|Home | About | Journals | Submit | Contact Us | Français|
Human β-defensin 3 (hBD3) is a highly charged (+11) cationic host defense peptide, produced by epithelial cells and neutrophils. hBD3 retains antimicrobial activity against a broad range of pathogens, including multiresistant Staphylococcus aureus, even under high-salt conditions. Whereas antimicrobial host defense peptides are assumed to act by permeabilizing cell membranes, the transcriptional response pattern of hBD3-treated staphylococcal cells resembled that of vancomycin-treated cells (V. Sass, U. Pag, A. Tossi, G. Bierbaum, and H. G. Sahl, Int. J. Med. Microbiol. 298:619-633, 2008) and suggested that inhibition of cell wall biosynthesis is a major component of the killing process. hBD3-treated cells, inspected by transmission electron microscopy, showed localized protrusions of cytoplasmic contents, and analysis of the intracellular pool of nucleotide-activated cell wall precursors demonstrated accumulation of the final soluble precursor, UDP-MurNAc-pentapeptide. Accumulation is typically induced by antibiotics that inhibit membrane-bound steps of cell wall biosynthesis and also demonstrates that hBD3 does not impair the biosynthetic capacity of cells and does not cause gross leakage of small cytoplasmic compounds. In in vitro assays of individual membrane-associated cell wall biosynthesis reactions (MraY, MurG, FemX, and penicillin-binding protein 2 [PBP2]), hBD3 inhibited those enzymes which use the bactoprenol-bound cell wall building block lipid II as a substrate; quantitative analysis suggested that hBD3 may stoichiometrically bind to lipid II. We report that binding of hBD3 to defined, lipid II-rich sites of cell wall biosynthesis may lead to perturbation of the biosynthesis machinery, resulting in localized lesions in the cell wall as demonstrated by electron microscopy. The lesions may then allow for osmotic rupture of cells when defensins are tested under low-salt conditions.
Host defense peptides (HDPs) are important effector molecules of the innate immune system, which provides effective barriers against a wide range of invading microorganisms. HDPs can be found in high concentrations mainly on epithelial surfaces and in the granules of neutrophils and have two key functions: (i) direct antimicrobial activities which lead to effective killing of microorganisms and (ii) immunomodulatory functions such as recruitment of immune cells, induction of cytokine release, and alteration of gene transcription (5, 6, 13, 14, 44).
Generally, HDPs are cationic peptides which can be found in all groups of organisms, from bacteria to vertebrates and mammals. They are gene encoded, consist of 10 to 50 amino acids (15, 50), and include structurally heterogenous groups such as the cecropins, magainins, bactenecins, protegrins, defensins, and cathelicidins (11, 15, 16, 21, 47), with the last two groups constituting the most relevant HDPs in mammals.
Defensins sensu stricto were discovered in mammals and are characterized by six conserved cysteine residues which form three disulfide bridge bonds. Two types are most prominent in humans: the α-defensins, which are produced primarily in neutrophils and in granules of Paneth cells of the small intestine, and the β-defensins, which are mainly expressed in epithelial cells (10, 22). Four human β-defensins have been described so far, with human β-defensin 3 (hBD3) possessing the highest positive charge of +11 (31, 35, 40, 43). In contrast to many other defensins, hBD3 is “salt insensitive” in that it kills microbes even at physiological salt concentrations (4, 15).
An important feature of cationic HDPs is the amphiphilic character, i.e., they are able to adopt three-dimensional structures in which polar and apolar residues are clustered on opposite sides of the molecule's surface. This structural feature appears to be important for their relatively selective interactions with microbial cell envelopes (48). It is generally assumed that, through patches of positive charges on the surface of the molecule, HDPs interact with negatively charged microbial cell envelopes and subsequently disrupt membrane barrier functions via pore formation or generalized perturbation of the bilayer (23, 45).
