Biochimica et Biophysica Acta 1828 (2013) 2370–2378

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bbamem

Surfactin production enhances the level of cardiolipin in thecytoplasmic membrane of Bacillus subtilis

Gabriela Seydlová a,⁎, Radovan Fišer a, Radomír Čabala b,Petr Kozlík b, Jaroslava Svobodová a,⁎, Miroslav Pátek c

a Department of Genetics and Microbiology, Faculty of Science, Charles University in Prague, Viničná 5, 128 44 Prague 2, Czech Republicb Department of Analytical Chemistry, Faculty of Science, Charles University in Prague, Hlavova 8, 128 40 Prague 2, Czech Republicc Institute of Microbiology v.v.i., Academy of Sciences of the Czech Republic, Vídeňská 1083, 142 20 Prague 4, Czech Republic

⁎ Corresponding authors at: Department of Genetics andCharles University in Prague, Viničná 5, 128 44 Prague 2, Cze738; fax: +420 221 951 724.

E-mail addresses: seydlova@natur.cuni.cz (G. Seyd(J. Svobodová).

0005-2736/$ – see front matter © 2013 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.bbamem.2013.06.032

a b s t r a c t

a r t i c l e i n f o

Article history:Received 1 March 2013Received in revised form 21 June 2013Accepted 28 June 2013Available online 8 July 2013

Keywords:SurfactinBacillus subtilisMembranePhospholipidCardiolipin

Surfactin is a cyclic lipopeptide antibiotic that disturbs the integrity of the cytoplasmic membrane. In this study,the role of membrane lipids in the adaptation and possible surfactin tolerance of the surfactin producer Bacillussubtilis ATCC 21332 was investigated. During a 1-day cultivation, the phospholipids of the cell membrane wereanalyzed at the selected time points, which covered both the early and late stationary phases of growth, whensurfactin concentration in the medium gradually rose from 2 to 84 μmol·l−1. During this time period, thephospholipid composition of the surfactin producer's membrane (Sf+) was compared to that of itsnon-producing mutant (Sf−). Substantial modifications of the polar head group region in response to thepresence of surfactin were found, while the fatty acid content remained unaffected. Simultaneously withsurfactin production, a progressive accumulation up to 22% of the stress phospholipid cardiolipin wasdetermined in the Sf+ membrane, whereas the proportion of phosphatidylethanolamine remained constant.At 24 h, cardiolipin was found to be the second major phospholipid of the membrane. In parallel, the Laurdangeneralized polarization reported an increasing rigidity of the lipid bilayer. We concluded that an enhancedlevel of cardiolipin is responsible for the membrane rigidification that hinders the fluidizing effect of surfactin.At the same time cardiolipin, due to its negative charge, may also prevent the surfactin-membrane interactionor surfactin pore formation activity.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Surfactin is a lipopeptide antibiotic produced by Bacillus subtilis. It isformed of a cyclic heptapeptide interlinked with a β-hydroxy fatty acidcomprising 12–16 carbon atoms. Surfactin interacts with the membraneand disturbs its integrity [1]. This effect can explain a wide range ofsurfactin biological activities such as antibacterial [2–4], antiviral [5],antifungal [6], antimycoplasma [7], antiparasitic [8] and anti-tumor [9].Surfactin as a powerful biosurfactant stimulates biofilm formation by itsproducer strains [10,11]; however, inhibits the biofilm formation ofpathogenic bacteria [12,13].

The high demand for new chemotherapeutics driven by the increas-ing drug resistance of pathogens has drawn attention to antibioticlipopeptides that represent potential new antimicrobial agents. Anumber of systematic in vitro studies were therefore devoted touncovering the molecular mechanism of surfactin-induced membrane

Microbiology, Faculty of Science,ch Republic. Tel.:+420 221 951

lová), jarsvob@natur.cuni.cz

l rights reserved.

destabilization [14–18]. It was found that surfactin interacts with thelipid membrane in three different ways: (i) it acts as a mobile cationcarrier [19,20], (ii) forms cationic channels [21,22], and (iii) destroys themembrane via a detergent effect [16]. Membrane leakage was observedat a surfactin concentration of 2 μmol·l−1 [23], i.e. far below its criticalmicelle concentration (CMC), which ranges from 9 to 50 μmol·l−1,depending on the experimental conditions [24–26]. Channel formation,membrane lysis or solubilization of the membrane to micelles has beenobserved above the CMC of surfactin [21,22,27,28]. Although themolecu-lar mechanism of these complex lipopeptide-membrane interactions hasbeen intensively studied, it is not yet fully understood [22,28].

Several studies on model membranes demonstrate the impact ofthe lipid composition on surfactin-membrane interaction and itspenetration into the bilayer. Both the polar heads and fatty acidchains play a role in the formation of complexes of surfactin withphospholipids [29,30]. Fatty acids with shorter chains, which have asimilar length to the surfactin hydrocarbon chains, interact betterwith surfactin than fatty acids with longer chains like stearic acid[31]. The polar head group composition profoundly affects thebehavior of surfactin at the phospholipid–water interface and itsmiscibility with phospholipids. Both the electrostatic properties [30]and the volume of the phospholipid head groups modulate these

2371G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

surfactin–phospholipid interactions [29,32]. Phosphatidylethanolamine(PE) promotes surfactin stabilization in the model bilayer; presum-ably the inverted cone molecular shape of surfactin complementscone-shaped PE [27,31]. With dipalmitoylphosphatidylserine(DPPS), surfactin decreases the electrostatic repulsions betweenthe negatively-charged head groups of DPPS, which results inDPPS–surfactin stability in a monolayer [29]. Similarly, the presenceof cations that neutralize membrane surface negative chargesfacilitates the penetration of surfactin into the lipid bilayer [33].

All antibiotic-producing bacteria ensure their self-resistance bycoding for various means of self-defense mechanisms. At the sametime, the systematic use of antibiotics selects resistant variants ofthe susceptible target pathogens. As for the resistance mechanism,the antibiotic-producing strains are the most convenient source ofinformation [34]. Surprisingly, almost no research has been focusedon the nature of the surfactin resistance of the surfactin producer. Itis most likely the cytoplasmic membrane which is targeted bysurfactin both from the cytoplasmic and extracellular side that isthe site of self-resistance against surfactin. In Bacilli, resistance toantimicrobial compounds is mediated by the genes controlled bythe extracytoplasmic function sigma factor σw [35]. Nevertheless,none of these resistance genes were proven to be engaged in surfactinresistance. The only gene plausibly involved in surfactin resistance isswrC [36,37]. It codes for the proton-dependent multidrug effluxpump belonging to the RND family and contributes to the secretionof surfactin. However, surfactin production was observed even in aswrC-deficient strains that persistently survived at concentrationshigher than 10 mmol·l−1 [37]. This finding suggests the existenceof other additional mechanisms that participate in the surfactinself-resistance of the producer.

In this study, we investigated the modifications of the membranelipid moiety induced by surfactin during its production in order toestimate the role of phospholipids in the membrane adaptationresulting in surfactin tolerance. These changes were further corre-lated with fluidity of the cytoplasmic membrane estimated bymeans of Laurdan generalized polarization.

2. Materials and methods

2.1. Bacterial strains and growth conditions

Bacterial strains used in this study were B. subtilis ATCC 21332 (sfp+)(wild type, American Type Culture Collection) and B. subtilis strain 0164(sfp0) constructed in this work (see Section 2.3). B. subtilis strains weregrown aerobically (120 rpm) in nutrient broth (1 g Lab-Lemco powder,2 g Bacto yeast extract, 5 g Bactopepton, 5 g NaCl per liter) at 30 °C.Culture growth was monitored turbidimetrically at 420 nm. The culturewas inoculated with the overnight inoculum, diluted in the morningwith fresh medium and grown to the exponential phase (OD420 of 0.5to 0.6). Finally, the cells were used to inoculate the particular cultivation.After 3 h, bacterial cultures were harvested by rapid filtration through aSynpor no. 5 filter (Pragochema, Czech Republic). In the stationaryphase of growth, the cells were harvested by centrifugation (4300 ×g,15 min, 4 °C). The number of viable vegetative cells (CFU) was countedby plating diluted cultures on nutrient agar. In parallel, thermoresistantspores were counted after heat inactivation of the culture (70 °C,15 min) and estimated as CFU.

