Two Types of Threonine-Tagged Lipopeptides Synergize in Host

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Two Types of Threonine-Tagged Lipopeptides Synergize in Host Colonization by Pathogenic Burkholderia Species Tawatchai Thongkongkaew, Wei Ding, Evgeni Bratovanov, Emilia Oueis, Maria Garcia-Altares, Nestor Zaburannyi, Kirsten Harmrolfs, Youming Zhang, Kirstin Scherlach, Rolf Müller, and Christian Hertweck ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.8b00221 • Publication Date (Web): 18 Apr 2018 Downloaded from http://pubs.acs.org on April 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Two Types of Threonine-Tagged Lipopeptides Synergize in Host Colonization by Pathogenic Burkholderia Species Tawatchai Thongkongkaew1, Wei Ding2, Evgeni Bratovanov1, Emilia Oueis2, Marίa GarcίaAltares1, Nestor Zaburannyi2, Kirsten Harmrolfs2, Youming Zhang3, Kirstin Scherlach1, Rolf Müller*2, Christian Hertweck*1,4 1

Leibniz Institute for Natural Product Research and Infection Biology, HKI, Beutenbergstr. 11a,

07745 Jena, Germany 2

Department of Microbial Natural Products, Helmholtz Institute for Pharmaceutical Research

Saarland, Helmholtz Centre for Infection Research and Department of Pharmaceutical Biotechnology, Saarland University, Campus E8.1, 66123 Saarbrücken, Germany 3

Shandong University–Helmholtz Institute of Biotechnology, State Key Laboratory of Microbial

Technology, School of Life Science, Shandong University, Shanda Nanlu 27, 250100 Jinan, People's Republic of China. 4

Chair for Natural Product Chemistry, Friedrich Schiller University, 07743 Jena, Germany

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

ABSTRACT Bacterial infections of agriculturally important plants and mushrooms pose a major threat to human food sources worldwide. However, structures of chemical mediators required by the pathogen for host colonization and infection remain elusive in most cases. Here, we report two types of threonine-tagged lipopeptides conserved among mushroom and rice pathogenic Burkholderia species that facilitate bacterial infection of hosts. Genome mining, metabolic profiling of infected mushrooms and heterologous expression of orphan gene clusters allowed the discovery of these unprecedented metabolites in the mushroom pathogen Burkholderia gladioli (haereogladin, burriogladin), and the plant pathogen Burkholderia glumae (haereoglumin, and burrioglumin). Through targeted gene deletions, the molecular basis of lipopeptide biosynthesis by non-ribosomal peptide synthetases was revealed. Surprisingly, both types of lipopeptides feature unusual threonine tags, which yield longer peptide backbones than one would expect based on the canonical co-linearity of the NRPS assembly lines. Both peptides play an indirect role in host infection as biosurfactants that enable host colonization by mediating swarming and biofilm formation abilities. Moreover, MALDI imaging mass spectrometry was applied to investigate the biological role of the lipopeptides. Our results shed light on conserved mechanisms that plant and mushroom pathogenic bacteria utilize for host infection and expand current knowledge on bacterial virulence factors that may represent a new starting point for the targeted development of crop protection measures in the future.

2 ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION The success of many pathogenic bacteria relies on their highly evolved strategies to colonize and weaken or even kill their hosts.1 To achieve this, pathogens may use an armory of virulence factors, including diverse inhibitors, lytic enzymes and toxins. Beyond these chemical mediators that are directly related to disease development, small molecules may indirectly contribute to the infection process by promoting cell motility and adhesion, a prerequisite to occupy diverse niches and adapt to hosts.2, 3 This ability is particularly pronounced in bacteria belonging to the genus Burkholderia, which occur in highly diverse environments and in a broad range of symbiotic associations.4,

5

Numerous Burkholderia species cause serious infections; for example, Burkholderia pseudomallei is the causative agent of melioidosis, which has high lethality rates in animals and humans,6 whereas members of the Burkholderia cepacia complex (BCC) cause serious pulmonary infections in immunocompromised patients.7 Apart from mammalian infection, other Burkholderia species occur in plant-pathogenic alliances with mold fungi or manage to invade plants individually.8 Burkholderia glumae, Burkholderia gladioli, and Burkholderia plantarii cause grain rot, sheath rot, and seedling blight in rice, respectively, and lead to huge economic losses in agriculture.9, 10 Likewise, Burkholderia species pose a major threat in the mushroom industry. The most widely consumed mushroom, the white button mushroom Agaricus bisporus, is infected by B. gladioli pv. agaricicola.3 In the early stage of infection, small brown spots are observed in the mushroom cap. Within 24 hours, these spots increase size to form brown holes, which has led to the term cavity disease (Figure 1A). Initially observed in farms in New Zealand and Europe,11 the same symptoms have been observed worldwide in other commercially important mushrooms such as Lentinula edodes in Japan and Pleurotus ostreatus in Korea.12 Notably, cavity disease develops very fast and the mushroom of the whole farm will be degraded in a short period. Furthermore, there is no efficient method to control the bacterial infection once it has set in.13 Yet, it is known that severe infections occur when relative humidity is high, and controlling air ventilation can minimize the spreading of the disease.14 Current knowledge on the molecular basis of cavity disease15 cannot explain these observations. Here we shed light on the molecular basis of host colonization by B. gladioli and elucidate two families of novel lipopeptide metabolites encoded in orphan gene clusters that are highly 3 ACS Paragon Plus Environment


Page 4 of 25

conserved among plant- and mushroom-pathogenic Burkholderia species. Furthermore, functional analyses and imaging mass spectrometry reveal that the lipopeptides play different roles in the infection process – and that unusual threonine tags are pivotal for the observed biological functions.

