N; P. elgii strains B69, BC34-6 and AC13: Identity; Primary Sequence; fatty-acyl side chain (R); Molecular Weight [M+H]+; Characteristics and Reference Literature…………………………………..26

Chapter II………………………………………………………………………………………..………36

Table 1. Average optical density (OD) at 600 nm for cultures of strain AC13 over exponential growth phase. Average Dry weight (g·L-1) of each given point was also measured for at least three biological replicates…………………………………………………………………….44

Table 2. Average Optical Density (600 nm) for cultures of P. elgii strain AC13 over 24 hours in nutrient broth at 37°C. Biomass (g·L-1) was calculated based on the correlation factor found (0,5075x). Measurements were done at each 2 or 2 and half hours. Starter population was settled as ~103 spores·mL-1…………………………………………………………………...44

Table 3. Strain AC13 was grown over 20 hours in nutrient broth, (NB); Mueller-Hinton Broth (MHB) and MHPB at 37°C. Starter population was settled as ~103 spores·mL-1. Biomass values from Log phase (t4 to t11 hours) were used to calculate Specific Growth Rate (µ); Generation Time (g) and Division Rate (v) from each culture tested………………………...46

Chapter III……………………………………………………………………………………………….53

Table 1. Samples’ identification. Sample ID, Cultures’ phenotype; Biomass (mg), and Standard Deviation of samples used in LC-MS/MS analysis………………………………...59

Table 2. Sample ID and Culture Phenotype expressed; Time (hours); Temperature (°C); Extraction (mg) and Final Extraction Concentration used (mg·mL-1) used in LC-MS/MS acquisition…………………………………………………………………………………….60

Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae (Appendix 1)……………………………………………………..90

Table 1. Minimum inhibitory concentration (MIC): Pelgipeptins A-D and reference antibiotics against tested strains (µg·mL-1)…………………………………………………..99

Table 2. Fractional Inhibitory Concentration (FIC) Index: Pelgipeptin A-C and other antibiotics against Klebsiella pneumoniae ATCC13883 and the MDR strain LACEN 3259271……………………………………………………………………………………….99 Supplementary Table 1. Checkerboard testing method: Pelgipeptins A-C and reference antibiotics against Klebsiella pneumoniae ATCC 13883 (µg·mL-1)………………………..100

Supplementary Table 2. Checkerboard testing method: Pelgipeptins A-C and reference antibiotics against MDR Klebsiella pneumoniae LACEN 3259271 (µg·mL-1)…………….100

Linear counterparts of the NRPs Pelgipeptins B and C naturally produced by Paenibacillus elgii strain AC13 in nutrient broth (Appendix 2)…………………………………………………101

Table 2. Fractions found in P. elgii strain AC13 nutrient broth cultures. Retention Time (RT); Molecular Weight [M+H]+; Lipid Tail (R); Amino acid sequence (MS/MS); Characteristic, and Identity………………………………………………………………………………….123

Table 3. Minimum inhibitory concentration values (µM) of Pelgipeptins A-D; B’, C’, and a Mix of Pelgipeptins A-D; and reference antibiotics against tested strains. Molecular Weight of molecules: [M+H]+…………………………………………………………………….……124

Table 4. Alignment of the nucleotide and protein sequences of pelgipeptin gene cluster of P. elgii AC13 in comparison to the one from P. elgii B69…………………………………….124

Sumário List of Abbreviations ...... 8

List of Figures ...... 10

List of Tables ...... 15

Introduction ...... 19

Cell density-dependent mechanisms ...... 20

Microbial Biofilm ...... 23

Motility expression ...... 24

Paenibacillus ...... 25

Non-ribosomal lipopeptides produced by Paenibacillus elgii ...... 25

Chapter I ...... 30

Draft Genome Sequence of the Antimicrobial-Producing Strain Paenibacillus elgii AC13. .. 30

Chapter II ...... 36

Pelgipeptin production from different Paenibacillus elgii strain AC13 inoculation methods . 36

Chapter III ...... 53

Molecular Networking analysis based on a proteomic collection from Paenibacillus elgii strain AC13 – DRAFT ...... 53

Discussion ...... 73

Conclusions ...... 76

Perspectives ...... 79

References ...... 81

Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae (Appendix 1) ...... 90

Linear counterparts of the NRPs Pelgipeptins B and C naturally produced by Paenibacillus elgii strain AC13 in nutrient broth - DRAFT (Appendix 2) ...... 101

Introduction

Cell density-dependent mechanisms For years, the scientific community considered microorganisms solely as self- sufficient individual beings, with uncomplicated self-replication and unable to communicate or organize in groups (Greenberg, 2003a; b; Njoroge and Sperandio, 2009). During an infectious process, "bacterial mass" was considered only as the sum of these individuals; and the idea that bacteria could be organized as a group and individuals within this group could orchestrate cellular changes in response to the population density was considered almost absurd (Greenberg, 2003a; b). However, it is currently known that bacteria and other prokaryotes have sophisticated communication networks based on metabolic changes that influence processes of basal behavior such as reproduction, sporulation, biofilm formation, virulence and antimicrobials expression among other toxic molecules synthesis (O'brien and Wright, 2011; Lee et al., 2013; Bodelon et al., 2016; Papenfort and Bassler, 2016). The communication system among bacteria is achieved through different processes involving production, release, detection, and responses to chemical compounds called self- inducers, small hormone-like molecules that allow the microorganism access the cellular density of a local population and coordinate the gene regulation in a population level (Njoroge and Sperandio, 2009; Lee et al., 2013). The information provided by self-inducers is critical in order to achieve a coordinated activity response involving a large group of cells. The chemical communication in which these molecules are involved is also present in eukaryotic cells, especially at the embryonic stage. Cell-to-cell communication systems can orchestrate and modify the expression of genes in response to changes in cell density or in response to the surrounding microbial community structure (Waters and Bassler, 2005; Papenfort and Bassler, 2016). The mechanism of chemical communication based on cellular density is called quorum sensing (QS) and it was initially observed in Vibrio fischeri, a bioluminescent marine bacterium that lives in colonies within specific organs of squid, its natural host. Vibrio fischeri and its squid host developed a symbiotic relationship. The expression of bioluminescence by these microorganisms provides prey and protection for the squid, since they are confused by the bioluminescence and the reflection of the moon over water line. Luciferase is the enzyme involved in bioluminescence for the Vibrio-squid symbiosis. The synthesis of luciferase enzyme by V. fischeri is inhibited when the bacterium is inoculated into sterile culture medium, but as cell density increases, bioluminescence is induced again. The culture in minimal medium revealed an activator released by the

20 bacterium, leading to the induction of luciferase; that later leads to the expression of bioluminescence (Figure 1) (Eberhard, 1972; Njoroge and Sperandio, 2009).

Figure 1: Lux IR bioluminescence expression system for Vibrio fischeri represented as quorum sensing mechanism model for diderm bacteria. Protein luxI is responsible to synthesizes auto-inducers (AHL). At high cell density, a critical threshold of auto-inducers triggers a chemical change in protein luxR. LuxR bounds to AHL and activates the transcription of the operon luxICDABE.

Quorum sensing communication is widely diffused among microorganisms and other biological processes are regulated by this mechanism such as biofilm formation, sporulation process and secretion of exoenzymes (Smith and Iglewski, 2003; Waters and Bassler, 2005; Bassler and Losick, 2006; Hawver et al., 2016; Papenfort and Bassler, 2016). QS-mediated cellular communication within pathogenic bacterial species also occurs. This communication is also responsible for coordinating a cooperative behavior that increases population survival probability under adverse conditions, like changes in host metabolism or facing the host's immune system (Gray et al., 2013). Among the QS signaling pathways, those regulated by N-acyl-homoserine lactones (AHLs) are well studied and produced by more than 70 different species of diderm bacteria, therefore, AHLs are considered as the typical communication system of this group (Kaufmann et al., 2006). There are four main steps found in all QS systems in diderm bacteria: self- inducers (AHLs or other similar molecules) are synthesized from S-adenosylmethionine, both capable of diffusing freely through bacterial membranes (a); these auto-inducers bind to specific receptors located either inside the membrane or in the cytoplasm (b); the self-inducers in these systems are responsible for changing from tens to hundreds of genes that support various biological processes (c); finally, self-inducers stimulate an increase of their own

21 synthesis, which promotes synchronized genetic expression within the population (d) (Hawver et al., 2016; Papenfort and Bassler, 2016). In contrast to QS pathways regulated by AHLs in diderm bacteria, monoderm bacteria are usually regulated by auto-inducing oligopeptides (AIPs) (Cakar, 2004; Papenfort and Bassler, 2016). These small extracellular peptides operate their regulation through a two-stage signaling system (Gray et al., 2013; Kalia, 2013). Initially, the signal molecule is synthesized by an AIP synthetase as a longer peptide precursor, which is exported and released in the extracellular environment, where carrier enzymes perform post-translational modifications. Then, these peptides are detected and modified by specific proteins. The detection of the auto- inducer triggers phosphate releases that control the QS response (Figure 2) (Hawver et al., 2016).

However, despite more than three decades of results related to the mechanics of these signaling pathways, it is only in recent years that quorum sensing has been used as a strategy to modulate a behavioral response that is dependent on bacterial cell concentration.

Microbial Biofilm Many infectious diseases are caused by bacteria that proliferate in the midst of biofilms. Biofilm formation and the QS communication systems are closely related since its formation is also regulated by QS and is one of the most common structures achieved in a cooperative way by a microorganism (Kalia, 2013; Lee et al., 2013). In most biofilms, the amount of microorganisms corresponds to only 10% of the total dry weight, while 90% corresponds to components of the extracellular matrix, consisting of numerous polymers of biological origin, extracellular DNA and polysaccharides (Costerton, 1999). Genetic studies in intraspecific biofilms indicate that the profile of genes transcribed from a population in a biofilm is different from the transcriptome of a population of the same planktonic form microorganism (Watnick and Kolter, 2000). Within a biofilm, the extracellular polymeric substances (EPS) are responsible for aggregate cells allowing a closer and more intense interaction between members of the population; which favors cell communication. Due to enzyme retention, an external and versatile digestive system is formed. In addition, EPS acts as a recycling center because it keeps all components of lysed cells available (Flemming and Wingender, 2010). Experiments on wild strains and mutants of Escherichia coli, Pseudomonas aeruginosa, and Vibrio cholerae indicate that the adhesion to surfaces performed by these organisms is accelerated by the expression of specific cell structures, such as flagella and pili type IV (O'toole and Kolter, 1998; Watnick and Kolter, 1999; Watnick and Kolter, 2000). However, once the microorganism is consolidated as a member of the biofilm, it must differentiate and repress the synthesis of flagella, since these structures are involved in motility and could destabilize the synthesis of exopolysaccharides, leading to the breakdown of the biofilm support (Watnick and Kolter, 2000). All those steps require a complex and refined communication system in order to achieve a population success. Biofilm production starts when bacterial cells sense a surface and changes its signal transduction mechanism, which directly involves the rotation of its flagellum. However, although there are many references on how microorganisms regulate and spin their flagella to move in liquid media, little is known about how this process occurs when bacteria are on solid surfaces and what are the mechanisms that lead to the conversion between "stick" to "swarm" behavior (Kearns, 2010; Belas, 2014).

Motility expression Many organisms coordinate their movement on a population scale, emerging as a self- organized swarming entity. This coordinated behavior can be observed in fish, birds, insects, and also in several microorganisms (Anyan et al., 2014). Swarming is a bacterial flagella- dependent behavior that allows cells to move on a surface in a coordinated way, favoring a quickly nutrients capture by a microbial population. It is also a well-established way of expansion and colonization of a population (Kearns, 2010; Belas, 2014). This collective movement behavior is different from swimming which is considered a single cell behavior. In addition, swarming is dependent on the differential synthesis of structures and biomolecules that lead to morphological changes within the cells. Swarm cells derive from swim vegetative cell elongation - a process caused by the inhibition of cellular septation, increase on the number of flagella, and stimulation of surfactant molecules synthesis (Ingham and Ben Jacob, 2008; Belas, 2014). This whole process will culminate into a synchronized and population- scaled expression of swarming. Briefly, swarming can be divided into four distinct phases:  Induction of cell differentiation;  Acclimatization (lag);  Expression of motility;  Consolidation, phase in which the migration ceases and the swarm cells return to the mobile vegetative cell state.

Among these stages, the acclimatization or adaptation phase attracts great scientific interest as it suggests the occurrence of intense intercellular communication. These processes involve the secretion of signaling compounds that, after reaching a threshold concentration, triggers the collective movement behavior (Belas et al., 1998; Belas, 2014). Despite all the differences between biofilm and swarming, those two population- scaled processes are closely related, since locomotion is often considered as a transition step prior to the formation of communities in stationary biofilms. It has been demonstrated, for Pseudomonas aeruginosa, that biofilm formation and swarming are inversely regulated by the intracellular concentration of cyclic cis-guanosine 3', 5’-monophosphate (c-di-GMP). Low concentrations of c-di-GMP promote swarming while high concentrations of the same compound signal leads to the production of extracellular polymer matrix (EPS), the structural foundations of a sessile biofilm (Anyan et al., 2014).

Paenibacillus Paenibacillus is a rod-shaped bacteria that is able to express quorum sensing communication resulting in cellular differentiation; surfactant and antimicrobial peptides production; motility behavior; and spore formation (Grady et al., 2016; Costa et al., 2019). This is a ubiquitous genus, although it can be easily found in soil, it has been described in the Antarctic plateau near the Concordia research station; found in rotting timber from Argentine forests; waste water from Taiwan and in the United States; and even in crude oil samples from a Brazilian oil field (Ko et al., 2007; Ghio et al., 2012; Van Houdt et al., 2013; Mathews et al., 2014; Gudina et al., 2015). Paenibacillus is commonly described as aerobic or facultative anaerobic; and sometimes expressing peritrichous flagella (Grady et al., 2016). Colonies are non-pigmented when grown in nutrient agar and show at least 89.6% sequence similarity on 16S ribosomal subunit gene (Grady et al., 2016). Recent data indicate that Paenibacillus species express enzymes capable of cellulose degradation, the main compound discarded in bioethanol production; and other enzymes of industrial importance such as xylanases and pectinases (Ko et al., 2007; Ghio et al., 2012). Furthermore, Paenibacillus species are able to produce antimicrobial compounds with broad spectrum activity such as bacteriocins (He et al., 2006; Lohans et al., 2012; Baindara et al., 2015) and non-ribosomal peptides (NRPs) (Selim et al., 2005; Canova et al., 2010; Ding, Wu, et al., 2011; Qian et al., 2011; Wen et al., 2011; Guo et al., 2012; Qian, C. D. et al., 2012; Niu et al., 2013).

Non-ribosomal lipopeptides produced by Paenibacillus elgii Translation of template mRNA is not the only mechanism for cells to build proteins or peptides (Ng et al., 2009). Nonribosomal Peptide Synthases (NRPS) are complexes of enzymes organized in amino acid-specific modules. Basically, a NRPS module exhibits three core domains: A-T-C, each one responsible for a biochemical step regarding peptide backbone formation and other structural modifications. Domain A is responsible to select a specific amino acid monomer; to perform a covalent bond between substrate and intermediate acyl chains during chain elongations, a thiol group is provided at T (thiolation) domain; finally, peptide-bond forming chain elongation occurs at C domain (Fischbach and Walsh, 2006). Besides, NRPS uses non-proteinogenic amino acids, not only the 20 possible building blocks in standard ribosomal peptides, which increases the number of possible templates to several hundred (Fischbach and Walsh, 2006; Ng et al., 2009). Non ribosomal peptides are therefore, natural products exhibiting surfactant properties; and antiviral, antitumor, or antibiotic activity against human and plant pathogens (Qian et al., 2011; Qian, C. et al., 25

2012). Hence, attracting the interests of biotechnological, pharmaceutical and agricultural industries (Ng et al., 2009; Cochrane and Vederas, 2016; Grady et al., 2016; Biniarz et al., 2017; Deng et al., 2017) Naturally produced by Paenibacillus elgii, the NRP group of Pelgipeptins (PGP) are five cyclic and cationic lipopeptides composed by nine amino acids and a fatty acid chain synthetized by a NRPS coded by the plp gene cluster (40.8 kb) (Wu et al., 2010; Ding, Wu, et al., 2011; Qian, C. et al., 2012; Kim et al., 2018). Pelgipeptins A (PGP-A), B (PGP-B), C (PGP-C), D (PGP-D) and E (PGP-E) molecular weights were determined to be 1072; 1100; 1086; 1086; and 1072, respectively (Ding, Wu, et al., 2011; Kim et al., 2018). Furthermore, all peptides within Pelgipeptin family have a β-hydroxy fatty acid chain with a variable tail of three or four carbon length at its N-terminal portion linked to the non-proteinogenic amino acid 2, 4-diaminobutyric acid (Dab) located at position 1 of the peptide backbone (Wu et al., 2010; Ding, Wu, et al., 2011; Cochrane and Vederas, 2016). Variations among PGPs A-E also include substitution on the second positioned amino acid, which may be either a Valine or Isoleucine (Wu et al., 2010; Qian, C. et al., 2012)(Table 1). Moreover, PGP-A and PGP-D exhibit a lipid tail with a methyl group linked on its penultimate (iso) carbon whereas PGP-B and PGP-C on the antepenultimate (anteiso) carbon (Kaneda, 1967; 1991; Qian, C. et al., 2012; Cochrane and Vederas, 2016). PGP-E was recently discovered and also displays broad- spectrum antimicrobial activity. It is produced by P. elgii strain BC34-6 isolated from red clay soil and has the same stereochemistry of amino acid and ester bond as PGP-A and PGP-C but different fatty acid moiety, exhibiting a straight lipid chain (-CH2CH2CH3) (Kim et al., 2018).

Table 1 Natural Lipopeptides Produced By Paenibacillus sp. strain OSY-N; P. elgii strains B69, BC34- 6 and AC13: Identity; Primary Sequence; fatty-acyl side chain (R); Molecular Weight [M+H]+; Characteristics and Reference Literature.

Primary Sequence Identity R [M+H]+ Characteristic References (H2N-Xn-COOH)

Dab1-Val2-Dab3-Phe4-Leu5- (Cochrane and Vederas, 2016; Kim et al., Pelgipeptin A (CH ) CH 1073 Natural, Cyclic Dab6-Val7-Leu8-Ser9 3 2 2018)

(Takeuchi et al., 1979; Wu et al., 2010); Pelgipeptin Dab1-Ile2-Dab3-Phe4-Leu5- CH CH CH(CH ) 1101 Natural, Cyclic (Cochrane and Vederas, 2016; Kim et al., B/Permetin A Dab6-Val7-Leu8-Ser9 3 2 3 2018)

(Sugawara et al., 1984; Ding, Wu, et al., Pelgipeptin Dab1-Val2-Dab3-Phe4-Leu5- CH CH CH(CH ) 1087 Natural, Cyclic 2011); (Cochrane and Vederas, 2016; Kim C/BMY-28160 Dab6-Val7-Leu8-Ser9 3 2 3 et al., 2018)

Dab1-Ile2-Dab3-Phe4-Leu5- Pelgipeptin D (CH ) CH 1087 Natural, Cyclic (Ding, Wu, et al., 2011) Dab6-Val7-Leu8-Ser9 3 2

Dab1-Val2-Dab3-Phe4-Leu5- Pelgipeptin E (CH ) CH 1073 Natural, Cyclic Kim et al. 2018 Dab6-Val7-Leu8-Ser9 2 2 3

Dab1-Val2-Dab3-Phe4-Leu5- Paenipeptin A (CH ) COOH 1105 Natural, Linear (Huang et al., 2017) Dab6-Val7-Leu8-Ser9 2 6

Dab1-Leu/Ile2-Dab3-Phe4-Leu5- Paenipeptin B (CH ) COOH 1119 Natural, Linear (Huang et al., 2017) Dab6-Val7-Leu8-Ser9 2 6

Dab1-Leu/Ile2-Dab3-Phe4-Leu5- Paenipeptin C (CH ) COOH 1133 Natural, Cyclic (Huang et al., 2017) Dab6-Val7-Leu8-Ser9 2 7

Paenibacillus elgii as a model for biotechnology and microbial physiology Natural products research has been focusing in isolating and testing new compounds. This approach is of great importance in regard to advances in Science. Biology and the associated fields (e.g. pharmacology, ecology, oncology, immunology, and physiology) however, do not work in isolation. Molecules and organisms interact and react in association with other molecules, cells, chemical compounds, and systemic beings. Understanding how a single molecule or a group of molecules communicates in complex systems is per se a challenge itself. Furthermore, communication-based processes can contribute to understand population-scaled behaviors. The relevance of this project can be exemplified by the economic importance of considering Paenibacillus species as biological models due to their ecological; biotechnological; industrial and human health importance (Morrissey et al., 2015; Kim et al., 2018). It is known, for example, that some species of the genus are pathogenic of invertebrates or even opportunistic pathogens in humans (Grady et al., 2016). Even though most interactions between members of Paenibacillus genus and humans are not harmful, colonization of these microorganisms has shown to be pathogenic in immunocompromised individuals, and a recent case of meningoencephalitis in a neonate revealed a P. alvei rare infection (Deleon and Welliver, 2016). In regard of its ecological and economical importance, the Apis honeybees’ lethal American foulbrood disease (AFB), for instance, occurs due to highly contagious and extremely resistant P. larvae endospores. Honeybees’ feeding is usually achieved by glandular secretion and processed honey (Morrissey et al., 2015). Once infected, adults transmit P. larvae spores to colonies larvae that die within 3 to 12 days. P. larvae spores, however, remain dormant and infective for more than 35 years, leaving unusable hives and causing substantial economic damages due to the latent possibility of new outbreaks and ecological unbalance, since honeybees pollination are responsible for approximately 90% of commercial crops.

