Potent antimicrobial peptides against Legionella

1 downloads 0 Views 2MB Size Report
The link must be accompanied by the following text: "The final ... Our data represent the ... pneumophila serogroup 1 is solely responsible for the great majority of ...
Potent antimicrobial peptides against Legionella pneumophila and its environmental host, Acanthamoeba castellanii Margot Schlusselhuber, Vincent Humblot, Sandra Casale, Christophe Méthivier, Julien Verdon, Matthias Leippe & Jean-Marc Berjeaud Applied Microbiology and Biotechnology ISSN 0175-7598 Appl Microbiol Biotechnol DOI 10.1007/s00253-015-6381-z

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Appl Microbiol Biotechnol DOI 10.1007/s00253-015-6381-z

ENVIRONMENTAL BIOTECHNOLOGY

Potent antimicrobial peptides against Legionella pneumophila and its environmental host, Acanthamoeba castellanii Margot Schlusselhuber & Vincent Humblot & Sandra Casale & Christophe Méthivier & Julien Verdon & Matthias Leippe & Jean-Marc Berjeaud

Received: 12 September 2014 / Revised: 20 November 2014 / Accepted: 31 December 2014 # Springer-Verlag Berlin Heidelberg 2015

Abstract Legionella pneumophila, the major causative agent of Legionnaires’ disease, is most often found in the environment in close association with free-living amoebae, leading to persistence, spread, biocide resistance, and elevated virulence of the bacterium. In the present study, we evaluated the antiLegionella and anti-Acanthamoeba activities of three alphahelical antimicrobial peptides (AMPs), namely, NK-2, CiMAM-A24, and Ci-PAP-A22, already known for the extraordinary efficacy against other microbes. Our data represent the first demonstration of the activity of a particular AMP against both the human facultative intracellular pathogen L. pneumophila and its pathogenic host, Acanthamoeba

castellanii. Interestingly, the most effective peptide, CiMAM-A24, was also found to reduce the Legionella cell number within amoebae. Accordingly, this peptide was immobilized on gold surfaces to assess its antimicrobial activity. Surfaces were characterized, and activity studies revealed that the potent bactericidal activity of the peptide was conserved after its immobilization. In the frame of elaborating anti-Legionella surfaces, Ci-MAM-A24 represents, by its direct and indirect activity against Legionella, a potent peptide template for biological control of the bacterium in plumbings. Keywords Legionella . Antimicrobial peptides . Amoebae . Surface modification

Electronic supplementary material The online version of this article (doi:10.1007/s00253-015-6381-z) contains supplementary material, which is available to authorized users. M. Schlusselhuber : J. Verdon : J.95 %. The primary structures are KILRGVCKKIMRTFLR RISKDILTGKK–NH2, WRSLGRTLLRLSHALKPLARRS GW–NH2, and ALRSAVRTVARVGRAVLPHVAI–NH2, respectively. Peptides were dissolved in 10 mM HCl to a peptide concentration of 1 mM and, for the experiments, were diluted further in the appropriate assay medium. Determination of anti-Legionella activity The bactericidal effect of peptides against L. pneumophila Lens in exponential growth phase was determined as described previously with minor modifications (Verdon et al. 2008). Briefly, L. pneumophila cells were grown to an OD600 between 0.4 and 0.8. Bacteria were then appropriately diluted in 10 mM sodium phosphate buffer (pH 6.9) to approximately 106 CFU/mL. Serial twofold dilutions of antimicrobial peptides were achieved in sterile 10-mM sodium phosphate buffer (pH 6.9) and added (25 μL) to bacterial suspension (475 μL) at a starting concentration of 0.1 μM. Prior to use, all tubes were coated with BSA 0.1 % (w/v) for 15 min to avoid peptide adsorption. The peptide solvent, 10 mM HCl, served as negative control in each experiment. Bacterial cells were incubated at 37 °C for 1 h and then washed by centrifugation (10,000×g, 5 min). The number of CFU was determined after 96 h of cultivation at 37 °C of tenfold serial dilutions of bacterial suspension on BCYE agar plates. Permeabilization of amoebae in the presence of peptides The effect of peptides on membrane integrity of amoebic trophozoïtes was assessed by monitoring the fluorescence of the DNA-intercalating dye SYTOX Green (Invitrogen, Carlsbad, CA, USA) in cells with compromised membranes (Hoeckendorf et al. 2012). Briefly, A. castellanii ATCC 30234 was cultured axenically in PYG medium at 30 °C and split the day before such that 2×104 trophozoites were present in each well of a 96-well black plate. Serial dilutions of peptides were done in PAS buffer (Page’s modified Neff’s amoeba saline), pH 6.5, containing 2 μM of the fluorescent dye SYTOX Green. The peptide solvent, 10 mM HCl, served as negative