RESULTS AND DISCUSSION Conserved gene loci in Burkholderia species To gain insight into gene loci that are potentially involved in virulence, we mined the genomes of diverse B. gladioli pathovars. The genome of B. gladioli pv. agaricicola was sequenced using 454 sequencing. The nucleotide sequence of this bacterium was aligned with the sequences of human and plant associated Burkholderia bacteria obtained from NCBI. Interestingly, all genomes of these B. gladioli strains shared two cryptic gene clusters that code for non-ribosomal peptide synthetase (NRPS) assembly lines that likely produce two different types of lipopeptides. These orphan gene clusters were considered to be promising starting points for functional analyses, since a) lipopeptides such as jagaricin and tolaasin have been found to play key roles in mushroom rot diseases,16, 17 b) one gene cluster (bgd) is located next to genes for tentatively virulence-related chitinase and a type II secretion system in B. gladioli pv. agaricicola,15 and c) homologous gene clusters are present in related plant-pathogenic Burkholderia species, Burkholderia glumae and Burkholderia plantarii (Figure 1B and 1C). In silico analysis by antiSMASH18 of the NRPS gene clusters provided first clues about the structures of the putatively encoded secondary metabolites. According to the bioinformatic predictions, the assembly line in charge of biosynthesis of haereogladin (hgd, see below) consists of 5 modules. The first module starts with a condensation domain (C), a hallmark of lipopeptide synthetases. Analysis of the A domain substrate specificity indicated that the first and the second domains incorporate threonine (Thr) as substrates. However, the remaining building blocks could not be deduced from the bioinformatic analysis. Thus, we expected a lipopeptide with the general architecture FA-Thr-Thr-X-X-X (FA is fatty acid). The second gene cluster (for biosynthesis of burriogladin, bgd, see below) in the B. gladioli genome tentatively codes for an NRPS consisting of seven modules with an N-terminal C domain. Through analysis of the A domain substrate specificity we predicted the product to represent a heptapeptide with an FA4 ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Thr-Pro-Glu-Ala-Val-X-Pro backbone. Similarly, the putative structures of lipopeptides from the homologous haereogladin and burriogladin gene clusters conserved in other Burkholderia spp. were predicted by in silico analyses (Table S1 –S6). We followed two complementary approaches to link these orphan gene clusters to products: 1) The genomics-based detection and isolation of lipopeptides from infected mushroom tissue was attempted for B. gladioli pv. agaricicola; 2) Since the NRPS gene clusters proved to be silent in B. glumae, we performed promotor exchange and heterologous expression experiments.

Detection of lipopeptides To probe the production of toxins during cavity disease, white button mushrooms (A. bisporus) were infected with B. gladioli pv. agaricicola. After 16 hours, yellow slimy holes were observed in the mushroom tissues. Analysis of the crude EtOAc extract by high-performance liquid chromatography (HPLC) and high-resolution mass spectrometry (HR-MS) indicated that several new compounds were formed during the infection (Figure 1D, i and ii). HR-MS and MS/MS analysis revealed that these compounds belong to two different families of lipopeptides that match well with the expected secondary metabolites. To unequivocally assign the hgd and bgd gene loci to the observed compounds, we created targeted knockout mutants in B. gladioli pv. agaricicola. Specifically, we employed a Red/ET recombination strategy to integrate an apramycin resistance cassette (aac(3)IV) within the respective NRPS-encoding genes, thus disrupting their normal function.19 HPLC monitoring of the cultures of the wild type and the mutants corroborated that the hgd gene locus codes for haereogladin, whereas burriogladin results from a pathway encoded by the bgd gene cluster (Figure 1D, iii– vi). In contrast to B. gladioli, metabolic profiling of B. glumae under various cultivation conditions did not show any related compounds. As it appeared that the corresponding NRPS gene clusters (hgm and bgm) are silent under these laboratory conditions, we attempted their activation via promoter exchange in the native host and heterologous expression. For this purpose, we first tried to directly exchange promoters in the native host (B. glumae). Inactivation and activation of the hgm gene cluster from Burkholderia PG1 were based on the Pseudomonas recombineering system (pBBR1-km-RhaB-GBAS)20 (Figure 2A). The resulting mutant of the activated cluster PG1:Papra-C4 was derived from the apramycin resistance gene in 5 ACS Paragon Plus Environment


Page 6 of 25

lieu of the original promoter (1,344,392–1,344,745). This resulted in the production of two lipopeptidic compounds, which are not produced by the wild type and are closely related to haereogladin, differing only in two amino acid residues (Figure 2E). Since this strategy has not led to the activation of the bgm gene cluster, we directly cloned the gene cluster from genomic DNA and connected the operon to the tetracycline inducible promoter Ptet (Figure 2B–2D). After heterologous expression in E. coli we detected production of two new lipopeptides, which are closely related to burriogladins, differing in two amino acid residues (Figure 2F).

Identification of two new lipopeptide families Since the lipopeptides were only produced in low amounts on rotten mushroom, we screened 15 types of media to increase the production and thus to obtain enough material for full structure elucidation. Eventually, potato dextrose agar (PDA) cultures were used for the largescale production of the B. gladioli metabolites. After incubation the agar plates were extracted, and the crude extract was separated by a combination of various chromatographic techniques. From a total volume of 2 L PDA, we obtained four new lipopeptides, haereogladins A (12.9 mg), B (2.2 mg), C (0.3 mg), and burriogladin A (3.5 mg) (Figure 3A). Besides these major metabolites, additional congeners, namely haereogladin D (0.8 mg), haereogladin E (0.9 mg) and burriogladin B (0.2 mg) were isolated from further culture experiments. The recombinant B. glumae mutant PG1:Papra-C4 was cultured and Amberlite XAD16 resin was added. The cells and XAD were extracted and purification afforded haereoglumins A and B (Figure 3A). In parallel, the heterologous host E. coli/pDMW3 was inoculated and the culture incubated before induction of the Ptet promoter with tetracycline. The XAD resin was added. The cells and resin were extracted, and purification afforded burrioglumins A and B (Figure 3A). Their structures were elucidated by analysis of 1D- and 2D-NMR, HR-MS, MS/MS data. The molecular formula of haereogladin A was deduced as C44H54N6O12 from HRESI-MS. The 13C and DEPT-135 spectra showed signals of 7 carbonyl, 2 olefinic methine, 12 aromatic methine, 8 quaternary olefinic carbons, 5 methine, 6 methylene, and 4 methyl protons. Analysis of the 1