Moreover, many Paenibacillus species are also known as antimicrobial-producing species, and since there is global concern over the rise of multi resistant bacterial strains; an urgent need to identify new antibiotics gives Paenibacillus genus great health importance (Cochrane and Vederas, 2016). Besides, enzymes with bioremediation potential are also attractive for bioprospecting within this group (Grady et al., 2016). P. elgii strain AC13 was first isolated from Cerrado’s soil samples. Initial colonies showed the presence of an inhibition halo around it after 20 days of incubation in different culture media (Kurokawa, 2006). Antimicrobial compounds were later described as Pelgipeptin, a group of non-ribosomal lipopeptides and their gene cluster were found after the genome annotation (Ortega et al., 2018; Costa et al., 2019). The main question leading this Project was to investigate if there was variation in the pool of non-ribosomal lipopeptides produced by P. elgii strain AC13. More questions emerged prior solid media cultures investigation: is the biomass of P. elgii strain AC13 higher when it forms a biofilm in comparison to the biomass when it expresses swarming motility? Are the natural products, such as pelgipeptin, produced differently when starting from spores and from vegetative cells? In this context, variations over the small group of natural produced non ribosomal lipopeptides expressing antimicrobial activity synthetized by P. elgii strain AC13 was chosen to be tracked throughout different cell-to-cell communication signaling release dependent life cycle stages and morphologically different cells. This thesis is composed by three chapters. The first one is the Draft Genome Sequence of the strain AC13 of P. elgii isolated from Cerrado’s soil samples. The second is an over-time experiment of different starter population in order to observe how it is changing in regard to Pelgipeptin production. Finally, a proteomic collection acquired by LC-MS/MS and molecular networking analysis of bioproducts produced by strain AC13 and a brief discussion about the implications in regard to cell-to-cell communication systems and its interruption. The hypothesis is that the pathway for the synthesis of non-ribosomal antimicrobial lipopeptides (second metabolism) is taking advantage of the catabolism of signaling compounds involved in quorum sensing determined behaviors (primary metabolism) as disposed in Figure 3. Besides, two appendixes are present in which the author was responsible to translate, write and edit the manuscripts. The first is the publication of antimicrobial lipopeptides produced by AC13 strain and a synergistic assay of the lipopeptides with classical antibiotics.

The second is a draft manuscript describing linear isoforms produced by P. elgii strain AC13 in a specific culture condition and partial characterization of the linear molecules produced.

Figure 3 Project’s hypothesis representation. P.elgii AC13 shows a complex life cycle; sporulation, germination, biofilm formation, and motility expression are represented here as Primary Metabolism (normal growth, development, and reproduction). Quorum-sensing signal molecules degradation is utilized on Second Metabolism to activate Non-ribosomal Peptide Synthetases genes and also utilized for building non ribosomal lipopeptides (NRPs).

Chapter I

Draft Genome Sequence of the Antimicrobial-Producing Strain Paenibacillus elgii AC13.

The content of this chapter is a manuscript entitled “Draft genome sequence of the antimicrobial-producing strain Paenibacillus elgii AC13” published in Genome Announcements on June 28th 2018. The author was responsible to genomic sequence to be annotation process, translate, write and submit the manuscript to the chosen journal.

Draft genome sequence of the antimicrobial-producing strain Paenibacillus elgii AC13

Daniel Barros Ortega1; Rosiane Andrade da Costa1; Allan da Silva Pires1; Thiago Fellipe de Araújo1; Janaina Fernandez de Araújo; Adriane Kurokawa; Beatriz Simas Magalhães; Octávio Luís Franco1; Ricardo Henrique Kruguer2; Georgios Joannis Pappas Junior; Cristine Chaves Barreto1.

¹ Universidade Católica de Brasília, Graduate Program in Genomic Sciences and Biotechnology, SGAN 916, Brasília – DF 70790-160, Brazil. 2 Universidade de Brasília, Graduate Program in Molecular Biology,

Corresponding author: Cristine Chaves Barreto. Phone: +55-61-3448-9217. Fax: +55-61-3347-4797 / e-mail: [email protected] Universidade Católica de Brasília, Programa de Ciências Genômicas e Biotecnologia. Brasília - Brazil. 70790-160.

ABSTRACT

A Paenibacillus elgii strain isolated from Cerrado soil samples shown antimicrobial activity. Its draft genome sequence was acquired (GS20 FLX Titanium 454 platform), comprises 126 contigs (N50 198.427 bp) and 6,810 predicted coding sequences. Antimicrobial genes, including a NRPS module identified as part of pelgipeptin synthesis gene cluster were found.

BACKGROUND Paenibacillus genus was defined after an extensive comparison between DNA sequences encoding the 16S rRNA (1). Production of lipopeptides with antimicrobial and surfactant properties have been reported (2–4). Lipopeptides are cyclic or linear peptides carrying a variable fatty acid in their N-terminal region. These molecules are a product of nonribosomal synthesis. Non Ribosomal Peptide Synthases (NRPS) use a mixture of D and L amino acids and non-proteinogenic amino acids as building blocks (2).

Paenibacillus sp. strain AC13 is a spore-forming, monoderm bacteria isolated from soil samples of the Brazilian Cerrado. This strain exhibits long-rods cells and motility by peritrichous flagella; growth occurs at pH values ranging from 4.7 to 8.0 at temperatures of 28 °C or 37 ºC. The Paenibacillus sp. AC13 strain also produces at least two groups of antimicrobial compounds (5–7). DNA of P. sp. AC13 was isolated from an overnight culture on Luria-Bertani medium (37°C). DNA was extracted using a modified CTAB protocol (8). Cells were lysed using a solution containing 10% of sodium dodecyl sulfate (SDS), CTAB 10% in 0.7 M of NaCl, and 1mg·mL-1 of lysozyme. Proteins were removed enzymatically with 100 ng·mL-1 (w/v) of proteinase K followed by a phenol:chloroform:isoamyl alcohol (25:24:1 v/v) extraction (9). DNA was suspended in distilled water and quantified. The genome sequence was acquired using a GS20 FLX Titanium 454 platform with a 20 X coverage yielding 455,876 reads. The assembly was performed using Newbler software (v2.3) yielding 126 contigs from which 108 had more than 200 bp with a N50 value of 198.427 bp, and 53.4% of G-C content. Genomic sequence annotation used the Prokaryotic Genome Annotation Process Pipeline (PGAP) (10); a modified NCBI Prokaryotic Genome Automatic Annotation Pipeline (PGAAP)5 protocol. PGAP determines structural annotation by comparing open reading frames (ORFs) to libraries of protein hidden Markov models (HMMs) (11–15), representative RefSeq proteins and proteins from well characterized reference genomes. GeneMarkS+ (v 4.4) (16) was used for best-placed reference protein set method. The 7,816,260 bp long genome contained 7,149 total genes and 6,810 coding sequences (CDS), including 111 RNA genes from which seven 5S rRNA genes, seven partial 16S rRNA genes, one partial 23S rRNA gene, and 92 tRNA genes. A total of 228 pseudo genes were identified, none of the pseudo genes had ambiguous residues; 119 of them were frameshifted and 123 had incomplete sequences. Forty putative NRPS genes were identified, and some of them might be part of the Pelgipeptin synthase gene cluster (17) that was identified in P. elgii strain B69 (18). Other antimicrobial genes were identified: a fusaricidin coding sequence, nine genes involved in polyketide gene cluster, eight genes related to bacteriocins, and two ABC transporter related with lantibiotics. The average nucleotide identity (ANI) comparison with other Paenibacillus genomes available at the NCBI genome database and the Joint Genome Institute (19) revealed ANI values of 70.95% with P. alvei TS-15; 74.96% with P. mucilaginosus K02; 96.45% with P.

32 elgii B69, and 97.62 % with P. elgii type strain M63; indicating that AC13 is a new Paenibacillus elgii strain. Nucleotide sequence accession numbers. The Paenibacillus elgii strain AC13 Whole Genome Shotgun project sequencing has been deposited at DDBJ/ENA/GenBank under the accession number PYHP00000000. The version described in this paper is version PYHP01000000.

ACKNOWLEDGMENTS We would like to thanks all the support from Alessandra Reis during the sequencing stage and from the GenBank staff for their assistance in genomic sequence data registration.

FUNDING INFORMATION Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for financial support our post graduate students and to the Project (CNPq n°: 560915/2010-1).

REFERENCES 1. Ash C, Priest FG, Collins MD. 1993. Molecular identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks and Collins) using a PCR probe test. Proposal for the creation of a new genus Paenibacillus. Antonie Van Leeuwenhoek 64:253–260. 2. Cochrane SA, Vederas JC. 2016. Lipopeptides from Bacillus and Paenibacillus spp.: A Gold Mine of Antibiotic Candidates. Med Res Rev 36:4–31. 3. Grady EN, MacDonald J, Liu L, Richman A, Yuan Z-C. 2016. Current knowledge and perspectives of Paenibacillus: a review. Microb Cell Fact 15:203. 4. Wu X-C, Shen X-B, Ding R, Qian C-D, Fang H-H, Li O. 2010. Isolation and partial characterization of antibiotics produced by Paenibacillus elgii B69. FEMS Microbiol Lett 310:32–38. 5. Kurokawa AS. 2006. Exploração bioecnológica de microrganismos do solo de Cerrado através da construção de bibliotecas metagenômicas e técnicas de cultivo. Universidade Católica de Brasília. 6. Araújo JF de. 2011. Diversidade Bacteriana do solo em diferentes fitofisionomias do bioma cerrado e perspectivas biotecnológicas. Universidade Católica de Brasília.

7. Costa RA da C. 2017. Paenibacillus elgii: ciclo celular, produção de peptídeos não ribossomais e potencial biotecnológico. Universidade Católica de Brasília. 8. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K. 1987. Preparation of genomic DNA from bacteriaCurrent Protocols in Molecular Biology. 9. Sambrook J, W Russell D. 2001. Molecular Cloning: A Laboratory Manual. Cold Spring Harb Lab Press Cold Spring Harb NY. 10. Angiuoli S V, Gussman A, Klimke W, Cochrane G, Field D, Garrity G, Kodira CD, Kyrpides N, Madupu R, Markowitz V, Tatusova T, Thomson N, White O. 2008. Toward an online repository of Standard Operating Procedures (SOPs) for (meta)genomic annotation. OMICS 12:137–141. 11. Klimke W, Agarwala R, Badretdin A, Chetvernin S, Ciufo S, Fedorov B, Kiryutin B, O’Neill K, Resch W, Resenchuk S, Schafer S, Tolstoy I, Tatusova T. 2009. The National Center for Biotechnology Information’s Protein Clusters Database. Nucleic Acids Res 37:D216-23. 12. Haft DH, Selengut JD, Richter RA, Harkins D, Basu MK, Beck E. 2013. TIGRFAMs and Genome Properties in 2013. Nucleic Acids Res 41:D387-95. 13. Finn RD, Coggill P, Eberhardt RY, Eddy SR, Mistry J, Mitchell AL, Potter SC, Punta M, Qureshi M, Sangrador-Vegas A, Salazar GA, Tate J, Bateman A. 2016. The Pfam protein families database: towards a more sustainable future. Nucleic Acids Res 44:D279-85. 14. Lowe TM, Chan PP. 2016. tRNAscan-SE On-line: integrating search and context for analysis of transfer RNA genes. Nucleic Acids Res 44:W54-7. 15. Nawrocki EP, Burge SW, Bateman A, Daub J, Eberhardt RY, Eddy SR, Floden EW, Gardner PP, Jones TA, Tate J, Finn RD. 2015. Rfam 12.0: updates to the RNA families database. Nucleic Acids Res 43:D130-7. 16. Borodovsky M, Lomsadze A. 2011. Gene Identification in Prokaryotic Genomes, Phages, Metagenomes, and EST Sequences with GeneMarkS Suite, p. 4.5.1- 4.5.17. In Current Protocols in Bioinformatics. John Wiley & Sons, Inc., Hoboken, NJ, USA. 17. Donadio S, Monciardini P, Sosio M. 2007. Polyketide synthases and nonribosomal peptide synthetases: the emerging view from bacterial genomics. Nat

Prod Rep 24:1073–1109. 18. Qian CD, Liu TZ, Zhou SL, Ding R, Zhao WP, Li O, Wu XC. 2012. Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiol. 19. Goris J, Konstantinidis KT, Klappenbach JA, Coenye T, Vandamme P, Tiedje JM. 2007. DNA-DNA hybridization values and their relationship to whole- genome sequence similarities. Int J Syst Evol Microbiol 57:81–91.

Chapter II

Pelgipeptin production from different Paenibacillus elgii strain AC13 inoculation methods

INTRODUCTION Non-ribosomal peptides (NRPs) are molecules produced by bacteria and fungi through a mechanism that involves large enzymatic complexes called non-ribosomal peptide synthetases (NRPS) (Wang et al., 2014). NRPs are extremely diverse molecules usually non limited by the 20 amino acid residues, and may have contain modified amino acids, lipids or glycogenic moieties attached to the peptide chain (Finking and Marahiel, 2004). In recent decades, lipopeptides from this class of biomolecules have attracted attention due to their antimicrobial potential as antibacterial or surfactant properties (Moon et al., 2017). Strain AC13 of Paenibacillus elgii was previously isolated from Cerrado soils samples showed antimicrobial activity (Costa et al., 2019). Moreover, in adverse situations this genus is able to produce metabolically inactive resistance structures called spores, which remains dormant and enduring intolerable environmental conditions until return to a vegetative cell (Paredes-Sabja et al., 2011). Recent work has discussed the importance of the inoculation method of a liquid culture and how it can influence growth dynamics, cell aggregation formation and growth dynamics and antimicrobial resistance (Kragh et al., 2018). The goal of the present work was to test the influence of two different inoculation methods (vegetative cells and spores) on growth dynamics and production of Pelgipeptins by P. elgii strain AC13.

MATERIAL AND METHODS The experiments were conducted between August, 2015; and July, 2018 at Universidade Católica de Brasília, Brasília/DF; Brazil. Strain AC13 (Kurokawa, 2006) of P. elgii was used throughout experimental procedures. For Mass Spectrometry (MS); a Bruker Daltonics (Billerica, MA, United States) Autoflex Matrix assisted laser desorption ionization time of flight (MALDI-TOF) mass spectrometer with linear and reflected detection mode was utilized. High Performance Liquid Chromatography was carried out using a Prominence (Shimadzu Corp., Kyoto; Japan) HPLC or UPLC. A Shimadzu (Kyoto, Japan) shim-pack C18 column VP-ODS (4.6 × 150 mm; 4.6 µm particle size) was used. Culture media used was either from HiMedia Laboratories (Mumbai; India) or Sigma-Aldrich Chemicals Company (St. Louis, MO, United States). Synchronization of strain AC13 starter population P. elgii strain AC13 spore preparation was obtained according to the protocol established by Huo et al. (2010). Initially, a -80 oC stock culture of P. elgii was inoculated in nutrient agar and incubated for 24 hours at 37 °C. Then, an isolated colony was inoculated

37 into 250 mL of concentrated (2×) Schaeffer’s Glucose (SG) broth and incubated into orbital shaker for 72 hours at 37 °C and 180 rpm shaking. The culture medium 2× SG is prepared by adding 16 g nutrient broth, 2 g KCl, and 0.5 g MgSO4. After adjusting the pH to 7.2, the medium was autoclaved, cooled to 55°C, and then mixed with the sterilized solutions of 1 -1 -1 -1 -1 mL·L of 1 M Ca(NO3)2, 1 mL·L of 0.1 M MnCl2, 1 mL·L of 1 mM FeSO4, and 2 mL·L of 50% (w/v) glucose (Nicholson, 1990; Huo et al., 2010). Cultures containing spores were transferred to sterile polypropylene tubes and centrifuged at 9000 × g for 15 minutes and washed with sterile deionized water four times to remove the culture media. Remaining vegetative cells and germinated spores were killed by incubation in a water bath at 80 °C for 15 minutes. The top layer of the cell pellet of gelatinous appearance was discarded and the remaining bottom layer was vigorously washed with sterile deionized water at 4 °C and centrifuged (9000 × g; 15 minutes) three times. Finally, the content obtained was homogenized in sterile deionized water and kept refrigerated until use. Strain AC13 stock solution quality control The spore purification protocol was validated by phase contrast microscopy and by mass spectrometry. To determine the efficacy of the procedure, samples of 1 mL were collected at three different time points throughout purification process. Two samples were collected during strain AC13 growth at 24 hours and 72 hours of incubation, and the last sample was obtained after the purification protocol was applied. Samples were washed with sterile deionized water and centrifuged at 9000 × g for 3 minutes, three times before mass spectra acquisition. Initially, samples were lyophilized and homogenized in 10 μL of 70% formic acid and 10 μL Acetonitrile (ACN) 100% under vigorous shaking for 30 seconds on a vortex for liquid extraction of proteins. Then, cells were collected by centrifugation (20,000 × g; 3 minutes) and 1 μL of the supernatant was transferred to a polypropylene tube (0.6 mL) containing 3 μL of α-cyano-4-hydroxycinnamic matrix. The samples were homogenized and applied onto stainless steel acquisition plate in triplicate. At least eight spectra were acquired from two or three different biological samples using MBT (2,000- 20,000 m/z) and RP900- 4,500 m/z methods. Strain AC13 samples were also placed on glass slides and covered with acrylic coverslips for microscope acquisition. Strain AC13 was observed under phase contrast. Microscopy images were obtained using an AxioCam ERc5s camera coupled to the microscope. Purified spores stock solution was quantified using a Neubauer's chamber.

Quantification of stock solution of spores Spores quantification was carried out by direct counting using a Neubauer chamber. After dilute (10-3) the spore stock solution in sterile deionized water, 10 μL of the dilution was applied on the top of the Neubauer chamber until the sample was completely absorbed by capillarity. Spores within five small quadrants of from the 0.1 μL volume square were directly counted. Finally, the equation 1 was applied:

푥 × 250.000 퐶 (푆푝표푟푒푠⁄푚퐿) = 푓

Where 푪 represents the final stock concentration; 풙 is the arithmetic mean of the direct count performed and 풇 is the dilution factor.

Growth curve and PGPs profile of strain AC13 population Strain AC13 was grown in different culture conditions in order to characterize and identify biomass and PGPs variations within growth curve. Optical density of cultures at 600 nm, dry weight content, mass spectrometry and HPLC were used in order to create PGPs accumulation profile.

Strain AC13 growth curve From the spore stock solution, 106 spores were inoculated in 100 mL of nutrient broth and incubated in an orbital shaker at 37 °C and 180 rpm. Absorbance at 600 nm was monitored throughout the growth curve. In addition, samples were collected to determine the dry weight at each point. Dry weight was measured in triplicates by filtering samples using a Büchner flask and vacuum pump system. Filters were placed in an oven at 105 ° C for 20 minutes. A standard curve was built using the mean from optical density at 600 nm found during exponential growth and average dry weight in order to determine the linear equation and correlation factor between biomass and absorbance. For that, a starter population was settled as ~103 spores·mL-1. Growth curves of P. elgii strain AC13 were then carried out into three different culture media.

Growth parameters of strain AC13 in three different culture media In order to compare growth in different media, 106 spores were inoculated into 100 mL of Mineral Medium for Paenibacillus (MMP); Nutrient broth (NB); Mueller-Hinton broth 39

(MHB), and 2% Peptone-enriched Mueller-Hinton broth (PMHB). Each tested culture was then incubated for 48 hours at 37° and 200 rpm shaking. Cell growth dynamics was measured by optical density at 600 nm for every 3-4 hours. After the incubation period, three growth parameters were determined for each treatment: Specific growth rate (µ), Generation time (g), and Division rate (v). For that, the following equations, described in Lehninger, A. L., Nelson, D. L., & Cox, M. M. (2000), were used:

Equation 1: 푙푛 푁 − 푙푛푁 µ = 0 푙푛 2

Where µ specific growth rate, 푵: end of log phase; 푵ퟎ: beginning of log phase.

Equation 2: 푙푛 2 푔 = µ Where 품: generation time, and µ: specific growth rate.