HDPs have retained their antimicrobial activity throughout evolution without selecting for high-level resistance, and yet pathogenic microbes have evolved mechanisms to reduce their susceptibility toward HDPs (33). A well-studied adaptation mechanism is based on reduction of the negative charge of the bacterial cell surface by introduction of d-alanine into teichoic acids or of l-lysine into phosphatidylglycerol of staphylococcal cell membranes (32, 34, 46). In Gram-negative organisms, e.g., in Salmonella enterica serovar Typhimurium, a two-component system, PhoPQ, senses the presence of cationic peptides and in response modulates lipopolysaccharide (LPS) structures and other membrane components. However, such adaptations did not lead to high-level resistance, as occurs for many human-designed antibiotics. More research on the molecular mode of action of HDPs on bacterial cells could help us to understand what has allowed these peptides to retain their activity over millions of years without eliciting high-level resistance. This could yield valuable information for the future design of new anti-infective drugs.
For such applications it is essential to better understand on the molecular level the mode of bactericidal activities of defensins. hBD3 probably interacts with membranes as a dimer forming a platform which remains floating on the surface with two long helices underneath sinking into the membrane interface (26). Consistent with such a model, we recently obtained evidence that hBD3 may not cause membrane disruption in Staphylococcus aureus but rather may interfere with the cell wall biosynthesis machinery (30, 38). This interpretation was also supported by early observations by Harder et al. (17), who first described hBD3 and reported cell wall perforations in hBD3-treated S. aureus cells reminiscent of those in penicillin-treated cells. To get further insight into such a mechanism, we here report on a series of cellular and biochemical in vitro cell wall synthesis experiments which suggest that inhibition of cell wall biosynthesis is a defined activity of hBD3. On the molecular level, it resembles cell wall biosynthesis inhibitors which directly interfere with the membrane-bound cell wall precursor lipid II.
Staphylococcus aureus SG511 was maintained on blood agar (Becton-Dickinson GmbH, Heidelberg, Germany). All whole-cell assays were performed in half-concentrated Mueller-Hinton (MH) broth (Oxoid, Basingstoke, United Kingdom), except in one experiment in which we used the assay medium as described by Dorschner et al. (9) to mimic in vivo-like growth conditions This medium contained 20% Trypticase soy broth (TSB), 10% fetal bovine serum (FBS; Invitrogen GmbH, Karlsruhe, Germany), and 150 mM NaCl in 70% minimal essential medium (MEM; Invitrogen GmbH, Karlsruhe, Germany) adjusted to pH 7.4. We used the medium with and without carbonate as described previously (9) and confirmed a 4-fold increase of hBD3 potency in the presence of carbonate. hBD3 and LL37 were synthesized and purified as described previously (4, 51).
Determination of MICs was performed in 96-well polypropylene microtiter plates (Nunc brand). Serial 2-fold peptide dilutions in half-concentrated MH broth were prepared from a stock solution of the respective peptide. Bacterial strains were grown to an optical density at 600 nm (OD600) of 1.0 and diluted 1:104-fold. Subsequently, 100 μl of this suspension was mixed with 100 μl of the peptide dilution in each well. After incubation for 18 h at 37°C, the MIC was read as the lowest concentration of antimicrobial agent resulting in the complete inhibition of visible growth, and results given are the means of at least two independent experiments performed in duplicate.
Antagonization of the antimicrobial activity of hBD3 by potential target molecules was monitored using polypropylene microtiter plates and half-concentrated Mueller-Hinton broth (Oxoid, Basingstoke, United Kingdom). To a conventional MIC determination putative target, molecules were added in a fixed concentration, yielding an increasing molar ratio of target to peptide. After 18 h of incubation at 37°C, the lowest ratio leading to complete antagonization of hBD3 activity was read.