H,1H-COSY correlations revealed a spin system of octanoate (Oct), two units of

dehydrobutyrine (Dhb), β-OH-tyrosine (β-OH-Tyr), p-hydroxyphenyl glycine (p-Hpg), p-amino benzoate (PABA) and threonine (Thr). Based on the HMBC correlations, the peptide sequence 6 ACS Paragon Plus Environment

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


was proposed as Oct/Dhb /Dhb/β-OH-Tyr/p-Hpg/PABA/Thr. Tandem MS measurements corroborated this data. The configuration of the Dhb units was determined as E by NOESY. Moreover, the NOE correlations of H-2/H-3, H-3/NH, and OH/NH in β-OH-Tyr subunit suggest R,R or S,S relative configurations (Figure 3B). The 1H NMR data of haereogladin B is similar to that of haereogladin A, except for the presence of the methylene protons at δH 3.02 and 2.91 ppm, respectively. These protons were assigned to the Tyr subunit based on HMBC correlations and this conclusion is in agreement with the MS/MS fragmentation pattern. Thus, haereogladin B contains Tyr instead of β-OH-Tyr. Similarly, NMR and MS/MS data revealed that haereogladin C differs from haereogladin A by lacking the terminal threonine subunit, whereas haereogladin D proved to be the Tyr-congener of haereogladin C. Haereogladin E showed the same molecular formula and MS/MS fragmentation pattern as haereogladin A indicating that both compounds are structural isomers. The 1H NMR chemical shifts of the threonine moiety of haereogladin E appeared in lower field compared to haereogladin A, and the NH amide proton (δH 7.87 ppm in haereogladin A) was missing indicating that the threonine in haereogladin E is connected via an ester bond. Haereoglumin A has the molecular formula C39H60N6O9, as deduced from HRESI-MS. 1H,13CHSQC spectra showed signals of 2 olefinic methine, 4 aromatic methine, 7 methine, 8 methylene, and at least 6 methyl protons. Analysis of the 1H,1H-COSY and 1H,1H-TOCSY correlations revealed a spin system of octanoate (Oct), two units of dehydrobutyrine (Dhb), two units of leucine (Leu), p-amino benzoate (PABA) and threonine (Thr). Based on the HMBC correlations

and

MS/MS

measurements,

the

peptide

sequence

was

proposed

as

Oct/Dhb/Dhb/Leu/Leu /PABA/Thr. The configuration of the Dhb units was determined as E from ROESY. The threonine linkage was found to be an amide with the NH correlating with the β-proton (1H,1H-TOCSY in DMSO) and further corroborated by 15N HSQC spectrum. The 1H NMR chemical shifts of the threonine Cα of haeroglumin A is within the same field range compared to haereogladin (A and B), whereas the Cα of threonine in haereogladin E with the ester linkage appears in lower field. The minor compound haereoglumin B was obtained in lower quantities and its structure was determined as Oct/Dhb/Dhb/Leu/Leu (supplemental information).



Page 8 of 25

The absolute configuration of the peptide backbone of haereogladins and haereoglumins was determined by Marfey’s analysis (Table S7–S11). In haereogladins, the peptide sequence was established using haereogladin B as L-Thr, L-p-Hpg, and D-Tyr. The absolute configuration at the α position of Tyr is R from Marfey’s analysis. Since the NOE data suggested either an R,R or S,S configuration of OH/NH in the β-OH-Tyr subunit, the β position in the β-OH-Tyr unit in haereogladins A, C, and E is proposed as R configuration. The presence of

D-Tyr

is in full

agreement with the in silico analysis of the haereogladin synthetase, despite the absence of designated epimerization domains. By a phylogenetic analysis of the condensation (C) domains in the haereogladin NRPS and 239 C domains of known NRPS 21, we found that the C domain of module 4 belongs to the group of dual E/C domains which perform condensation and epimerization (Figure 3C). The C4 of haereogladin NRPS converts L-Tyr from module 3 to D-Tyr and catalyzed a peptide bond formation between D-Tyr and L-p-Hpg, which comes from the downstream module (Figure 4A). Similarly, the absolute configuration of peptide backbones of both haereoglumins A and B was determined as D- and L- Leu (in A and B) and L-Thr (in A). The distinction between the D- and L-Leu within the sequence was proposed by the phylogenetic analysis of C domains in haereoglumin NRPS, where the C4 belongs to the group of dual E/C domains. p-Aminobenzoate is well known to be a component of folic acid,22 but only rarely used as a building block in non-ribosomal peptides.23 The few known PABA-derived peptides include the antibacterial cystobactamides from the bacterium Cystobacter sp.24 and the DNA gyrase inhibitor