Equation 3: 1 푣 = ⁄푔 Where 풗: Division rate and 품: generation time.

Pelgipeptin production in cultures starting from spores or from vegetative cells Synthesis and accumulation of PGPs were monitored over time, comparing two different starter populations: from spores and from vegetative cells. Strain AC13 spores or vegetative cells were inoculated in a chemically defined medium at an initial concentration of 103 cells/spores·mL-1 and both culture methods were then incubated at the same conditions (37 ºC; 180 rpm). The cellular biomass and PGPs concentration were evaluated for 120 hours. Samples were collected at every 24 hours and analyzed by HPLC in order to detect released PGPs.

RESULTS Synchronization of strain AC13 starter population. Strain AC13 synchronization was achieved by purifying spores and removing remaining vegetative cells and cultures’ supernatant. A group of peptides ranging from 1,500 to 2,000 m/z were not visualized by mass spectrometry at cultures comprised in their majority of vegetative cells and ions with m/z between 2,000 and 5,000 appeared only after spores’ purification protocol (Figures 1-3).

Figure 1. MALDI-ToF spectra of strain AC13 after 24 hours of incubation at 37°C in 2× SG broth. Peptides were extracted from pellet with 10 µL of formic acid 70% and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method (A). Phase-contrast Microscopy of strain AC13 after 24 hours of incubation in 2× SG broth at 37°C. Elongated rods are vegetative cells. Amplification of 100× (B).

Figure 2. MALDI-ToF spectra of strain AC13 after 24 hours of incubation at 37 °C in 2× SG broth. Peptides were extracted from pellet with 10 µL of formic acid 70 % and 10 µL of acetonitrile (ACN) 100 %. Spectra acquired using RP900-4500 m/z method (A). Phase-contrast Microscopy of strain AC13 after 24 hours of incubation in 2× SG broth at 37 °C. Elongated rods are vegetative cells. Amplification of 100× (B).

The absence of signal to ionize monoisotopic masses from a group of peptide ranging from 1,660 to 1,720 m/z from cultures with vegetative cells as majority population (Figure 4A) indicates that these group of molecules are characteristic of AC13 spores, since those molecules were ionized and exhibited similar ionization pattern (Figure 4B and 4C).

Besides, when observed by mass spectrometry, cultures’ protein profile at different time-points indicates a shift between proteins associated with vegetative cells or spores (Figure 5A-C). Also at least more seven groups of proteins were properly ionized after the spore purification protocol was adopted (Figure 5C).

Figure 5. MALDI-ToF spectra of P. elgii strain AC13. Protein profile comparison of a population compressed of free vegetative cells (A); free spores and a few vegetative cells (B) and purified spores (C). Protein content was extracted from pellet using 10 µL of formic acid 70% and 10 µL of acetonitrile (ACN) 100 %. Spectra were acquired using MBT (2,000-20,000 m/z) method.

Strain AC13 growth curve Dry Weight and Optical Density measurements were used in order to build a Standard curve prior the growth curve experiment (Table1; Figure 6). Growth was monitored during exponential phase. The linear equation was determined as 풚:

푦 = ퟎ. ퟓퟎퟕퟓ푥 + 0.0756 R² = 0.9577

Where 풙 is the value of optical density found; and 0.5075 is the correlation factor between dry weight and optical density. Values found show 95% of probability to be correct (R² = 0.9577).

thours 6 7 8 9 10

ODAC13 (600nm) 0.097 0.255 0.349 0.497 0.717 Standard Deviation 0.001247 0.023921 0.017442 0.012284 0.025382 Dry weight (g/L) 0.13 0.17 0.28 0.33 0.43 Standard Deviation 0.013608 0.079737 0.068493 0.020787 0.106284

0.50 0.45

0.40 0.35 0.30 0.25 0.20 0.15 Dry (g/L) Dry weight 0.10 0.05 0.00 0 0.2 0.4 0.6 0.8 600 nm

The overall growth profile of strain AC13 was visualized based on optical density values found at 600 nm for cultures in Nutrient Broth at 37 °C (Table 2). The correlation factor found within the linear equation was used to convert Optical Density values in Biomass. Strain AC13 growth was monitored for 24 hours (Figure 7).

thours 0 2 4 6 8.5 11.5 13.5 15.5 17.5 19.5 21.5 23.5

ODAC13 (600nm) 0 0.01 0.01 0.02 0.02 0.15 0.40 0.59 0.72 0.76 0.70 0.72 Standard Deviation 0 0.01 0.01 0.00 0.00 0.03 0.04 0.02 0.02 0.02 0.02 0.03 Biomass (g/L) 0 0.01 0.00 0.01 0.01 0.08 0.20 0.30 0.37 0.39 0.36 0.37

Figure 7. Growth Curve of P. elgii strain AC13 over 24 hours in nutrient broth at 37 °C. Starter 3 -1 population was settled as ~10 spores·mL . Lag phase from t0 to t9 hours. Exponential growth from t9 to t20 -1 hours. Stationary phase from t20 to t24 hours. Biomass (g·L ) in axis y and time (hours) in axis x.

0.45 0.4 0.35

0.3 0.25 0.2

Biomass Biomass (g/L) 0.15 0.1 0.05 0 0 5 10 15 20 Time (hours)

Determination of growth rates of strain AC13 at three different culture media. Strain AC13 was grown into three different culture media. Exponential growth phase started within 4 hours on NB and after 8 hours on MHB and MHPB (Figure 8).

Figure 8. Growth Curve of strain AC13 over 20 hours in nutrient broth (NB); Mueller-Hinton Broth (MHB) or Mueller-Hinton Peptone Broth (MHPB) at 37°C. Starter population was settled as ~103 spores·mL-1. Biomass (g·L-1) in axis y and time (hours) in axis x.

0.70

0.60

) 1 - 0.50

g·L 0.40 0.30 0.20

Biomass Biomass ( 0.10 0.00 0 5 10 15 20 Time (Hours) Nutrient Broth (NB) Biomass Muellher-Hinton Broth (MHB) Biomass Mueller-Hinton Peptone Broth (MHPB) Biomass

Variations of microbial growth parameters were observed. During exponential growth, a division rate of 0.5677 generations per hour was reached in NB. When grown in MHPB, this rate was 0.3001generations per hour and in MHB, 0.3222 generations per hour (Table 3).

Growth rate was higher in NB (0.3935 h-1) followed by MHB (0.2233 h-1) and MHPB (0.2080 h-1). While in NB AC13 showed a generation time of 1.7 hours, in MHB it was 3.3 and in MHB it was 3.1 hours (Table 3).

Table 3. Strain AC13 was grown over 20 hours in nutrient broth, (NB); Mueller-Hinton Broth (MHB) 3 -1 and MHPB at 37°C. Starter population was settled as ~10 spores·mL . Biomass values from Log phase (t4 to t11 hours) were used to calculate Specific Growth Rate (µ); Generation Time (g) and Division Rate (v) from each culture tested.

NB MHB MHPB

Specific Growth Rate µ (h-1) 0.3935 0.2233 0.2080 Generation Time g (hours) 1.7615 3.1041 3.3324 Division Rate v (generation/hour) 0.5677 0.3222 0.3001

PGPs production between two different starter populations of strain AC13 Cultures of strain AC13 were grown in mineral medium MMP from two different starter populations. Vegetative cells population entered log-growth phase after 15 hours while cultures from spores started an exponential growth after 24 hours. The biomass became stable within the experiment duration for both conditions, and reached a maximum value after day 2 for vegetative cell starter population and day 3 for spores (Figure 9).

Figure 9. Comparison of average biomass of P. elgii strain AC13 at stationary phase from two different starter populations. No meaningful differences were found (n=18).

1.2

1.0 Cultures from spore stock solution inoculum at day 3

0.8 Cultures from spore stock solution inoculum at day 4 Cultures from spore stock solution 0.6 inoculum at day 5 Cultures from vegetative cell Biomass Biomass (g/L) 0.4 inoculum at day 2 Cultures from vegetative cell 0.2 inoculum at day 3 Cultures from vegetative cell inoculum at day 4 0.0

Biomass at stationary phase of both methods did not shown meaningful variation and average biomass found for three time points during stationary phase ranged from 0.8 to 1.2 g·L-1 (Figure 10).

Figure 10. Comparison of average biomass at stationary phase from two different starter population of strain AC13. Both conditions ended with an average biomass of ~1 g·L-1 (n=18).

1.2

1.0

0.8

0.6

Biomass Biomass (g/L) 0.4

0.2

0.0 Vegetative Cells Spore Stock Solution

The final biomass reached at both treatments had no significant difference. PGPs production, however, showed variations between the two starter populations tested (Figure 11). Figure 11. Comparison of strain AC13 PGP production at stationary phase from two different inoculum methods (n=18)

180 160

140 Cultures from spore stock solution 120 inoculum at day 3 Cultures from spore stock solution 100 inoculum at day 4 Cultures from spore stock solution 80 inoculum at day 5 60 Cultures from vegetative cell

inoculum at day 2 Pelgipeptins (µg/mL) Pelgipeptins 40 Cultures from vegetative cell inoculum at day 3 20 Cultures from vegetative cell inoculum at day 4 0 47

Average PGPs production at stationary phase from a spore starter population was 161.14 μg·mL-1 while the vegetative cell starter population yielded 101.33 μg·mL-1 of PGPs during the stationary phase (p<0.05) (Figure 12).

Figure 12. Comparison of average Pelgipeptin production at stationary phase from two different starter population of P. elgii strain AC13. ~100 µg·mL-1 (n=18) produced by vegetative cell starter population and ~160 µg·mL-1 produced by a spore starter population.

180

160

) 1

- 140 mL · 120 100 80

60 Pelgipeptin (µg Pelgipeptin 40 20 0 Vegetative Cells Spore Stock Solution

When plotted together, the variations of PGPs production over three time points within stationary phase for both, a spore starter population and vegetative cells starter population are easily seen and does not correspond to the same variations over biomass (Figure 13).

from spores was higher.

) 200 1.2

1 -

180 )

1.0 1 160 - 140 0.8 120 100 0.6 80

60 0.4 Biomass Biomass (g.L 40 0.2 Pelgipeptin (µg.mL 20 0 0.0 0 20 40 60 80 100 120 140 Time (hours) Pelgipeptin production obtained from the spore stock solution inoculum Pelgipeptin production obtained from the vegetative cells pre-inoculum Biomass of cultures from the spore stock solution inoculum Biomass of cultures from vegetative cells pre-inoculum 48

DISCUSSION Spores purification The rapid advancement in mass spectrometry techniques associated with high- performance liquid chromatography has triggered the interest of the scientific community in developing new rapid and accurate techniques for identifying microorganisms (Dybwad et al., 2013). The creation of fingerprint mass spectra and subsequent proteomic analysis have been used to detect and characterize small acid-soluble proteins (SASPs) that are characteristic of spores of the genus Bacillus sp. It is known that SASP proteins are responsible to confer resistance to DNA, preventing their deterioration. SASPs also serve as an amino acid reserve during spore germination. Moreover, due to their alkaline nature, they can be solubilized in acid and easily protonated, emitting stronger signals when ionized by MALDI-ToF (Dybwad et al., 2013). In addition, the application of MALDI-ToF mass spectrometry associated with other techniques proved to be effective in recognizing patterns related to the different phases of microbial growth, resulting in the correct classification of Escherichia coli samples. It also was able to reveal witch ions contributed to the correct grouping of the samples. Ribosomal proteins, for example, are determinants for the identification of the exponential phase cultures (Momo et al., 2013). In many microbial species there are signs of fluctuations among the groups of proteins produced as a function of the life stage. Therefore, it is believed that using MALDI-ToF mass spectrometry was helpful to determine the degree purification of strain AC13 spore stock solution, even though it was associated with results from a less sensitive technique like optical microscopy. Definitions of terms like “production”, “productivity or “yield” have not always been concise when dealing with biological systems. Thienemann (1931) stated that modern biologists use the term “production” referring to ‘the amount of organic matter produced in a given time from an organism’ (Macfadyen, 1948). That quantity, however, cannot be accessed in practice because of intrinsic difficulties in making such measurements from a continuous process. “Productivity” is defined as the quantity of matter derived from a given area or biological community. “Yield” is the productivity over a established period of time, or the difference of two energy contents (Macfadyen, 1948). Here it is assumed that “production” is the correlation between the product of interest, Pelgipeptin and the resources used; the biomass. Even though it is not the ideal measurement it resulted in different correlation between those two variables (Biomass and Pelgipeptin quantification) when the starter population is compared. Interestingly, the maximum growth rate was found in nutrient broth.

It was not expected, since peptone enriched Mueller Hinton broth was settled as a more nutritious media. Even though a maximum growth rate in theory should reach 1, that is not possible since biochemical reactions are not 100% efficient and energy is lost as activation heat and carbon atoms are used in cellular metabolism or released as CO2. In regard of different starter populations (Spores × Vegetative Cells), P. elgii AC13 grows at different growth rates. From a spore starter population its lag phase last longer, but the pelgipeptin quantification was higher afterwards. Biomass from both conditions showed similar values, whereas PGPs quantification from spores was higher. However, more data need to be acquired in order to define if a spore starter population yields a higher amount of pelgipeptin. Yield, however, is a parameter that needs to be further investigated.

CONCLUSIONS Strain AC13 can modulate its Pelgipeptin production when the starter population of a culture is changed. However, in order to verify Pelgipeptin productivity or yield, strain AC13 must be cultured in a more controlled environment like fermentation systems.

REFERENCES

ASH, C.; PRIEST, F. G.; COLLINS, M. D. Molecular identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks and Collins) using a PCR probe test. Proposal for the creation of a new genus Paenibacillus. Antoinie van Leeuwenhoek, v. 64, p. 253-260, 1993.

COCHRANE, S. A.; VEDERAS, J. C. Lipopeptides from Bacillus and Paenibacillus spp.: A Gold Mine of Antibiotic Candidates. Med Res Rev, v. 36, n. 1, p. 4-31, Jan 2016. ISSN 1098-1128 (Electronic) 0198-6325 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24866700 >.

COSTA, R. et al. Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae – a short communication. Biotechnology Research and Innovation, 2019/01/16/ 2019. ISSN 2452-0721. Available at: < http://www.sciencedirect.com/science/article/pii/S2452072118300613 >.

DYBWAD, M. et al. Rapid identification of Bacillus anthracis spores in suspicious powder samples by using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS). Appl Environ Microbiol, v. 79, n. 17, p. 5372-83, Sep

2013. ISSN 1098-5336 (Electronic) 0099-2240 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23811517 >.

GRADY, E. N. et al. Current knowledge and perspectives of Paenibacillus: a review. Microb Cell Fact, v. 15, n. 1, p. 203, Dec 01 2016. ISSN 1475-2859 (Electronic) 1475-2859 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/27905924 >.

FINKING, R.; MARAHIEL, M. A. BIOSYNTHESIS OF NONRIBOSOMAL PEPTIDES. Annual review of Microbiology, v. 58, p. 453-488, 2004.

HUO, Z. et al. Investigation of factors influencing spore germination of Paenibacillus polymyxa ACCC10252 and SQR-21. Applied microbiology and biotechnology, v. 87, n. 2, p. 527-536, 2010. ISSN 0175-7598.

KRAGH, K. N. et al. The inoculation method could impact the outcome of microbiological experiments. Appl. Environ. Microbiol., v. 84, n. 5, p. e02264-17, 2018. ISSN 0099-2240.

KUROKAWA, A. S. 2006. Exploração bioecnológica de microrganismos do solo de Cerrado através da construção de bibliotecas metagenômicas e técnicas de cultivo. Universidade Católica de Brasília.

LEHNINGER, A. L., NELSON, D. L., & COX, M. M. (2000). Lehninger principles of biochemistry; 5th Edition. New York: Worth Publishers.

MACFADYEN, A. The Meaning of Productivity in Biological Systems. Journal of Animal Ecology, v. 17, n. 1, p. 75-80, 1948. ISSN 00218790, 13652656. Available at: < http://www.jstor.org/stable/1611 >.

MOON, S. H. et al. Novel Linear Lipopeptide Paenipeptins with Potential for Eradicating Biofilms and Sensitizing Gram-Negative Bacteria to Rifampicin and Clarithromycin. Journal of Medicinal Chemistry, v. 60, n. 23, p. 9630-9640, 2017/12/14 2017. ISSN 0022-2623. Available at: < https://doi.org/10.1021/acs.jmedchem.7b01064 >.

MOMO, R. A. et al. MALDI-ToF mass spectrometry coupled with multivariate pattern recognition analysis for the rapid biomarker profiling of Escherichia coli in different growth phases. Anal Bioanal Chem, v. 405, n. 25, p. 8251-65, Oct 2013. ISSN 1618-2650 (Electronic) 1618-2642 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23942565 >.

ORTEGA, D. B. et al. Draft Genome Sequence of the Antimicrobial-Producing Strain Paenibacillus elgii AC13. Genome Announc, v. 6, n. 26, Jun 28 2018. ISSN 2169-8287 (Electronic). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/29954895 >.

PAREDES-SABJA, D.; SETLOW, P.; SARKER, M. R. Germination of spores of Bacillales and Clostridiales species: mechanisms and proteins involved. Trends in microbiology, v. 19, n. 2, p. 85-94, 2011. ISSN 0966-842X.

RODGERS, K. J. Non-protein amino acids and neurodegeneration: the enemy within. Exp Neurol, v. 253, p. 192-6, Mar 2014. ISSN 1090-2430 (Electronic) 0014-4886 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24374297 >.

WANG, H. et al. Atlas of nonribosomal peptide and polyketide biosynthetic pathways reveals common occurrence of nonmodular enzymes. Proc Natl Acad Sci U S A, v. 111, n. 25, p. 9259-64, Jun 24 2014. ISSN 1091-6490 (Electronic) 0027-8424 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24927540 >.

WU, X. C. et al. Isolation and partial characterization of antibiotics produced by Paenibacillus elgii B69. FEMS microbiology letters, v. 310, n. 1, p. 32-8, Sep 1 2010. ISSN 1574-6968 (Electronic) 0378-1097 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/20618851 >.

Chapter III

Molecular Networking analysis based on a proteomic collection from Paenibacillus elgii strain AC13 – DRAFT

Working Title: Non ribosomal lipopeptides produced by Paenibacillus elgii strain AC13 may interfere on quorum sensing communication signaling system for biofilm formation or swarming behavior

BACKGROUND

Phenotypical variations within Paenibacillus

From Bacillaceae family (phylum, Firmicutes), Paenibacillus genus is a spore-forming bacteria, expressing intriguing phenotypical changes and high degree of morphological alterations observed throughout cellular development (Kearns, 2010; Mandic-Mulec et al., 2016; Mahdi and Fisher, 2018). Spores allows bacteria to disperse as far as between continents, thereby; understanding sporulation and germination processes are crucial in microbiology, since those characteristics are of great economic and ecological concern (Morrissey et al., 2015; Mandic-Mulec et al., 2016). The Apis honeybees’ lethal American foulbrood disease (AFB), for instance, occurs due to highly contagious and extremely resistant Paenibacillus larvae endospores. Honeybees’ feeding is usually achieved by glandular secretion and processed honey (Morrissey et al., 2015). Once infected, adults transmits spores to bee larvae that die within 3 to 12 days; P. larvae spores, however, remain dormant and infective for more than 35 years, leaving empty hives due to the latent possibility of new outbreaks. Moreover, it causes economic and ecological damages, since honeybees pollination are responsible for approximately 90% of commercial crops (Schafer et al., 2014; Morrissey et al., 2015; Grady et al., 2016; Mahdi and Fisher, 2018).

Despite intriguing, spores formation and germination process, however, are not the only phenotypical variations expressed by Paenibacillus. Another interesting feature is the motility behavior occurrence and biofilm formation (Alasil et al., 2015), which has been reported as of great importance regarding virulence factors transmission and antibiotic resistance promotion. The expression of swarming is a common behavior of Bacillus and Paenibacillus genus. Swarming is a type of social-level motility behavior in which the morphological differentiation of specific groups within a population occurs. Hyper-expressed flagella, surfactants production and type IV pilli formation are usually related with the expression of swarming (Anyan et al., 2014). At a macroscopic level, it was found that colonies capable of swarming present greater resistance to antimicrobial compounds when compared to colonies grown in liquid medium (Be'er et al., 2013), which makes it even more

54 intriguing and challenging studies concerning the behavior of microorganisms on solid or semi-solid culture.

Bio surfactants are amphipathic molecules that can increase the availability of insoluble carbon sources and, in general, microorganisms produce it to dissolve and use insoluble carbon in order to survive (Liang et al., 2014). In addition, it is also known that many species of microorganisms capable of expressing swarming motility do so by releasing surfactant compounds. These molecules reduce the tension between the substrate and the bacterial cell and allow the microorganism to spread on the surfaces (Kearns, 2010). Pelgipeptins, a group of non-ribosomal lipopeptides synthesized by Paenibacillus elgii; are homologous to Surfactin, a similar lipopeptide exhaustively described in literature as surfactant compound with antimicrobial activity that induces biofilm formation (Kearns, 2010; Romero et al., 2011; Ines and Dhouha, 2015; Maunders and Welch, 2017).