For analysis of the cytoplasmic peptidoglycan nucleotide precursor pool of S. aureus SG511, we used the method of Kohlrausch and Holtje (18), elaborated for Bacillus subtilis, with slight modifications. Cells were grown in 5 ml of half-concentrated Mueller-Hinton broth to an OD600 of 0.5 and then supplemented with 130 μg/ml of chloramphenicol and incubated for 15 min. Chloramphenicol is necessary to prevent, under the impact of the antibiotic under investigation, induction of autolytic process and de novo synthesis of enzymes hydrolyzing the nucleotide-activated sugars interfering with determination of the soluble precursor (e.g., the work of Dai and Ishiguro ). Then peptides were added at 5× MIC as determined under standard conditions described above, and the cells were incubated for another 30 min. Subsequently, cells were rapidly cooled on ice and spun down (15,000 × g, 5 min, 4°C), resuspended in cold water, and, under stirring, treated with 2 volumes of boiling water. Cell debris was removed (48,000 × g, 30 min), and the supernatant was lyophilized. For C18 reverse-phase high-pressure liquid chromatography (HPLC), lyophilizates were dissolved in water and acidified to pH 2 with 20% (vol/vol) phosphoric acid; insoluble material was removed (15,000 × g, 5 min); aliquots of the supernatants, adjusted to identical cell wet weights for the differently treated cultures, were applied to the column. Separation was achieved under isocratic conditions with 50 mM sodium phosphate, pH 5.2, as solvent; the identity of the UDP-linked wall precursor was confirmed by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry using the negative mode and 6-aza-2-thiothymine dissolved in 50% (vol/vol) ethanol-20 mM ammonium citrate as matrix.
In vitro lipid II synthesis was performed using membranes of Micrococcus luteus as previously described (7, 42). In short, synthesis was performed in a total volume of 150 μl containing 300 to 400 μg of membrane protein, 10 nmol of undecaprenylphosphate (C55P), 100 nmol of UDP-MurNAc-pp, 100 nmol of UDP-N-acetylglucosamine (UDP-GlcNAc) in 60 mM Tris-HCl, 5 mM MgCl2, pH 7.5, and 0.5% (wt/vol) Triton X-100. UDP-MurNAc-pp was purified as described elsewhere (18). Bactoprenol-containing products were extracted with butanol-pyridine acetate (2:1, vol/vol, pH 4.2), purified as described previously, and analyzed by thin-layer chromatography (TLC), using phosphomolybdic acid (PMA) staining (42). For testing the impact of hBD3, the defensin was added in molar ratios with respect to C55P. For purification of milligram quantities of lipid II to be used as substrate in the enzyme assay, the analytical procedure was scaled up by a factor of 500 and lipid II was purified as described previously (42). Radiolabeled lipid II was synthesized using [14C]UDP-GlcNAc as substrate. Purification of lipid I followed the same protocol except that UDP-GlcNAc was omitted.
Cloning, expression, and purification of the cell wall biosynthesis enzymes penicillin-binding protein 2 (PBP2), MraY, MurG, and FemX of S. aureus NCTC 8325 were performed as described previously without any modifications (41).
To determine the enzymatic activity of purified MraY-His6, assays were carried out in a total volume of 50 μl containing 5 nmol C55P or [3H]C55P (14.8 GBq/mmol; Biotrend, Cologne, Germany), 50 nmol of UDP-MurNAc-pp in 100 mM Tris-HCl, 30 mM MgCl2, pH 7.5, and 10 mM N-lauroylsarcosine. The reaction was initiated by the addition of 7.5 μg of the enzyme, and the reaction mixture was incubated for 1 h at 37°C.
The MurG activity assay was performed in a final volume of 30 μl containing 2.5 nmol purifed lipid I, 25 nmol UDP-GlcNAc or [14C]UDP-GlcNAc in 200 mM Tris-HCl, 5.7 mM MgCl2, pH 7.5, and 0.8% Triton X-100 in the presence of 0.45 μg of purified MurG-His6 enzyme. The reaction mixture was incubated for 30 min at 30°C.
The assay for the synthesis of lipid II-Gly1 catalyzed by FemX was performed as described previously without modifications, including the presence of 0.8% Triton X-100 (42). In a second series of experiments Triton X-100 was omitted in order to test the effect of detergent on the action of hBD3.
Enzymatic activity of PBP2 was determined by incubating 2.5 nmol lipid II or 1 nmol [14C]lipid II in 20 mM morpholineethanesulfonic acid (MES), 5 mM MgCl2, pH 5.5, in a total volume of 50 μl. The reaction was initiated by the addition of 7.5 μg PBP2-His6, and the reaction mixture was incubated for 1.5 h at 30°C. In all in vitro assays hBD3 was added in molar ratios ranging from 0.5 to 2 with respect to the amounts of C55P, lipid I, and lipid II, respectively.
