|Home | About | Journals | Submit | Contact Us | Français|
Antimicrobial peptides (AMPs) form an important part of the innate host defense. In contrast to most AMPs, human dermcidin has an anionic net charge. To investigate whether bacteria have developed specific mechanisms of resistance to dermcidin, we screened for mutants of the leading human pathogen, Staphylococcus aureus, with altered resistance to dermcidin. To that end, we constructed a plasmid for use in mariner-based transposon mutagenesis and developed a high-throughput cell viability screening method based on luminescence. In a large screen, we did not find mutants with strongly increased susceptibility to dermcidin, indicating that S. aureus has no specific mechanism of resistance to this AMP. Furthermore, we detected a mutation in a gene of unknown function that resulted in significantly increased resistance to dermcidin. The mutant strain had an altered membrane phospholipid pattern and showed decreased binding of dermcidin to the bacterial surface, indicating that dermcidin interacts with membrane phospholipids. The mode of this interaction was direct, as shown by assays of dermcidin binding to phospholipid preparations, and specific, as the resistance to other AMPs was not affected. Our findings indicate that dermcidin has an exceptional value for the human innate host defense and lend support to the idea that it evolved to evade bacterial resistance mechanisms targeted at the cationic character of most AMPs. Moreover, they suggest that the antimicrobial activity of dermcidin is dependent on the interaction with the bacterial membrane and might thus assist with the determination of the yet unknown mode of action of this important human AMP.
Antimicrobial peptides (AMPs) form an important part of the innate host defense in neutrophil phagosomes, where they contribute to the elimination of ingested bacteria, and on epithelia, where they control the proliferation of commensal and invading microorganisms (9). Many bacteria have developed mechanisms of resistance to AMPs, such as inactivation of AMPs by secreted proteases, the repulsion or sequestration of AMPs by surface polymers, and the expulsion of AMPs from the cytoplasmic membrane (19). Electrostatic forces govern many AMP-bacterium interactions, and frequently, bacterial mechanisms of resistance specifically target the typical cationic character of AMPs (19). For example, the introduction of positive charges in bacterial surface polymers lowers the overall negative charge of the bacterial surface and minimizes the attraction of AMPs to their target.
As a result of the long interaction during evolution, hosts have invented ways to circumvent bacterial resistance, such as by stabilizing AMPs via intramolecular bridges or by increasing the density of positive charges to increase the interaction with the negatively charged bacterial surface (23). The active processed forms of the human AMP dermcidin, DCD-1 and DCD-1L, represent peptides with noticeable alternatives to these strategies aimed at circumventing bacterial resistance, inasmuch as in these peptides, the typical cationic charge of the AMPs is replaced by a negative net charge (26). One clear advantage of the anionic character of dermcidin is the ability to evade resistance mechanisms that specifically target cationic AMPs. Possibly for that reason, dermcidin is a main effector of the innate host defense against bacterial pathogens, particularly those on human skin (25). While dermcidin is likely subject to nonspecific resistance mechanisms, such as degradation by secreted proteases (13, 28), it is not known whether specific bacterial resistance mechanisms are targeted at dermcidin. Furthermore, the anionic character comes with the price that the electrostatic attraction to the bacterial surface, which is considered crucial for antimicrobial activity, is mechanistically challenging. Previous research has established that dermcidin binds to the bacterial surface (29), but the molecular basis of binding and the bactericidal mechanism of dermcidin are unknown.