control in each experiment. A. castellanii was treated with the peptides in a total volume of 100 μL at 30 °C for 2 h. For spontaneous lysis of the amoebae (0 % value), incubation was done in buffer, and maximum permeabilization (100 % value) was determined when 0.5 % Triton X-100 (v/v) was added to the buffer. After incubation, fluorescence of the samples was measured at 535-nm emission after excitation at 488 nm with a spectrofluorometer (Tristar2 LB 942 Multimode Reader, Berthold Technologies). Hence, the intensity of fluorescence at 535 nm is directly proportional to the amount of permeabilized cells. Membrane-permeabilizing activity of peptides was expressed as percentage of permeabilized amoebae. Experiments were done in triplicate and data are expressed as mean ± standard deviation.

Activity of peptides against Legionella replicating inside amoebae The activity of peptides against intra-amoebic L. pneumophila was examined as described previously by Harrison et al. (2013), with some modifications. Briefly, 2×104 trophozoites were cultured in PYG medium and split the day prior to infection such that 2×104 cells were present in each well of a 96well black plate. Cultures of GFP-expressing L. pneumophila Lens (Bigot et al. 2013) were resuspended from a plate to a starting OD600 of 0.1 in BYE medium and grown at 37 °C and under constant shaking at 180 rpm overnight. Bacteria were washed and diluted in PAS buffer, and 100 μL of suspension was added to each well such that it contains 1×106 bacteria (MOI 50). Infections were synchronized by centrifugation at 1500 rpm, 30 °C for 10 min. The wells were washed 3 h postinfection to discard extracellular bacteria. Antimicrobial peptides, diluted in PAS buffer, were added 6 h post-infection. Wells containing only Legionella in PAS buffer were included in each experiment to ensure that the bacteria did not proliferate in the buffer. Infected cultures were incubated in a 30-°C incubator, and fluorescence was measured 48 h post-infection at 535-nm emission after excitation at 488 nm with a spectrofluorometer (Tristar2 LB 942 Multimode Reader, Berthold Technologies). Experiments were performed in triplicates, and the results are expressed as means; bar scales indicate standard deviation. For confocal microscopic observations, trophozoïtes were seeded at the same density as above in eight-well Lab-Tek II chambered coverglass (Thermo Fisher, Rochester, NY, USA). Cells were infected, treated with peptides, and incubated as described above. After incubation, chambers were directly examined with a confocal FV-1000 station installed on an inverted microscope IX-81 (Olympus, Tokyo, Japan). Images were acquired with an Olympus UplanSapo X60 water, 1.2 NA, objective lens (800 × 800 pixels images with 0.13 mm/pixel corresponding to Nyquist criteria for optimal sampling).

Author's personal copy Appl Microbiol Biotechnol

Gold surface preparation Surfaces were prepared as previously described (Humblot et al. 2009). Briefly, surfaces constituting of glass substrates (11×11 mm), coated successively with a 50-Å-thick layer of chromium and a 200-nm-thick layer of gold, were purchased from Arrandee (Werther, Germany). The substrates were immersed in a pure solution at 0.01 M of 11mercaptoundecanoic acid (MUA) in 10 mL of absolute ethanol for 3 h in order to ensure optimal homogeneity of the adlayer (Briand et al. 2006; Hobara and Kakiuchi 2001). Immobilization of Ci-MAM-A24 (10 μg/mL in sterile sodium phosphate buffer, pH 6.9) on gold surfaces was carried out by depositing a 150-μL drop of peptide solution on the Au-modified substrates at room temperature for 2 h. After the immobilization step, the surfaces were vigorously rinsed in PBS with agitation and dried under a flow of dry nitrogen. For each step, four series of identical experiments were conducted: one series of samples was then characterized by PM-RAIRS and XPS, the second one was characterized by PM-RAIRS, SEM-FEG, and AFM, and the two others were also checked by PM-RAIRS before performing the bacterial adhesion analysis.