albicidin

from

Xanthomonas

albilineans.25

Another

uncommon

feature

of

haereogladins A-C is the presence of two moieties of configurationally instable Edehydrobutyrine (E-Dhb). Moreover, the tandem E-Dhb structure is unprecedented for natural products.26 The molecular formula of burriogladin A was deduced by HRESI-MS as C49H68N8O12 . Analysis of 13C and DEPT-135 spectral data revealed signals of 9 carbonyls, 1 olefinic methine, 9 aromatic methines, and 3 quaternary olefinic carbons which account for 20 degrees of unsaturation suggesting that burriogladin A is a linear peptide. Analysis of the 1H,1H-COSY correlations revealed the presence of β-OH-decanoate (β-OH-Dec), dehydrobutyrine (Dhb), glutamine (Gln), alanine (Ala), p-hydroxyphenyl glycine (p-Hpg) and phenylalanine (Phe). In addition, two units of proline (Pro) were proposed based on the characteristic chemical shifts of 8 ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


nitrogen substituted methine protons at δH 4.29 and 4.20 ppm, respectively. The peptide sequence was deduced as β-OH-Dec/Dhb/Pro/Gln/Ala/p-Hpg/Phe/Pro through HMBC measurements and confirmed by MS/MS fragmentation studies. The configuration of the Dhb unit was determined as Z from NOESY correlations. The minor metabolite burriogladin B was only obtained in minute amounts, which prevented its full structure elucidation by NMR. Therefore, the structure was inferred based on HRMS and MS/MS fragmentation studies (supplemental information). Burrioglumin A has the molecular formula C48H74N8O13, as deduced from HRESI-MS. The 13C and DEPT-135 spectra showed signals of 8 carbonyl, 2 quaternary olefinic carbons, 5 aromatic methine, 1 olefinic methine, 10 methine, 15 methylene, and 6 methyl protons. Analysis of the 1

H,1H-COSY and 1H, 1H-TOCSY correlations revealed the presence of β-OH-decanoate (β-OH-

Dec), dehydrobutyrine (Dhb), serine (Ser), alanine (Ala), valine (Val), phenylalanine (Phe), and threonine (Thr), as well as two units of proline (Pro). Based on HMBC correlations and MS/MS measurements,

the

peptide

sequence

was

proposed

to

be

β-OH-

Dec/Dhb/Pro/Ser/Ala/Val/Phe/Pro/Thr. The configuration of the Dhb units was determined as Z from NOESY correlations. Burrioglumin B shows a similar 1H NMR data as burrioglumin A, except for the presence of a leucine instead of the valine. This was also corroborated by HRESI-MS and MS/MS fragmentation studies. By Marfey’s analysis, we determined the configuration of the burriogladin A building blocks to be

D-

and L-Pro, L-Phe,

D-p-Hpg, L-Ala,

and

D-Gln.

In analogy to the

haereogladin synthetase, the burriogladin NRPS does not contain any designated epimerization domains. However, the C domain phylogeny revealed that the C domains of modules 3, 4, and 6 are dual E/C domains, which is in line with the finding of D-Pro, D-Gln, and D-p-Hpg. From the phylogeny the position of the two enantiomeric Pro bricks could be inferred (Figure 4B). Likewise, the burrioglumins A and B building blocks were determined as D- and L-Pro, D-Ser, LAla, L-Phe, L-Thr, and

D-Val

for burrioglumin A and

D-Leu

for burrioglumin B. From the

phylogeny analysis, the position of the two enantiomeric Pro bricks could be identified as only the C3 domain belongs to the group of dual E/C domains (Figure 3C). The absolute configuration of the OH group of β-hydroxydecanoic acid moiety in burriogladin A was determined R by GC-MS analysis (Figure 3D). Similarly, the absolute configuration of fatty acid in burrioglumins A and B was established as R from LC-MS analysis of 9 ACS Paragon Plus Environment


Page 10 of 25

R- and S-Mosher acid chloride derivatives of an authentic R-sample and enantiomeric mixtures (Figure S1).

Non-colinearity in lipopeptide assembly lines It is remarkable that both families of linear lipopeptides include congeners with longer peptide backbones than expected. According to the rule of colinearity in multimodular NRPS assembly lines, the number of amino acid building blocks correlates with the number of A domain containing modules. In contrast, haereogladins A, B, E and haereoglumin A are composed of six amino acids, while the corresponding NRPS has only five adenylation domains (Figures 2A and 4A). Likewise, the heptamodular burriogladin and burrioglumin NRPSs give not only rise to the heptapeptides, but also to octapeptides. In all cases, the expected size of the backbone and the produced peptides differ by one C-terminal threonine unit. The possibility of a missed NRPS module encoded somewhere else in the genomes could be ruled out since burrioglumins obtained from heterologous expression also featured the additional Thr unit. The discrepancy between the number of adenylation domains in NRPS and the number of amino acids in peptide backbones could be rationalized by the repetitive use of a single module. Such iterative modules have been identified in some NRPS assembly lines.27-29 However, as to the identified Burkholderia lipopeptides, this scenario is rather unlikely because of the different structures of the building blocks incorporated. Furthermore, in most lipopeptides the Thr units are linked through amide bonds, whereas ester linkage was observed only in haereogladin E. An alternative scheme would involve a post-NRPS modification mediated by a freestanding enzyme such as a ligase or trans-acting C domain in analogy to, for example, peptidyl nucleoside biosynthetic pathways, such as pacidamycin.30 However, the hgd/hgm and bgd/bgm gene clusters do not contain genes that could potentially code for similar enzymes. Furthermore, the heterologous expression experiments clearly rule out this possibility. Consequently, the most plausible scenario is that terminal threonines are introduced during offloading. It is well conceivable that the thioester bond is cleaved by nucleophilic attack of the threonine hydroxy or amino groups, thus yielding ester- and amide-linked Thr tags. Although chain termination by water is commonly observed, the termination by an amine is rare. Only one example was reported in the biosynthesis of pacidamycin, where the TE domain has been shown to recruit the external nucleophile 3΄-deoxy-4΄, 5΄-enamino-uridine for peptide 10 ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


chain termination.31 Therefore, it is possible that the TE domains use threonine in lieu of H2O (Figure 4). Irrespective of the exact mechanism, the Thr tags are pivotal for the biological function of the lipopeptides, as outlined below.