In the present study, we aimed to identify the bacterial molecules involved in resistance to dermcidin. To that end, we chose Staphylococcus aureus as a model organism. Staphylococci are part of the normal microflora on the human skin and mucous membranes, where dermcidin is predominantly produced (24, 26). Furthermore, they are important pathogens and thus have a particularly essential need for protection from AMPs in their natural habitat and during infection (15, 31). We developed a luminescence-based assay to screen for mutants with altered resistance to dermcidin. Furthermore, we constructed a plasmid for use in mariner-based transposon mutagenesis to avoid the reported integration bias of the previously used transposon, Tn917 (1). Notably, we did not detect mutants with strongly increased susceptibility to dermcidin and no mutations in genes with a previously established function in antimicrobial resistance. These findings lend support to the idea that dermcidin is an evolutionary adaptation of the human innate immune system formed to evade bacterial resistance mechanisms. In addition, we detected an S. aureus mutant with significantly increased resistance to dermcidin, and characterization of that mutant suggested that the antimicrobial activity of dermcidin is dependent on a specific interaction with bacterial membrane phospholipids.
The bacterial strains and plasmids used in this study are summarized in Table Table1.1. Escherichia coli strain DH5α was used in the cloning experiments. S. aureus strain RN4220 was used as a gateway strain prior to propagation of the plasmids into S. aureus Xen36. The bacteria were routinely grown in tryptic soy broth (TSB) with 0.25% glucose or agar plates at 37°C, unless otherwise indicated. Antibiotics were used at the following concentrations, unless otherwise noted: ampicillin, 100 μg/ml; chloramphenicol, 10 μg/ml; erythromycin, 5 μg/ml; kanamycin 50 μg/ml, and tetracycline, 25 μg/ml. Dermcidin (in its biologically active, processed form [DCD-1]) and human β-defensin-3 (hBD3) were synthesized by the Peptide Synthesis Unit of the Research Technologies Branch, National Institute of Allergy and Infectious Diseases. LL-37 was purchased from GenScript Corp., and nisin was purchased from Sigma.
The erm gene was amplified from pEC2 by PCR with primers IRXba and IRPst (Table (Table2).2). The 1.3-kb product was digested with XbaI and PstI and ligated into XbaI- and PstI-digested pBT2, yielding plasmid pBTIR. A 1.9-kb XbaI-BamHI fragment containing the xylose-inducible promoter and the xylR regulator gene of pTX15 was ligated into XbaI-BamHI-digested pBTIR, yielding plasmid pBIRxy. The 1.2-kb himar1 transposase gene was amplified from pBADA7 with primers TnBam and TnKpn (Table (Table2)2) and ligated into BamHI-KpnI-digested pBIRxy. The ligation mixture was electroporated into E. coli strain DH5α. The new transposon plasmid, pBTn, was confirmed by PCR, restriction analysis, and sequencing and was transferred by electroporation via S. aureus RN4220 into S. aureus Xen36.
The delivery of IRerm into the S. aureus Xen36 chromosome was achieved by a temperature shift. A stationary-phase culture (24 h) of S. aureus Xen36 containing plasmid pBTn, grown at the permissive temperature (30°C) in TSB (without glucose) containing 0.5% xylose, chloramphenicol, and erythromycin, was diluted 1:100 into fresh TSB (without glucose) with 0.5% xylose and erythromycin; and the mixture was incubated at 42°C for 24 h. This was repeated twice with the antibiotics and once without the antibiotics. After S. aureus Xen36 containing plasmid pBTn was plated onto TSB agar containing erythromycin, the putative transposon insertion mutants (Ermr Cms) were selected by plating the Ermr colonies onto TSB agar containing chloramphenicol. Clones with an Ermr Cms phenotype were then preserved in glycerol stocks at −80°C for future screening.
Southern blot analysis was performed with digoxigenin high prime DNA labeling and detection starter kit I, according to the manufacturer's protocol (Roche), and erm as the probe. Preparations of chromosomal DNA from 10 randomly chosen transposon mutants were digested with EcoRI.
Single clones were inoculated in microtiter plates from precultures grown overnight (200 μl per well) with basic medium without glucose (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 0.1% K2HPO4) and grown with shaking at 37°C for 3 h to an optical density at 600 nm (OD600) of 2.5 to 3.0. The plates were centrifuged at 3,700 × g for 10 min, the supernatant was discarded, and wells were washed twice with phosphate-buffered saline (PBS; 10 mM sodium phosphate, pH 6.5, 100 mM NaCl). The cells were diluted to a concentration of 106/ml with PBS. To 100 μl cells, 5 μl of basic medium without glucose was added. The cells were incubated in the presence of 25 μg/ml dermcidin at 37°C for 3 h in 200 μl PBS. Bioluminescence was measured with a MicroLumat Plus LB 96V instrument.