I C1s ¼ I Au4 f

5ρMAM M Au σC1s T C1s λMAM C1s

1−exp −

d

!!

λMAM C1s sinθ ! d Au ρAu M MAM σAu4 f T Au4 f λAu4 f exp − MAM λAu4 f sinθ

where θ is the photoelectron collection angle. TC 1s and TAu4f are the relative sensitivity factors of C and Cu, respectively, provided by the spectrometer manufacturer. The Scofield photoionization cross-sections σ are equal to 1 for C 1s and 17.1 for Au 4f (Scofield 1976). λyx is the inelastic mean free paths of electrons x in the matrix y. They were calculated using the Quases program based on the TPP2M formula (Tanuma et al. 1994). ρM AM and ρAu are the density of Ci-MAM-A24 and gold, respectively. MMAM and MAu are the molecular weight of Ci-MAM-A24 and Au, respectively. The number of Ci-MAM-A24 molecules per square centimeter nMAM could be estimated by using the following equation:   ρ d  NA nMAM molecules cm‐2 ¼ MAM M MAM

where NA is the Avogadro number. XPS characterization

PM-RAIRS measurements The gold samples were placed in the external beam of the FT-IR instrument (Nicolet Nexus 5700 FT-IR spectrometer), and the reflected light was focused on a nitrogen-cooled HgCdTe wide band detector. The infrared spectra were recorded at 8-cm−1 resolution, with coaddition of 128 scans. A ZnSe grid polarizer and a ZnSe photoelastic modulator to modulate the incident beam between p and s polarizations (HINDS Instruments, PM90, modulation frequency = 36 kHz) are placed prior to the sample. The detector output is sent to a two-channel electronic device that generates the sum and difference interferograms. Those were processed and undergo Fourier transformation to produce the PM-RAIRS signal (ΔR/R0) = (Rp−Rs)/( Rp + Rs). Using modulation of polarization enabled us to perform rapid analyses of the sample after treatment in various solutions without purging the atmosphere or requiring a reference spectrum.

Thickness and number of molecules on gold surfaces Assuming a homogeneous layer, the thickness of the layer was estimated from the C 1s over Au4f intensity ratio using the following equation:

XPS analyses were performed using a PHOIBOS 100 X-ray photoelectron spectrometer from SPECS GmbH (Berlin, Germany) with a monochromated Al Kα X-ray source (hν= 1486.6 eV) operating at P=1×10–10 Torr or less. Spectra were carried out with 50-eV pass energy for the survey scan and 10-eV pass energy for the C 1s, O 1s, N 1s, and Ti 2p regions. High-resolution XPS conditions have been fixed: “fixed analyzer transmission” analyses mode, a 7×20-mm entrance slit, leading to a resolution of 0.1 eV for the spectrometer, and an electron beam power of 150 W (15 kV and 10 mA). A takeoff angle of 90° from the surface was employed for each sample. Element peak intensities were corrected by Scofield factors (Scofield 1976). The spectra were fitted using Casa XPS v.2.3.15 Software (Casa Software Ltd.) and by applying a Gaussian/Lorentzian ratio G/L equal to 70/30. Adhesion of bacteria on gold slides Adhesion of L. pneumophila Lens in exponential growth phase on gold surfaces was analyzed by CFU counting and confocal microscopy. Gold samples were washed successively in 70 % ethanol and sterile water and then dried in sterile environment. To avoid desiccation, the sample was cautiously deposited with the gold face upwards in a small Petri dish placed in a large Petri dish filled with sterile water. Bacteria,