2.2. Transformation of B. subtilis

A total of 10 ml of pre-transformation medium (0.2% (NH4)2SO4,1.4% K2HPO4, 0.6% KH2PO4, 0.1% sodium citrate, 0.02% MgSO4, 0.5%glucose, 5% tryptophane, 0.02% casamino acids, 75 mmol·l−1 CaCl2and 2 mmol·l−1 MnSO4) was inoculated with a single colony andgrown at 37 °C until the culture reached the stationary phase ofgrowth. The culture was diluted 10-fold in fresh transformation

medium(0.2% (NH4)2SO4, 1.4%K2HPO4, 0.6%KH2PO4, 0.1% sodiumcitrate,0.02% MgSO4, 0.3% glucose, 0.0005% casamino acids). 0.2 μg of plasmidDNA and 1 ml of warm transformation medium were added to 100 μlof the diluted culture, and the tubes were incubated with shaking at37 °C for 1 h. The tubes were centrifuged for 3 min at 13,000 ×g, thepellet was resuspended in 300 μl of LB medium and the cells were platedon LB agar with chloramphenicol (5 μg·ml−1).

2.3. Construction of the sfp0 (Sf−) strain

Surfactin is produced by nonribosomal synthesis by the surfactinsynthetase complex coded by the srfA operon [38] and sfp gene. ThesrfA operon is also found in the chromosome of non-producing strainssuch as B. subtilis 168. However, the sfp gene, which is essential forsurfactin synthesis, carries a frameshift mutation in B. subtilis 168yielding a non-functional protein [39]. An isogenic strain defectivesurfactin synthesis (sfp0 mutant) based on the parental strain B.subtilis ATCC 21332 was constructed as follows. An internal fragmentcovering 784 bp of the sfp gene carrying the frameshift mutation wasamplified by PCR using B. subtilis 168 genomic DNA as a template andthe primers sfpF1 (TAGAATTCTTGTGGAAGTATGATAGGAT) and sfpR1(GAGAATTCAAGATATT GAGCGAGGTG). The amplified fragment wascloned in the EcoRI site of the vector pK19cat [40]. The resultingplasmid pK19cat-sfp0 was used to transform the B. subtilis ATCC21332 strain. The clones with an abolished ability to producesurfactin were identified by examination on blood agar according tothe absence of erythrocyte lysis. The presence of the frameshiftmutation within the sfp gene in the chromosome of these cloneswas verified by sequencing. The parental B. subtilis ATCC 21332 strainand its surfactin non-producing derivative are here designated theSf+ and Sf− strain, respectively.

2.4. Surfactin analysis

Samples of the culture were mixed with methanol (1:1, v/v) andcentrifuged at 10 000 ×g for 10 min to remove the cell biomass. Surfactinconcentration was determined in the cell-free supernatants byreverse-phase C18 HPLC using a HPLC Agilent 1200–Agilent 6460 TripleQuadrupole MS system equipped with a Zorbax Eclipse XDB-C18 column(5 μm, Agilent Technologies). The mobile phase consisted of (a) 0.1%formic acid in acetonitrile and (b) aqueous 0.1% formic acid at a ratio ofa/b 20%:80% (v/v) and the mobile phase flow rate used for the analysiswas 1 ml/min. Sample size was 50 μl. Five surfactin isoforms (C12–C16surfactin and sodium adduct ions) were detected at m/z 994.6 and1016.6, m/z 1008.6 and 1030.6, m/z 1022.6 and 1044.6, m/z 1036.6 and1058.6, m/z 1050.6 and 1072.6, respectively. Commercial surfactin(Sigma) served as a standard. The data were analyzed by AgilentMassHunter Workstation Data Acquisition and Agilent MassHunterData Analyses.

2.5. Membrane isolation

Cytoplasmicmembraneswere isolatedusing enzymedigestion,whichenables the membrane fraction to be purified from spore contamination[41]. The cells were incubated with lysozyme at the cultivation temper-ature of the initial culture (30 °C) for 30 min with exponential cells orfor 60 min with stationary cells. The cell lysate was centrifuged for10 min at 3000 ×g and 4 °C to remove the crude cell debris and spores,which remained unaffected by lysozyme. The cell-free supernatant wasthen centrifuged (25 000 ×g, 25 min, 4 °C) to separate the membranevesicles from the cytoplasmic fraction. The membrane sediment wasused directly for lipid extraction or resuspended in phosphate buffer(100 mmol·l−1, pH 6.6) and stored at −80 °C. Pierce BCA ProteinAssay (Pierce Biotechnology) was used for protein determination.

2372 G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

2.6. Phospholipid extraction and analysis

Membrane phospholipids (PLs) were extracted from B. subtilismembrane preparations with hexane-isopropanol (3:2, v/v)mixture [42]. After evaporation of the solvent in vaccuo at 40 °C,the PLs were dissolved in chloroform, filtered through glass fiberfilters (Whatman) and concentrated under a stream of nitrogen.The lipid samples were stored at −80 °C and used for polar headgroup or fatty acid analysis. Six PL classes – phosphatidylglycerol(PG), phosphatidylethanolamine (PE), phosphatidylserine (PS),cardiolipin (CL), lysylphosphatidylglycerol (lysylPG), and phospha-tidic acid (PA) – were separated by thin-layer chromatography(TLC, silica gel 60 G plates; Merck) in chloroform-methanol–water(65:25:4, v/v/v) as the mobile phase, collected from the plates andthen quantified spectrophotometrically, as described by Rouser etal. [43]. The average ± S.D. of results from three independentexperiments is presented.

2.7. Fatty acid analysis

Membrane lipids were transesterified to fatty acid methyl estersby incubation in sodium methoxide at room temperature [44].After neutralization by the addition of methanolic HCl, thefatty acid methyl esters (FAMEs) were extracted three times with200 μl of pentane and dried under the flow of nitrogen. Methylesters were dissolved in 100 μl of heptane and analyzed by gaschromatography–mass spectrometry (GC–MS) (Shimadzu GC17A–QP5050A gas chromatograph–mass spectrometer) in a DB-35column (Agilent Technologies; 32 m by 0.25 mm, film thickness1 μm). Electron impact spectra were obtained at 70 eV of electronenergy. The following operating conditions were used: injectiontemperature of 250 °C and initial oven temperature of 50 °C, risingat 7.5 °C/min to 250 °C, and isothermally for 20 min at 250 °C. TheFAMEs were identified with the aid of FAMEs standards (Sigma)and the identity was confirmed using the NIST Mass Spectral Library(2006). The average ± S.D. of results from three independentexperiments is presented below. The percentage of each FA wascalculated by the ratio of peak area/sum of total identified peakareas.

2.8. Laurdan generalized polarization

Laurdan generalized polarization (GP) measurements werecarried out using a confocal microscope (Leica TCS SP2). 2-Dimethylamino-6-lauroylnaphthalene (Laurdan, Molecular Probes) wasadded to the membrane samples (protein concentration of 75 μg·ml−1)in a phosphate buffer (100 mmol·l−1, pH 7) from a DMSO solution togive a final probe concentration of 10−6 mol·l−1. After incubation inthe dark at 41 °C for 45 min, the pelleted membranes were placed on acoverslip and GP was measured under the confocal microscope at 25 °Cusing 405-nm diode laser for excitation. Emission intensities of bluefluorescence (415–460 nm) and green fluorescence (460–540 nm)were acquired simultaneously into two separated channels using HCXPL APO CS 63.0 × 1.40 oil objective, 512 × 512 pixel format with400 Hz scan speed, 4× line averaging and 12-bit picture depth. Highvoltage of both photomultiplier tubes was kept constant at 600 V. Theintensity of recorded fluorescence was modified only by laser power.Approximately 50fluorescence photographswere collected for each sam-ple and processed using ImageJ software (http://rsbweb.nih.gov/ij/,http://www.uhnresearch.ca/facilities/wcif/) as follows: separated, greenand blue channels were corrected by subtraction of background intensi-ties (about 10%) deduced from autofluorescence of unlabeled samples.GP was calculated on a pixel by pixel basis from the blue and greenchannel intensities GP = (Iblue − Igreen) / (Iblue + Igreen). The imageareas containing membranes were selected by intensity thresholding

and mean values of GP were calculated for each image. The graphshows average GP value for all analyzed images ± S.D.

2.9. Statistics

Student's t test was used to establish differences between twomean values of PL and FA levels, GP and growth parameters. P valuesof b0.05 were considered significant.

3. Results

3.1. Growth of Sf+ and Sf− B. subtilis strains and surfactin production

In antibiotic-producing cells, the self-resistance mechanisms areactivated simultaneously with the biosynthesis of antibiotics. Theexpression of the defense systems subsequently increases as theproduction progresses. In order to characterize the adaptationmechanisms during surfactin production, the growth characteristicsof the wild-type surfactin producer (B. subtilis ATCC 21332, Sf+) andits non-producing mutant (B. subtilis 0164, Sf−) were examinedduring a 1-day cultivation in parallel with the determination ofsurfactin concentration in the growth medium.