The primers used for inverse PCR are listed in Table Table2.2. The first round of PCR was performed in a final volume of 30 μl. One of the arbitrary primers (primer arb1 or arb2; 40 pmol/reaction mixture) was paired with an external erm primer (primer erm-5.3, erm-5.2, erm-3.3, or erm-3.2; 20 pmol/reaction mixture). One microliter of the chromosomal DNA preparation was used as the template; and PCR was performed under the following conditions: 95°C for 5 min; 6 cycles of 94°C for 30 s, 30°C for 30 s, and 72°C for 1 min; 30 cycles of 94°C for 30 s, 45°C for 30 s, and 72°C for 1 min; and 72°C for 5 min. The second round of PCR was performed in a final volume of 30 μl with 3 μl of the PCR product from round 1 as the template (PureTaq Ready-To-Go TM PCR beads; Amersham Biosciences). Primer arb3 (40 pmol/reaction) was paired with the respective internal erm primer (primer erm-5.2, erm-5.1, erm-3.2, or erm-3.1; 20 pmol/reaction); and PCR was performed under the following conditions: 30 cycles of 94°C for 30 s, 45°C for 30 s, and 72°C for 1 min, followed by 72°C for 5 min. Nucleotide sequence analysis was performed by using the product from the second PCR directly with the respective erm internal primers. For identification of the IRerm insertion sites, a BLASTn search with the sequence representing the chromosomal DNA of the amplified fragments was performed.
To delete the fecCD gene from the genome of S. aureus Xen36, the homologous recombination procedure with plasmid pKOR1 was used (2). To that end, sequences of ~1 kb in length up- and downstream of the fecCD gene were amplified by PCR from the genomic DNA of S. aureus MW2, with BamHI restriction sites being introduced next to the gene to be deleted and att recombination sites being introduced at the distal ends, respectively (primers MW2101att1, MW2101att2, MW2101rev1, and MW2101rev2). After cloning of the PCR-amplified fragments in pKOR1, the resulting plasmid was transformed into RN4220, isolated, and then transformed into Xen36. The allelic recombination procedure was then performed as described previously (2). For genetic complementation, the fecCD gene was amplified by PCR with primers MW2101pst and MW2101sal and cloned in plasmid pT181mcs. For genetic complementation of the dak2 transposon mutant, the dak2 gene was amplified by PCR with primers dakBam and dakPst and cloned in plasmid pT181mcs.
Killing assays were performed as described previously (13). In brief, S. aureus cells were grown to exponential growth phase (OD600, 2.5 to 3.0); and the cultures were harvested, washed twice with PBS, and resuspended in the same buffer. The cells in each sample were diluted to a final concentration of 106 cells. The bacteria were exposed to a range of peptide concentrations and incubated at 37°C for 3 h, and appropriate dilution series of the samples were plated on TSB agar plates. The samples were incubated at 37°C for 24 h, and the surviving S. aureus cells were counted.
Cytochrome c binding assays were performed as described previously (22). Briefly, bacteria were grown to mid-exponential phase, washed twice with morpholinepropanesulfonic acid buffer (20 mM, pH 7.0), and incubated with 0.5 mg/ml cytochrome c for 10 min. After centrifugation, the amount of cytochrome c in the supernatant was determined photometrically by determination of its absorption at 530 nm.