Global Natural Products Social Molecular Networking (GNPS) Molecular networking is a tandem mass spectrometry (MS/MS) data organizational approach that organizes the MS/MS data as a relational spectral network and it was recently introduced in the drug discovery, metabolomics, and medical fields. GNPS implements dereplication of Natural Products MS/MS data by querying newly acquired MS/MS spectra against all the accumulated reference spectra in GNPS spectral libraries (Wang et al., 2016). The platform is being used to decode the metabolomic ‘dark matter’ in a variety of human and environmental samples by propagating spectral library-based annotation and showing that there are chemical relationships between detected molecules across many sample types (Quinn et al., 2017). Until 2016, more than 93 million MS/MS spectra have been searched at GNPS, yielding putative dereplication matches of 7.7 million spectra to 15,477 compounds (Wang et al., 2016). The curation and dereplication processes are based on the fact that two related molecules are likely to display similar fragment ion spectra because the chemistry of molecules dictates how they will be fragmented by MS/MS in the gas phase (Quinn et al., 2017). Molecular networking is thereby, able to map the chemistry that was detected in an MS/MS-based metabolomics experiment. Besides, the development of integral resources, such as the US National Center for Biotechnology Information (NCBI; Bethesda, MD, USA) and UniProt Knowledge Base (UniProtKB) provide robust platforms for Global genomics and proteomics research and facilitates data sharing and knowledge dissemination (Wang et al., 2016). In biology, the 55 visualization of networks using Cytoscape (www.cytoscape.org) enable the direct observation of similarities as well as differences between two or more conditions in which similar entities within the network are clustered together while disparate or unique entities are grouped separately (Watrous et al., 2012). Here it is suggest that establishing Paenibacillus elgii strain AC13 metabolomics profile by LC-MS/MS and its molecular networking analysis may help to better understand microbial population processes, since several characteristics such as biofilm formation or motility expression (e.g. antimicrobial production, toxins release, virulence) are described as been regulated by intercellular quorum sensing communication factors.

General Goal: Create a collection of natural products synthesized by Paenibacillus elgii strain AC13 using LC-MS/MS and investigates the chemical relationships between detected molecules from strain AC13 and molecules across many sample types using global genomics and proteomics research databases.

Specific goals: 1. Time-based and phenotypical extraction of strain AC13 bioproducts produced on solid agar; 2. Identification of lipopeptides naturally produced by strain AC13 using LC- MS/MS; 3. Molecular Networking analysis of identified biomolecules produced by strain AC13.

MATERIALS AND METHODS

The Project was conducted between August, 2018; and February, 2019 at Universidade Católica de Brasília, Brasília/DF; Brazil and at University of California San Diego, La Jolla/CA; United States. Strain AC13 (Ortega et al., 2018) of Paenibacillus elgii was used throughout experimental procedures. Culture media used was either from HiMedia Laboratories (Mumbai; India) or Sigma-Aldrich Chemicals Company (St. Louis, MO, United States). A Thermo qExactive Hybrid Quadrupole-Orbitrap (ThermoFisher Scientific; Waltham, MA, United States) of High resolution quadrupole-orbitrap with maximum scan rate of 12Hz and 1ppm accuracy was used to perform LC-MS analysis. Culture conditions: molecules synthesized by strain AC13 were extracted from populations of P. elgii AC13

56 expressing and not expressing motility behavior. For that, lyophilized spores of strain AC13 were suspended in 1 mL of sterile deionized water. Then, a starter amount was grown in Luria-Bertani (LB) agar (48 h; 37°C) to verify cells viability and contamination - determined by the presence of colony-forming units (CFU) expressing same phenotype. When in LB agar AC13 did not express motility. After that, 106 spores per plate were suspended in deionized water at a final volume of 10 µL and used as inoculum for cultures in LB or Nutrient agar. Bacterial cells and associated metabolites were harvested by scraping the colonies from agar plates after 24; 48 and 72 hours of incubation at 37°C. Biomass was determined using a precision scale and natural products were extracted in three different time points throughout P. elgii strain AC13 stationary growth phase; and analyzed by means of LC MS/MS and molecular networking analysis. In order to keep statistical balance, a total of 153 cultures were used to obtain samples of extracted material. Average biomass for each morphological type at different time points was determined. Two batches of P. elgii strain AC13 cultures were acquired Lipopeptides extraction: Metabolome extraction was achieved by resuspension of bacterial lawn harvested from Petri dishes in 500 µL of 100% methanol at room temperature for three hours followed by ultrasonic water bath for 10 minutes. Bacterial cells were removed by centrifugation at 7.000 × g for 10 min. Resultant supernatant was transferred to micro tubes previously weighted and dried. Samples were stored in freezer until the day of LC-MS/MS analysis. Feature Based Molecular Networking: Feature Based Molecular Networking Analysis was carried out using MZmine 2 version .33. Thermo RAW data was used as project’s raw data file import format. Exact mass detection module was used for both MS and MS/MS scans (RT = 0.00 – 9.01 minutes). Noise level established was 1.0E3 for MS and 1.0E2 for MS/MS. Chromatogram builder step kept mass detection parameters. Time span - over which the same ion must be observed to be recognized - was 0.01 minutes. Minimum intensity for highest data point in the chromatogram was 3.0E3. Maximum m/z difference of data points in consecutive scans to be connected to the same chromatogram was established as 0.05 or 20.0 ppm. Chromatograms deconvolution step was carried using Baseline cut-off algorithm. Minimum acceptable peak height regarding absolute intensity was established as 1.0E3, peak duration range from 0.01 – 9.01 minutes and base line settled as 1.0E3. In order to enable feature based molecular analysis, m/z range for MS level 2 scan pairing was determined to be 0.02 Da and the RT shift for the MS level 2 scan pairing 0.1 minutes. Isotope pattern was found using isotopic peak grouper module. For this and for the alignment step, m/z tolerance was 0.001 m/z or 20.0 ppm. Maximum distance in RT from the

57 expected location of a peak was 0.1 minutes; and maximum charge for samples was settled to 3. The following step consisted in data alignment. Weight for m/z was settled in 75% and weight for RT was 25%. Retention time tolerance was 0.1 minutes. Only peaks with MS/MS scan; minimum peaks in a row and minimum peaks in an isotope pattern of two were settled as filtering criteria. Files were converted (.CSV) and sent to GNPS analysis. After, the export format file was analyzed using Cytoscape (v3.6.1). Hierarchy Layout: Retention Time was chosen to be the node column to be considered. A node label indicates Precursor Mass. Molecular Networking Analysis was done based on Cosine Scores (0.7

RESULTS AND DISCUSSION

Strain AC13 cultures extraction Biomass obtained from swarming cultures was consistently higher than in non- swarming cultures (Figure 1, Figure 2 and Table 1). Non swarming cultures exhibited higher extracted products when compared to swarming cultures (Figure 2, Figure 3 and Table 2). This result is in accordance to previously results described in literature for other species (Costerton, 1999); it is said that the amount of microorganisms within a biofilm corresponds to only 10% of the total dry weight, while 90% corresponds to components of the extracellular matrix, consisting of numerous polymers of biological origin, extracellular DNA and polysaccharides (Costerton, 1999; Watnick and Kolter, 2000). Figure 1 Average biomass of P. elgii strain AC13 cultures at 37 °C expressing motility (swarming) or forming biofilm (non-swarming). Cultures were scraped from agar and weighted (n=147).

140 Swarming cultures after 24 hours Swarming cultures after 48 hours 120 Swarming cultures after 72 hours Non-swarming cultures after 24 hours 100 Non-swarming cultures after 48 hours Non-swarming cultures after 72 hours

40 Biomass Biomass (mg)

Table 1. Sample ID; Injection was carried out using 100 µL of sample. The Injection volume used was 5 µL Biomass (mg), and Standard Deviation of samples used in LC-MS/MS analysis.

Sample ID Samples’ Identification Biomass (mg) Standard Deviation Empty Sample (MeOH 80 % + Blank 3 - - 1 % Formic acid) QC 1 Control group - - 166 ±50 Sample 3 Swarming after 24 hours 213 ±44 211 ±4 40 ±13 Sample 5 Non swarming after 24 hours 16 ±3 28 ±4 165 ±12 Sample 7 Swarming after 48 hours 146 ±20 203 ±7 70 ±0.4 Sample 10 Non swarming after 48 hours 60 ±1 84 ±10 192 ±3 Sample 13 Swarming after 72 hours 187 ±3 172 ±10 96 ±12 Sample 16 Non swarming after 72hours 101 ±1 121 ±26 98,5 - Sample 20 Swarming after 24 hours 153,5 ±1 153,6 - 148,8 ±2 Sample 23 Non swarming after 24 hours 139,7 ±1 129,4 ±1 141,5 ±1 Sample 25 Swarming after 48 hours 236,5 ±1 278,5 ±7 267,2 ±9 229,8 ±17 Sample 28 Non swarming after 48 hours 197,9 ±9 192,3 ±6 192,3 ±10 190,7 ±6 Sample 33 Swarming after 72 hours 193,3 ±11 122,3 - 206,3 ±8 236,8 ±7 270,4 ±14 Sample 36 Non swarming after 48 hours 225,4 ±19 258 ±13 236,8 ±8

Figure 2 Cultures’ phenotype of P. elgii strain AC13 expressing motility (swarming) in nutrient agar (superior) and only forming biofilms in Luria-Bertani 1.5 % agar (inferior, right) after 24 hours of incubation at 37 °C.

Figure 3 Average extracted natural products from P. elgii strain AC13 cultures at 37°C expressing °C expressing motility (swarming) or forming biofilm (non-swarming). Cultures were scraped from agar and weighted (n=147). Overall extracted bioproducts ranged between 0.4 and 10 milligrams.

12 Swarming 24 hours Swarming 48 hours 10 Swarming 72 hours

Non-Swarming 24 hours Non-swarming 48 hours 8 Non-swarming 72 hours

4 Average Average Extraction(mg)

Table 2. Samples’ information. Sample ID and Culture Phenotype expressed; Time (hours); Temperature (°C); Extraction (mg) and Final Extraction Concentration used (mg·mL-1) used in LC-MS/MS acquisition.

Sample ID Culture Phenotype Time Temperature (°C) Extraction (mg) Final Concentration (mg/mL) Blank 3 - - - - - QC 1 - - - - - Sample 3 Swarming 24 hours 37 2 1 Sample 5 Non-Swarming 24 hours 37 1,6 1 Sample 7 Swarming 48 hours 37 4,1 1 Sample 10 Non-swarming 48 hours 37 7,4 1 Sample 13 Swarming 72 hours 37 4,6 1 Sample 16 Non-swarming 72 hours 37 7,3 1 Sample 20 Swarming 24 hours 37 2,1 1 Sample 23 Non-Swarming 24 hours 37 2,6 1 Sample 25 Swarming 48 hours 37 2,1 1 Sample 28 Non-swarming 48 hours 37 6,3 1 Sample 33 Swarming 72 hours 37 2,6 1 Sample 36 Non-swarming 72 hours 37 10 1

Featured based networking analysis of strain AC13. Initial results from featured based networking analysis of strain AC13 show that several precursor masses including all precursor masses related to Pelgipeptins were clustered together Molecular networking analysis indicates a relationship between P. elgii strain AC13 60 network based on the precursor ion 1,318.8176 with the Pathway ErbB1, a downstream signaling pathway. The precursor ion cited was chosen because it has a lot of interactions with other precursor ions and was considered as an important node. In addition, pelgipeptin cluster is connected to the L-leucine biosynthesis pathway, and a cluster found within strain AC13 collection is connected to an enzyme responsible to synthesize O-acetyl-L-homoserine from L-homoserine. The precursor mass of 1,101.7052 Da – Pelgipeptin B/Permetin A – is connected to the gene Leu B pathway, responsible to code the protein 3-isopropylmalate dehydrogenase described for Paenibacillus dendritiformis C454 (https://www.uniprot.org; query 63727; entry H3SAZ4). Similar proteins with 90% of identity also indicate 3-isopropylmalate dehydrogenase that were described for Paenibacillus selenitireducens, Paenibacillus yonginensis, Paenibacillus amylolyticus, Paenibacillus sp. CAA11, and Paenibacillus thiaminolyticus. The enzyme is part of leucine biosynthetic pathway and is responsible for catalyze the oxidation of 3-isopropylmalate to 3-carboxy-4-methyl-2-oxopentanoate that becomes 4-methyl-2-oxopentanoate by a decarboxylation (Figure 4B). Leucine is one of the amino acids from Pelgipeptin backbone that appear at least once. Interestingly, when first described, the identification of the fifth positioned amino acid was unclear between Leucine and Isoleucine (Wu et al., 2010). A third cluster suggests that the enzyme Homoserine O-acetyltransferase (gene metAA) expressed by Lysinibacillus sp. strain FJAT-14222 is connected to the precursor mass of 416.7348 Da. The enzyme is involved in the catalytic activity during the first step of the sub pathway responsible to synthesize O-acetyl-L-homoserine from L-homoserine. The sub pathway is within the L-methionine biosynthesis de novo pathway, which is part of Amino- acid biosynthesis (Figure 4C). Among the QS signaling pathways, N-acyl-homoserine lactones are present as chemical signaling molecules for more than 70 different species of diderm bacteria.

acetyltransferase

omoserine O

acid biosynthesis: h

sub pathway of Amino

MS/MS (A). Pelgipeptin cluster and Molecular networking analysis of ion 1,101.7052 Da

homoserine (C).

on on 416.7368 Da connects to a

homoserine, acetylforming

AC13 proteomic collection obtained from LC

CoA to L

Paenibacillus elgii

Figure 4.

connected to an enzyme from leucine biosynthetic pathway (B).

is acetyl transfers an group from acetyl 62

CONCLUSIONS

Strain AC13 is able to express swarming motility when cultured in nutrient broth. Besides, biomass for swarming cultures was significantly higher than biomass found for non- swarming cultures. Extracted bioproducts, however, seems to follow an opposite logic: higher on non-swarming cultures and lower on swarming ones. Those are in accordance with data found on literature. Relationship among molecules and different pathways for different organisms found by molecular networking allows a more holistic view of bioproducts produced by strain AC13. Similar molecules for quorum sensing signaling systems and Amino-acid biosynthesis pathways can be connected and indicates how intricate the cellular environment is.

ACKNOWLEDGEMENT

The authors thank to University of California San Diego and all Dorrestein Lab team for providing support on the development of this research. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001; Project 88882.182971/2018-01 and by a grant from Conselho Nacional de Desenvolvimento Científico e Tecnológico – (CNPq).

Supplementary Figure 1. Leucine biosynthetic pathway is responsible to catalyze the oxidation of 3- carboxy-2-hydroxy-4-methylpentanoate (3-isopropylmalate) to 3-carboxy-4-methyl-2-oxopentanoate. The product formed decarboxylates to 4-methyl-2 oxopentanoate.

REFERENCES

ALASIL, S. M. et al. Inhibition of Quorum Sensing-Controlled Virulence Factors and Biofilm Formation in Pseudomonas aeruginosa by Culture Extract from Novel Bacterial Species of Paenibacillus Using a Rat Model of Chronic Lung Infection. Int J Bacteriol, v. 2015, p. 671562, 2015. ISSN 2356-6957 (Print) 2314-596X (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/26904749 >.

ANYAN, M. E. et al. Type IV pili interactions promote intercellular association and moderate swarming of Pseudomonas aeruginosa. Proc Natl Acad Sci U S A, v. 111, n. 50, p. 64

18013-8, Dec 16 2014. ISSN 1091-6490 (Electronic) 0027-8424 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/25468980 >.

BAINDARA, P. et al. Characterization of the Antimicrobial Peptide Penisin, a Class Ia Novel Lantibiotic from Paenibacillus sp. Strain A3. Antimicrobial Agents Chemotherapy, v. 60, n. 1, p. 580-591, 2015.

BAYM, M. et al. Spatiotemporal microbial evolution on antibiotic landscapes. Science, v. 353, n. 6304, p. 1147-1151, 2016. ISSN 0036-8075.

BE'ER, A. et al. Periodic reversals in Paenibacillus dendritiformis swarming. J Bacteriol, v. 195, n. 12, p. 2709-17, Jun 2013. ISSN 1098-5530 (Electronic)0021-9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23603739 >.

BELAS, R. Biofilms, flagella, and mechanosensing of surfaces by bacteria. Trends Microbiol, v. 22, n. 9, p. 517-27, Sep 2014. ISSN 1878-4380 (Electronic) 0966-842X (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24894628 >.

BELAS, R.; SCHNEIDER, R.; MELCH, M. Characterization of Proteus mirabilis precocious swarming mutants: identification of rsbA, encoding a regulator of swarming behavior. J Bacteriol, v. 180, n. 23, p. 6126-39, Dec 1998. ISSN 0021-9193 (Print) 0021-9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/9829920 >.

BIRD, A. DNA methylation patterns and epigenetic memory. Genes & development, v. 16, n. 1, p. 6-21, 2002. ISSN 0890-9369.

CANOVA, S. P. et al. Characterization of lipopeptides from Paenibacillus sp. (IIRAC30) suppressing Rhizoctonia solani. World Journal of Microbiology and Biotechnology, v. 26, p. 2241-2247, 2010.

COSTERTON, J. W. Introduction to biofilm. International journal of antimicrobial agents, v. 11, n. 3-4, p. 217-221, 1999. ISSN 0924-8579.

DAVIES, D. G. et al. The involvement of cell-to-cell signals in the development of a bacterial biofilm. Science, v. 280, n. 5361, p. 295-298, 1998. ISSN 0036-8075.

DING, R. et al. Isolation and Identification of Lipopeptide Antibiotics from Paenibacillus elgii B69 with Inhibitory Activity Against Methicillin-Resistant Staphylococcus aureus. The Journal of Microbiology, v. 49, n. 6, p. 942-949, 2011.

FLEMMING, H. C.; WINGENDER, J. The biofilm matrix. Nat Rev Microbiol, v. 8, n. 9, p. 623-33, Sep 2010. ISSN 1740-1534 (Electronic) 1740-1526 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/20676145 >.

FLEITAS MARTÍNEZ, O. et al. Interference With Quorum-Sensing Signal Biosynthesis as a Promising Therapeutic Strategy Against Multidrug-Resistant Pathogens. Frontiers in Cellular and Infection Microbiology, v. 8, n. 444, 2019-February-05 2019. ISSN 2235-2988. Available at: < https://www.frontiersin.org/article/10.3389/fcimb.2018.00444 >

GHIO, S. et al. Isolation of Paenibacillus sp. and Variovorax sp. strains from decaying woods and characterization of their potential for cellulose deconstruction. Int J Biochem Mol Biol, v. 3, n. 4, p. 352-64, 2012. ISSN 2152-4114 (Print) 2152-4114 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23301200 >.

GUDINA, E. J. et al. Novel bioemulsifier produced by a Paenibacillus strain isolated from crude oil. Microb Cell Fact, v. 14, p. 14, Jan 31 2015. ISSN 1475-2859 (Electronic) 1475-2859 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/25636532 >.

GUO, Y. et al. Isolation of a Paenibacillus sp. Strain and Structural Elucidation of Its Broad-Spectrum Lipopeptide Antibiotic. Applied and Environmental Microbiology, v. 78, n. 9, p. 3156-3165, 2012.

HE, Z. et al. Isolation and Identification of a Paenibacillus polymyxa Strain That Coproduces a Novel Lantibiotic and Polymyxin. Applied and environmental microbiology, v. 73, n. 1, p. 168-178, 2006.

HUANG, E. et al. New Paenibacillus strain produces a family of linear and cyclic antimicrobial lipopeptides: Cyclization is not essential for their antimicrobial activity. FEMS Microbiol Lett, Feb 25 2017. ISSN 1574-6968 (Electronic) 0378-1097 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/28333237 >.

INES, M.; DHOUHA, G. Lipopeptide surfactants: Production, recovery and pore forming capacity. Peptides, v. 71, p. 100-12, Sep 2015. ISSN 1873-5169 (Electronic) 0196- 9781 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/26189973 >.

INGHAM, C. J.; BEN JACOB, E. Swarming and complex pattern formation in Paenibacillus vortex studied by imaging and tracking cells. BMC Microbiol, v. 8, p. 36, Feb 25 2008. ISSN 1471-2180 (Electronic) 1471-2180 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/18298829 >.

KALIA, V. C. Quorum sensing inhibitors: an overview. Biotechnol Adv, v. 31, n. 2, p. 224-45, Mar-Apr 2013. ISSN 1873-1899 (Electronic) 0734-9750 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23142623 >.

KANEDA, T. Fatty acids in the genus Bacillus. I. Iso- and anteiso-fatty acids as characteristic constituents of lipids in 10 species. J Bacteriol, v. 93, n. 3, p. 894-903, Mar 1967. ISSN 0021-9193 (Print) 0021-9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/4960925 >.