Roles of lipopeptides in host colonization The ecological functions of lipopeptides were tested in a mushroom infection model. To evaluate their potentially damaging effects, we applied solutions of the lipopeptides to the mushroom tissue (Figure 5A). As positive control, we used tolaasin I, a known virulence factor in mushroom brown blotch disease that causes brown holes in the tissue.16 Notably, neither haereogladin A-C nor burriogladin A caused damage to the mushroom tissue (Figures 5A and S3). Consequently, haereogladin and burriogladin do not function like the potent antifungal lipopeptides tolaasin32 and jagaricin.17 This finding was also corroborated by antimicrobial screening assays, in which neither peptide displayed any antifungal activity (supplemental information). Alternatively, the B. gladioli lipopeptides may function as surfactants that contribute indirectly to the development of mushroom cavity disease by promoting host colonization, bacterial cell movement and biofilm formation.

To test their surfactant properties, we first measured the decrease of the contact angle of a water droplet on a hydrophobic surface, which correlates with the surfactant's ability to reduce surface tension. According to a standardized setup, we determined the contact angles for each lipopeptide and plotted the results in a graph as a function of concentration (Figure 5B). This graph shows that haereogladins A, B and burriogladin A have surfactant properties. Haereogladin C, which lacks the threonine tag, did not display tensioactive activity, indicating that the threonine tag is critical for the surfactant property of haereogladin. In addition, critical micelle concentrations (CMC), a characteristic of lipopeptides, were determined from the point where the discontinuity occurred. The CMC values for haereogladins A, B, and burriogladin are 0.64, 1.12, and 0.99 mM, respectively. The lower CMC value indicates the stronger surfactant property of the lipopeptide. In general, the CMC value is inversely proportional to the ability to penetrate cell membranes. Thus, haereogladins A, B and burriogladin have moderate membrane penetration ability when they were compared to those 11 ACS Paragon Plus Environment


Page 12 of 25

reported of membrane pore forming lipopeptides tolaasin (CMC 0.24 mM), cormycin A (CMC 138 μM), and surfactin (CMC 7.5 μM).33-35 To determine the possible impact of the lipopeptides on the motility of the bacterial cells, we performed a swarming assay on soft PDA (0.5% agar) using the wild type and the two knockout mutants deficient in each lipopeptide. Whereas the wild type swarmed out to colonize a large area, the haereogladin-deficient mutant occupied a substantially smaller area. In stark contrast, the burriogladin mutant completely lacked the ability to swarm (Figure 5C and 5D). These findings unequivocally show that burriogladin plays a pivotal role for swarming motility of B. gladioli. Swarming is the first step in the colonization of the mushroom and a prerequisite for biofilm formation36. With the wild type and null mutants in hand, we next investigated biofilm formation ability in 96-well plates under static condition. The biofilm attached to the well plate was stained with crystal violet, dissolved in aqueous acetic acid, and quantified by measuring the absorbance. We found that the burriogladin mutant which lacks the ability to swarm, produced the highest amount of biofilm, whereas the haereogladin null mutant produced the lowest amount of biofilm (Figure 5E). Interestingly, the biofilm formation rate of the burriogladin mutant was even higher than the one of the wild type. Thus, haereogladin promotes biofilm formation. To visualize the molecular basis of the colonization process, we used MALDI imaging mass spectrometry. In the wild type the production of haereogladin A and burriogladin A is concentrated specifically at the edge of the swarming colony (Figure 6A–6D). The bacterial cells at the edge need biosurfactants for movement and colonization of new surface area. A similar scenario was described for the production of the biosurfactant surfactin, which accumulates at the edge of the swarming colony of Bacillus subtilis.37

Taken together, these results unequivocally show that haereogladins and burriogladins play important, indirect roles in the mushroom infection process, as they are a prerequisite of swarming motility and promote biofilm formation. Indeed, biosurfactants have been known to mediate various plant-bacteria interactions with both, mutualistic or pathogenic outcome. The production of tensioactive compounds may increase the ecological fitness of the bacteria enabling movement on hydrophobic plant surfaces such as the leaf cuticles or changing the 12 ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


viscosity of the surface through a wetting effect. Biosurfactants may support colonization processes of roots and alter the availability of nutrients.38 The plant pathogenic bacterium B. glumae produces rhamnolipids with surfactant properties that are essential for its swarming ability.39 Similarly, B. plantarii was shown to produce a rhamnolipid that is structurally related to rhamnolipid exotoxins produced by pathogenic bacteria such as Pseudomonas aeruginosa and Burkholderia pseudomallei40 suggesting that biosurfactants also play important roles in the interaction with other higher organisms. In addition, the bioinformatic analysis of the conserved haereogladin and burriogladin gene clusters in Burkholderia bacteria shows that the rice pathogenic bacteria B. glumae, B. plantarii and B. gladioli have the ability to produce the same types of lipopeptides (Figure 1B and Table S1 and S2). Most likely, haereoglumins and burrioglumins also play similar roles in plant infections. As for the mushroom pathogens, we showed for the first time that the lipopeptides have individual functions in a bacterial pathogen. Haereogladin supports biofilm formation, while burriogladin facilitates bacterial cells’ movement by swarming. Our results support the observation that the mushroom infection relates to high humidity. The movement and attachment of the pathogen require biosurfactants for lowering surface tension of water droplet.41, 42 When the water droplet becomes flat and covers the mushroom surface, the bacteria can spread from an infected mushroom to a healthy mushroom in order to search for new food sources and colonize new habitats.