Bacteria were grown to stationary growth phase and washed once in 20 mM sodium acetate buffer (pH 4.5), and the cells were extracted as a wet pellet by a method modified from that of Bligh and Dyer (3). The cell pellets were resuspended in chloroform-methanol-water (10:20:8, vol/vol/vol) at 4°C, and glass beads (Sigma) at four times the weight of the pellets were added to the mixture. The cells were disrupted by vortexing three times for 2 min each time at 4°C. After centrifugation at 3,700 × g for 15 min, the mixture was filtered and divided into two phases by adding chloroform and water to a final relation of chloroform-methanol-water of 20:20:18. After the cells settled (24 h), the lower phase was collected and the aqueous phase was reextracted with chloroform. The two organic phases were added and rotoevaporated under reduced pressure. The lipid extract was fractionated on a silica gel Sep-Pack cartridge (1 g). The sample was loaded on the top of the cartridge and eluted successively with chloroform (10 ml), acetone (10 ml), acetone-methanol (10 ml; 95:5, vol/vol), and methanol (30 ml). The last fraction, which contained purified phospholipids, was dried under nitrogen and stored at −20°C.
Thin-layer chromatography (TLC) was carried out as described previously (20). In brief, equal amounts of phospholipid extracts were spotted onto silica 60 F254 HPTLC plates (Merck) and developed with chloroform-methanol-water (65:25:4, by volume) in the first direction and chloroform-acetic acid-methanol-water (80:15:12:4, by volume) in the second direction. Phospholipids were selectively stained with molybdenum blue (Sigma-Aldrich). For one-dimensional TLC, only the first dimension was run. For quantification of the phospholipid species in the wild-type, dak2 mutant, and complemented strains, three different concentrations of each sample were spotted and the experiment was repeated twice. The nine values obtained were compared to those obtained with the wild-type strain, the mean for which was set equal to 100%. Densitometry was performed using Image Quant software.
Modification of fatty acids to their respective methyl ester derivatives was performed by incubating the phospholipid fractions with 1.2 N methanolic hydrochloric acid (Sigma-Aldrich) at 55°C for 4 h. Following modification, fatty acid methyl esters (FAMEs) were extracted with n-hexane and analyzed by gas chromatography-mass spectrometry (GC-MS; 6850 gas chromatograph, 5975C MSD; Agilent) on an HP-5ms capillary (Agilent). Fatty acids were identified by their mass spectra and retention times.
To quantify the binding of AMPs to phospholipids, phospholipids were isolated by the method described above. Phospholipids were dissolved in 10 mM sodium phosphate buffer (pH 6.5) with 100 mM NaCl. Aliquots (5 μl) of the samples were applied to a nitrocellulose membrane, which was then air dried and blocked with 0.5% skim milk in TBS buffer (150 mM NaCl, 10 mM Tris-HCl [pH 7.4]) overnight. The nitrocellulose membrane was washed with TBS buffer three times, incubated for 3 h with 100 μg/ml AMP, washed three times with TBS buffer, and incubated for 3 h with antidermcidin antiserum (1:200) (29) or commercial antisera against the other AMPs: rabbit anti-hBD3 (1:500; Leinco Technologies) and rabbit anti-human LL-37 (1:200; Cell Sciences). After the membrane was washed with TBS buffer three times, it was incubated for 2 h with an anti-immunoglobulin G (anti-IgG) alkaline phosphatase conjugate (1:5,000; Sigma). Spots were detected by addition of 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium (Sigma). The degree of immunoreactivity was measured with a photodetection system and quantitated (Quantity One 4.6; Bio-Rad).