______. Iso- and anteiso-fatty acids in bacteria: biosynthesis, function, and taxonomic significance. Microbiol Rev, v. 55, n. 2, p. 288-302, Jun 1991. ISSN 0146-0749 (Print) 0146- 0749 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/1886522 >.

KAUFMANN, G. F. et al. Antibody interference with N-acyl homoserine lactone- mediated bacterial quorum sensing. J Am Chem Soc, v. 128, n. 9, p. 2802-3, Mar 08 2006. ISSN 0002-7863 (Print) 0002-7863 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/16506750 >.

KEARNS, D. B. A field guide to bacterial swarming motility. Nat Rev Microbiol, v. 8, n. 9, p. 634-44, Sep 2010. ISSN 1740-1534 (Electronic) 1740-1526 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/20694026 >.

KIM, J. et al. Isolation and structural elucidation of pelgipeptin E, a novel pore- forming pelgipeptin analog from Paenibacillus elgii with low hemolytic activity. The Journal 67 of Antibiotics, 2018/08/22 2018. ISSN 1881-1469. Available at: < https://doi.org/10.1038/s41429-018-0095-2 >.

KO, C. H. et al. Paenibacillus campinasensis BL11: a wood material-utilizing bacterial strain isolated from black liquor. Bioresour Technol, v. 98, n. 14, p. 2727-33, Oct 2007. ISSN 0960-8524 (Print) 0960-8524 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/17123815 >.

LEE, J. et al. A cell-cell communication signal integrates quorum sensing and stress response. Nat Chem Biol, v. 9, n. 5, p. 339-43, May 2013. ISSN 1552-4469 (Electronic) 1552-4450 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23542643 >.

LIANG, T. W. et al. Exopolysaccharides and antimicrobial biosurfactants produced by Paenibacillus macerans TKU029. Appl Biochem Biotechnol, v. 172, n. 2, p. 933-50, Jan 2014. ISSN 1559-0291 (Electronic) 0273-2289 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24122708 >.

LOHANS, C. et al. Structural characterization of the highly cyclized lantibiotic paenicidin A via a partial desulfurization/reduction strategy. J Am Chem Soc., v. 134, n. 48, p. 19540-3, 2012.

MAHDI, O. S.; FISHER, N. A. Sporulation and Germination of Paenibacillus larvae Cells. Current Protocols in Microbiology, v. 48, n. 1, p. 9E.2.1-9E.2.10, 2018. Available at: < https://currentprotocols.onlinelibrary.wiley.com/doi/abs/10.1002/cpmc.46 >.

MANDIC-MULEC, I.; STEFANIC, P.; VAN ELSAS, J. D. Ecology of Bacillaceae. In: (Ed.). The Bacterial Spore: from Molecules to Systems: American Society of Microbiology, 2016.

MARAHIEL, M. A. Working outside the protein‐synthesis rules: insights into non‐ ribosomal peptide synthesis. Journal of peptide science: an official publication of the European Peptide Society, v. 15, n. 12, p. 799-807, 2009. ISSN 1075-2617.

MARAHIEL, M. A.; STACHELHAUS, T.; MOOTZ, H. D. Modular peptide synthetases involved in nonribosomal peptide synthesis. Chemical reviews, v. 97, n. 7, p. 2651-2674, 1997. ISSN 0009-2665.

MATHEWS, S. L.; PAWLAK, J. J.; GRUNDEN, A. M. Isolation of Paenibacillus glucanolyticus from pulp mill sources with potential to deconstruct pulping waste. Bioresour Technol, v. 164, p. 100-5, Jul 2014. ISSN 1873-2976 (Electronic) 0960-8524 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24841577 >.

MAUNDERS, E.; WELCH, M. Matrix exopolysaccharides; The sticky side of biofilm formation. FEMS Microbiol Lett, Jun 12 2017. ISSN 1574-6968 (Electronic) 0378-1097 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/28605431 >.

MORRISSEY, B. J. et al. Biogeography of Paenibacillus larvae, the causative agent of American foulbrood, using a new multilocus sequence typing scheme. Environ Microbiol, v. 17, n. 4, p. 1414-24, Apr 2015. ISSN 1462-2920 (Electronic) 1462-2912 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/25244044 >.

NG, J. et al. Dereplication and de novo sequencing of nonribosomal peptides. Nat Methods, v. 6, n. 8, p. 596-9, Aug 2009. ISSN 1548-7105 (Electronic) 1548-7091 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/19597502 >.

NIU, B. et al. Polymyxin P is the active principle in suppressing phytopathogenic Erwinia spp. by the biocontrol rhizobacterium Paenibacillus polymyxa M-1. BMC Microbiology, v. 13, n. 137, p. 1-13, 2013.

NJOROGE, J.; SPERANDIO, V. Jamming bacterial communication: new approaches for the treatment of infectious diseases. EMBO Mol Med, v. 1, n. 4, p. 201-10, Jul 2009. ISSN 1757-4684 (Electronic) 1757-4676 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/20049722 >.

POEPPEL, D.; EMBICK, D. Defining the relation between linguistics and neuroscience. In: (Ed.). Twenty-first century psycholinguistics: Routledge, 2017. p.103-118.

QIAN, C. et al. Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiology, v. 12, n. 197, p. 1-7, 2012. Available at: < http://www.biomedcentral.com/1471-2180/12/197 >. 69

QIAN, C.-D. et al. Battacin (Octapeptin B5), a new cyclic lipopeptide antibiotic from Paenibacillus tianmuensis active against multidrug-resistant Gram-negative bacteria. Antimicrobial agents and chemotherapy, v. 56, n. 3, p. 1458-65, Mar 2011. ISSN 1098-6596 (Electronic) 0066-4804 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/22183171 >.

QIAN, C. D. et al. Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiology, v. 12, n. 197, p. 1-7, 2012. Available at: < http://www.biomedcentral.com/1471-2180/12/197 >.

ROMERO, D. et al. Antibiotics as signal molecules. Chemical reviews, v. 111, n. 9, p. 5492-505, Sep 14 2011. ISSN 1520-6890 (Electronic) 0009-2665 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/21786783 >.Available at: < http://pubs.acs.org/doi/abs/10.1021/cr2000509 >.

SCHAFER, M. O. et al. Rapid identification of differentially virulent genotypes of Paenibacillus larvae, the causative organism of American foulbrood of honey bees, by whole cell MALDI-TOF mass spectrometry. Vet Microbiol, v. 170, n. 3-4, p. 291-7, Jun 04 2014. ISSN 1873-2542 (Electronic) 0378-1135 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/24613082 >.

SELIM, S. et al. Isolation and Partial Characterization of Antagonistic Peptides Produced by Paenibacillus sp. Strain B2 Isolated from the Sorghum Mycorrhizosphere. APPLIED AND ENVIRONMENTAL MICROBIOLOGY, v. 71, n. 11, p. 6501–6507, 2005.

SMITH, R.; IGLEWSKI, B. H. P. aeruginosa quorum-sensing systems and virulence. Current Opinion in Microbiology, v. 6, n. 1, p. 56-60, 2003. ISSN 13695274.

SUGAWARA, K.; KONISHI, M.; KAWAGUCHI, H. BMY-28160, a new peptide antibiotic. The Journal of antibiotics, v. 37, n. 10, p. 1257-1259, 1984. ISSN 0021-8820.

TAKEUCHI, Y. et al. The structure of permetin A, a new polypeptin type antibiotic produced by Bacillus circulans. The Journal of antibiotics, v. 32, n. 2, p. 121-129, 1979. ISSN 0021-8820.

TRAXLER, M. F. et al. Interspecies interactions stimulate diversification of the Streptomyces coelicolor secreted metabolome. MBio, v. 4, n. 4, Aug 20 2013. ISSN 2150- 7511 (Electronic). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23963177 >.

VAN GESTEL, J.; VLAMAKIS, H.; KOLTER, R. From cell differentiation to cell collectives: Bacillus subtilis uses division of labor to migrate. PLoS biology, v. 13, n. 4, p. e1002141, 2015. ISSN 1545-7885.

VAN HOUDT, R. et al. Characterization of culturable Paenibacillus spp. from the snow surface on the high Antarctic Plateau (DOME C) and their dissemination in the Concordia research station. Extremophiles, v. 17, n. 4, p. 565-73, Jul 2013. ISSN 1433-4909 (Electronic) 1431-0651 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23588647 >.

WANG, M. et al. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nature biotechnology, v. 34, n. 8, p. 828, 2016. ISSN 1546-1696.

WATERS, C. M.; BASSLER, B. L. Quorum sensing: cell-to-cell communication in bacteria. Annu Rev Cell Dev Biol, v. 21, p. 319-46, 2005. ISSN 1081-0706 (Print) 1081- 0706 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/16212498 >.

WATNICK, P.; KOLTER, R. Biofilm, city of microbes. J Bacteriol, v. 182, n. 10, p. 2675-9, May 2000. ISSN 0021-9193 (Print) 0021-9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/10781532 >.

WATNICK, P. I.; KOLTER, R. Steps in the development of a Vibrio cholerae El Tor biofilm. Mol Microbiol, v. 34, n. 3, p. 586-95, Nov 1999. ISSN 0950-382X (Print) 0950- 382X (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/10564499 >.

WATROUS, J. et al. Mass spectral molecular networking of living microbial colonies. Proceedings of the National Academy of Sciences, v. 109, n. 26, p. E1743-E1752, 2012. ISSN 0027-8424.

WEN, Y. et al. Identification and analysis of the gene cluster involved in biosynthesis of paenibactin, a catecholate siderophore produced by Paenibacillus elgii B69. Environmental Microbiology, v. 13, n. 10, p. 2726-2737, 2011.

YANG, J. Y. et al. Molecular networking as a dereplication strategy. Journal of natural products, v. 76, n. 9, p. 1686-1699, 2013. ISSN 0163-3864.

Discussion

Paenibacillus elgii AC13, chemical signaling and quorum sensing inhibition The genus Paenibacillus was defined in 1993 after extensive comparison between DNA sequences encoding the 16s ribosomal subunit (Ash et al., 1993). Since then, different species of this genus have been described as biological control agents. Paenibacillus species are able to form highly resistant spores and are producers of several lipopeptides with antimicrobial activity and molecules with surfactant properties (Wu et al., 2010; Cochrane and Vederas, 2016; Grady et al., 2016). Cyclic or linear peptides usually bear a variable fatty acid in their N-terminal region. Within Paenibacillus genus, these molecules are considered secondary metabolites produced by NRPS and usually contain a mixture of D and L amino acids and non-proteinogenic amino acids (Cochrane and Vederas, 2016). Non-proteinogenic amino acids (NPAAs) are those that do not occur naturally in the genetic code of an organism, however, they are of biological importance since many are used as metabolic intermediates or are found in proteins with posttranslational modifications. Many NPAAs are synthesized by plants and, in many cases, are part of plants’ chemical defense arsenal against pathogens and predators. NPAAs can also be helpful since they give advantage over other plant species that compete for the same resources (Rodgers, 2014). Thus, NPAAs are of great biotechnological relevance due to their agricultural and pharmacological potential. In 2010, a new strain of Paenibacillus sp. was isolated from soil samples in China and, after physicochemical characterization and genetic sequencing; it was identified as P. elgii strain B69, the holotype for the species of microorganism. Two lipopeptides with antimicrobial activity, named Pelgipeptins A and B were also identified. The molecular masses of Pelgipeptin A and B are respectively 1072 and 1100 Da and their sequences are: Dab-Val-Leu / Ile-X1-Dab-Val-Dab- Phe-Leu / Ile (Pelgipeptin A) and Dab-Val-Leu / Ile-X2- Dab-Leu / Ile-Dab-Phe-Leu / Ile (Pelgipeptin B), X indicates an substitution position and the ambiguity concerning identification Leu / Ile still remains (Wu et al., 2010). "Dab" is the non- proteinogenic amino acid, 2, 4-diaminobutyric acid and its presence in molecules with antimicrobial potential is consistent with the large ecological role of Paenibacillus species, since this genus is associated with plant roots in a symbiotic relationship. Interestingly, part of the compounds synthesized by Paenibacillus can directly influence the growth of plants through the production of indole-3-acetic acid and other phytohormones. Besides Paenibacillus can solubilize phosphates and make it available for absorption (Grady et al., 2016).

Strain AC13 is able to form highly viscous and difficult to dissociate biofilms. The interaction between a host and a pathogenic bacterium is mostly controlled by the size of the microorganism population. Since a bacterial cell is able to "perceive" other members of the same species and respond according to population density, molecules that interfere in QS communication are being considered as promising agents in the search for new strategic routes of prophylactic measures or infections. In addition, antibodies and receptors that act as a decoy of QS communication signaling molecules are also suggested as possible approaches for anti-infective therapies (Kaufmann et al., 2006). The QS communication can be interrupted in different ways, either by reducing the activity of the AHL receptor protein or the AHLs synthetases, which leads to the inhibition of the QS signaling molecules synthesis, or by synthetic molecules analogous to the QS signaling compounds. Another possibility to interrupt QS is by the degradation of AHLs (Kaufmann et al., 2006; Kalia, 2013). Bacillus subtilis cultures have shown that biofilms are composed of different groups of cells and that these cell groups remain differentiated over time (Romero et al., 2011). Interestingly, another behavior crucial to the evolutionary success of several microorganisms depends on the morphological modification of a portion of the population of microorganisms is the ability to swarm. Pseudomonas aeruginosa is a species of Gram-negative bacillus that exhibits motility and is highly versatile. Found in soils, fresh or salt water, it is able to tolerate low concentrations of oxygen, grow at the most varied temperatures (from 4 to 42 ° C) and survive with minimal amounts of nutrients. Due to its ability to stick and survive in medical equipment, it is considered an opportunistic pathogen of public health importance due to its potential to promote several infections mainly in immunosuppressed patients, burn victims or neonates (Kaufmann et al., 2006). P. aeruginosa uses a hierarchical quorum-sensing communication network to coordinate the expression of virulence genes; however, clinical isolates have demonstrated a loss of function of some gene clusters responsible for this behavior. Recent findings indicate that disruption in certain QS signaling pathways lead to attenuation of virulence expression in P. aeruginosa, which has led researchers to better understand the complex QS signaling system (Smith and Iglewski, 2003; Njoroge and Sperandio, 2009; Lee et al., 2013). Besides, the odds of selecting resistant strains by targeting QS signaling network are weaker than using conventional antibiotics (Fleitas Martínez et al., 2019).

Conclusions

NRPS were already identified from strain AC13 genome; some of them are probably part of the Pelgipeptin synthase gene cluster, a group of lipopeptides with high biotechnological potential. Lipopeptides from the Paenibacillus spp. present in its structure both, D and L amino acids, resulting in higher stability over proteolytic enzymes and increasing the number of possible templates. Growth parameters, pelgipeptins quantification, and partial characterization of lipopeptides produced by P. elgii strain AC13 suggests that strain AC13 is able to modulate its lipopeptides production pattern based on variations of its environment or initial population, however, a more detailed analysis is indicated. A cluster found within strain AC13 collection is connected to an enzyme responsible to synthesize O-acetyl-L-homoserine from L-homoserine. Among the QS signaling pathways, N-acyl-homoserine lactones are present as chemical signaling molecules for more than 70 different species of diderm bacteria. Furthermore, NRPs pathways are part of second metabolism. Leucine biosynthetic pathway is part of the amino-acid biosynthesis, crucial to bacterial development. Lactone ring within Pelgipeptins synthesized in nutrient broth are interrupted. Quorum sensing lactone ring binding proteins, however, were not yet found within strain AC13 proteomic collection. Thus, further analysis involving surfactants variations and motility expression or biofilm formation are encouraged.

Figure 4 Thesis’ representation. Based on results found, Analogues of Pelgipeptin/Permeatin A may be needed to triggers a social response. A myriad of compounds and signaling molecules synthetized can be recycled to the first metabolism to the synthesis of enzymes involved in amino acid synthesis pathway (e.g. 3- isopropylmalate dehydrogenase). Those enzymes are needed to catalyze reactions to build proteins analogues to agr B. An acetylation step might be responsible to add a lactone ring modification to the agr B protein analogues; activating the quorum sensing signaling process for surfactants release involved in social response like swarming behavior.

Perspectives

It is remarkable how data analysis processes have been changing the way biological sciences are perceived. In addition, new discoveries regarding how non-traditional proteins’ pathway synthesis, epigenetic, and cell-to-cell signaling processes occurs have been giving more information over biological complex systems (Marahiel et al., 1997; Davies et al., 1998; Bird, 2002; Waters and Bassler, 2005; Marahiel, 2009). However, despite the new tools available, traditional methods like applied microbiology and theoretical knowledge like ecology and evolution are still needed in order to create, test, and to support insights within Biology, Biochemistry, Proteomics, Metabolomics and Genomic Sciences (Traxler et al., 2013; Van Gestel et al., 2015; Baym et al., 2016). Those insights in science will, inexorably, lead Biotechnology into new, practical and sustainable discoveries. A currently mismatch among experimental results, database processing rates, and dereplication of previously acquired data has been a common issue in order to organize, interpret and give proper explanations to complex questions that lies ahead Neurosciences, System Biology and Biotechnology itself (Ng et al., 2009; Yang et al., 2013; Poeppel and Embick, 2017). Moreover, in order to match experimental results, theoretical knowledge and previously acquired data; information needs to be public, easily accessed and clear. As a step forward on strain AC13 investigation, it is suggest that morphological alterations within a bacterial strain might be identified by specific metabolomics profile specially related to the presence or absence of secondary metabolites, lipopeptides and/or communication factors, guiding research over that matter. Furthermore, a more detailed research involving the relationship between specific amino acids (e.g. leucine) may be useful to better understand and address key points on quorum sensing expressed features. P. elgii AC13 molecular networking is available at GNPS, what can bring more insights over this strain ecological and metabolism features.

References

ANYAN, M. E. et al. Type IV pili interactions promote intercellular association and moderate swarming of Pseudomonas aeruginosa. Proc Natl Acad Sci U S A, v. 111, n. 50, p. 18013-8, Dec 16 2014. ISSN 1091-6490 (Electronic) 0027-8424 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/25468980 >.

BASSLER, B. L.; LOSICK, R. Bacterially speaking. Cell, v. 125, n. 2, p. 237-46, Apr 21 2006. ISSN 0092-8674 (Print) 0092-8674 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/16630813 >.

BELAS, R.; SCHNEIDER, R.; MELCH, M. Characterization of Proteus mirabilis precocious swarming mutants: identification of rsbA, encoding a regulator of swarming behavior. J Bacteriol, v. 180, n. 23, p. 6126-39, Dec 1998. ISSN 0021-9193 (Print) 0021- 9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/9829920 >.

BINIARZ, P.; ŁUKASZEWICZ, M.; JANEK, T. Screening concepts, characterization and structural analysis of microbial-derived bioactive lipopeptides: a review. Critical Reviews in Biotechnology, v. 37, n. 3, p. 393-410, 2017/04/03 2017. ISSN 0738-8551. Available at: < https://doi.org/10.3109/07388551.2016.1163324 >.

BODELON, G. et al. Detection and imaging of quorum sensing in Pseudomonas aeruginosa biofilm communities by surface-enhanced resonance Raman scattering. Nat Mater, v. 15, n. 11, p. 1203-1211, Nov 2016. ISSN 1476-1122 (Print) 1476-1122 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/27500808 >.

CAKAR, A. [Bacterial communication: quorum-sensing]. Mikrobiyol Bul, v. 38, n. 3, p. 273-84, Jul 2004. ISSN 0374-9096 (Print) 0374-9096 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/15490849 >.

CANOVA, S. P. et al. Characterization of lipopeptides from Paenibacillus sp. (IIRAC30) suppressing Rhizoctonia solani. World Journal of Microbiology and Biotechnology, v. 26, p. 2241-2247, 2010.

DELEON, S. D.; WELLIVER, R. C., SR. Paenibacillus alvei Sepsis in a Neonate. Pediatr Infect Dis J, v. 35, n. 3, p. 358, Mar 2016. ISSN 1532-0987 (Electronic) 0891-3668 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/26866854 >.

DENG, Q. et al. A sensitive method for simultaneous quantitative determination of surfactin and iturin by LC-MS/MS. Analytical and bioanalytical chemistry, v. 409, n. 1, p. 179-191, 2017. ISSN 1618-2642.

DING, R. et al. Draft Genome Sequence of Paenibacillus elgii B69, a Strain with Broad Antimicrobial Activity. Journal of Bacteriology, v. 193, n. 17, p. 4537, 2011. Available at: < http://jb.asm.org/content/193/17/4537.abstract >.

______. Isolation and Identification of Lipopeptide Antibiotics from Paenibacillus elgii B69 with Inhibitory Activity Against Methicillin-Resistant Staphylococcus aureus. The Journal of Microbiology, v. 49, n. 6, p. 942-949, 2011.