CONCLUSION Bacterial colonization of eukaryotic hosts is a common scenario in many ecosystems and represents the prerequisite of various mutualistic and pathogenic relationships among different organisms. Despite the importance of these processes in the environment, little is known about the mechanisms and chemical mediators involved. Our studies identified two different families of lipopeptides conserved among mushroom- and plant-pathogenic Burkholderia species that facilitate host colonization by enabling swarming and adhesion. Through genome mining, heterologous expression and gene deletion experiments, previously cryptic biosynthetic gene clusters could be linked to these novel lipopeptides. Surprisingly, both peptide families are derived from a non-colinear NRPS and decorated with unusual threonine tags that are essential for their biological activity. Whereas one family of linear lipopeptides mediates bacterial 13 ACS Paragon Plus Environment


Page 14 of 25

swarming, the other one enables biofilm formation. The functional role of these compounds was further visualized through MALDI imaging mass spectrometry. Our results may open new possibilities to control bacterial infections of agriculturally important crops.


Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURES AND CAPTIONS

Figure 1. Mushroom cavity disease and analysis of orphan gene clusters. a) Infection of white button mushroom with the bacterium B. gladioli pv. agaricicola. b) 16S rRNA–based phylogenetic tree showing conserved haereogladin and burriogladin gene clusters. c) Comparative genomes analysis and conserved biosynthetic gene clusters. d) HPLC chromatograms at λmax 254 nm of crude extract from i) healthy mushroom, ii) infected mushroom, iii) potato dextrose agar (PDA), iv) bacterial culture on PDA, v) haereogladin deficient mutant on PDA, and vi) burriogladin deficient mutant on PDA.



Page 16 of 25

Figure 2. Genetic manipulation of NRPS gene clusters of B. glumae. a) The map of haereoglumin gene cluster in chromosome 1. b) Burrioglumin biosynthesis gene cluster. Arrows indicate the region that should be cloned. Restriction sites marked with red color were used to release this cluster. c) Linear plus linear recombination to get construct pDMW2 containing the burrioglumin gene cluster. d) Linear plus circular recombination in E. coli GBdir to modify pDMW2 to pDMW3. e) Production of haereoglumin i) with inducible promoter, ii) without promoter, iii) wild type. f) Production of burrioglumins, i) extracted mass trace of burrioglumin A in extract of E. coli + bgm, ii) extracted mass trace of burrioglumin B in extract of E. coli + bgm, iii) extracted mass trace of burrioglumin A in extract of E. coli + ∆bgm, iv) extracted mass trace of burrioglumin B in extract of E. coli + ∆bgm.


Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Structure elucidation of new lipopeptides. a) Haereogladins A-E (1–5) and burriogladins (6–7) from B. gladioli pv. agaricicola. Haereoglumins A-B (8–9) and burrioglumins (10–11) from B. glumae. b) Key COSY, HMBC, and NOESY correlations of 1 and 6. c) Phylogenetic analysis of condensation domains (C) of haereogladin (hgd, filled triangles), burriogladin (bgd, filled squares), haereoglumin (hgm, empty triangles) and burrioglumin (bgm, empty squares) NRPSs and known C domain subtypes (LCL, modified amino acid, starter, epimerization, heterocyclization, dual E/C, DCL). d) GC–MS analysis of fatty acid in burriogladin A; i) Methyl (R,S)-3-hydroxyldecanoate; ii) Methyl (R,S)-3-hydroxyldecanoate; iii) Hydrolysate from burriogladin A.



Page 18 of 25

Figure 4. NRPS module architecture and model of biosynthetic pathways. a) Haereogladin assembly line. b) Burriogladin assembly line. The TE domains can use both water and threonine as external nucleophile for lipopeptide chain release. The same biosynthetic logic takes place in the homologous haereoglumin and burrioglumin NRPSs.


Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Biological functions of lipopeptides. a) Mushroom pathogenicity test. Tolaasin I is a positive control. b) Contact angle measurement of aqueous solutions of lipopeptides. Tween 80 is a positive control. c) Swarming phenotypes of wild type (WT) as well as haereogladin (Δhgd) and burriogladin (Δbgd) deficient mutants. d) Swarming diameter. e) Biofilm formation. *p value < 0.05 is considered as significantly different.



Page 20 of 25

Figure 6. MALDI imaging of the swarming colony. a) Optical image of the swarming colony. The measurement area is within black triangle. b) Unsupervised spatial segmentation of two regions. The region of interest is in red. c) A representative MS spectra from region of interest. d) Visualization of haereogladin A and burriogladin A ([M+K]+ ion). The relative intensity is color-coded according to the legend. n = Number of mass spectra


Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ASSOCIATED CONTENT Supporting Information The supporting information including experimental detail and compounds characterization is available free of charge on the ACS publications website at DOI: AUTHOR INFORMATION Corresponding Authors *[email protected]; [email protected]

Author contributions K.S., R.M., C.H. designed and guided all aspects of this work. In the B. gladioli project T.T. performed chemical experiments and biological assays; E.B. performed genetic manipulation and bioinformatics analysis; M.G-A. performed MALDI imaging experiments and supported statistical analyses. In the B. glumae project, W.D. and Y.Z. performed genetic manipulation; N.Z. performed bioinformatics analysis; K.H and E.O. performed chemical studies. All authors contributed to manuscript preparation. C.H. wrote the final version. Conflict of interest The authors declare no competing financial interest.