S. aureus cells were grown to exponential growth phase (OD600, 2.5 to 3.0); and the cultures were harvested, washed twice with PBS, and resuspended in the same buffer. The cells in each sample were diluted to a final concentration of 106 cells. One milliliter of the bacteria was exposed to 100 μg/ml dermcidin and incubated at 37°C for 3 h. The cells were harvested and washed with PBS buffer once. The cells were resuspended in 200 μl of antidermcidin antiserum (1:200), and the mixture was incubated at 4°C for 12 h. Samples were washed with PBS and resuspended in the same buffer. The cell suspensions (25 μl) were attached to coverslips, and the cells were subsequently incubated with goat anti-rabbit IgG conjugated with 20 nm of gold (BB International) at 37°C for 2 h and then washed twice with 0.1 M CaCl2. Following antibody labeling, the cells were fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate and postfixed with 1% osmium tetroxide in 0.1 M sodium cacodylate. The samples were washed with distilled water, dehydrated in a graded ethanol series, dried to the critical point under CO2 with a Bal-Tec model cpd 030 drier (Balzers), mounted on aluminum studs, and sputter-coated with 100 Å of chromium in a model IBS/TM200S ion beam sputterer (South Bay Technologies), before they were viewed at 10 kV on a Hitachi S-4500 field emission scanning electron microscope (Hitachi) in backscatter imaging mode.
Membrane permeability was determined by using the membrane potential-sensitive fluorescent dye 3,3′-dipropylthiadicarbocyanine iodide [DiSC3(5); Invitrogen], as described previously (32). The bacteria were grown at 37°C in Mueller-Hinton broth (Difco) supplemented with 25 mg/liter of Ca2+ and 12.5 mg/liter of Mg2+ (CA-MHB) with shaking to mid-logarithmic phase (OD600, 0.5 to 0.6). The cells were harvested, washed once with buffer A (5 mM HEPES, pH 7.2, 5 mM glucose), and resuspended in the same buffer with 100 mM KCl and 2 μM of DiSC3(5) to an OD600 of 0.05; and the mixture was incubated for 15 min at room temperature to allow dye uptake and fluorescence quenching. The change in fluorescence (excitation = 622 nm, emission = 670 nm) was measured immediately after the addition of the desired concentration of the peptides tested on a SpectraMax fluorescence microplate reader (Molecular Devices). The maximal increase in fluorescence due to the disruption of the cytoplasmic membrane was recorded. A blank with only cells and the dye was used to subtract the background fluorescence.
Cells were grown in CA-MHB to mid-logarithmic phase, harvested, washed once with buffer A, and resuspended in the same buffer to an OD600 of 0.05. Different concentrations of the peptides tested were added to the cells; and the mixtures were incubated for 15, 30, and 60 min. Cells were removed by centrifugation, and the supernatant was assayed for ATP with a PKlight assay kit (Cambrex). Light production was measured with a Microlumatplus LB 96V instrument (Berthold). All assays were repeated at least three times.
Statistical analyses were performed by using GraphPad Prism (version 5.0) software and two-tailed t tests or one-way analysis of variance (ANOVA), where applicable. Error bars always show the standard error of the mean. Prediction of the membrane-spanning domains was performed using the TopPred program at http://mobyle.pasteur.fr/cgi-bin/MobylePortal/portal.py?form=toppred.
Resistance to dermcidin is usually determined by traditional killing assays, which are laborious and not amenable to high-throughput analysis (26). On the other hand, the analysis of MICs has proven difficult, as the peptide does not inhibit bacterial growth in media to a sufficient extent and over the necessary incubation time, and thus, very high concentrations of dermcidin must be used. Therefore, we developed a novel method to screen an S. aureus transposon bank for altered resistance to dermcidin in a high-throughput manner using a 96-well plate format. To that end, we used bioluminescent S. aureus strain Xen36, in which the lux genes that confer bioluminescence are integrated into a native plasmid and bioluminescence can be used to measure bacterial viability (7). Over 12 h of growth, there was a good correlation between growth and luminescence, and for cell concentrations of 105/ml to 107/ml, we detected a linear correlation between the numbers of CFU and luminescence (Fig. 1A and B). Furthermore, we used mariner-based transposon mutagenesis, which, in contrast to the frequently used Tn917, does not exhibit a bias for integration into specific sequences of the S. aureus genome (1). For this, we constructed a plasmid that contained the backbone of temperature-sensitive E. coli/Staphylococcus shuttle plasmid pBT2 (4), the himar1 transposase under the control of a xylose-inducible promoter (21), and an erythromycin resistance cassette between the horn fly transposon inverted repeats. By sequencing of the insertion locus in 50 randomly selected mutants, using inverted PCR, we determined that there was no apparent bias for mariner transposon integration in S. aureus Xen36, as most of the insertions (86%) were in different loci. Moreover, we performed Southern blot analysis with 10 randomly selected mutants to ascertain that transposon insertion occurred in only one locus per clone. This was the case in all mutants tested, which showed a single Southern blot signal (data not shown).