As shown in Fig. 1, the doubling times of the surfactin producerand its non-producing mutant were both T = 37 ± 2 min. Bothcultures entered the stationary phase of growth (t0) after 4 h ofcultivation at log2 (OD420 × 1000) = 10.1 ± 0.1. These data indicat-ed that the sfp0 mutation resulting in the Sf− phenotype did not affectthe growth of the non-producing strain in the exponential phase.However within the stationary phase of growth, while surfactin wasaccumulating in the cultivation medium, wild-type B. subtilis ATCC21332 exhibited a slightly but reproducibly higher viability than theSf− strain. After 24 h of cultivation (t20), the cell biomass of the Sf+

and Sf− strains reached log2 (OD420 × 1000) 12.5 ± 0.1 and12.2 ± 0.2 (P b 0.009), respectively.

The slight difference in the growth capacity of the two strainsalso resulted in different total vegetative cell counts, which at t20reached 4.3 ± 0.2 × 108 ml−1 (Sf+) and 3.7 ± 0.1 × 108 ml−1

(Sf−) (P b 0.006), respectively. The proportion of thermoresistantspores was almost the same value in both cultures, 27% (Sf+) and26% (Sf−), respectively.

Based on the course of the growth curves of the two strains, thetime points for sampling and determining the surfactin concentrationin the cultivation medium were chosen. As documented in Table 1,the presence of surfactin at a concentration of 2.2 mg·l−1 was firstdetected after 4.5 h of cultivation (t0.5), i.e. as early as 30 min afterentry into the stationary phase, and it gradually accumulated up to87 mg·l−1 after 24 h (t20). Four surfactin isoforms with a fatty acidchain length of 13 to 16 carbon atoms were identified. TheC15-surfactin isoform was found to be the predominant constituent,whereas there were only traces of the C16 isoform. The relativeproportions of the surfactin isoforms remained stable over thewhole cultivation period. No surfactin was detected in the culturebroth of the Sf− strain.

The capacity of the cells to produce surfactin during the increasingnutritional stress of the stationary phase is illustrated by the amount ofsurfactin produced per vegetative cell (Table 1). Surfactin productionsteadily rose throughout the early stationary phase of growth (t0–t3).During this period, the amount of surfactin rose from 0.02 to 0.17 pgper cell and reached 0.21 pg per cell at t20.

The data presented above allowed us to select a time-scale for thesubsequent membrane analyses of cells adapted to their own toxicproduct which covered the whole range of cultivation stages andsurfactin concentrations. These were as follows: t−1, when theculture grew exponentially and surfactin was not yet synthetized. Atabout t0.5 the surfactin synthesis was initiated and progressed further(t1.5). We supposed that during this period, the defense mechanism

Fig. 1. Growth of B. subtilis ATCC 21332 (A) and 0164 (B) strains during 1-day cultivation.Cultures were grown in nutrient broth at 30 °C with shaking. Total bacterial counts wereassessed by plating on agar medium and expressed as colony forming units (CFU). Theproportion of thermoresistant spores in the samples was determined by colony countsafter exposure to a temperature of 70 °C for 15 min. The time point of entry into thestationary phase of growth is designated as t0. The data represent the means and standarddeviations from five (growth) or three (CFU) independent experiments.

2373G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

against the deleterious effect of surfactin was also being induced.Finally, at t3 and t20, the surfactin concentration was high enough tobe harmful to the producing cells and the resistance response shouldtherefore have been fully expressed. Since only growing vegetativecells produce surfactin, the spores present in the culture wereseparated prior to each analysis in order to exclude biochemicalcontamination of the membranes (see Section 2.5).

Table 1Concentration of surfactin in culture broth of B. subtilis ATCC 21332 during 1-day cultivatio

Time point in stationary phase Surfactin isoforms (mg·l−1)a

C13 C14 C15

t−1 0.0 0.0 0.0t0.5 0.2 ± 0.0 0.5 ± 0.0 1.4 ± 0.0t1.5 0.8 ± 0.0 0.9 ± 0.0 10.6 ± 0.1t3 3.7 ± 0.1 4.3 ± 0.2 34.5 ± 0.9t20 8.9 ± 0.2 8.1 ± 0.2 69.1 ± 0.2

Surfactin concentration in the cultivation mediumwas measured by LC/MS as described in Sperformed in duplicate.

a Surfactin isoform with fatty acid chain length of 13, 14, 15 and 16 carbon atoms, respeb Surfactin production in time expressed as the amount of surfactin (in pg) produced pe

3.2. Membrane phospholipid profiles

The polar head groups of membrane phospholipids can profoundlymodify the surfactin–membrane interaction. To estimate the role of themembrane surface in cell protection against surfactin, the changes in PLcomposition were monitored in the producing cells throughout the24-h cultivation. A parallel analysis of the Sf− mutant provided data onthe PL modifications that were induced during the stationary phase ofgrowth merely by increasing nutritional and energy depletion, withoutthe effect of surfactin. A detailed analysis of the PLprofiles of the Sf+ strainshowed that the adaptive response of the cells to surfactin accumulationin the culture medium developed gradually.

In the exponential phase of growth (t−1) in the absence of surfactin,the composition of membrane PLs (Table 2) of the two B. subtilis strainswas nearly identical. Phosphatidylglycerol, a principal lipid componentof the membrane, amounted to approximately 40% of the total. Thisbarrel-shaped phospholipid that stabilizes the lamellar phase ofmembrane bilayer predominated until t20, when the heavy stress of thelate stationary growth was combined with the rising concentration ofsurfactin in the producer strain. At this point, a drop in the PG contentfrom 38 to 16% was observed in the Sf+ strain, while the PG level wasreduced to a lesser extent (28%) in the Sf− strain. This 12% differencesuggested the involvement of PG in the adaptive response of the surfactinproducer, as its decrease correlated with the enhanced level of CL, whichis synthesized from two PG molecules.

Remarkably, in the Sf+ strain the CL content gradually increasedthroughout the stationary phase of growth and peaked at t20 whenthe surfactin stress culminated. The proportion of CL in themembrane increased from 13 to 22% (P b 0.01) at t20. Cardiolipinwas the second most abundant PL in the membrane of the Sf+ strainin the final stage of CL cultivation. In contrast, in the Sf− strain theproportion of CL was maintained at a constant level until t20, whenit fell to as little as 8%. This opposite tendency indicated that theincrease in CL content, similarly to the decrease of PG, was a specificresponse of the cell to surfactin.

During surfactin production the proportion of phosphatidylethanol-amine, another major representative of B. subtilis membrane PL, wasmaintained at a constant level of 22% until t3 in the Sf+ strain and at t20the PE content decreased to 11%. In contrast, in the Sf− strain the amountof PE rose to 30% at t3 and then fell to 16% at t20. Thus, the PE content in themembrane of the Sf+ strain was held under control during surfactin pro-duction. The level of the other aminophospholipid, phosphatidylserine,remained stable throughout the whole cultivation, including t20.

At t20, the polar head group composition of the membrane sampleschanged sharply, probably as a consequence of the continuing presenceof the antimicrobial concentration of surfactin combined with the deepnutritional exhaustion of the late stationary phase. Accordingly, thebiosynthesis of PL was strictly reduced and the content of PA, i.e. thecommon precursor of PL, dramatically rose from trace levels to 36 and32% (P b 0.009) in the Sf+ and Sf− strain, respectively. In fact, the fall inthe proportions of PG + PE in the membranes of the two strains

n.

Total Surfactin (pg) per vegetative cellb

C16

0.0 0.0 0.00.1 ± 0.0 2.2 ± 0.0 0.02 ± 0.000.2 ± 0.0 12.5 ± 0.1 0.08 ± 0.000.6 ± 0.0 43.1 ± 1.1 0.17 ± 0.011.1 ± 0.0 87.2 ± 0.2 0.21 ± 0.01

ection 2.4. The data represent the means ± S.D. from three independent measurements

ctively.r B. subtilis ATCC 21332 vegetative cell (CFU) during the stationary growth phase.

Table 2Phospholipid composition of B. subtilis cytoplasmic membrane.