ACKNOWLEDGEMENT We thank the collaborators at HKI for the following tasks: H. Heinecke and A. Perner for NMR and MS measurements, F. Kloss for GC-MS measurements and R. Hermenau for providing tolaasin I. The DAAD is gratefully acknowledged for a Ph.D. fellowship (for T. Thongkongkaew). This work was financially supported by the DFG (SFB 1127 ChemBioSys, and Leibniz Prize to C.H.). M. Garcίa-Altares is grateful for financial support from the ERC for a Marie SklodowskaCurie Individual Fellowship (IF-EF) Project reference 700036. Funding for the work in R. Müller group is from the Budesministerium für Bildung und Forschung (BMBF 0315820).



Page 22 of 25

REFERENCES [1] Beceiro, A., Tomás, M., and Bou, G. (2013) Antimicrobial resistance and virulence: a successful or deleterious association in the bacterial world?, Clin. Microbiol. Rev. 26, 185230. [2] Venturi, V., and Fuqua, C. (2013) Chemical signaling between plants and plant-pathogenic bacteria, Annu. Rev. Phytopathol. 51, 17-37. [3] Scherlach, K., Graupner, K., and Hertweck, C. (2013) Molecular bacteria-fungi interactions: effects on environment, food, and medicine, Annu. Rev. Microbiol. 67, 375-297. [4] Compant, S., Nowak, J., Coenye, T., Clément, C., and Ait Barka, E. (2008) Diversity and occurrence of Burkholderia spp. in the natural environment, FEMS Microbiol. Rev. 32, 607-626. [5] Vial, L., Groleau, M.-C., Dekimpe, V., and Déziel, É. (2007) Burkholderia diversity and versatility: an inventory of the extracellular products, J. Microbiol. Biotechnol. 17, 14071429. [6] Wiersinga, W. J., Currie, B. J., and Peacock, S. J. (2012) Melioidosis, N. Engl. J. Med. 367, 10351044. [7] Mahenthiralingam, E., Urban, T. A., and Goldberg, J. B. (2005) The multifarious, multireplicon Burkholderia cepacia complex, Nat. Rev. Microbiol. 3, 144-156. [8] Angus, A. A., Agapakis, C. M., Fong, S., Yerrapragada, S., Estrada-de los Santos, P., Yang, P., Song, N., Kano, S., Caballero-Mellado, J., de Faria, S. M., Dakora, F. D., Weinstock, G., and Hirsch, A. M. (2014) Plant-associated symbiotic Burkholderia species lack hallmark strategies required in mammalian pathogenesis, Plos One. 9, e83779. [9] Ham, J. H., Melanson, R. A., and Rush, M. C. (2011) Burkholderia glumae: next major pathogen of rice?, Mol. Plant. Pathol. 12, 329-339. [10] Seo, Y.-S., Lim, J. Y., Park, J., Kim, S., Lee, H.-H., Cheong, H., Kim, S.-M., Moon, J. S., and Hwang, I. (2015) Comparative genome analysis of rice-pathogenic Burkholderia provides insight into capacity to adapt to different environments and hosts, BMC Genomics. 16, 349. [11] Gill, W., and Cole, A. (1992) Cavity disease of Agaricus bitorquis caused by Pseudomonas cepacia, Can. J. Microbiol. 38, 394-397. [12] Lee, C., Yun, H., Jhune, C., Cheong, J., and Yoo, Y. (2010) Occurrence of bacterial soft rot of Pleurotus ostreatus caused by Burkholderia gladioli pv. agaricicola in Korea, J. Plant Pathol. 92, 235-240. [13] Largeteau, M. L., and Savoie, J. M. (2010) Microbially induced diseases of Agaricus bisporus: biochemical mechanisms and impact on commercial mushroom production, Appl. Microbiol. Biotechnol. 86, 63-73. [14] Fletcher, J. T., and Gaze, R. H. (2007) Mushroom pest and disease control: a colour handbook, pp 104-105, CRC Press. [15] Chowdhury, P. R., and Heinemann, J. A. (2006) The general secretory pathway of Burkholderia gladioli pv. agaricicola BG164R is necessary for cavity disease in white button mushrooms, Appl. Environ. Microbiol. 72, 3558-3565. [16] Andolfi, A., Cimmino, A., Cantore, P. L., Iacobellis, N. S., and Evidente, A. (2008) Bioactive and structural metabolites of Pseudomonas and Burkholderia species causal agents of cultivated mushrooms diseases, Persp. Med. Chem. 2, 81-112. [17] Graupner, K., Scherlach, K., Bretschneider, T., Lackner, G., Roth, M., Gross, H., and Hertweck, C. (2012) Imaging mass spectrometry and genome mining reveal highly 22 ACS Paragon Plus Environment