Synthesized bactoprenol-containing intermediates were extracted from the reaction mixtures with n-butanol-pyridine acetate, pH 4.2 (1:1, vol/vol), and analyzed by TLC. Quantification of enzyme reactions was routinely performed by PMA staining and subsequent evaluation of spot density using ImageQuant TL software (Amersham); all experiments were performed at least three times. To increase accuracy, the Fem and PBP2 assays were confirmed at least once using 14C-labeled glycine (for the Fem reactions) and lipid II with a 14C-labeled GlcNAc moiety; radiolabeled bands were quantified using a storage phosphor screen to visualize radioactivity in a Storm imaging system (GE Healthcare). For quantification of the PBP2-catalyzed lipid II polymerization, reaction mixtures were applied directly onto TLC plates developed in solvent B (butanol-acetic acid-water-pyridine [15:3:12:10, vol/vol/vol/vol]), and quantification of residual radiolabeled free lipid II was determined using phosphorimaging.
Mid-logarithmic-phase S. aureus SG511 (approximately 5 × 108 CFU/ml) was exposed to 15 μM hBD3 in 10 mM sodium phosphate buffer, pH 7.4, with 20% Mueller-Hinton broth for 10, 30, or 60 min at 37°C. At intervals, bacterial cells were collected by centrifugation at 2,000 × g, resuspended in 1 ml of freshly prepared 2.5% glutaraldehyde in 0.2 M sodium cacodylate buffer, pH 7.4, and fixed for 2.5 h at room temperature. Samples were then washed with cacodylate buffer and postfixed in 1% osmium tetroxide in 0.1 M cacodylate buffer, pH 7.4 (the osmolarity of the buffer was adjusted to 450 osM by addition of sucrose), for 40 min. Samples were washed three times in cacodylate buffer, and cell pellets were placed in 1% agarose and then dehydrated in a graded (50 to 100%) ethanol series and embedded in Araldite-Epon resin mix using acetone as an intermediate solvent; polymerization was carried out subsequently at 37, 45, and 60°C for 3 days. Ultramicrotome sections were contrasted with uranyl acetate and lead citrate and examined with a JEM 1011 electron microscope operated at 80 kV with digital image acquisition.
It is generally accepted that, at the very high concentrations locally occurring in the host, cationic antimicrobial peptides may destroy bacterial cells by first binding to the surface and subsequently disrupting the membrane bilayer structure through mechanisms such as that described by the carpet model (28, 36, 45). However, we recently demonstrated, using micromolar concentrations in the MIC range, that hBD3 hardly affected the membrane integrity but strongly induced genes of the core cell wall stress stimulon of staphylococci (38) as defined by McAleese et al. (24). Such genes included those for the two-component system VraRS, a cell wall stress sensor system (12, 20), and putative detoxification ABC transporters such as VraDE (38).
To obtain evidence that inhibition of cell wall biosynthesis is a relevant component of the antibiotic activity of hBD3, we first determined the cytoplasmic levels of UDP-N-acetylmuramyl pentapeptide (UDP-MurNAc-pp) in Staphylococcus aureus SG511. Antibiotics such as vancomycin, bacitracin, or beta-lactams, which interfere with the late, membrane-bound stages of peptidoglycan synthesis, trigger an accumulation of the final soluble peptidoglycan precursor (Fig. (Fig.1)1) in the cytoplasm (18). After hBD3 treatment (27.8 mg/liter, corresponding to 5× MIC as determined earlier ) for 30 min, significant accumulation of the precursor was observed (Fig. (Fig.2D),2D), similar to the accumulation seen with vancomycin-treated controls (Fig. (Fig.2B).2B). Comparable results were obtained both in the standard test medium (half-concentrated MH broth) and in the carbonate-containing medium of Dorschner et al. (9) which mimics in vivo-like growth conditions (Fig. (Fig.2E).2E). In contrast, treatment of cells with LL37, a linear cationic HDP with a stronger impact on the bilayer structure (27), did not lead to an accumulation of UDP-MurNAc-pp (Fig. (Fig.2C).2C). This result is significant and indicative in two ways: (i) hBD3 obviously does not depolarize the cytoplasmic membrane to such an extent that cells are unable to perform biosynthesis reactions and (ii) hBD3 should not induce gross membrane leakage since the precursor was retained in the cytoplasm and did not leak from treated cells. This interpretation is further substantiated by previously published data that hBD3 (38) and a template β-defensin (37) only slightly affected the membrane potential of growing staphylococcal cells (reduction by 15 to 20 mV at 5× MIC). In contrast, in control experiments with 5× MIC of LL37 a substantial decrease of the membrane potential by 50 to 60 mV was observed (data not shown), indicating that this peptide intrinsically may have a higher capacity for bilayer perturbation than hBD3 or may directly interfere with energy generation.