Time (h) % of total phospholipid content (mean ± S.D.)

t−1 t0.5 t1.5 t3 t20

Strain ATCC 21332 0164 ATCC 21332 0164 ATCC 21332 0164 ATCC 21332 0164 ATCC 21332 0164

LysylPG 3.3 ± 0.3 2.8 ± 0.6 7.6 ± 1.1 5.6 ± 1.3 5.3 ± 1.0 4.5 ± 0.5 3.4 ± 0.6 5.5 ± 0.4 2.8 ± 0.6 1.5 ± 0.1PS 10.8 ± 1.0 11.1 ± 1.1 13.2 ± 0.0 13.3 ± 1.3 11.5 ± 1.0 11.8 ± 1.1 14.0 ± 0.8 11.6 ± 0.8 13.9 ± 0.3 13.5 ± 0.7PG 42.6 ± 0.6 43.5 ± 0.4 37.2 ± 0.7 38.7 ± 1.4 39.5 ± 0.6 43.3 ± 1.7 37.9 ± 1.4 36.0 ± 0.6 16.1 ± 1.3 28.3 ± 1.6PE 25.7 ± 1.5 25.1 ± 1.1 22.1 ± 1.8 21.4 ± 0.7 21.7 ± 1.1 26.5 ± 1.2 22.2 ± 1.1 29.9 ± 1.6 11.0 ± 1.2 16.4 ± 0.8CL 13.4 ± 0.7 13.8 ± 1.2 12.8 ± 0.7 13.3 ± 0.0 17.9 ± 1.3 13.6 ± 1.0 19.7 ± 0.7 14.1 ± 0.7 21.9 ± 0.5 8.4 ± 1.1PA 4.2 ± 0.5 3.8 ± 0.8 7.2 ± 1.2 7.8 ± 1.1 4.2 ± 0.8 3.2 ± 0.7 3.0 ± 0.1 2.9 ± 0.2 35.6 ± 0.5 31.9 ± 1.3

Values represent means ± S.D. from three determinations. LysylPG—lysylphosphatidylglycerol, PS—phosphatidylserin, PG—phosphatidylglycerol, PE—phosphatidylethanolamine,CL—cardiolipin, PA—phoshatidic acid.

2374 G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

accompanied by the drop in CL in the Sf− strain was quantitativelyreplacedwith PA. Nevertheless, despite the energy depletion and ongoingsurfactin production, the pathway leading to CL was still active in thesurfactin-producing cells and CL was the second most abundantmembrane PL at the end of the cultivation. This fact indicates itsimportance in the membrane bilayer of the Sf+ strain.

3.3. Membrane fatty acid composition

Phospholipid analyses revealed specific modifications in the polarhead groups of membrane phospholipids in the surfactin producer,suggesting the impact of the membrane surface composition on the cellprotection against surfactin. Similarly to PL head groups, the fatty acidsof themembrane, particularly the aliphatic chain length of phospholipids,can also strongly influence the surfactin–membrane interactions.Therefore, the FA profiles in the different stages of surfactin productionin the Sf+ strain were compared with its non-producing counterpart atthe same cultivation time points (Table 3).

At t−1, similarly to the polar head group of PL, the FA composition ofboth strains was the same with almost 90% proportion of thebranched-chain FAs. Of these, anteiso- and iso-branching FA accountedfor 60% and 30%, respectively. This distribution of FA in the cytoplasmicmembrane was almost unchanged in both strains until t3.

The late stationary phase (t20) was accompanied by alterations in fourmajor FAs in the membrane. These were branched-chain FAs withaliphatic chain lengths of 15 and 17 carbon atoms. The proportion ofhigh-melting rigidizing iso-series increased with a correspondingdecrease in low melting anteiso-branched-chain FAs (Fig. 2). At t20, the

Table 3Fatty acid composition of B. subtilis cytoplasmic membrane.

FA Tm (°C) % of total

t−1 t0.5 t1.5

ATCC 21332 0164 ATCC 21332 0164 ATCC

i-14:0 53 1.4 ± 0.1 2.4 ± 0.1 1.0 ± 0.1 1.4 ± 0.1 0.8 ±14:0 54 0.6 ± 0.1 0.7 ± 0.0 0.4 ± 0.0 0.3 ± 0.4 0.3 ±i-15:0 52 13.0 ± 0.5 12.3 ± 0.4 15.3 ± 1.1 13.4 ± 1.1 15.3a-15:0 23 42.0 ± 1.0 38.3 ± 0.5 41.7 ± 0.7 41.9 ± 0.1 40.515:0 53 0.2 ± 0.0 0 0.1 ± 0.1 0 0i-16:0 62 6.2 ± 0.1 9.9 ± 0.1 4.5 ± 0.3 6.5 ± 0.2 4.0 ±16:1 −0.5 1.2 ± 0.1 1.7 ± 0.4 1.4 ± 0.2 3.4 ± 0.2 1.2 ±16:0 63 5.2 ± 0.8 7.2 ± 0.1 3.9 ± 0.4 3.9 ± 0.2 3.9 ±i-17:0 60 9.3 ± 0.7 9.4 ± 0.3 11.0 ± 0.7 9.9 ± 0.1 11.1a-17:0 37 16.7 ± 1.4 14.6 ± 0.8 17.2 ± 0.7 16.2 ± 1.2 17.217:0 61 0 0.3 ± 0.2 0.2 ± 0.2 0 2.3 ±18:1 5 1.8 ± 0.2 1.4 ± 0.3 1.1 ± 0.8 0 1.0 ±18:0 71 2.1 ± 0.8 1.7 ± 0.1 1.9 ± 0.5 3.1 ± 0.2 2.4 ±

RatiosIso/anteiso 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ±Branched/straight 7.8 ± 0.8 6.6 ± 0.7 9.9 ± 1.6 8.3 ± 0.5 8.0 ±

Values represent means ± S.D. from three determinations. Iso and Anteiso stand for the re

total iso-branched-chain FAs rose from 36 to 47% in the Sf+ strain,whereas the content of these FAs rose from 39 to 56% of the total in theSf− strain. In the surfactin producer, the decline in the overall level ofanteiso-branched-chain FAs was caused by an 8% drop in the content ofa-15:0 to 33% and by an increase in the proportion of i-15:0 from 17 to24%. In the surfactin non-producing mutant the observed changes weremore profound, as the proportion of a-15:0 fell to 28% and i-15:0 roseto 30%. These changes led to a slight prevalence of the iso-branchingpattern in the Sf− strain and anteiso-branched chain FAs in the Sf+ strain.To a large extent these shifts in FA composition reflected the general latestationary phase stress, as a similar response was observed in bothB. subtilis strains irrespective of the surfactin production.

GC–MS analyses revealed no substantial differences between thesurfactin producing and non-producing B. subtilis strain in terms ofthe presence of acyl chain length profiles, the proportion of straightsaturated or unsaturated FAs. In contrast to the polar head groups ofmembrane PL, the FA composition in the surfactin producerdeveloped in a similar manner as in its non-producing counterpart,and the slight differences in the levels of iso-FAs are unlikely to bedue to the adaptation to surfactin.

3.4. Fluidity of the cytoplasmic membrane

The fluidity of cytoplasmic membranes isolated from the Sf+ andSf− strains was measured by means of Laurdan generalized polariza-tion. The fluorescent membrane probe Laurdan incorporates at thehydrophilic–hydrophobic interface of the phospholipid bilayer. It isknown to be sensitive to the polarity of the environment, exhibiting

t3 t20

21332 0164 ATCC 21332 0164 ATCC 21332 0164

0.0 0.8 ± 0.0 0.9 ± 0.2 1.2 ± 0.0 1.6 ± 0.1 1.3 ± 0.10.0 0.4 ± 0.2 0.2 ± 0.1 0.2 ± 0.0 0.4 ± 0.3 0.1 ± 0.0

± 1.6 15.1 ± 1.6 16.9 ± 0.5 19.8 ± 0.7 24.0 ± 0.6 30.2 ± 0.8± 0.6 39.5 ± 0.5 40.8 ± 1.9 39.7 ± 0.4 32.7 ± 1.4 27.5 ± 0.9

0 0.1 ± 0.1 0 0 2.1 ± 0.80.1 5.0 ± 0.2 4.0 ± 0.4 5.3 ± 0.4 6.6 ± 1.2 5.4 ± 0.30.0 1.2 ± 0.3 1.2 ± 0.8 1.1 ± 0.0 1.0 ± 1.0 0.5 ± 0.20.9 4.8 ± 0.4 3.2 ± 0.2 2.8 ± 0.0 4.6 ± 2.0 2.0 ± 0.2

± 0.0 11.3 ± 0.1 14.3 ± 0.1 13.1 ± 0.6 14.6 ± 1.4 18.3 ± 1.2± 0.5 17.7 ± 0.1 17.0 ± 0.8 15.6 ± 0.4 11.7 ± 1.9 9.9 ± 0.40.0 0 0 0 0 1.6 ± 0.70.2 0.9 ± 0.3 0.4 ± 0.1 0.5 ± 0.2 1.0 ± 1.4 0.2 ± 0.10.9 3.4 ± 0.3 1.0 ± 0.1 0.8 ± 0.0 1.8 ± 0.8 0.8 ± 0.6

0.0 0.6 ± 0.0 0.6 ± 0.1 0.7 ± 0.0 1.1 ± 0.0 1.5 ± 0.00.9 8.4 ± 0.5 16.9 ± 1.1 17.6 ± 0.8 14.6 ± 1.4 16.2 ± 1.4

spective branching pattern for iso- and anteiso-branched fatty acids, respectively.