A screen with 7,000 mutants at the previously determined optimal concentration of 25 μg/ml dermcidin (Fig. (Fig.1C)1C) yielded 51 mutants with at least an ~10% increased susceptibility to dermcidin and 18 mutants with at least an ~10% decreased susceptibility to dermcidin. We selected 12 mutants, 6 with increased susceptibility and 6 with decreased susceptibility, for further screening on the basis of the frequency of mutations found in the respective genes and the degree by which susceptibility was changed (Table (Table3).3). Traditional killing assays confirmed the increased susceptibility for five of the six mutants and decreased susceptibility for three of the six mutants (Table (Table33).
For most mutants with increased susceptibility to dermcidin, the 50% lethal doses were only slightly different from that for the wild-type strain. We focused on one mutant in which the transposon had inserted in a gene with similarity to the permease component of ABC-type Fe3+-siderophore transport systems (FecCD). To confirm the increased susceptibility of the transposon mutant, we produced a deletion mutant of the gene in strain S. aureus Xen36 by allelic replacement and constructed a plasmid for genetic complementation. The results of the bioluminescence and killing assays with wild-type, deletion mutant, complemented, and control strains are shown in Fig. Fig.2.2. While genetic complementation confirmed the observed effect, the difference in the 50% lethal dose was less than 2 even for this mutant, for which the difference, as determined by killing assays, was the greatest among all mutants with increased susceptibility. Although the screen was likely not saturating, the fact that we did not detect mutant strains with strongly increased susceptibility to dermcidin underpins the notion that S. aureus does not have genes that confer high-level resistance to this AMP.
Among the three mutant strains for which killing assays confirmed increased resistance to dermcidin, one mutant had considerably increased resistance in the bioluminescence and the traditional killing assays (Fig. (Fig.3).3). In this S. aureus mutant strain, the transposon had integrated into a gene with homologues in a wide variety of bacteria. The corresponding gene product is similar to a family of dihydroxyacetone kinase-like proteins, collectively called DAK2 domain proteins, whose function is unknown. The DAK2 domain proteins contain an N-terminal predicted kinase/phosphatase domain (DAK2-like domain). Complementation of the transposon mutant with the entire DAK2 domain gene restored the original resistance to dermcidin, demonstrating that the gene is responsible for the observed dramatic change in dermcidin resistance (Fig. (Fig.3).3). Interestingly, we were not able, after several attempts, to construct an entire deletion mutant of the gene by allelic replacement, suggesting that the N terminus of the gene product, which contains the DAK2 domain and which is encoded upstream of the transposon insertion, has an essential function and that the function that determines altered resistance to dermcidin is located in the deleted C-terminal part of the gene product.
The appearance of the dak2 mutant strain in SEM was indistinguishable from that of the wild-type strain, and in vitro growth was only slightly affected (data not shown), indicating that deletion of the C-terminal part of the dak2 gene does not have major disadvantages for bacterial viability. Furthermore, we tested whether dermcidin induces transcription of the dak2-like gene by quantitative reverse transcription-PCR in the presence and absence of dermcidin but did not find altered amounts of transcript (data not shown), indicating that dermcidin does not induce the transcription of dak2.
One possible explanation for the observed altered resistance to dermcidin in the dak2 mutant would be that the interrupted gene encodes a membrane or surface protein that is crucial for the binding of dermcidin to the bacterial surface. However, on the basis of the predicted membrane-spanning regions and the absence of a signal peptide, the transposon-interrupted dak2-like gene does not code for a membrane or secreted protein.