When growing cells of S. aureus strain SG511 were inspected by transmission electron microscopy, small membrane protrusions filled with cytoplasmic contents were observed after 30 min of treatment with hBD3. Such evaginations were not detectable after 10 min and became more frequent after 60 min of treatment (Fig. (Fig.3).3). The protrusions were rather well defined in shape and obviously occurred through small lesions in the cell wall, in agreement with observations made by Harder et al. (17). Generalized and rapid cell wall lysis as mediated by autolytic enzymes, particularly in the septum area, was not detectable; such lytic events were described for Staphylococcus simulans after treatment with cationic lantibiotics (3) and with a template defensin (37). Only after 60 min of treatment could cells with progressive cell wall damage be identified (Fig. (Fig.3,3, panel 9).
To obtain information on the molecular activities of hBD3, we performed a series of in vitro cell wall biosynthesis assays using membrane preparations for the overall reaction and purified enzymes for testing the individual steps of the membrane-bound synthesis cycle (Fig. (Fig.1).1). Membrane preparations of Micrococcus luteus contain sufficient MraY and MurG activity for the in vitro formation of the cell wall precursors lipid I and lipid II when the soluble precursor UDP-MurNAc-pp is added (49). Addition of hBD3 to such a test system did not lead to inhibition of lipid I and II biosynthesis (data not shown). Similarly, testing the individual reactions of MraY (by adding UDP-MurNAc-pp to the bactoprenol carrier to form lipid I), MurG (by adding GlcNAc to form lipid II), and FemX, which adds to lipid II the first glycine residue of the interpeptide bridge, did not indicate any inhibition by hBD3 (data not shown). All of these assays have in common that 0.8% (by volume) Triton X-100 is routinely included in the assay mixture, which we suspected to interfere with hBD3 activity. Among those assays, we found that only with FemX can Triton be omitted without affecting enzyme activity, allowing us to study a possible detergent effect on hBD3 inhibitory activity (Fig. (Fig.4).4). Indeed, in the absence of Triton, hBD3 almost completely inhibited the FemX reaction when added in equimolar concentrations with regard to lipid II, indicating that the defensin may stoichiometrically bind to the substrate of the enzyme (Fig. (Fig.4).4). In contrast, no inhibition was observed in the presence of Triton X-100, even with a 2-fold molar excess of hBD3, whereas vancomycin and nisin were fully inhibitory under these conditions (Fig. (Fig.4).4). The reasons for the difference between hBD3 and the antibiotics are unclear; besides a lower affinity of hBD3 for lipid II, it may be possible that Triton also affects dimerization of hBD3, which appears necessary for antibiotic activity (40).
For further analysis of the inhibitory nature of hBD3, we chose another lipid II-consuming reaction, the transformation of monomeric lipid II into a polymeric peptidoglycan as catalyzed by the bifunctional penicillin-binding protein 2 (PBP2) of S. aureus, which occurs on the outside of the membrane and therefore is the likely target reaction of hBD3. Using this assay, both the disappearance of lipid II and the appearance of one product, the liberated bactoprenol carrier C55P, can be detected simultaneously. Purified, recombinant PBP2-His6 was incubated with lipid II and hBD3, and the lipids were subsequently extracted and separated by TLC (Fig. (Fig.5A).5A). For reliable quantification of residual lipid II, we also used radiolabeled lipid II as substrate (Fig. (Fig.5B).5B). In the presence of hBD3, the PBP2-catalyzed reaction was clearly blocked. Again, at a molar ratio of hBD3 to lipid II of 1:1, only 19% lipid II was transglycosylated; at a ratio of 2:1, inhibition of the reaction was complete. Thus, the inhibitory potency of hBD3 was comparable to that of nisin (96% inhibition at a nisin/lipid II molar ratio of 2:1).