Fig. 2. Proportions of iso- (circles) and anteiso-branched FAs (squares) in B. subtiliscytoplasmicmembrane during24-h cultivation. The time point of entry into the stationaryphase of growth is designated as t0. The data represent averages and S.D.

2375G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

a 50 nm red shift in emission spectrum over the gel to liquid-crystallinephase transition. Laurdan spectral shifts are usually quantified in theform of generalized polarization (GP), which is inversely proportional tomembrane fluidity (see Section 2.8). We used this technique in order tofind how the changes in the composition of the polar head groups ofmembrane PLs changed the respective biophysical characteristics ofmembranes isolated from B. subtilis.

From the data in Fig. 3, it is apparent that in the exponential phase ofgrowth (t−1) cytoplasmic membranes derived from both B. subtilisstrains showed the same values of Laurdan GP in line with the samePL composition. As the PL profile remained almost unchanged at t0.5,the same GP values were also obtained in the early stationary phase,and this tendency continued until t1.5 despite the subsequent changesin the polar head group composition. However, at t3 and t20, i.e. thetime points, at which major changes in lipid composition of the Sf+

strain were detected, GP values increased from about −0.02 to 0.02(P b 0.02) and 0.06 (P b 0.002), respectively. The increasing GP param-eter of the Sf+ membranes thus indicated a substantial rigidization ofthe membrane of the surfactin producer compared to the Sf− strain.At t20, this trend was much more profound than at t3. On the otherhand, the membrane fluidity of the Sf− strain remained almostunchanged despite the observed PL modifications. This indicates that

Fig. 3. GP parameter during different cultivation periods measured in B. subtilis ATCC21332 (full symbols) and 0164 (open symbols) membranes. The time point of entryinto the stationary phase of growth is designated as t0. The data represent averagesand S.D.

the cytoplasmic membrane rigidity significantly increased concomitantlywith the accumulation of surfactin.

4. Discussion

The aim of this study was to investigate the role of membranelipids in adaptation of B. subtilis ATCC 21332 to its own toxic productsurfactin. Surfactin is synthesized in cytoplasm and is transportedthrough the membrane by an unknown mechanism. This impliesthat it can interact with the membrane both from the cytoplasmicand extracellular side. However, surfactin clearly did not impair itsproducer, since the multiplication of bacteria continued in the culturein parallel with surfactin accumulation for almost 20 h. Within thisperiod the cell number increased almost fivefold while the surfactinconcentration increased from 2 to 84 μmol·l−1. Furthermore, themost potent C15 surfactin isoform [17] was also the most abundantone, amounting to 80% of the surfactin present. Surprisingly, duringthe production period, the Sf+ culture grew even faster than thecontrol Sf− population. This stimulatory effect of surfactin couldprobably stem from its quorum-sensing role which includes thetriggering of cannibalism, as observed recently during the biofilmdevelopment of B. subtilis [45]. The producer cells survivedlong-term exposure to surfactin concentrations that can perturb thebarrier properties in model membranes [23,28]. This is even morestriking when we consider that the minimum inhibitory concentrationsof surfactin determined for Salmonella enteritidis, Proteus vulgaris,Enterobacter cloacae, Bacillus pumilus and Escherichia coli are all withinthe range 6–30 μmol·l−1 [3,4,46]. These data suggested that bacteriaproducing surfactin are equipped with an efficient self-protectivemechanism. In searching for this mechanism, our primary attention wasfocused on membrane phospholipids, which are a well-documentedsurfactin target [16,29,33].

The interaction of surfactin with membrane PLs is hypothesized toinitiate as the insertion of individual surfactin molecules. This stepdoes not highly disorganize the PL bilayer, however after theinsertion of several other surfactin molecules in the membrane,pores can be formed and the mixed micelles of surfactin with PLscould lead to bilayer solubilization [17,22,32]. This model was alsosupported by the study of Nazari et al. [47] where surfactin wasthought to segregate within the membrane into surfactin-richclusters that disrupted the membrane locally. In our study, theamount of surfactin molecules in the medium per B. subtilis cell ofthe Sf+ strain ranged from 1.1 × 107 at t0.5 to 1.2 × 108 at t20.Therefore, during 1-day cultivation, we should also expect all of theabove-mentioned modes of surfactin action in the cells of thesurfactin producer, which however obviously resisted the adverseeffect of surfactin.

In the membrane of the surfactin producer, two specific featuresfound in the polar head group composition indicated their possiblerole in the tolerance to surfactin. First, cardiolipin was the only PL toprogressively accumulate over the whole 24-h cultivation (Table 2).At t20, the CL concentration reached 22% of the total and CL becamethe second most abundant PL. The significance of this preferentialsynthesis was underlined by the parallel decline of CL content down to8% in the Sf− strain in the same time point. Second, in contrast to CL,the level of PE was maintained strictly constant in the producer'smembrane, perhaps due to PE having a stabilizing effect on surfactinin the membrane [31]. This might be unfavorable as surfactin canaggregate in the membrane [47] and tilts the acyl chains of lipids [48].On the other hand, in the non-producing strain the proportion of PEsteadily increased, which is typical of the stationary phase of growth ofB. subtilis [49].

In contrast to the polar head groups of membrane PL, almost nochanges in the fatty acid composition were found. The observedtendency of the lower content of iso-branched-chain FAs in the Sf+

strain (Fig. 3) was unlikely to be due to an adaptive response to

2376 G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

surfactin. As an explanation of this effect we suggest a competitionfor leucine and valine between branched-chain FA synthesisand biosynthesis of surfactin. Leu and Val are precursors ofiso-branched-chain FAs [50], whereas surfactin contains 4 Leuand 1 Val residues in its heptapeptide cycle [51]. Such a weakadaptive response at the level of FAs found in the surfactinproducer was also published in the recent study on B. subtilisresistance to lipopeptide daptomycin, where the major determinant ofDapR was the composition of the polar head groups [52].

Cardiolipin is an unusual anionic lipid carrying four acyl chains andtwo phosphatidyl moieties. Its molecule consists of a large hydrophobicregion and strongly charged relatively small polar head group, whichimply that CL favors negative curvature and forms both lamellar andinverted non-lamellar lipid phases [53]. At the same time, the mobilityand conformational flexibility of CL should be severely restricted [54],which probably enhances the structural rigidity of CL-containingmembranes [55]. In bacteria, CL is regarded as a stress phospholipid[56]. Enhanced levels of it were observed in different bacteria in responseto various adverse conditions such as high salinity [57–59], alkaline pH[60] or the presence of chloramphenicol and tetracycline [61]. CL alsospecifically interacts with key protein components of the cell cycle[62,63]. Together with PE, it is involved in forming a mosaic ofmicrodomains in polar and septal membranes, i.e. in regions of increasedmembrane curvature [64,65]. Recently, a coordinated regulation of CL andPE levels was discovered in the inner mitochondrial membrane of yeastcells [66].

Different roles of increased CL content in the membrane of surfactinproducer could be thus hypothesized. Since CL is an anionic phospholipidbearing two negative charges, elevated concentration of CL may increasethe net negative charge of the membrane [62]. As the peptide cycle ofsurfactin also bears two negative charges, an electrostatic repulsioncould hinder the interaction of surfactin with the phospholipid headgroups [67]. The opposite coulombic mechanism is involved in the actionof cationic antimicrobial peptides, which preferentially bind to negativelycharged phospholipid membranes and thus increase their permeability[68]. Consequently, resistance of cells to these peptides is mediated byenhancing the net positive charge [69]. This mechanism has also beendescribed in B. subtilis resistant to positively charged lipopeptidedaptomycin, where reduced level of anionic phospholipid PG reducedthe net negative surface charge and weakened the interaction withdaptomycin [52].