EBERHARD, A. Inhibition and activation of bacterial luciferase synthesis. J Bacteriol, v. 109, n. 3, p. 1101-5, Mar 1972. ISSN 0021-9193 (Print) 0021-9193 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/5011244 >.

FISCHBACH, M. A.; WALSH, C. T. Assembly-line enzymology for polyketide and nonribosomal Peptide antibiotics: logic, machinery, and mechanisms. Chem Rev, v. 106, n. 8, p. 3468-96, Aug 2006. ISSN 0009-2665 (Print) 0009-2665 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/16895337 >.

GRAY, B.; HALL, P.; GRESHAM, H. Targeting agr- and agr-Like quorum sensing systems for development of common therapeutics to treat multiple gram-positive bacterial infections. Sensors (Basel), v. 13, n. 4, p. 5130-66, Apr 18 2013. ISSN 1424-8220 (Electronic) 1424-8220 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23598501 >.

GREENBERG, E. P. Bacterial communication and group behavior. J Clin Invest, v. 112, n. 9, p. 1288-90, Nov 2003a. ISSN 0021-9738 (Print) 0021-9738 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/14597752 >.

______. Bacterial communication: tiny teamwork. Nature, v. 424, n. 6945, p. 134, Jul 10 2003b. ISSN 1476-4687 (Electronic) 0028-0836 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/12853935 >.

HAWVER, L. A.; JUNG, S. A.; NG, W. L. Specificity and complexity in bacterial quorum-sensing systems. FEMS Microbiol Rev, v. 40, n. 5, p. 738-52, Sep 2016. ISSN 1574-6976 (Electronic) 0168-6445 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/27354348 >.

HE, Z. et al. Isolation and Identification of a Paenibacillus polymyxa Strain That Coproduces a Novel Lantibiotic and Polymyxin. Applied and environmental microbiology, v. 73, n. 1, p. 168-178, 2006.

HUANG, E. et al. New Paenibacillus strain produces a family of linear and cyclic antimicrobial lipopeptides: Cyclization is not essential for their antimicrobial activity. FEMS

Microbiol Lett, Feb 25 2017. ISSN 1574-6968 (Electronic) 0378-1097 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/28333237 >.

KIM, J. et al. Isolation and structural elucidation of pelgipeptin E, a novel pore- forming pelgipeptin analog from Paenibacillus elgii with low hemolytic activity. The Journal of Antibiotics, 2018/08/22 2018. ISSN 1881-1469. Available at: < https://doi.org/10.1038/s41429-018-0095-2 >.

KUROKAWA, A. S. Exploração bioecnológica de microrganismos do solo de Cerrado através da construção de bibliotecas metagenômicas e técnicas de cultivo. 2006. 249p. (Dissertação de Mestrado). Pós Graduação em Ciências Genômicas e Biotecnologia, Universidade Católica de Brasília - UCB, Brasília.

LEE, J. et al. A cell-cell communication signal integrates quorum sensing and stress response. Nat Chem Biol, v. 9, n. 5, p. 339-43, May 2013. ISSN 1552-4469 (Electronic)1552-4450 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/23542643 >.

LOHANS, C. et al. Structural characterization of the highly cyclized lantibiotic paenicidin A via a partial desulfurization/reduction strategy. J Am Chem Soc., v. 134, n. 48, p. 19540-3, 2012.

MACFADYEN, A. The Meaning of Productivity in Biological Systems. Journal of Animal Ecology, v. 17, n. 1, p. 75-80, 1948. ISSN 00218790, 13652656. Available at: < http://www.jstor.org/stable/1611 >.

NICHOLSON, W. Sporulation, germination, and outgrowth. Molecular biological methods for Bacillus, 1990.

O'BRIEN, J.; WRIGHT, G. D. An ecological perspective of microbial secondary metabolism. Curr Opin Biotechnol, v. 22, n. 4, p. 552-8, Aug 2011. ISSN 1879-0429 (Electronic) 0958-1669 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/21498065 >.

O'TOOLE, G. A.; KOLTER, R. Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol, v. 30, n. 2, p. 295-304, Oct 1998. ISSN 0950-382X (Print) 0950-382X (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/9791175 >.

PAPENFORT, K.; BASSLER, B. L. Quorum sensing signal-response systems in Gram-negative bacteria. Nat Rev Microbiol, v. 14, n. 9, p. 576-88, Aug 11 2016. ISSN 1740- 1534 (Electronic) 1740-1526 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/27510864 >.

QIAN, C.-D. et al. Battacin (Octapeptin B5), a new cyclic lipopeptide antibiotic from Paenibacillus tianmuensis active against multidrug-resistant Gram-negative bacteria. Antimicrobial agents and chemotherapy, v. 56, n. 3, p. 1458-65, Mar 2011. ISSN 1098- 6596 (Electronic) 0066-4804 (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/22183171 >.

SMITH, R.; IGLEWSKI, B. H. P. aeruginosa quorum-sensing systems and virulence. Current Opinion in Microbiology, v. 6, n. 1, p. 56-60, 2003. ISSN 13695274.

SUGAWARA, K.; KONISHI, M.; KAWAGUCHI, H. BMY-28160, a new peptide antibiotic. The Journal of antibiotics, v. 37, n. 10, p. 1257-1259, 1984. ISSN 0021-8820.

TAKEUCHI, Y. et al. The structure of permetin A, a new polypeptin type antibiotic produced by Bacillus circulans. The Journal of antibiotics, v. 32, n. 2, p. 121-129, 1979. ISSN 0021-8820.

WATNICK, P. I.; KOLTER, R. Steps in the development of a Vibrio cholerae El Tor biofilm. Mol Microbiol, v. 34, n. 3, p. 586-95, Nov 1999. ISSN 0950-382X (Print) 0950-382X (Linking). Available at: < http://www.ncbi.nlm.nih.gov/pubmed/10564499 >.

Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae (Appendix 1)

The content of this appendix is a manuscript entitled “Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae” published in Biotechnology Research and Innovation Available online on January 16th 2019.

Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae

R. A. da Costaa, D. B. Ortegaa, D. L. A. Fulgêncioa, F. S. Costaa, T. F. de Araújoa; C. C. Barretoa*. a Universidade Católica de Brasília, Graduate Program in Genomic Sciences and Biotechnology, SGAN 916, Brasília – DF 70790-160, Brazil. * Corresponding author: Cristine Chaves Barreto. Phone: +55-61-3448-9217. Fax: +55-61-3347-4797 / e-mail: [email protected] Universidade Católica de Brasília, Programa de Ciências Genômicas e Biotecnologia. Brasília - Brazil. 70790-160.

Abstract Multidrug resistant bacteria infections have motivated the search for new therapeutics. The synergic effect of conventional antibiotics and lipopeptides of the pelgipeptin family was evaluated by the Checkerboard method. Results indicate that the association of Pelgipeptin B and C or chloramphenicol has a synergic effect against a multidrug resistant bacteria strain; which may be due to variations on hydrophobicity and mechanisms of action of the molecules tested. Keywords: Paenibacillus, pelgipeptin, synergism, multidrug resistance; checkerboard.

Introduction The emergence of multidrug-resistant (MDR) bacteria has reduced the efficacy of conventional antibiotics1-3. The rapid dissemination of Gram-negative (diderm) bacteria, such as Klebsiella pneumoniae and Escherichia coli, is responsible for an emerging crisis due to their carbapenemase coding gene that can be carried by plasmids4, 5. In addition, infections caused by β-lactam resistant Gram-positive (monoderm) bacteria (e.g. Staphylococcus aureus) have increased in incidence5-8. Drug combinations that present synergic effects

91 emerge as a therapeutic option aiming to reduce the drug administration if compared to classical single antibiotic treatments2, 8, and 9. Synergy testing methods uses susceptibility techniques to measure a cumulative efficacy of two or more drugs in which the resulted activity occurs at a lower concentration than the sum of each of them isolated1. Synergic results may indicate an increase in efficacy and/or reduced toxicity and side effects due to lower therapeutic dosage administration10. The Paenibacillus genus is known for producing antimicrobial compounds with broad spectrum activity such as bacteriocins11-13 and non-ribosomal lipopeptides14-21. Pelgipeptin is a group of cyclic cationic lipopeptides that contain the non-proteinogenic amino acid 2, 4- diaminobutyric acid (Dab) on its N-terminal position linked to a fatty acid chain21-23. Pelgipeptin presents antimicrobial activity against monoderm and diderm bacteria21, 23. Major variations among Pelgipeptin A-E isoforms include a substitution on the second positioned amino acid and on the fatty acid length23-25. Pelgipeptins A and D present a methyl group linked on its penultimate (iso) position of its lipid tail whereas Pelgipeptins B and C show the methyl group on the antepenultimate (anteiso-) position of its lipid tail22,24,26,27. Pelgipeptin E shows same stereochemistry of amino acid and ester bond of Pelgipeptin A and Pelgipeptin C but attached to a straight lipid chain25. Although the Minimal Inhibitory Concentrations (MICs) of each Pelgipeptin isoforms have been previously studied1, 2, here we investigated the antimicrobial effects of combined isoforms, aiming to describe synergistic effects among them.

Materials and Methods Pelgipeptins were obtained from Paenibacillus elgii strain AC13, isolated from Brazilian Cerrado soil samples28, grown on chemically defined medium MMP (Patent BR102017018881-7). Purification was attained by high performance liquid chromatography (HPLC; Shimadzu) using a reverse phase column Shimadzu Shim-pack VP-ODS (4.6 µm, 150 x 4.6 mm). The mobile phase consisted by HPLC-grade water containing 0.1 % of TFA (mobile phase A) and acetonitrile with 0.1 % of TFA (mobile phase B). The molecular masses of the lipopeptides were confirmed by MALDI-ToF (Autoflex Speed; Bruker Daltonics) on reflected-positive mode (700 to 3500 m/z). The minimal inhibitory concentrations (MIC) for Pelgipeptins isoforms A-D were evaluated against Staphylococcus aureus ATCC 14458, Escherichia coli ATCC 11229, Klebsiella pneumoniae ATCC 13883; and two MDR strains obtained from the Central

Laboratory of Brasilia (LACEN): K. pneumoniae LACEN3259271 and E. coli LACEN 3789319. Both strains were isolated from hospitalized patients’ blood samples in Brasilia, Brazil and all tested strains were used in a suspension concentration of 5×105 UFC/mL. The Checkerboard testing method was used to evaluate synergism among Pelgipeptins A, B and C or combined with Chloramphenicol or Penicillin against the MDR K pneumoniae (LACEN3259271). The Checkerboard method uses the combination of two compounds in increasing concentrations to provide a final classification of the combined compounds based on a Fractional Inhibitory Concentration (FIC) Index (FICI) 1,29: S, synergy (FIC ≤ 0,5); A, additive (FIC > 0.50 and < 1); I, indifferent (FIC >1 and ≤ 4). Reference antibiotics were used as controls. Interpretation of sensitivity towards an antimicrobial compound was based on values found for Polymyxin B (≤ 8µg·mL-1), a pure commercial antibiotic first isolated from Paenibacillus polymyxa and used to fight infections caused by tested strains.

Results and Discussion All pelgipeptins were active even at lower concentrations, indicating potential sensitivity of tested microorganisms. Pelgipeptin A showed lower MICs against both K. pneumoniae strains whereas pelgipeptins C and D were the most active against multidrug- resistant E. coli. MICs for Pelgipeptins A; C and D were 8 µg·mL1 whereas Pelgipeptin B showed MICs varying from 64 to 16 µg·mL1 (Table 1). Synergism among Pelgipeptins against two Klebsiella pneumoniae strains, by means of the Checkerboard testing method1, 29. Pelgipeptin D purification yield was the lowest and that isoform was not used to perform the synergy test. The combination Chloramphenicol + Pelgipeptin A and Pelgipeptins A + B resulted in indifferent effect over the growth inhibition of K. pneumoniae ATCC 13883 (FIC = 1.5 and 1.125, respectively). However, FIC values lower than 0.5 obtained for Pelgipeptin A + C; B + C; Penicillin + Chloramphenicol; and Pelgipeptin C + Penicillin or Chloramphenicol indicates synergism. When combined molecules were tested against MDR strain LACEN 3259271 the results were slightly different. Chloramphenicol + Pelgipeptin A, B or C; and Pelgipeptin C + Penicillin showed indifferent result (FIC = 1). The association of pelgipeptins A + B or A + C had an additive effect with FIC values varying from 0.5 and 0.75. Moreover, FIC values lower than 0.5 obtained on Pelgipeptin B + C or chloramphenicol associations indicates synergic effect (Table 2, Supplementary Tables 1 and 2).

The observed synergic effect against K. pneumoniae ATCC13883 of Pelgipeptin isoforms and the control antimicrobials was expected, due to the differences in the targeted cell structure of each antimicrobial. Lipopeptides from the Paenibacillus spp. present in its structure both, D and L amino acids, resulting in higher stability over proteolytic enzymes22, 30, and 31. The presence of the non proteinogenic amino acid Dab results an overall cationic charge of pelgipeptins, suggesting that its primary mechanism of action may be the membrane disruption of targeted organisms22. Penicillin is a bactericidal molecule that inhibits trans peptidase, enzymes that catalyzes reactions between peptidoglycan chains of the bacterial cell wall32, 33; leading to the cell wall stiffness suppression and turning bacterial cells susceptible to osmotic variations. Chloramphenicol binds to the ribosome 50S sub-unit and inhibits protein synthesis34. Furthermore, no synergic effect against the MDR strain was observed between Penicillin G and Pelgipeptin isoforms since the antimicrobial effect is solely due to the Pelgipeptins mechanism of action. Despite the different cell target of each drug tested, only some combinations with Pelgipeptins were synergic (Table 2). This fact may be a function of the pelgipeptins structural difference on the fatty acid chains, since its length and structural formation may change the hydrophobicity level, directly interfering over bacterial cell membrane interaction22; and pelgipeptin B seem to be the best synergic agent.

Conclusions The synergic effect of combinations with pelgipeptins was demonstrated by means of Checkerboard testing method, which classifies molecules interactions as synergistic, additive, indifferent or antagonistic based on FICI values acquired. It is important to state that checkerboard is one of many synergy testing methods. However, despite an ongoing discussion whereas combined antibiotic administration is or is not an efficient therapeutic method, it is important to make synergism studies available in order to create a larger database on that therapy option. The presence of multidrug resistant bacteria has been increasing the use and necessity of stronger and more expensive antimicrobial compounds with possible side effects, but after validation by in vivo experiments, synergism could be explored for developing novel therapeutic options since this approach could be used to reduce the needed dosage of a compound, and consequently reduce toxicity (e.g. chloramphenicol), or even rescue compounds that lost activity in the past (e.g. penicillin). Besides, the overall lipopeptides fatty

94 acid tail structure and the mechanism of action of a molecule seem to be crucial to better understand synergy effects variations.

Acknowledgments The authors thank to Célio de Faria Júnior and Octávio Luiz Franco for providing the multidrug resistant strains. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001 and by a grant from Conselho Nacional de Desenvolvimento Científico e Tecnológico – (CNPq).

References 1 Doern, C. D. When does 2 plus 2 equal 5? A review of antimicrobial synergy testing. Journal of clinical microbiology 52, 4124-4128, doi:10.1128/JCM.01121-14 (2014). 2 Chi, H. & Holo, H. Synergistic Antimicrobial Activity between the Broad Spectrum Bacteriocin Garvicin KS and Nisin, Farnesol and Polymyxin B against Gram- Positive and Gram-Negative Bacteria. Current microbiology 75, 272-277, doi: 10.1007/s00284-017-1375-y (2018). 3 Tanwar, J., Das, S., Fatima, Z. & Hameed, S. Multidrug resistance: an emerging crisis. Interdisciplinary perspectives on infectious diseases 2014, 541340, doi:10.1155/2014/541340 (2014). 4 Ribeiro, S. M. et al. Antibiofilm peptides increase the susceptibility of carbapenemase-producing Klebsiella pneumoniae clinical isolates to beta-lactam antibiotics. Antimicrob Agents Chemother 59, 3906-3912, doi:10.1128/AAC.00092-15 (2015). 5 Tzouvelekis, L. S., Markogiannakis, A., Psichogiou, M., Tassios, P. T. & Daikos, G. L. Carbapenemases in Klebsiella pneumoniae and Other Enterobacteriaceae: an Evolving Crisis of Global Dimensions. Clinical Microbiology Reviews 25, 604-660 (2012). 6 Croxen, A. M. et al. Recent Advances in Understanding Enteric Pathogenic Escherichia coli. Clinical Microbiology Reviews 26, 822-880 (2013). 7 Tong, S. Y. C., Davis, J. S., Eichenberger, E., Holland, T. L. & Fowler Jr., V. G. Staphylococcus aureus Infections: Epidemiology, Pathophysiology, Clinical Manifestations, and Management. Clinical Microbiology Reviews 28, 682-707 (2015). 8 Narasimhaswamy, N., Bairy, I., Shenoy, G. & Bairy, L. In vitro activity of recombinant lysostaphin in combination with linezolid, vancomycin and oxacillin against

95 methicillin-resistant Staphylococcus aureus. Iranian journal of microbiology 9, 208-212 (2017). 9 Montero, M. et al. Evaluation of Ceftolozane-Tazobactam in Combination with Meropenem against Pseudomonas aeruginosa Sequence Type 175 in a Hollow-Fiber Infection Model. Antimicrob Agents Chemother 62, doi:10.1128/AAC.00026-18 (2018). 10 Kumar, S. N., Siji, J. V., Nambisan, B. & Mohandas, C. Activity and synergistic interactions of stilbenes and antibiotic combinations against bacteria in vitro. World J Microbiol Biotechnol 28, 3143-3150, doi:10.1007/s11274-012-1124-0 (2012). 11 He, Z. et al. Isolation and Identification of a Paenibacillus polymyxa Strain That Coproduces a Novel Lantibiotic and Polymyxin. Applied and environmental microbiology 73, 168-178, doi:10.1128/AEM.02023-06 (2006). 12 Lohans, C. et al. Structural characterization of the highly cyclized lantibiotic paenicidin A via a partial desulfurization/reduction strategy. J Am Chem Soc. 134, 19540- 19543, doi:10.1021/ja3089229 (2012). 13 Baindara, P. et al. Characterization of the Antimicrobial Peptide Penisin, a Class Ia Novel Lantibiotic from Paenibacillus sp. Strain A3. Antimicrobial Agents Chemotherapy 60, 580-591 (2015). 14 Niu, B. et al. Polymyxin P is the active principle in suppressing phytopathogenic Erwinia spp. by the biocontrol rhizobacterium Paenibacillus polymyxa M-1. BMC Microbiology 13, 1-13 (2013). 15 Selim, S., Negrel, J., Govaerts, C., Gianinazzi, S. & Tuinen, D. v. Isolation and Partial Characterization of Antagonistic Peptides Produced by Paenibacillus sp. Strain B2 Isolated from the Sorghum Mycorrhizosphere. Applied and environmental microbiology 71, 6501–6507, doi:10.1128/AEM.71.11.6501–6507.2005 (2005). 16 Canova, S. P. et al. Characterization of lipopeptides from Paenibacillus sp. (IIRAC30) suppressing Rhizoctonia solani. World Journal of Microbiology and Biotechnology 26, 2241-2247 (2010). 17 Guo, Y., Huang, E., Yuan, C., Zhang, L. & Yousef, A. E. Isolation of a Paenibacillus sp. Strain and Structural Elucidation of Its Broad-Spectrum Lipopeptide Antibiotic. Applied and environmental microbiology 78, 3156-3165 (2012). 18 Wen, Y. et al. Identification and analysis of the gene cluster involved in biosynthesis of paenibactin, a catecholate siderophore produced by Paenibacillus elgii B69. Environmental microbiology 13, 2726-2737 (2011).