As an alternative explanation of how the dak2 gene product might affect dermcidin binding, we hypothesized that it may affect nonprotein membrane components, namely, phospholipids. Thus, we determined the phospholipid patterns of the wild-type, dak2 mutant, and dak2-complemented strains by one- and two-dimensional TLC of membrane extracts. We found that, first, the total relative amount of diphosphatidylglycerol (DPG) in the dak2 mutant strain was significantly reduced to a mere 31% of the amount of DPG in the wild-type strain. The wild-type DPG level was restored in the dak2-complemented strain (Fig. 4A and C). In contrast, there were no significant changes in the other two phospholipid species, phosphatidylglycerol (PG) and lysyl-phosphatidylglycerol (LPG). Unchanged amounts of the cationic phospholipid LPG, an important component that provides resistance to cationic AMPs (20), indicate that the observed differential susceptibility of the mutant strain was not simply due to electrostatic repulsion. This was further supported by the lack of differential binding of the cationic protein cytochrome c to cells of the wild-type and dak2 mutant strains (Fig. (Fig.4D).4D). Second, both the PG and the DPG spot patterns showed differences in the two-dimensional TLC, mainly in retention along the first dimension. For example, the main PG spot was clearly shifted to the right. This possibly indicates different fatty acid compositions (Fig. (Fig.4B).4B). The spot pattern of LPG was unchanged. To determine whether the differences in the PG and DPG spot patterns were due to an overall change in the phospholipid fatty acid composition, we performed FAME GC-MS of phospholipid preparations. We observed slight, yet reproducible changes in the phospholipid fatty acid chain types (Fig. (Fig.5).5). Namely, there was a reduction of C17 and C19 branched-chain (iso and anteiso) fatty acids, which appeared to be compensated for by an increase in some straight-chain fatty acids (C14, C20). Thus, the dak2 gene product has a significant influence on the S. aureus phospholipid pattern that is accompanied by a change in the phospholipid fatty acid chain type.
To investigate whether the changed phospholipid pattern causes the increased resistance to dermcidin by preventing the binding of dermcidin to the bacterial surface, we performed immuno-SEM with antidermcidin antisera (Fig. (Fig.6).6). We observed the binding of dermcidin to the cell surface of the wild-type and the complemented mutant strains in a way similar to what has been described previously (29). In contrast, there was no binding of dermcidin to the surfaces of mutant strain cells, indicating that the changed phospholipid pattern impairs the binding of dermcidin to the bacterial surface.
To analyze whether the phospholipid pattern affects the binding of dermcidin in a direct or an indirect way, we purified phospholipids from wild-type, mutant, complemented mutant, and control strains and analyzed the interaction between dermcidin and phospholipids by the use of immuno-dot blots (Fig. (Fig.7).7). We detected significantly decreased levels of binding of dermcidin to the phospholipids from the mutant strain, which was restored in the complemented mutant, indicating that dermcidin directly interacts with these phospholipids. The interaction was specific for dermcidin, as the cationic AMPs LL-37 and hBD3 showed no or only small differences in binding to the phospholipid preparations from the different strains. These results indicate that dermcidin directly interacts with the phospholipids of S. aureus.
It is known that the antimicrobial activity of AMPs may be influenced by the different phospholipid compositions of the bacterial cytoplasmic membrane. For example, the acyl chain length affects the activity of AMPs via membrane fluidity (10, 16). Additionally, the activity of AMPs may also be influenced in a more specific fashion dependent on the charge and spatial arrangements of the phospholipid types (6, 19). To gain insight into how the change in the phospholipid pattern in the dak2 mutant strain affects the binding of dermcidin, we analyzed its susceptibility to other, cationic AMPs known to interact with bacterial membranes using killing assays. The susceptibilities to hBD3 3, LL-37, and nisin were not altered in the dak2 mutant strain (data not shown), indicating a specific interaction between dermcidin and membrane phospholipids. This specificity may be partially due to the anionic charge of dermcidin. However, it is not likely caused by a simple electrostatic interaction, because this would have also resulted in different susceptibilities to the cationic AMPs.