Another possible mechanism of the membrane protection againstsurfactin can be deduced from the rigidizing effect of CL [55] on themembrane. In vitro surfactin induces fluidization of the phospholipidbilayer [27]. Therefore, it could be expected that an adaptationstrategy of the surfactin producer would include the increase of therigidity of its membrane. When we observed the parallel control ofCL and PE during surfactin production, we decided to study thechanges in the fluidity of the upper layer of the membrane as aprimary target of the interaction with surfactin. A membranefluorescence probe Laurdan was used for this. Both the fluorescentmoiety of Laurdan and surfactin are located at the glycerol backboneof the PL head groups [25,70]. It has been shown that the Laurdanfluorescence spectrum is sensitive to the polarity [71] and phasestate of the phospholipid bilayers [72]. From t1.5 onwards, our resultsshowed an increase in the general polarization (GP) of Laurdan duringsurfactin production, which documented an intense progressive increasein rigidity of the membrane (Fig. 3). This kinetics followed thedevelopment of the polar head group, which reflected the response tothe gradually rising surfactin stress in the producer cells. The LaurdanGP data were also confirmed by steady-state fluorescence anisotropy,rss, using TMA-DPH and DPH probe molecules that monitor the polar[73] andhydrophobic [74] regions of the lipid bilayer. The resulting valuesof rss TMA-DPH and rss DPH were significantly higher in the membranesof the Sf+ strain than those determined in the Sf− membranes, andindicated a higher rigidity of the producer membranes, both in the polar

head and aliphatic chain bilayer moieties (data not shown). On theother hand, the observed decrease of membrane fluidity was apparentlynot high enough to impair cell growth as the number of viable cellsincreased and surfactin production progressed throughout the stationaryphase of the Sf+ strain cultivation.

At t20, the key lipid component that contributed to the substantial riseof the rigidity in producer membrane was most likely CL, along with asevere decrease in PG, which is one of the PLs that has a fluidizing effect[75]. Accumulation of PA, the common precursor of PLs up to one thirdof the total suggested an imbalance in PA de novo synthesis and itsconversion into other lipid species [76] in themembranes of both strains,perhaps as a consequence of the energy depletion of cells. This notionwassupported by the substantial exhaustion of the PG pool for direct CLsynthesis, which surprisingly remained active in the Sf+ cells despitethe double surfactin and nutritional stress at t20 and indicated the specificneed for this PL.

The presence of the CL head group in the lipid membrane canresult in an increase in the order and tighter packing even in lipidacyl chains [77]. Hence, the increase in rss DPH in the PL aliphaticchain region that we observed in the Sf+ strain can also be ascribedto CL, although the FA composition only exhibited minor alterations.

The rigidizing effect of CL on the membrane bilayer has beendocumented both in computational simulations and in experimen-tal studies of model membranes using a wide variety of techniques[77–80]. The presence of CL in the membrane may also enhance theorder of the lipid bilayer [81]. The accumulation of CL with theconcomitant decrease in the proportion of PE is crucial for thelong-term adaptation of Pseudomonas putida to the presence oftoluene. These parallel changes in CL and PE lead to increased cellmembrane rigidity and should be regarded as physical mecha-nisms that prevent solvent penetration [82,83]. Thus, increasedCL content might be a general cell response due to membraneadjustment [57]. From this point of view, the rigidization of themembrane bilayer brought about by CL can be considered notonly as a side-effect of surfactin repulsion but also as a compensa-tory response of the cell preventing fluidization [27,84], disorder ofits barrier properties and solubilization [18] caused by surfactin.

Finally, one more function of CL in the adaptation of B. subtilisATCC 21332 to surfactin could be assumed from the complementaryshapes of CL and surfactin molecules. The cone-shaped CL can resultin the formation of aggregates with negative spontaneous curvature[55,85] which can counteract the positive curvature stress introducedby the inverted cone shape of surfactin [17,27]. CL was recentlyconsidered to force inverse cone-shaped lipids away from pores,thus stabilizing the bilayer [85,86]. An analogous CL mode of actionagainst surfactin pore expansion and rupture could be proposed asan additional mechanism increasing the membrane resistance ofsurfactin producer.

Despite systematic experimental effort being focused on CL, itsunique physical and chemical characteristics in cell membranes arenot well understood [55]. Moreover, a novel role of CL as a specificregulator of fundamental processes occurring at biomembranes hasemerged. P. putida defective in CL synthesis exhibited a compromisedfunctioning of the RND efflux pumps that were involved in antibioticand solvent resistance [61]. In mitochondria, the CL molecule candisrupt the supramolecular complex of the voltage-dependent anionchannel that plays a central role in apoptosis [87].

In this study, both the changes in the phospholipid compositionand physical properties of the membrane showed that an adaptiveresponse was induced during surfactin production in B. subtilis ATCC21332. We conclude that in this process, the enhanced level ofcardiolipin was the key factor which possibly brought about themembrane protection. Our preliminary data obtained with modelmembranes show that the presence of cardiolipin increasesmembrane stability after surfactin challenge (data not shown).However, the precise mechanism of the effect of cardiolipin on

2377G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

protection against surfactin and its role in the dynamic organizationof the cytoplasmic membrane of the surfactin producer remains tobe elucidated.

Acknowledgements

This work was supported by grants 13-18051P and P207/12/P890 from the Czech Science Foundation, by Institutional ResearchProject RVO61388971 of Institute of Microbiology and SVV project2013-267215. The authors thank Ivo Konopásek for critical readingof the manuscript.

References

[1] A.W. Bernheim, L.S. Avigad, Nature and properties of a cytolytic agent producedby Bacillus subtilis, J. Gen. Microbiol. 61 (1970) 361–366.

[2] P.A.V. Fernandes, I.R. de Arruda, A.F.A.B. dos Santos, A.A. de Araujo, A.M.S. Maior,E.A. Ximenes, Antimicrobial activity of surfactants produced by Bacillus subtilisR14 against multidrug-resistant bacteria, Braz. J. Microbiol. 38 (2007) 704–709.

[3] X.Q. Huang, X.P. Gao, L.Y. Zheng, G.Z. Hao, Optimization of sterilization ofSalmonella enteritidis in meat by surfactin and iturin using a response surfacemethod, Int. J. Pept. Res. Ther. 15 (2009) 61–67.

[4] P. Das, S. Mukherjee, R. Sen, Antimicrobial potential of a lipopeptide biosurfactantderived from a marine Bacillus circulans, J. Appl. Microbiol. 104 (2008)1675–1684.

[5] D. Vollenbroich, M. Ozel, J. Vater, R.M. Kamp, G. Pauli, Mechanism of inactivationof enveloped viruses by the biosurfactant surfactin from Bacillus subtilis, Biologi-cals 25 (1997) 289–297.

[6] P.I. Kim, J. Ryu, Y.H. Kim, Y.T. Chi, Production of biosurfactant lipopeptides iturin A,fengycin and surfactin A from Bacillus subtilis CMB32 for control of Colletotrichumgloeosporioides, J. Microbiol. Biotechnol. 20 (2010) 138–145.

[7] M.H. Hwang, M.H. Kim, E. Gebru, B.Y. Jung, S.P. Lee, S.C. Park, Killing rate curveand combination effects of surfactin C produced from Bacillus subtilis complexBC1212 against pathogenic Mycoplasma hyopneumoniae, World J. Microbiol.Biotechnol. 24 (2008) 2277–2282.

[8] I. Geetha, A.M. Manonmani, K.P. Paily, Identification and characterization of amosquito pupicidal metabolite of a Bacillus subtilis subsp. subtilis strain,Appl. Microbiol. Biotechnol. 86 (2010) 1737–1744.

[9] X.H. Cao, A.H. Wang, C.L. Wang, D.Z. Mao, M.F. Lu, Y.Q. Cui, R.Z. Jiao, Surfactin inducesapoptosis in human breast cancer MCF-7 cells through a ROS/JNK-mediatedmitochondrial/caspase pathway, Chem. Biol. Interact. 183 (2010) 357–362.

[10] D. Lopez, H. Vlamakis, R. Losick, R. Kolter, Paracrine signaling in a bacterium,Genes Dev. 23 (2009) 1631–1638.

[11] D. Lopez, M.A. Fischbach, F. Chu, R. Losick, R. Kolter, Structurally diverse naturalproducts that cause potassium leakage trigger multicellularity in Bacillus subtilis,Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 280–285.

[12] M. Nitschke, L.V. Araujo, S.G. Costa, R.C. Pires, A.E. Zeraik, A.C. Fernandes, D.M. Freire,J. Contiero, Surfactin reduces the adhesion of food-borne pathogenic bacteria to solidsurfaces, Lett. Appl. Microbiol. 49 (2009) 241–247.

[13] F. Rivardo, R.J. Turner, G. Allegrone, H. Ceri, M.G. Martinotti, Anti-adhesion activity oftwo biosurfactants produced by Bacillus spp. prevents biofilm formation of humanbacterial pathogens, Appl. Microbiol. Biotechnol. 83 (2009) 541–553.