Author's personal copy Appl Microbiol Biotechnol

1×105 in 100 μL of buffer, were deposited on the gold surface. After 3 h at 37 °C, each sample was washed three times with sodium phosphate buffer (pH 6.9). To determine the percentage of inhibition, samples were then transferred into a sterile tube containing 2 mL of sterile sodium phosphate buffer and then sonicated for 3 min at 51 W in a bath sonicator (Bioblock Scientific 86480, Fisher, Illkirch, France). Bacteria were pelleted by centrifugation at 10,000×g for 5 min. Subsequently, 1.5 mL of supernatant was removed cautiously, and the bacteria were re-suspended by vortexing. The bacterial suspension was tenfold serially diluted, and 100 μL of each dilution was deposited on agar plates, in duplicate for each dilution. Plates were incubated at 37 °C for 96 h before CFU counting. To determine the viability of bacteria attached to the surfaces, confocal microscopy was performed. After an incubation period of 3 h in the presence of bacteria, gold samples were washed as described above. Ten microliters of BacLight™ mixture (LIVE/DEAD® BacLight™ Bacterial Viability Kit) was deposited on the surface of the sample and incubated for 10 min in the dark at room temperature prior to microscopic analysis. Samples were examined by confocal microscopy as described above. Multiple fluorescence signals were acquired sequentially to avoid cross-talk between image channels. Fluorophores were excited with the 488-nm line of an argon laser (for Syto9) and the 543-nm line of an HeNe laser (for propidium iodide). The emitted fluorescences were detected through spectral detection channels between 500– 530 and 555–655 nm for green and red fluorescence, respectively. AFM characterization The AFM images of dried surfaces were recorded using a commercial di Caliber AFM microscope from Bruker Instruments Inc. In order to avoid tip and sample damages, topographic images were taken in the non-contact dynamic mode, also known as tapping® mode. Silicon nitride tips (resonance frequency of 280–400 kHz, force constant of 40–80 N/ m) have been used. Images were obtained at a constant speed of 2 Hz with a resolution of 512 lines of 512 pixels each. The raw data were processed using the imaging processing software di SpmLabAnalysis v.7.0. from Bruker Instruments Inc. AFM analyses were carried out at least at three different locations on each surface, with a minimum of 100 bacteria observed. SEM-FEG characterization SEM-FEG images were obtained using a SEM-FEG Hitachi SU-70 scanning electron microscope with an accelerating voltage of 1 kV and working distance of around 3.1 or 3.2 mm; in lense secondary electron detector SE(U) was used.

A hundred microliters of a suspension of bacteria at 109 cfu/ mL was deposited onto a gold substrate and dried at room temperature. The gold substrate was fixed on an Alumina SEM support with a carbon adhesive tape. SEM-FEG analyses were carried out at least at four different locations on each surface, with a minimum of 100 bacteria observed. Statistics Results were evaluated statistically using one-way analysis of variance (ANOVA) with Bonferroni correction post-test using GraphPad Prism version 5.04 for Windows (GraphPad Software, San Diego, CA, USA).

Results Anti-Legionella activity The anti-Legionella activities of NK-2, Ci-MAM-A24, and Ci-PAP-A22 were studied. L. pneumophila cells in exponential growth phase were incubated with various concentrations of peptides for 1 h, and bacterial viability was subsequently assessed by counting the colony-forming units (CFU) after a growth phase of 96 h. All three peptides exhibited potent bactericidal activity in a range of micromolar concentrations (Fig. 1). They killed 50 % of the bacterial population (EC50) at a concentration below 0.5 μM (Table 1). However, Ci-MAMA24 appears to be the most potent in L. pneumophila killing as the minimal bactericidal concentration (MBC) was 15 times lower than that of Ci-PAP-A22 (1.6 vs. 25 μM), and its efficient concentration for 90 % of the bacterial population (EC90) was four times lower than that of NK-2 (0.2 vs. 0.8 μM). The effect of Ci-MAM-A24 against L. pneumophila was then observed by 2D scanning electron microscopy. At a concentration of 1.6 μM, the peptide apparently resulted in the permeabilization of the bacterial membrane and loss of cellular content as depicted by the circular halo observed around the bacteria (Fig. 2).tgroup Cytotoxic effect of peptides on A. castellanii Cytotoxicity of NK-2, Ci-MAM-A24, and Ci-PAP-A22 against A. castellanii, a natural amoebic host for Legionella in the environment, was assessed by the Sytox green assay, which relies on the intercalation of the fluorescent dye into DNA of cells with compromised membranes. Fluorescence measurements of A. castellanii trophozoïtes treated with various concentrations of peptides indicated that NK-2 and CiPAP-A22 did not exert a significant cytotoxic effect at the maximal concentration tested (25 μM), while Ci-MAM-A24 permeabilized about 85 % of trophozoïtes (Fig. 3). Below this