On the basis of the results presented here, which indicate a direct interaction of dermcidin with the bacterial cytoplasmic membrane, it is tempting to speculate that the bactericidal mode of action of dermcidin involves pore formation, a very common mechanism used by AMPs. In contrast, previous findings achieved with artificial liposomes, E. coli, and electron microscopy suggested that dermcidin does not act by pore formation (29). To determine whether dermcidin forms pores in S. aureus by biochemical analysis, we analyzed the release of ATP and membrane depolarization using the fluorescent dye DiSC3(5) (Fig. (Fig.8).8). There was no ATP release after addition of dermcidin, and membrane depolarization occurred only with extremely high concentrations of dermcidin, which most likely represents a secondary effect due to cell death. Accordingly, the level of membrane depolarization was higher in the wild-type strain than in the dak2 mutant at these high dermcidin concentrations, confirming the effect of the dak2-like gene on the susceptibility to dermcidin. Together, these studies indicate that dermcidin does not act by perturbing the cytoplasmic membrane by either pore formation or more global permeabilization, as shown for other host defense AMPs.
During coevolution with their hosts, bacteria have developed mechanisms of resistance to host defenses, which have a crucial influence on the outcome of a bacterial infection. AMPs form a key part of host defenses, particularly on the human skin, and have frequently been suggested to be lead substances for novel antimicrobial agents (8, 17, 18, 30). To exert antimicrobial activity, AMPs must bind to the bacterial surface, whether they act by the inhibition of biosynthetic processes on the bacterial surface, pore formation in the cytoplasmic membrane, or yet other mechanisms. The bacterial surface is negatively charged owing to the production of anionic polymers. Moreover, the bacterial cytoplasmic membrane has a negative charge on its outside due to phospholipid head groups and the charge gradient present in living bacterial cells. Thus, to interact with these structures, most AMPs have a positive charge. The anionic peptide dermcidin represents an important exception to this rule and is believed to have evolved to circumvent bacterial resistance mechanisms that are based on the cationic character of most AMPs.
In the present study we used a low-bias, large transposon-based screen to identify mutants of the human pathogen S. aureus with altered susceptibility to dermcidin. Although our screen was likely not saturating, it needs to be compared to more limited screens with Tn917, which has a higher integration bias, that have previously led to the identification of key mechanisms of resistance to cationic AMPs in staphylococci (20, 22). Thus, the fact that we did not find mutants with strongly increased susceptibility in our screen suggests that S. aureus does not have efficient specific mechanisms of resistance to dermcidin. This finding underpins the value of dermcidin as a key part of the innate host defense and confirms the notion that the anionic charge of dermcidin is an evolutionary adaptation to bacterial resistance mechanisms based on the electrostatic repulsion of cationic AMPs.
The mode of action of dermcidin is not known, but it has been reported to bind to the bacterial surface (29). However, due to its negative net charge, dermcidin cannot make use of the anionic charge of the bacterial membrane or surface to bind and, thus, likely needs different docking structures for that purpose. Our finding that dermcidin binding is dependent on the phospholipid pattern and that the binding to phospholipids is specific and direct will assist efforts aimed at pinpointing the docking molecules for dermcidin on the bacterial surface. In addition, while our results confirm that dermcidin does not act by pore formation, these results may also help in providing an understanding of dermcidin's mode of action. Finally, it will be an important task for future research to elucidate the enzymatic function of the DAK2 protein to determine whether it is directly involved in controlling the phospholipid composition.
We thank Philip Stewart for plasmid pBADA7.
This work was supported by the Intramural Research Program of the NIAID, National Institutes of Health (to M.O.) and the German Research Council (grant Schi 510/3-3 to B.S. and grant SFB 766 to B.S. and A.P.).
Published ahead of print on 13 July 2009.