[14] R. Maget-Dana, M. Ptak, Interfacial properties of surfactin, J. Colloid Interface Sci.153 (1992) 285–291.

[15] Y. Ishigami, M. Osman, H. Nakahara, Y. Sano, R. Ishiguro, M.Matsumoto, Significance ofbeta-sheet formation for micellization and surface-adsorption of surfactin, ColloidsSurf. B Biointerfaces 4 (1995) 341–348.

[16] H. Heerklotz, J. Seelig, Detergent-like action of the antibiotic peptide surfactin onlipid membranes, Biophys. J. 81 (2001) 1547–1554.

[17] M.Deleu, O. Bouffioux, H. Razafindralambo,M. Paquot, C. Hbid, P. Thonart, P. Jacques,R. Brasseur, Interaction of surfactin with membranes: a computational approach,Langmuir 19 (2003) 3377–3385.

[18] M. Deleu, J. Lorent, L. Lins, R. Brasseur, N. Braun, K. El Kirat, T. Nylander, Y.F. Dufrêne,M.-P. Mingeot-Leclercq, Effects of surfactin on membrane models displaying lipidphase separation, Biochim. Biophys. Acta-Biomembr. 1828 (2013) 801–815.

[19] L. Thimon, F. Peypoux, J. Wallach, G. Michel, Ionophorous and sequestering propertiesof surfactin, a biosurfactant from Bacillus subtilis, Colloids Surf. B Biointerfaces 1(1993) 57–62.

[20] C. Déjugnat, O.Diat, T. Zemb, Surfactin self-assembles intodirect and reverse aggregatesin equilibrium and performs selective metal cation extraction, Chemphyschem 12(2011) 2138–2144.

[21] J.D. Sheppard, C. Jumarie, D.G. Cooper, R. Laprade, Ionic channels induced by surfactin inplanar lipid bilayer membranes, Biochim. Biophys. Acta 1064 (1991) 13–23.

[22] O.S. Ostroumova, V.V. Malev, M.G. Ilin, L.V. Schagina, Surfactin activity dependson the membrane dipole potential, Langmuir 26 (2010) 15092–15097.

[23] H. Heerklotz, J. Seelig, Leakage and lysis of lipid membranes induced by thelipopeptide surfactin, Eur. Biophys. J. 36 (2007) 305–314.

[24] M. Morikawa, Y. Hirata, T. Imanaka, A study on the structure–function relationship oflipopeptide biosurfactants, Biochim. Biophys. Acta-Biomembr. 1488 (2000) 211–218.

[25] H.H. Shen, R.K. Thomas, P. Taylor, The location of the biosurfactant surfactin inphospholipid bilayers supported on silica using neutron reflectometry, Langmuir 26(2010) 320–327.

[26] A. Zou, J. Liu, V.M. Garamus, Y. Yang, R. Willumeit, B. Mu, Micellization activity of thenatural lipopeptide [Glu1, Asp5] surfactin-C15 in aqueous solution, J. Phys. Chem. B114 (2010) 2712–2718.

[27] C. Carrillo, J.A. Teruel, F.J. Aranda, A. Ortiz, Molecular mechanism of membranepermeabilization by the peptide antibiotic surfactin, Biochim. Biophys. Acta1611 (2003) 91–97.

[28] H.H. Shen, R.K. Thomas, J. Penfold, G. Fragneto, Destruction and solubilization ofsupported phospholipid bilayers on silica by the biosurfactant surfactin, Langmuir 26(2010) 7334–7342.

[29] O. Bouffioux, A. Berquand, M. Eeman, M. Paquot, Y.F. Dufrene, R. Brasseur, M. Deleu,Molecular organization of surfactin–phospholipid monolayers: effect of phospholipidchain length and polar head, Biochim. Biophys. Acta 1768 (2007) 1758–1768.

[30] S. Buchoux, J. Lai-Kee-Him, M. Garnier, P. Tsan, F. Besson, A. Brisson, E.J. Dufourc,Surfactin-triggered small vesicle formation of negatively charged membranes: anovel membrane-lysis mechanism, Biophys. J. 95 (2008) 3840–3849.

[31] A. Grau, J.C. Gomez Fernandez, F. Peypoux, A. Ortiz, A study on the interactions ofsurfactin with phospholipid vesicles, Biochim. Biophys. Acta-Biomembr. 1418(1999) 307–319.

[32] G. Francius, S. Dufour, M. Deleu, M. Papot, M.P. Mingeot-Leclercq, Y.F. Dufrene,Nanoscale membrane activity of surfactins: Influence of geometry, charge andhydrophobicity, Biochim. Biophys. Acta-Biomembr. 1778 (2008) 2058–2068.

[33] M. Eeman, A. Berquand, Y.F. Dufrene, M. Paquot, S. Dufour, M. Deleu, Penetrationof surfactin into phospholipid monolayers: nanoscale interfacial organization,Langmuir 22 (2006) 11337–11345.

[34] D.A. Hopwood, How do antibiotic-producing bacteria ensure their self-resistance be-fore antibiotic biosynthesis incapacitates them? Mol. Microbiol. 63 (2007) 937–940.

[35] B.G. Butcher, J.D. Helmann, Identification of Bacillus subtilis sigma(W)-dependentgenes that provide intrinsic resistance to antimicrobial compounds produced byBacilli, Mol. Microbiol. 60 (2006) 765–782.

[36] D.B. Kearns, F. Chu, R. Rudner, R. Losick, Genes governing swarming in Bacillussubtilis and evidence for a phase variation mechanism controlling surface motility,Mol. Microbiol. 52 (2004) 357–369.

[37] K. Tsuge, Y. Ohata, M. Shoda, Gene yerP, involved in surfactin self-resistance inBacillus subtilis, Antimicrob. Agents Chemother. 45 (2001) 3566–3573.

[38] P. Cosmina, F. Rodriguez, F. de Ferra, G. Grandi, M. Perego, G. Venema, D. vanSinderen, Sequence and analysis of the genetic locus responsible for surfactinsynthesis in Bacillus subtilis, Mol. Microbiol. 8 (1993) 821–831.

[39] M.M. Nakano, N. Corbell, J. Besson, P. Zuber, Isolation and characterization of sfp:a gene that functions in the production of the lipopeptide biosurfactant, surfactin,in Bacillus subtilis, Mol. Gen. Genet. 232 (1992) 313–321.

[40] R.D. Pridmore, New and versatile cloning vectors with kanamycin-resistancemarker, Gene 56 (1987) 309–312.

[41] G. Seydlova, J. Svobodova, Rapid and effective method for the separation ofBacillus subtilis vegetative cells and spores, Folia Microbiol. 57 (2012) 455–457.

[42] A. Hara, N. Radin, Lipid extraction of tissues with a low-toxicity solvent, Anal.Biochem. 90 (1978) 420–426.

[43] G. Rouser, S. Fleische, A. Yamamoto, Two dimensional thin layer chromatographicseparation of polar lipids and determination of phospholipids by phosphorusanalysis of spots, Lipids 5 (1970) 494–496.

[44] R.L. Glass, Alcoholysis, saponification and preparation of fatty acid methyl esters,Lipids 6 (1971) 919–925.

[45] D. Lopez, H. Vlamakis, R. Losick, R. Kolter, Cannibalism enhances biofilm developmentin Bacillus subtilis, Mol. Microbiol. 74 (2009) 609–618.

[46] X. Huang, Z. Wei, G. Zhao, X. Gao, S. Yang, Y. Cui, Optimization of sterilization ofEscherichia coli in milk by surfactin and fengycin using a response surfacemethod, Curr. Microbiol. 56 (2008) 376–381.

[47] M. Nazari, M. Kurdi, H. Heerklotz, Classifying surfactants with respect to theireffect on lipid membrane order, Biophys. J. 102 (2012) 498–506.

[48] H. Heerklotz, T. Wieprecht, J. Seelig, Membrane perturbation by the lipopeptidesurfactin and detergents as studied by deuterium, J. Phys. Chem. B 108 (2004)4909–4915.

[49] J.A.F. Op den Kamp, I. Redai, L.L. Van Deene, Phospholipid composition of Bacillussubtilis, J. Bacteriol. 99 (1969) 298–303.

[50] T. Kaneda, Biosynthesis of branched-chain fatty acids. V. Microbial strereospecificsynthesis of D-12-methyltetradecanoic and D-14-methylhexadecanoicacids,Biochim. Biophys. Acta 125 (1966) 43–54.

[51] F. Peypoux, J.M. Bonmatin, J. Wallach, Recent trends in the biochemistry ofsurfactin, Appl. Microbiol. Biotechnol. 51 (1999) 553–563.