Author's personal copy Appl Microbiol Biotechnol

Fig. 1 Bactericidal activity of various AMPs against L. pneumophila Lens. A bacterial suspension of 1×106 CFU/mL was incubated with various concentrations of peptide in sodium phosphate buffer (pH 6.9) 1 h at 37 °C. Results are expressed as survival rate in comparison to the

negative control (initial peptide solvent, 10 mM HCl). The dashed line indicates 50 % survival. Experiments were performed in duplicates, and the results are expressed as means. Bar scales indicate standard deviation

concentration (≤12.5 μg/mL), this peptide also did not induce substantial permeabilization of amoebae.

located within amoebae are both detected by this method. In order to monitor whether the observed activity of Ci-MAMA24 was related to intracellular killing of bacteria and not only to destruction of bacteria that were released from amoebae in the buffer after infection, confocal microscopy was performed (Fig. 5). Intriguingly, the great majority of amoebae treated with Ci-MAM-A24 did not carry bacteria anymore; only few cells were still infected (Fig. 5a), while all the untreated cells carried numerous bacteria (Fig. 5b).

Effect of exogenous addition of peptides on intra-amoebic L. pneumophila We examined whether exogenous addition of NK-2, CiMAM-A24, or Ci-PAP-A22 could kill Legionella replicating inside A. castellanii. Trophozoites were infected with GFPexpressing L. pneumophila for 3 h. Infected cells were subsequently incubated for 42 h with concentrations of up to 12.5 μM of each peptide, a concentration evaluated to be non-toxic for the host cells. After incubation, the amount of legionellae was determined by detecting GFP fluorescence emission at 535 nm. At concentrations below 12.5 μM, the peptides did not affect the number of GFP-expressing Legionella (data not shown). However, incubation with 12.5 μM of Ci-MAM-A24 led to a significant decrease of legionellae residing in amoebae when compared to untreated trophozoïtes and NK-2- or Ci-PAP-A22-treated trophozoites (Fig. 4). Interestingly, this effect was observed at 48 h postinfection, but not at 24 h, a time point at which Legionella replication was observed (data not shown). Legionella is not able to grow in PAS buffer; therefore, bacteria released from amoebae in the buffer and bacteria Table 1

Antimicrobial activity of selected peptides against Legionella

Ci-MAM-A24 Ci-PAP-A22 NK-2

MBC (μM)

EC50 (μM)

EC90 (μM)

1.6 25 1.6

≤0.1 0.4 0.2

0.2 1.6 0.8

MBC minimal bactericidal concentration (99.9 % of bacterial population killed), EC50 concentration that kills 50 % of bacterial population, EC90 concentration that kills 90 % of bacterial population

Characterization of functionalized and peptide-coated surfaces by polarization modulation reflection absorption infrared spectroscopy and X-ray photoelectron spectroscopy analyses The gold substrate was first functionalized with a selfassembled monolayer of 11-mercaptoundecanoic acid (AuMUA). The acid functions were then activated into esters (Au-MUAact) in order to let Ci-MAM-A24 peptide react via one of its amino groups (Au-MUA–Ci-MAM-A24). Figure S1 (provided as supporting information) shows the polarization modulation reflection absorption infrared spectroscopy (PM-RAIRS) spectra recorded after the successive steps of gold surface functionalization. The Au-MUA spectrum (Fig. S1 A) presented a rather intense νC = O band at 1718 cm−1, characteristic of carboxylic groups (Bain et al. 1989; Tielens et al. 2008), showing the presence of the carboxylic acid-terminated thiol. The asymmetric and symmetric νCH bands of the CH2 chains are recorded at 2855 and 2921 cm−1, respectively. The broad massif between 1400 and 1550 cm−1 likely included contributions from the scissor mode of CH2 groups and from the symmetric and asymmetric stretching of COO− groups. On another spectrum (Fig. S1 B), a slight upward shift and increase of the νC = O band to 1741 cm−1 confirm the transformation of acid into ester