19 Qian, C.-D. et al. Battacin (Octapeptin B5), a new cyclic lipopeptide antibiotic from Paenibacillus tianmuensis active against multidrug-resistant Gram-negative bacteria. Antimicrob Agents Chemother 56, 1458-1465, doi:10.1128/AAC.05580-11 (2011). 20 Qian, C. D. et al. Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiology 12, 1-7, doi:10.1186/1471-2180-12-197 (2012). 21 Ding, R. et al. Isolation and Identification of Lipopeptide Antibiotics from Paenibacillus elgii B69 with Inhibitory Activity against Methicillin-Resistant Staphylococcus aureus. The Journal of Microbiology 49, 942-949, doi: 10.1007 (2011). 22 Cochrane, S. A. & Vederas, J. C. Lipopeptides from Bacillus and Paenibacillus spp.: A Gold Mine of Antibiotic Candidates. Medicinal research reviews 36, 4-31, doi:10.1002/med.21321 (2016). 23 Wu, X. C. et al. Isolation and partial characterization of antibiotics produced by Paenibacillus elgii B69. FEMS Microbiol Lett 310, 32-38, doi:10.1111/j.1574- 6968.2010.02040.x (2010). 24 Qian, C. et al. Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiology 12, 1-7, doi: 10.1186/1471-2180-12-197 (2012). 25 Kim, J. et al. Isolation and structural elucidation of pelgipeptin E, a novel pore- forming pelgipeptin analog from Paenibacillus elgii with low hemolytic activity. The Journal of Antibiotics, doi: 10.1038/s41429-018-0095-2 (2018). 26 Kaneda, T. Iso- and anteiso-fatty acids in bacteria: biosynthesis, function, and taxonomic significance. Microbiological reviews 55, 288-302 (1991). 27 Kaneda, T. Fatty acids in the genus Bacillus. I. Iso- and anteiso-fatty acids as characteristic constituents of lipids in 10 species. Journal of bacteriology 93, 894-903 (1967). 28 Ortega, D. B. et al. Draft Genome Sequence of the Antimicrobial-Producing Strain Paenibacillus elgii AC13. Genome Announcements 6, doi:10.1128/genomeA.00573-18 (2018). 29 Balouiri, M., Sadiki, M. & Ibnsouda, S. K. Methods for in vitro evaluating antimicrobial activity: A review. Journal of Pharmaceutical Analysis 6, 71-79, doi: 10.1016/j.jpha.2015.11.005 (2016). 30 Murai, A., Amino, Y. & Ando, T. Absolute configuration of the beta-hydroxyl fatty acid constituent of permetin A. The Journal of antibiotics 38, 1610-1613 (1985).

31 Huang, E. & Yousef, A. E. Lipopeptide antibiotic, paenibacterin, binds to bacterial outer membrane and exerts bactericidal activity through cytoplasmic membrane damage. Applied and environmental microbiology 80, 2700-2704 (2014). 32 Tipper, D. J. & Strominger, J. L. Mechanism of action of penicillins: a proposal based on their structural similarity to acyl-D-alanyl-D-alanine. Proceedings of the National Academy of Sciences of the United States of America 54, 1133-1141 (1965). 33 Cushnie, T. P., O'Driscoll, N. H. & Lamb, A. J. Morphological and ultrastructural changes in bacterial cells as an indicator of antibacterial mechanism of action. Cellular and molecular life sciences: CMLS 73, 4471-4492, doi: 10.1007/s00018-016-2302-2 (2016). 34 von Nussbaum, F., Brands, M., Hinzen, B., Weigand, S. & Habich, D. Antibacterial natural products in medicinal chemistry--exodus or revival? Angew Chem Int Ed Engl 45, 5072-5129, doi:10.1002/anie.200600350 (2006).

Tables

Table 1. Minimum inhibitory concentration (MIC): Pelgipeptins A-D and reference antibiotics against tested strains (µg·mL-1).

Pelgipeptin Pelgipeptin Pelgipeptin C Pelgipeptin D Chloramphenicol Penicillin G A B Escherichia coli 8 32 32 8 NT* NT* ATCC 11229 Escherichia coli 16 16 8 8 NT* NT* LACEN 3789319 Staphylococcus aureus 128 16 32 16 8 NT* ATCC 14458 Klebsiella pneumoniae 8 64 64 8 16 0.5 ATCC 13883 Klebsiella pneumoniae 8 32 16 16 32 >1024 LACEN 3259271 * NT: not tested.

Table 2. Fractional Inhibitory Concentration (FIC) Index: Pelgipeptin A-C and other antibiotics against Klebsiella pneumoniae ATCC13883 and the MDR strain LACEN 3259271.

Combinations K. pneumoniae ATCC13883 K. pneumoniae LACEN 3259271 Molecule A Molecule B FIC Interpretation* FIC Interpretation* Pelgipeptin A Chloramphenicol 1.5 I 2 I Pelgipeptin B 1.125 I 0.75 A

Pelgipeptin C 0.375 S 0.562 A

Pelgipeptin B Chloramphenicol 0.25 S 0.312 S Pelgipeptin C 0.265 S 0.5 S

Penicillin G 0.5 S 1.5 I

Pelgipeptin C Chloramphenicol 0.265 S 1.25 I Penicillin G 0.5 S 1.5 I

*S, synergy (FIC ≤ 0, 5); A, additive (FIC > 0.50 and < 1); I, indifferent (FIC >1 and ≤ 4)

Supplementary Material

Supplementary Table 1: Checkerboard testing method: Pelgipeptins A-C and reference antibiotics against Klebsiella pneumoniae ATCC 13883 (µg·mL-1).

Pelgipeptin A Pelgipeptin B Pelgipeptin C Chloramphenicol Penicillin G

Pelgipeptin A 8 8 2 8 ND* Pelgipeptin B 8 64 16 8 16 Pelgipeptin C 8 1 64 16 16 Chloramphenicol 8 2 0.25 16 ND* Penicillin G ND* 0.125 0.125 ND* 0.5 * ND: non-determined.

Supplementary Table 2: Checkerboard testing method: Pelgipeptins A-C and reference antibiotics against MDR Klebsiella pneumoniae LACEN 3259271 (µg·mL-1).

Pelgipeptin A Pelgipeptin B Pelgipeptin C Chloramphenicol Penicillin G

Pelgipeptin A 16 8 1 16 ND* Pelgipeptin B 8 32 8 8 16 Pelgipeptin C 8 4 16 8 8 Chloramphenicol 32 2 32 32 ND* Penicillin G ND* >1024 >1024 ND* >1024 * ND: non-determined.

100

Linear counterparts of the NRPs Pelgipeptins B and C naturally produced by Paenibacillus elgii strain AC13 in nutrient broth - DRAFT (Appendix 2)

101

Working title: Linear counterparts of the NRPs Pelgipeptins B and C are naturally produced by Paenibacillus elgii strain AC13 in nutrient broth.

Daniel Barros Ortega1; Rosiane Andrade da Costa1; Thiago Fellipe de Araújo1; Débora

Luiza Albano Fulgencio1; Marise Leite Mendonça1; Flávio Silva Costa1; Michel Lopes Leite1;

Cristine Chaves Barreto1*.

¹ Universidade Católica de Brasília, Genomic Sciences and Biotechnology Graduate

Program, SGAN 916, Brasília – DF 70790-160, Brazil.

Corresponding author: Cristine Chaves Barreto. Phone: +55-61-3448-9217. Fax:

+55-61-3347-4797 / e-mail: [email protected] Universidade Católica de Brasília,

Programa de Ciências Genômicas e Biotecnologia. Brasília - Brazil. 70790-160.

ORCID: 0000-0003-0721-7010 (Daniel Barros Ortega); 0000-0002-1739-3731

(Rosiane Andrade da Costa).

Abstract Non-ribosomal peptides (NRPs) are widely produced by Paenibacillus strains.

Previously isolated from Cerrado soil samples, Paenibacillus elgii strain AC13 is able to synthetize at least four cyclic lipopeptides named Pelgipeptins (PGP A-D). However, variations over microbial culture conditions may reveal hidden naturally produced NRPs.

Here we present the partial characterization of two linear lipopeptides synthetized by strain

AC13 in nutrient broth. Initially, Paenibacillus elgii strain AC13 was grown in nutrient broth.

Cultures’ supernatants were separated by RP-HPLC, and fractions were identified by mass spectrometry (MALDI-ToF MS and MS/MS). Then, a microplate dilution method was used to determine Minimum Inhibitory Concentrations (MICs) of purified lipopeptides. Cytotoxicity 102 and cell viability was performed by MTT colorimetric assay using human fibroblast cells.

Results show that P. elgii AC13 can produce two molecules with m/z ratio of 1105 and 1119 in addition to the four known PGP isoforms (A-D). Partial characterization indicates that those molecules are linear isoforms of PGP C (named PGP-C’) and PGP B (named PGP B’).

Besides, lower cytotoxicity was found for linear lipopeptides counterparts when compared to cyclic ones. In addition, a genomic analysis of P. elgii AC13 indicates that all molecules found are related to the same DNA gene cluster; thus, here we assume that Paenibacillus elgii

AC13 is able to produce not only cyclic PGPs but also linear counterparts of PGP-B and C.

Linear antimicrobial lipopeptides produced in liquid media are highly interesting in regard to biotechnological and economic advantages.

Keywords: Paenibacillus elgii; lipopeptides; pelgipeptin; NRPs.

Introduction Template mRNA translation is not the only mechanism for cells to build proteins or peptides (Ng et al. 2009). Non-ribosomal Peptide Synthases (NRPS) are complexes of enzymes organized in amino acid-specific modules. Basically, a NRPS module exhibits three core domains, namely A-T-C, each one responsible for a biochemical step regarding peptide backbone formation and other structural modifications. Domain A (adenylation) is responsible for selecting a specific amino acid monomer; to perform a covalent bond between substrate and intermediate acyl chains during chain elongations; a thiol group is provided at T

(thiolation) domain; finally, peptide-bond-forming chain elongation occurs at C

(condensation) domain (Fischbach and Walsh 2006). Besides, NRPS uses non-proteinogenic amino acids, which increase the number of possible templates when compared to the synthesis process of standard ribosomal peptides (Fischbach and Walsh 2006; Ng et al. 2009). Non- ribosomal peptides (NRPs) are natural products exhibiting surfactant properties and antiviral,

103 antitumor, or antibiotic activity against human and plant pathogens (Qian et al. 2011; Qian et al. 2012). Therefore, attracting interest of biotechnological, pharmaceutical and agricultural industries (Ng et al. 2009; Cochrane and Vederas 2016; Grady et al. 2016; Biniarz et al. 2017;

Deng et al. 2017).

Naturally produced by Paenibacillus elgii, the NRP family of Pelgipeptins (PGPs) consists of five cyclic and cationic lipopeptides comprising nine amino acids and a fatty acid chain synthetized by a NRPS coded by the plp gene cluster (40.8 kb) (Wu et al. 2010; Ding et al. 2011b; Qian et al. 2012; Kim et al. 2018). The molecular weights of Pelgipeptin A (PGP-

A), B (PGP-B), C (PGP-C), D (PGP-D) and E (PGP-E) were determined to be 1072; 1100;

1086; 1086; and 1072, respectively (Ding et al. 2011b; Kim et al. 2018). Furthermore, all peptides within the PGP family have a β-hydroxy fatty acid chain with a variable tail of three or four carbon lengths at its N-terminal portion, linked to the non-proteinogenic amino acid 2,

4-diaminobutyric acid (Dab) by a lactone ring (Wu et al. 2010; Ding et al. 2011b; Cochrane and Vederas 2016). Variations among PGPs A-D also include substitution at the second amino acid position, which may be either a Valine or Isoleucine (Wu et al. 2010; Qian et al.

2012). Moreover, PGP-A and PGP-D exhibit a lipid tail with a methyl group linked at its penultimate (iso) carbon, whereas PGP-B and PGP-C show this at the antepenultimate

(anteiso) carbon (Kaneda 1967; 1991; Qian et al. 2012; Cochrane and Vederas 2016). PGP-E was recently described as a natural product of Paenibacillus elgii strain BC34-6, isolated from red clay soil, and it also exhibits broad-spectrum antimicrobial activity. The stereochemistry of published PGPs is as follows: L-Dab1;L-Val2/L-Ile2;L-Dab3;D-Phe4;L-Leu5;L-Dab6;D-

Val7;L-Leu8;L-Ser9 (Kim et al. 2018). PGP-E shows same stereochemistry as the amino acid and ester bond of both, PGP-A and PGP-C; but attached to a straight lipid chain (Table 1)

(Kim et al. 2018).

104

Cyclic lipopeptides, however, are not the only group of molecules with high biotechnological potential produced by Paenibacillus. Naturally-produced linear lipopeptides with antimicrobial, antifungal or surfactant properties are also described (Huang et al. 2017;

Moon et al. 2017; De Zoysa et al. 2018). It was recently reported that linearization of Battacin

- a cyclic lipopeptide with eight amino acids produced by Paenibacillus tiamunensis - improved antifungal activity against pathogenic Candida albicans and reduced hemolytic activity against mouses’ red blood cells (De Zoysa et al. 2015; De Zoysa et al. 2018).

Moreover, linear Battacin also exhibits antibiofilm activity against Staphylococcus aureus and

Pseudomonas aeruginosa (De Zoysa et al. 2015).

Additionally, a non-related study reports that Paenibacillus sp. strain OSY-N was described as being able to synthesize three lipopeptides belonging to a new polypeptide family named Paenipeptin A-C (Huang et al. 2017). Within Paenipeptin A-C, only

Paenipeptin C is a cyclic lipopeptide, exhibiting molecular mass of 1133 [M+H]+. Paenipeptin

A-B are, in fact, linear counterparts of Pelgipeptin B/Permetin A and Pelgipeptin C/BMY-

28160, respectively (Huang et al. 2017). Paenipeptins’ synthetic linear isoforms have been tested and their potential to eradicate biofilm has been reported as an alternative to combat multidrug-resistant (MDR) bacteria or fungi (Moon et al. 2017).

Paenibacillus elgii strain AC13 was isolated from Cerrado soil samples (Kurokawa,

2006). When cultured into the chemically defined medium MMP (Patent BR102017018881-

7); culture supernatant produces PGPs A-D, which exhibited antimicrobial activity against

Escherichia coli and Staphylococcus aureus (Costa et al. 2019). However, different culture conditions may reveal hidden natural products. In this context, the present study aimed to investigate the occurrence of naturally produced lipopeptides isoforms by Paenibacillus elgii strain AC13 in complex media. Liquid cultures’ bioproducts are preferable since it makes extraction and production processes more affordable.

105

Material and Methods Bacteria strains Antimicrobial compounds were isolated and purified from Paenibacillus elgii strain

AC13. Antimicrobial assays were performed using Escherichia coli ATCC11229 and

Staphylococcus aureus ATCC14458.

Strain AC13 cultures A single colony from Paenibacillus elgii strain AC13 was incubated in nutrient broth

(Difco) at 37 °C with shaking at 180 rpm. Growth was monitored by absorbance at 600 nm until reaching exponential phase. Then, 1 mL was transferred to 100 mL of sterile nutrient broth and incubated at 37 °C with shaking at 180 rpm. Lipopeptides extraction and purification were carried out after 40 hours.

Purification and quantification of antimicrobial compounds

In order to obtain purified lipopeptides, cultures were centrifuged at 9000 × g for 10 min to remove cells; then, the supernatant was extract (twice) with n-butanol and water (1:1; v/v). The organic layer was lyophilized and the dry content was suspended in ultrapure water, filtered (0.22 µm) and purified by reversed-phase high-performance liquid chromatography

(RP-HPLC) using a Shim-pack Shimadzu C18 column VP-ODS (4.6 × 150 mm; 4.6 µm particle size, Shimadzu, Japan). Mobile phases were HPLC-grade water and acetonitrile

(ACN), both containing 0.1 % trifluoracetic acid (TFA).Solvents gradient was used as follows: 0-10min, 30% ACN; 10-25 min, 30-50% ACN; 25-30min, 50-95% ACN .Flow rate was kept as 1.0 mL•min-1. Elution was monitored by measuring the absorbance at 216 nm using a UV detector (Prominence, Shimadzu, Japan). Fractions containing purified lipopeptides were collected and purified a second time. Alternatively was made a stock of a mixture of cyclic pelgipeptins (Mix A-D), once it is easier and cheaper purifying the four cyclic isoforms together. The pure peptides stock solutions were quantified by Murphy’s method (Murphy and Kies 1960). Matrix Assisted Laser Desorption Ionization Mass

106

Spectrometry (MALDI-MS; Autoflex Speed, Bruker Daltonics) was used to confirm m/z ratio of stock solutions. Spectra were acquired in reflected positive mode (RP 700-3500 m/z).

Concentration of lipopeptides isoforms within nutrient broth cell free supernatant was determined by RP-HPLC using a standard curve made with dilutions of stock solutions. The quantitative chromatography of standards and supernatant (after organic extraction) were performed in linear gradient of 5-95% ACN during 50 minutes with a flow rate of 0.6 mL•min-1 using the same column and detection.

Characterization of lipopeptides’ primary structure

Purified cyclic lipopeptides had its lactone ring hydrolyzed with a 1 M NaOH solution

(1:1, v/v) at room temperature. After 12 hours, the reaction was neutralized with 1 M HCl

(1:1, v/v) and desalted using a C18 tip (ZipTip® Pipette Tips, Millipore). Amino acid sequence of PGPs was determined using tandem mass spectrometry (MALDI-ToF MS/MS)

LIFT method in positive ion mode. Final analysis for the primary sequence of the peptides was performed using FlexAnalysis software (Bruker Daltonics®).

Determination of Minimal Inhibitory Concentration (MIC)

Minimal inhibitory concentration assay of purified lipopeptides was measured in 96- well microplates according to the Clinical and Laboratory Standards Institute (CLSI 2009)

(M07-A6). The peptide solutions were prepared to final concentrations of 1.5-100 µM.

Commercial polymyxin B (Sigma Aldrich), Penicillin G (Sigma-Aldrich), and

Chloramphenicol (Sigma-Aldrich) were used as positive control; and sterile ultra-pure water was used as negative control. Tested bacterial strains and peptide solution mixtures were incubated at 37 ºC for 16 hours. Absorbance (595nm) was measured once per hour in a

BioTek (Winooski, VT, USA) microplate reader. MICs were defined as the lowest peptide concentration that inhibited bacterial growth.

107

Genomic analysis NRPS cluster sequences were identified using antiSMASH (Weber et al. 2015) and compared to Paenibacillus elgii B69 Pelgipeptin Synthase sequences (Qian et al. 2012) using

BLAST. Contigs identified as related to PGP were mapped to the biosynthetic gene cluster of

P. elgii B69 using Geneious 11.1.4.

Cytotoxic activities of PGPs A, B, C, D and B’ The cytotoxicity of PGPs A, B, C, D and B’ was evaluated using human fibroblast cells. Initially, fibroblasts were grown in Dulbecco’s Modified Eagle’s Medium (DMEM;

Gibco, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum

(Invitrogen, Carlsbad, CA, USA) and incubated at 5% CO2 and 95% humidity at 37ºC until the cytotoxicity assays were carried out.

Cell viability tests were performed on technical replicates and were assessed by the

MTT (3-[4,5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide; Sigma-Aldrich, USA) colorimetric assay (Mosmann 1983). Briefly, cells were seeded onto 96-well plates (103 cells/well; 100 µL) and allowed to adhere and grow for 24 hours. Then, cells were incubated with fresh medium prepared and lipopeptide or Polymyxin B to final concentrations of 3–50

µM. After treatment, 90 µL was added to 10µL of MTT (50 µg) was added to each well, and the cells were incubated for 4 h. Then, 100 µL of dimethyl sulfoxide (DMSO) was added to all wells and mixed to dissolve the dark blue crystals (formazan). Finally, the microplate was read on a microplate spectrophotometer at 570 nm. Cells treated with 50 µL of the lysis solution sodium hypochlorite (2.5%) were used as negative control; and untreated cells were used as the positive control group.

Data are expressed as the mean ± standard deviation (SD) of three independent experiments. Multiple comparisons were performed by one-way analysis variance (ANOVA) followed by Bonferroni post hoc test using GraphPad Prism (GraphPad Software, San Diego,

CA). The P values of <0.05 were considered statistically significant. 108

Results Partial characterization of antimicrobial compounds RP-HPLC of strain AC13 cultured into nutrient broth resulted in six fractions with RT varying between 25 and 37 minutes. Spectral analysis of separated compounds showed masses ranging from 1073 to 1119 [M+H]+ (Table 2). Fractions I and II show retention time

(RT) of 25.5 and 28 minutes respectively (Figure1). These fractions corresponded to compounds with monoisotopic masses [M+H] + of 1105.7 (I) and 1119.7 (II). Fraction I and II exhibited higher quantification values in comparison to PGPs A-D altogether (Figure 1 and

2). MS were acquired before alkaline hydrolysis. Amino acid sequences were deduced by

MALDI-ToF MS/MS de novo sequencing after alkaline hydrolysis and are presented in Table

2.Fraction I and II are linear counterparts of PGPs C and B. Predicted sequence of fraction I and II are the same as those found for PGP-C and PGP-B: Dab1, Val2, Dab3, Phe4, Leu5,

Dab6, Val7, Leu8, and Ser 9 for fraction I and Dab1, Ile2, Dab3, Phe4, Leu5, Dab6, Val7,

Leu8, and Ser 9 for fraction II (Figures 3 and 4). Molecular weight differences between cyclic and linear isoforms are due to lactone ring hydrolysis (Figures 5 and 6). Thus, purified lipopeptides obtained from strain AC13 cultures in nutrient broth were identified as follows: fractions I and II are linear counterparts of PGP-C (PGP C’) and -B (-B’), respectively; and fractions III to VI contain antimicrobial cyclic lipopeptides PGPs A-D.

MIC values of tested molecules were found for Escherichia coli ATCC11229 and

Staphylococcus aureus ATCC14458 and are summarized in Table 3. PGPs A-D were effective against tested microorganisms.