[52] A.B. Hachmann, E. Sevim, A. Gaballa, D.L. Popham, H. Antelmann, J.D. Helmann,Reduction in membrane phosphatidylglycerol content leads to daptomycinresistance in Bacillus subtilis, Antimicrob. Agents Chemother. 55 (2011)4326–4337.

[53] M. Dahlberg, Polymorphic phase behavior of cardiolipin derivatives studied bycoarse-grained molecular dynamics, J. Phys. Chem. B 111 (2007) 7194–7200.

[54] P.R. Allegrini, G. Pluschke, J. Seelig, Cardiolipin conformation and dynamics inbilayer membranes as seen by deuterium magnetic resonance, Biochemistry 23(1984) 6452–6458.

[55] R. Lewis, R.N. McElhaney, The physicochemical properties of cardiolipin bilayersand cardiolipin-containing lipid membranes, Biochim. Biophys. Acta-Biomembr.1788 (2009) 2069–2079.

[56] A. Petersohn, M. Brigulla, S. Haas, J.D. Hoheisel, U. Volker, M. Hecker, Globalanalysis of the general stress response of Bacillus subtilis, J. Bacteriol. 183(2001) 5617–5631.

[57] C.S. López, A.F. Alice, H. Heras, E.A. Rivas, C. Sánchez-Rivas, Role of anionicphospholipids in the adaptation of Bacillus subtilis to high salinity, Microbiology152 (2006) 605–616.

2378 G. Seydlová et al. / Biochimica et Biophysica Acta 1828 (2013) 2370–2378

[58] L. Catucci, N. Depalo, V.M.T. Lattanzio, A. Agostiano, A. Corcelli, Neosynthesis ofcardiolipin in Rhodobacter sphaeroides under osmotic stress, Biochemistry 43(2004) 15066–15072.

[59] M. Tsai, R.L. Ohniwa, Y. Kato, S.L. Takeshita, T. Ohta, S. Saito, H. Hayashi, K. Morikawa,Staphylococcus aureus requires cardiolipin for survival under conditions of high salinity,BMC Microbiol. 11 (2011) 13.

[60] M.S. Muntyan, I.V. Popova, D.A. Bloch, E.V. Skripnikova, V.S. Ustiyan, Energetics ofalkalophilic representatives of the genus Bacillus, Biochem. Mosc. 70 (2005)137–142.

[61] P. Bernal, J. Munoz-Rojas, A. Hurtado, J.L. Ramos, A. Segura, A Pseudomonas putidacardiolipin synthesis mutant exhibits increased sensitivity to drugs related totransport functionality, Environ. Microbiol. 9 (2007) 1135–1145.

[62] F. Kawai,M. Shoda, R.Harashima, Y. Sadaie,H.Hara, K.Matsumoto, Cardiolipin domainsin Bacillus subtilisMarburg membranes, J. Bacteriol. 186 (2004) 1475–1483.

[63] K. Sekimizu, A. Kornberg, Cardiolipin activation of DnaA protein, the initia-tion protein of replication of Escherichia coli, J. Biol. Chem. 263 (1988)7131–7135.

[64] A. Nishibori, J. Kusaka, H. Hara, M. Umeda, K. Matsumoto, Phosphatidylethanol-amine domains and localization of phospholipid synthases in Bacillus subtilismembranes, J. Bacteriol. 187 (2005) 2163–2174.

[65] E. Mileykovskaya, W. Dowhan, Cardiolipin membrane domains in prokaryotesand eukaryotes, Biochim. Biophys. Acta-Biomembr. 1788 (2009) 2084–2091.

[66] C. Osman, M. Haag, C. Potting, J. Rodenfels, P.V. Dip, F.T. Wieland, B. Brugger,B. Westermann, T. Langer, The genetic interactome of prohibitins: coordi-nated control of cardiolipin and phosphatidylethanolamine by conservedregulators in mitochondria, J. Cell Biol. 184 (2009) 583–596.

[67] R. Maget-Dana, M. Ptak, Interactions of surfactin with membrane models,Biophys. J. 68 (1995) 1937–1943.

[68] N. Papo, Y. Shai, Can we predict biological activity of antimicrobial peptidesfrom their interactions with model phospholipid membranes? Peptides 24(2003) 1693–1703.

[69] M. Kovacs, A. Halfmann, I. Fedtke, M. Heintz, A. Peschel, W. Vollmer, R.Hakenbeck, R. Bruckner, A functional dlt operon, encoding proteins required forincorporation of D-alanine in teichoic acids in gram-positive bacteria, confersresistance to cationic antimicrobial peptides in Streptococcus pneumoniae, J.Bacteriol. 188 (2006) 5797–5805.

[70] P.L.G. Chong, P.T.T. Wong, Interactions of Laurdan with phosphatidylcholine lipo-somes—a high-pressure FTIR study, Biochim. Biophys. Acta 1149 (1993) 260–266.

[71] T. Parasassi, M. Distefano, M. Loiero, G. Ravagnan, E. Gratton, Influence of choles-terol on phospholipid bilayers phase domains as detected by Laurdan fluores-cence, Biophys. J. 66 (1994) 120–132.

[72] T. Parasassi, G. De Stasio, G. Ravagnan, R.M. Rusch, E. Gratton, Quantitation oflipid phases in phospholipid vesicles by the generalized polarization ofLaurdan fluorescence, Biophys. J. 60 (1991) 179–189.

[73] V. Borenstain, Y. Barenholz, Characterization of liposomes and other lipidassemblies by multiprobe fluorescence polarization, Chem. Phys. Lipids 64(1993) 117–127.

[74] M. Adler, T.R. Tritton, Fluorescence depolarization measurements on orientedmembranes, Biophys. J. 53 (1988) 989–1005.

[75] P. Garidel, A. Blume, Miscibility of phosphatidylethanolamine–phosphatidylglycerolmixtures as a function of pH and acyl chain length, Eur. Biophys. J. 28 (2000) 629–638.

[76] K. Athenstaedt, G. Daum, Phosphatidic acid, a key intermediate in lipid metabolism,Eur. J. Biochem. 266 (1999) 1–16.

[77] F. Etienne, Y. Roche, P. Peretti, S. Bernard, Cardiolipin packing ability studied bygrazing incidence X-ray diffraction, Chem. Phys. Lipids 152 (2008) 13–23.

[78] M. Dahlberg, A. Maliniak, Molecular dynamics simulations of cardiolipin bilayers,J. Phys. Chem. B 112 (2008) 11655–11663.

[79] T. Rog, H. Martinez-Seara, N. Munck, M. Oresic, M. Karttunen, I. Vattulainen, Roleof cardiolipins in the inner mitochondrial membrane: insight gained throughatom-scale simulations, J. Phys. Chem. B 113 (2009) 3413–3422.

[80] E.V. Brink-van der Laan, J.A. Killian, B. de Kruijff, Nonbilayer lipids affect peripheraland integral membrane proteins via changes in the lateral pressure profile, Biochim.Biophys. Acta-Biomembr. 1666 (2004) 275–288.

[81] F.L. Hoch, Cardiolipins and biomembrane function, Biochim. Biophys. Acta 1113 (1992)71–133.

[82] J.L. Ramos, E. Duque, J.J. RodriguezHerva, P. Godoy, A. Haidour, F. Reyes, A.FernandezBarrero, Mechanisms for solvent tolerance in bacteria, J. Biol. Chem. 272(1997) 3887–3890.

[83] P. Bernal, A. Segura, J.L. Ramos, Compensatory role of the cis-trans-isomerase andcardiolipin synthase in the membrane fluidity of Pseudomonas putida DOT-T1E,Environ. Microbiol. 9 (2007) 1658–1664.

[84] R. Brasseur, N. Braun, K. El Kirat, M. Deleu, M.P. Mingeot-Leclercq, Y.F. Dufrene, Thebiologically important surfactin lipopeptide induces nanoripples in supported lipidbilayers, Langmuir 23 (2007) 9769–9772.

[85] M. Dahlberg, A. Maliniak, Mechanical properties of coarse-grained bilayersformed by cardiolipin and zwitterionic lipids, J. Chem. Theory Comput. 6(2010) 1638–1649.

[86] J. Wohlert, W.K. den Otter, O. Edholm, W.J. Briels, Free energy of a trans-membranepore calculated from atomistic molecular dynamics simulations, J. Chem. Phys. 124(2006) 154905.

[87] V. Betaneli, E.P. Petrov, P. Schwille, The role of lipids in VDAC oligomerization,Biophys. J. 102 (2012) 523–531.