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


antifungal virulence factor of mushroom soft rot pathogen, Angew. Chem. Int. Ed. 124, 13350-13354. [18] Weber, T., Blin, K., Duddela, S., Krug, D., Kim, H. U., Bruccoleri, R., Lee, S. Y., Fischbach, M. A., Müller, R., and Wohlleben, W. (2015) antiSMASH 3.0—a comprehensive resource for the genome mining of biosynthetic gene clusters, Nucl. Acids Res. 43, W237-W243. [19] Zhang, Y., Buchholz, F., Muyrers, J. P., and Stewart, A. F. (1998) A new logic for DNA engineering using recombination in Escherichia coli, Nat. Genet. 20, 123-128. [20] Swingle, B., Bao, Z., Markel, E., Chambers, A., and Cartinhour, S. (2010) Recombineering Using RecTE from Pseudomonas syringae, Appl. Environ. Microbiol. 76, 4960-4968. [21] Rausch, C., Hoof, I., Weber, T., Wohlleben, W., and Huson, D. H. (2007) Phylogenetic analysis of condensation domains in NRPS sheds light on their functional evolution, BMC Evol. Biol. 7, 78. [22] Bermingham, A., and Derrick, J. P. (2002) The folic acid biosynthesis pathway in bacteria: evaluation of potential for antibacterial drug discovery, Bioassays 24, 637-648. [23] Walsh, C. T., Haynes, S. W., and Ames, B. D. (2012) Aminobenzoates as building blocks for natural product assembly lines, Nat. Prod. Rep. 29, 37-59. [24] Baumann, S., Herrmann, J., Raju, R., Steinmetz, H., Mohr, K. I., Hüttel, S., Harmrolfs, K., Stadler, M., and Müller, R. (2014) Cystobactamids: myxobacterial topoisomerase inhibitors exhibiting potent antibacterial activity, Angew. Chem. Int. Ed. 53, 14605-14609. [25] Cociancich, S., Pesic, A., Petras, D., Uhlmann, S., Kretz, J., Schubert, V., Vieweg, L., Duplan, S., Marguerettaz, M., and Noëll, J. (2015) The gyrase inhibitor albicidin consists of paminobenzoic acids and cyanoalanine, Nat. Chem. Biol. 11, 195-197. [26] Siodłak, D. (2015) α, β-Dehydroamino acids in naturally occurring peptides, Amino Acids 47, 1-17. [27] Süssmuth, R. D., and Mainz, A. (2017) Nonribosomal peptide synthesis—principles and prospects, Angew. Chem. Int. Ed. 56, 3770-3821. [28] Juguet, M., Lautru, S., Francou, F. X., Nezbedova, S., Leblond, P., Gondry, M., and Pernodet, J. L. (2009) An iterative nonribosomal peptide synthetase assembles the pyrrole-amide antibiotic congocidine in Streptomyces ambofaciens, Chem. Biol. 16, 421-431. [29] Jaitzig, J., Li, J., Sussmuth, R. D., and Neubauer, P. (2014) Reconstituted biosynthesis of the nonribosomal macrolactone antibiotic valinomycin in Escherichia coli, ACS Synth. Biol. 3, 432-438. [30] Zhang, W., Ntai, I., Bolla, M. L., Malcolmson, S. J., Kahne, D., Kelleher, N. L., and Walsh, C. T. (2011) Nine enzymes are required for assembly of the pacidamycin group of peptidyl nucleoside antibiotics, J. Am. Chem. Soc. 133, 5240-5243. [31] Zhang, W., Ostash, B., and Walsh, C. T. (2010) Identification of the biosynthetic gene cluster for the pacidamycin group of peptidyl nucleoside antibiotics, Proc. Natl. Acad. Sci. U.S.A. 107, 16828-16833. [32] Scherlach, K., Lackner, G., Graupner, K., Pidot, S., Bretschneider, T., and Hertweck, C. (2013) Biosynthesis and mass spectrometric imaging of tolaasin, the virulence factor of brown blotch mushroom disease, ChemBioChem 14, 2439-2443. [33] Hutchison, M. L., and Johnstone, K. (1993) Evidence for the involvement of the surface active properties of the extracellular toxin tolaasin in the manifestation of brown blotch disease symptoms by Pseudomonas tolaasii on Agaricus bisporus, Physiol. Mol. Plant Pathol. 42, 373-384. [34] Hamley, I. W. (2015) Lipopeptides: from self-assembly to bioactivity, Chem. Commun. 51, 8574-8583. 23 ACS Paragon Plus Environment


Page 24 of 25

[35] Scaloni, A., Dalla Serra, M., Amodeo, P., Mannina, L., Vitale, R. M., Segre, A. L., Cruciani, O., Lodovichetti, F., Greco, M. L., Fiore, A., Gallo, M., D'Ambrosio, C., Coraiola, M., Menestrina, G., Graniti, A., and Fogliano, V. (2004) Structure, conformation and biological activity of a novel lipodepsipeptide from Pseudomonas corrugata: cormycin A, Biochem. J. 384, 25-36. [36] Ron, E. Z., and Rosenberg, E. (2001) Natural roles of biosurfactants, Environ. Microbiol. 3, 229-236. [37] Debois, D., Hamze, K., Guérineau, V., Le Caër, J. P., Holland, I. B., Lopes, P., Ouazzani, J., Séror, S. J., Brunelle, A., and Laprévote, O. (2008) In situ localisation and quantification of surfactins in a Bacillus subtilis swarming community by imaging mass spectrometry, Proteomics 8, 3682-3691. [38] D'Aes, J., De Maeyer, K., Pauwelyn, E., and Hofte, M. (2010) Biosurfactants in plantPseudomonas interactions and their importance to biocontrol, Environ. Microbiol. Rep. 2, 359-372. [39] Nickzad, A., Lépine, F., and Déziel, E. (2015) Quorum sensing controls swarming motility of Burkholderia glumae through regulation of rhamnolipids, Plos One 10, e0128509. [40] Andra, J., Rademann, J., Howe, J., Koch, M. H., Heine, H., Zahringer, U., and Brandenburg, K. (2006) Endotoxin-like properties of a rhamnolipid exotoxin from Burkholderia (Pseudomonas) plantarii: Immune cell stimulation and biophysical characterization, Biol. Chem. 387, 301-310. [41] Angelini, T. E., Roper, M., Kolter, R., Weitz, D. A., and Brenner, M. P. (2009) Bacillus subtilis spreads by surfing on waves of surfactant, Proc. Natl. Acad. Sci. U.S.A. 106, 18109-18113. [42] Hennes, M., Tailleur, J., Charron, G., and Daerr, A. (2017) Active depinning of bacterial droplets: The collective surfing of Bacillus subtilis, Proc. Natl. Acad. Sci. U.S.A. 114, 59585963.


Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphical abstract 54x39mm (300 x 300 DPI)

ACS Paragon Plus Environment

Two Types of Threonine-Tagged Lipopeptides Synergize in Host

Recommend Documents