Author's personal copy Appl Microbiol Biotechnol

Fig. 2 SEM-FEG micrographs of L. pneumophila Lens cells treated with Ci-MAM-A24. Exponential-growth-phase Legionella cells were incubated with a 1.6 μM of the antimicrobial peptide Ci-MAM-A24 or

b the peptide solvent (10 mM HCl), as a negative control in sodium phosphate buffer (pH 6.9) for 2 h. Representative micrographs are shown

terminal groups; two other weak bands appear at 1791 and 1820 cm−1 as the signatures of ester groups of the succinimide moiety (Duevel 1992). The successful binding of Ci-MAMA24 is indicated by the appearance of two new intense bands on the last spectrum (Fig. S1 C) arising at 1652 and 1541 cm−1 and assigned respectively to the amide I (νC = O) and amide II (δNH+ νC–N) vibrations of the peptide backbone (Briand et al. 2006). One can also note the presence of several other infrared features coming up from the lateral groups of the amino acids constituting the Ci-MAM-A24 peptide. X-ray photoelectron spectroscopy (XPS) analyses after CiMAM-A24 immobilization on Au-MUA-modified surface provided complementary information. The C 1s, N 1s, O 1s, S 2p, and Au 4f high-resolution regions were recorded, although only the N 1s, O 1s and C 1s regions are presented in Figure S2. Adsorption of Ci-MAM-A24 was shown by the appearance of a nitrogen peak, N 1s, centered at 400.3± 0.1 eV, mainly due to the amide functions of the peptide (Fig. S2 A). Two other small contributions were observed at 402.0±0.1 and 390.0±0.1 eV, assigned respectively to protonated amine group and to nitrogen atoms involved in double

bonds with carbon or nitrogen atoms from tryptophan, arginine and histidine moieties. The O 1s of the Ci-MAM-A24modified surface (Fig. S2 B) shows a main contribution at 531.2±0.1 eV assigned to the C = O of the amide bonds from the peptide backbone. Two other contributions were observed at higher energies, at 532.3 ± 0.1 and at 533.3 ± 0.1 eV, assigned to –OH of the different amino acid lateral chains of the Ci-MAM-A24 peptide. After Ci-MAM-A24 adsorption, the C 1s peak could be fitted with four components (Fig. S2 C), in agreement with previous studies of peptide (Magainin I) adsorption on mixed thiols SAMs (Humblot et al. 2009). The first peak, at the lowest binding energy (BE) of 285.0±0.1 eV was assigned to the carbon C-C,C-H; the second peak, at 286.4±0.1 eV was attributed to the C-C-N carbon + C-OH atoms and to the carbon atoms in C-C-O + Caromatic bonds; and finally, the third and fourth contribution at 288.0±0.1 and 289.2±0.1 eV, included the C atoms involved in peptide bonds, O = C-N and acidic groups (H)O-C = O. Finally, the S2p peak (data not shown), presented a very small contribution (1.3 atomic percent) with a main contribution, S2p3/2, at 162.1±0.1 eV coming from the thiol spacer between the gold

Fig. 3 Effect of NK-2, Ci-PAP-A22, and Ci-MAM-A24 on A. castellanii trophozoïtes. A trophozoïte suspension of 2×104 cells was incubated at 30 °C for 2 h with various concentrations of peptide in PAS buffer (pH 6.5). Results are expressed as the percentage of cell permeabilization in

comparison to the negative control (peptide solvent, 10 mM HCl). Experiments were performed in triplicates, and the results are expressed as means. Bar scales indicate standard deviation. ***p