Genomic analysis The PGP cluster in the AC13 genome is fragmented within contigs 118, 35, 85 and 17

(Figure 7). Sequences for genes plp A, B, C, D, F, G and H are complete, while the sequence for plpE is divided into contigs 118, 85 and 35, and presents all the domains of condensation, thiolation and epimerization (Figure 8). Adenylation domains for Dab in positions 3 and 5 are

109 missing, while contig 17 presents high similarity with these regions; however, there is no overlap with any other contigs, therefore making it impossible to determine its position in the gene cluster.

Cytotoxicity of Pelgipeptins A-D and B’ Cell viability was determined by the MTT colorimetric assay after 24 hours incubation of human fibroblast cells with different concentrations of PGPs and Polymyxin B. Among tested molecules, PGP- B’ exhibited the lowest cytotoxic activity, while PGP-B, D and PGP

A-D (Mix) exhibited high cytotoxicity against tested cells (Figure 9).

Discussion Antimicrobial compounds partial characterization Antimicrobial lipopeptides have been studied as a source of efficient compounds against emerging MDR bacterial infections, which is a primary concern for human and animal health industries. Paenibacillus elgii AC13 naturally produces linear isoforms of PGP-C and

PGP-B when cultured into nutrient broth. Linear peptides show high biotechnological potential since those molecules are usually preferable to be tested and bioengineered due to its structure (Moon et al., 2017). It is important to state that fractions identified as PGP-C’ and

PGP-B’ were present before and after alkaline hydrolysis (Figure 7a and 7b).

Recently described, Paenipeptin A and B are linear lipopeptides produced by a

Paenibacillus strain (OSY-N) with the same molecular weight found for PGP-C’ and PGP-B’.

Paenipeptins A-B were described as isolates after culturing Paenibacillus sp. strain OSY-N in

300 TSA plates at 37°C for 72 hours. Paenibacillus elgii strain AC13, however, naturally produces molecules with same primary characteristics in nutrient broth at 37°C, which indicates how diverse are strains isolated from different places.

Minimal Inhibitory Concentration (MIC)

110

Despite its higher MIC (50 µM), PGP-B’ was able to interact with E. coli

ATCC11229, indicating that the PGP cyclization process is not a crucial step for antimicrobial activity against diderm bacteria, and suggesting that the PGP-B linearization process did not prevent that molecule from acting on diderm cell envelopes. However, interaction with monoderm cells was reduced. When tested against S. aureus ATCC14458, both PGP-B’ and C’ resulted in MICs higher than 100 µM. A study produced Paenipeptins A and B analogues by solid phase peptide synthesis and tested antimicrobial activity against 19 bacterial strains (Huang et al. 2017). Synthetic Paenipeptin MICs are in accordance with results found in our study, exhibiting the same trend as that in our findings. Furthermore,

Paenipeptins’ mechanisms of action have already been elucidated (Moon et al. 2017). It was found that synthetic linear analogues of a cyclic Paenipeptin (Paenipeptin C) neutralizes antimicrobial activity when tested against purified lipopeptides from E. coli O111:B4

(Sigma), suggesting that Paenipeptins have a high affinity to outer membrane compounds of diderm bacteria; this means that these molecules probably promote chemical instability towards the structure of the bacterial envelope (Moon et al. 2017).

Linear isoforms of lipopeptides produced by Paenibacillus elgii strain AC13 were observed only in nutrient broth cultures. These linear isoforms may have been caused by the non-cyclization of PGP-B, a process that occurs in the thioesterase domain of the NRPS (Qian et al. 2012). Another hypothesis suggests that the lactone ring opening process may have occurred after NRP cyclization during the organic extraction with butanol, since n-butanol presents a pKa= 16.1 and that alkaline conditions can cause lactone rings to open and change lipopeptide structure (Huang et al. 2013); the butanol hydrolyzation effect over cyclic PGPs was tested (Supplementary Figure S1); and after 24 hours of PGPs exposure to n-butanol they did not undergo to any variations. Besides, nutrient broth buffering leads to minimal pH variations (data not show), and linearization was not originated by organic extraction, since

111 linear molecules were present in crude supernatant before butanol extraction (Supplementary

Figure S2). As stated before in other research, it is common to observe a preferential synthesis of certain molecules to the detriment of others, as a result of changes in the culture medium

(Akpa et al. 2001; Ding et al. 2011c), and that sounds more likely to had happened.

Genomic analysis The AC13 genome contains 40 fragments of putative NRPS (Ortega et al. 2018). The sequence of plpA in AC13 is identical to the one in B69 strain, but the other genes present small modifications in the amino acids sequence (Table 4). There is no other gene described for NRPS that could synthesize a peptide with a similar primary structure to the peptides described here in the AC13 genome.

Cytotoxicity of Pelgipeptins A-D and B’

Naturally produced on solid medium by a Paenibacillus sp strain; Paenipeptins exhibit high antimicrobial activity. Synthetic linear analogues of this very similar Pelgipeptin family of molecules were tested in regard hemolytic activity against defibrinated rabbit blood cells and was observed the increase in antimicrobial activity accompanied by the elevation of hemolytic activity (Huang et al. 2017). Most isoforms found for Paenibacillus elgii AC13 cultures are not toxic to the primary culture of human fibroblasts tested, which opens up the possibility of exploring these molecules.

It is interesting to note that, even with the similarities between both groups of molecules (Pelgipeptins and Paenipeptins) variations over the cytotoxicity of the lipopeptides produced by Paenibacillus species is strongly related to the cell lineage tested, leading to a more detailed investigation over this matter.

Conclusions It was identified that Paenibacillus elgii AC13 naturally produces PGPs A-D (Wu et al. 2010; Ding et al. 2011b). Besides, strain AC13 also produces two lipopeptides with

112 monoisotopic masses [M+H]+ of 1105.7 (PGP-C’) and 1119.7 (PGP-B’) when cultured into nutrient broth.

Despite the similarities between Paenipeptins and PGPs linear isoforms found in our study, variations along the fatty acid structure or different amino acid configurations needs to be investigates in order to evaluate if the molecules found at this work and the Paenipeptins are in fact the same. Thus, a more detailed characterization needs to be conducted.

Liquid media cultures are more affordable and facilitate production scaling for fermentation systems. Paenipeptins, were obtained by scraping colonies of 300 plates cultured on TSA agar (Huang et al. 2017; Moon et al. 2017). Paenibacillus elgii strain AC13, however, is able to produce similar molecules in liquid media, presenting suitable conditions for the synthesis of linear molecules which can be economically exploited due to its biotechnological potential.

Acknowledgments This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de

Nível Superior – Brasil (CAPES) – Finance Code 001; Project 88882.182971/2018-01 and by a grant from Conselho Nacional de Desenvolvimento Científico e Tecnológico – (CNPq).

References Akpa, E., Jacques, P., Wathelet, B., Paquot, M., Fuchs, R., Budzikiewicz, H. and

Thonart, P. (2001) Influence of culture conditions on lipopeptide production by Bacillus subtilis. Applied biochemistry and biotechnology 91, 551-561.

Biniarz, P., Łukaszewicz, M. and Janek, T. (2017) Screening concepts, characterization and structural analysis of microbial-derived bioactive lipopeptides: a review.

Critical Reviews in Biotechnology 37, 393-410.

113

CLSI (2009) Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria

That Grow Aerobically. p.64. Pennsylvania - USA: Clinical and Laboratory Standards

Institute.

Cochrane, S.A. and Vederas, J.C. (2016) Lipopeptides from Bacillus and Paenibacillus spp.: A Gold Mine of Antibiotic Candidates. Medicinal research reviews 36, 4-31.

Costa, R., Ortega, D.B., Fulgêncio, D.L.A., Costa, F.S., Araújo, T.F. and Barreto, C.C.

(2019) Checkerboard testing method indicates synergic effect of pelgipeptins against multidrug resistant Klebsiella pneumoniae – a short communication. Biotechnology Research and Innovation.

De Zoysa, G.H., Cameron, A.J., Hegde, V.V., Raghothama, S. and Sarojini, V. (2015)

Antimicrobial Peptides with Potential for Biofilm Eradication: Synthesis and Structure

Activity Relationship Studies of Battacin Peptides. Journal of Medicinal Chemistry 58, 625-

639.

De Zoysa, G.H., Glossop, H.D. and Sarojini, V. (2018) Unexplored antifungal activity of linear battacin lipopeptides against planktonic and mature biofilms of C. albicans.

European Journal of Medicinal Chemistry 146, 344-353.

Deng, Q., Wang, W., Sun, L., Wang, Y., Liao, J., Xu, D., Liu, Y., Ye, R. and

Gooneratne, R. (2017) A sensitive method for simultaneous quantitative determination of surfactin and iturin by LC-MS/MS. Analytical and bioanalytical chemistry 409, 179-191.

Ding, R., Li, Y., Qian, C. and Wu, X. (2011a) Draft Genome Sequence of

Paenibacillus elgii B69, a Strain with Broad Antimicrobial Activity. Journal of Bacteriology

193, 4537.

Ding, R., Wu, X.-C., Qian, C.-D., Teng, Y., Li, O., Zhan, Z.-J. and Zhao, Y.-H.

(2011b) Isolation and Identification of Lipopeptide Antibiotics from Paenibacillus elgii B69

114 with Inhibitory Activity Against Methicillin-Resistant Staphylococcus aureus. The Journal of

Microbiology 49, 942-949.

Ding, R., Wu, X.C., Qian, C.D., Teng, Y., Li, O., Zhan, Z.J. and Zhao, Y.H. (2011c)

Isolation and identification of lipopeptide antibiotics from Paenibacillus elgii B69 with inhibitory activity against methicillin-resistant Staphylococcus aureus. J Microbiol 49, 942-

949.

Fischbach, M.A. and Walsh, C.T. (2006) Assembly-line enzymology for polyketide and nonribosomal Peptide antibiotics: logic, machinery, and mechanisms. Chem Rev 106,

3468-3496.

Grady, E.N., MacDonald, J., Liu, L., Richman, A. and Yuan, Z.C. (2016) Current knowledge and perspectives of Paenibacillus: a review. Microb Cell Fact 15, 203.

Huang, E., Yang, X., Zhang, L., Moon, S.H. and Yousef, A.E. (2017) New

Paenibacillus strain produces a family of linear and cyclic antimicrobial lipopeptides:

Cyclization is not essential for their antimicrobial activity. FEMS Microbiol Lett.

Huang, Z., Hu, Y., Shou, L. and Song, M. (2013) Isolation and partial characterization of cyclic lipopeptide antibiotics produced by Paenibacillus ehimensis B7. BMC Microbiology

13, 1-8.

Kaneda, T. (1967) Fatty acids in the genus Bacillus. I. Iso- and anteiso-fatty acids as characteristic constituents of lipids in 10 species. Journal of bacteriology 93, 894-903.

Kaneda, T. (1991) Iso- and anteiso-fatty acids in bacteria: biosynthesis, function, and taxonomic significance. Microbiological reviews 55, 288-302.

Kim, J., Il Kim, P., Bong, K.M., Il Kim, J., Shin, S.Y., Song, J., Min, H.J. and Lee,

C.W. (2018) Isolation and structural elucidation of pelgipeptin E, a novel pore-forming pelgipeptin analog from Paenibacillus elgii with low hemolytic activity. The Journal of

Antibiotics.

115

Moon, S.H., Zhang, X., Zheng, G., Meeker, D.G., Smeltzer, M.S. and Huang, E.

(2017) Novel Linear Lipopeptide Paenipeptins with Potential for Eradicating Biofilms and

Sensitizing Gram-Negative Bacteria to Rifampicin and Clarithromycin. Journal of Medicinal

Chemistry 60, 9630-9640.

Mosmann, T. (1983) Rapid colorimetric assay for cellular growth and survival:

Application to proliferation and cytotoxicity assays. Journal of Immunological Methods 65,

55-63.

Ng, J., Bandeira, N., Liu, W.T., Ghassemian, M., Simmons, T.L., Gerwick, W.H.,

Linington, R., Dorrestein, P.C. and Pevzner, P.A. (2009) Dereplication and de novo sequencing of nonribosomal peptides. Nat Methods 6, 596-599.

Ortega, D.B., Costa, R.A., Pires, A.S., Araújo, T.F., Araújo, J.F., Kurokawa, A.S.,

Magalhães, B.S., Reis, A.M.M., Franco, O.L., Kruger, R.H., Pappas, G.J. and Barreto, C.C.

(2018) Draft Genome Sequence of the Antimicrobial-Producing Strain Paenibacillus elgii

AC13. Genome Announcements 6.

Qian, C., Liu, T., Zhou, S., Ding, R., Zhao, W., Li, O. and Wu, X. (2012)

Identification and functional analysis of gene cluster involvement in biosynthesis of the cyclic lipopeptide antibiotic pelgipeptin produced by Paenibacillus elgii. BMC Microbiology 12, 1-

Qian, C.-D., Wu, X.-C., Teng, Y., Zhao, W.-P., Li, O., Fang, S.-G., Huang, Z.-H. and

Gao, H.-C. (2011) Battacin (Octapeptin B5), a new cyclic lipopeptide antibiotic from

Paenibacillus tianmuensis active against multidrug-resistant Gram-negative bacteria.

Antimicrobial agents and chemotherapy 56, 1458-1465.

Sugawara, K., Konishi, M. and Kawaguchi, H. (1984) BMY-28160, a new peptide antibiotic. The Journal of antibiotics 37, 1257-1259.

116

Takeuchi, Y., Murai, A., Takahara, Y. and Kainosho, M. (1979) The structure of permetin A, a new polypeptin type antibiotic produced by Bacillus circulans. The Journal of antibiotics 32, 121-129.

Weber, T., Blin, K., Duddela, S., Krug, D., Kim, H. U., Bruccoleri, R., Lee, S. Y.,

Fischbach, M. A., Müller, R., Wohlleben, W., Breitling, R., Takano, E. & Medema, M. H

(2015). antiSMASH 3.0 — a comprehensive resource for the genome mining of biosynthetic gene clusters. Nucleic Acids Research (2015) doi: 10.1093/nar/gkv437.

Wu, X.C., Shen, X.B., Ding, R., Qian, C.D., Fang, H.H. and Li, O. (2010) Isolation and partial characterization of antibiotics produced by Paenibacillus elgii B69. FEMS microbiology letters 310, 32-38.

117

Figures

Figure 1. RP-HPLC chromatogram of pelgipeptins’ isoforms showing the supernatant of P. elgii AC13 cultured in Nutrient broth, after organic extraction with butanol. Analyses were conducted using a C18 Shim- pack VP-ODS column (150×4.6 mm, 4.6 µm; Shimadzu). Flow rate of 1 ml·min-1, after optimized method. The elution was monitored at 216 nm for 50 minutes. Solvent A: water + 0.1 % TFA. Solvent B: acetonitrile + 0.1 % TFA. [M + H]+ 1105 (I); [M + H]+ 1119 (II); [M + H]+ 1073 (III); [M + H]+ 1087 (IV); [M + H] + 1087 (V); and [M + H]+ 1101 (VI).

118

Figure 3. MALDI-ToF MS/MS spectra from the precursor ion [M+H]+ 1105 (Pelgipeptin C’). Sequences found for -b and -y series.

Figure 4. MALDI-ToF MS/MS spectra from the precursor ion [M+H]+ 1119 (Pelgipeptin B’). Sequences found for series -b and -y.

119

Figure 6. MALDI-ToF/MS spectra of Pelgipeptin C before alkaline hydrolysis (A) and after alkaline hydrolysis (B). [M+H]+ 1087.9 m/z; [M+Na]+ 1109.9 m/z; [M+K]+ 1125.9.

120

Figure 8. Domain organization in pelgipeptin synthase gene plpE of P. elgii AC13 compared to the complete gene in P. elgii B69 presented at Ding et al. (2011).

121

122

Tables

Primary Sequence Identity R [M+H]+ Characteristic References (H2N-Xn-COOH)

Dab1-Val2-Dab3-Phe4-Leu5- (Cochrane and Vederas, 2016; Kim et al., Pelgipeptin A (CH ) CH 1073 Natural, Cyclic Dab6-Val7-Leu8-Ser9 3 2 2018)

Dab1-Ile2-Dab3-Phe4-Leu5- Pelgipeptin D (CH ) CH 1087 Natural, Cyclic (Ding, Wu, et al., 2011) Dab6-Val7-Leu8-Ser9 3 2

Dab1-Val2-Dab3-Phe4-Leu5- Pelgipeptin E (CH ) CH 1073 Natural, Cyclic Kim et al. 2018 Dab6-Val7-Leu8-Ser9 2 2 3

Paenipeptin A/ Dab1-Val2-Dab3-Phe4-Leu5- (CH ) COOH 1105 Natural, Linear (Huang et al., 2017); this work Pelgipeptin C’ Dab6-Val7-Leu8-Ser9 2 6

Paenipeptin B/ Dab1-Leu/Ile2-Dab3-Phe4-Leu5- (CH ) COOH 1119 Natural, Linear (Huang et al., 2017); this work Pelgipeptin B’ Dab6-Val7-Leu8-Ser9 2 6

Dab1-Leu/Ile2-Dab3-Phe4-Leu5- Paenipeptin C (CH ) COOH 1133 Natural, Cyclic (Huang et al., 2017) Dab6-Val7-Leu8-Ser9 2 7

Table 2. fractions found in P. elgii strain AC13 nutrient broth cultures. Retention Time (RT); Molecular Weight [M+H] +; Lipid Tail (R); Amino acid sequence (MS/MS); Characteristic, and Identity.

Fraction RT [M+H]+ Lipid Tail MS/MS Characteristic Identity

Dab1-Val2-Dab3-Phe4-Leu5- I 25’30’’ 1105 aiC Linear Pelgipeptin C’/ Paenipeptin A 4 Dab6-Val7-Leu8-Ser9

Dab1-Ile2-Dab3-Phe4-Leu5-Dab6- II 28’ 1119 aiC Linear Pelgipeptin B’/ Paenipeptin B 4 Val7-Leu8-Ser9

iC Dab1-Val2-Dab3-Phe4-Leu5- III 31’ 1073 3(Cochrane and Vederas, Cyclic Pelgipeptin A(Wu et al., 2010) 2016) Dab6-Val7-Leu8-Ser9

aiC Dab1-Val2-Dab3-Phe4-Leu5- Pelgipeptin C (BMY-28160)(Ding, IV 32’40’’ 1087 4(Cochrane and Cyclic Vederas, 2016) Dab6-Val7-Leu8-Ser9 Wu, et al., 2011)

iC Dab1-Ile2-Dab3-Phe4-Leu5-Dab6- Pelgipeptin D(Ding, Wu, et al., V 34’ 1087 3(Cochrane and Vederas, Cyclic 2016) Val7-Leu8-Ser9 2011)

aiC Dab1-Ile2-Dab3-Phe4-Leu5-Dab6- Pelgipeptin B (Permetin A)(Wu et VI 35’40’’ 1101 4(Cochrane and Cyclic Vederas, 2016) Val7-Leu8-Ser9 al., 2010)

123

Linear Cyclic pelgipeptin Controls pelgipeptin Strains A B C D A-D B’ C’ Ampicillin Chloramphenicol Polymyxin 1073 1101 1087 1087 Mix 1119 1105

E. coli ATCC11229 6.25 12.5 6.25 6.25 12.5 50 >100 25 ND* 3.1

S. aureus 12.5 12.5 12.5 12.5 6.25 >100 >100 ND* 12.5 25 ATCC14458 *non-determined

Table 4. Alignment of the nucleotide and protein sequences of pelgipeptin gene cluster of P. elgii AC13 in comparison to the one from P. elgii B69.

Nucleotide Protein gene Function E-value Identity E-value Identity

plpA 0.0 1270/1281 (99%) 0.0 425/426 (99%) diaminobutyrate-2-oxoglutarate aminotransferase

plp B 0.0 1529/1563(98%) 0.0 512/520 (98%) esterase and lipase

plp C 0.0 720/735(98%) 0.0 242/244(99%) 4'-phosphopantetheinyl transferase

plp D 0.0 4454/4527(98%) 0.0 1484/1508(98%) nonribosomal peptide synthetase

plpE ND* ND* ND* ND* nonribosomal peptide synthetase

plp F 0.0 3263/3318(98%) 0.0 1091/1105(99%) nonribosomal peptide synthetase

plp G 0.0 1758/1785(98%) 0.0 344/347(99%) ABC transporter

plp H 0.0 1666/1719(97%) 0.0 563/572(98%) ABC transporter

* Non- determined

124

Supplementary Figures

Figure S1. Chromatogram of Pelgipeptin Mix A-D stock solution before (A) and after (B) treatment with butanol. Analysis were conducted using a C18 Shim-pack VP-ODS column (4.6µm, 150×4.6mm; Shimadzu). Flow rate of 0,6 ml·min-1, linear gradient 5-95% B. Solvent A: water + 0.1% TFA. Solvent B: acetonitrile + 0.1% TFA. The elution was monitored at 216nm. There are no variations in retention time or chromatography profile.

125

126