Since the Triton experiments indicated that hBD3 may bind lipid II less tightly than nisin or vancomycin, we tried to obtain information on the specificity of the inhibition and the involvement of ionic interactions between lipid II with two phosphate groups and hBD3 (net charge, +11). Interactions of hBD3 and lipid II solely based on electrostatics should be weakened by other cationic peptides which could also interact with the negative charges in the lipid II molecule. We chose the lantibiotic Pep5 with overall similar physicochemical properties (molecular mass of 3,488, net charge of +8), which effectively kills staphylococci through pore formation (19) and which does not interfere with cell wall biosynthesis, as well as the synthetic tripeptide Lys-Lys-Lys (net charge, +3), and conducted the same assay. Neither Pep5 (molar ratio, 2:1) nor the tripeptide (molar ratios of 20:1 and 50:1) was able to inhibit the PBP reaction and therefore may not bind to lipid II with the same affinity as that with hBD3. On the other hand, both were able to some extent to interfere with the inhibitory activity of hBD3 when incubated with lipid II before hBD3 and the enzyme were added (Fig. (Fig.6).6). These data suggest that ionic forces are involved in hBD3 binding to lipid II and yet are not sufficient to fully describe the interaction between the defensin and the cell wall precursor.
To obtain further information on the lipid II-hBD3 interaction, we performed antagonization assays, based on conventional broth microdilution MIC determination in the presence of lipid II and its structural components and related compounds. For this, the potential antagonist was added in different molar ratios with respect to hBD3 to a cell suspension of S. aureus SG511 and supplemented with a bactericidal concentration of hBD3 (1× to 8× MIC). Growth of bacteria after 18 h of incubation indicated an antagonizing effect of the putative target structure by binding of hBD3, thus making it unavailable for interaction with staphylococcal cells. As potential antagonists, we used lipid II, undecaprenylphosphate (C55P, the lipid carrier of the disaccharide-pentapeptide unit of lipid II), two derivatives of C55P with shorter lipid tails (C20P and C15P, consisting of 4 and 3 isoprenoid units, respectively), pure phospholipids (dioleoylphosphatidylglycerol [DOPG]), UDP-GlcNAc, UDP-MurNAc, and glucosamine-6-phosphate, respectively. Of these putative antagonists, only those with negative charges and the potential to form micelles in aqueous solutions were able to interfere with hBD3 activity but, however, with different potencies as indicated by different molar ratios necessary for antagonization. Phospholipids (DOPG), C55P, and the shortened carrier C20P antagonized at a molar excess of 2 to 5, whereas only lipid II was able to bind hBD3 at a 1:1 molar ratio. Activated cell wall sugars such as UDP-GlcNAc, UDP-MurNAc, and glucosamine-6-phosphate, as well as C15P, which is water soluble and unable to form micelles, did not have any antagonizing effect. Obviously, besides a negatively charged membrane-like environment which is known to attract amphiphilic cationic peptides, lipid II provides additional structural elements, possibly in the disaccharide moiety, for intensified interactions with hBD3.
Amphiphilic cationic host defense peptides are generally assumed to kill microbes by disruption of the membrane bilayer. Here we present evidence that for hBD3 in concentrations close to its MIC inhibition of cell wall biosynthesis is a major component of its antibiotic activity against staphylococci and that interactions with the membrane-bound cell wall building block lipid II appear to be involved in inhibition. These conclusions are based on transcriptional response analysis as published earlier (38) and on experiments with intact cells, including electron microscopy and a series of in vitro assays with lipid II-consuming enzyme reactions. Electron microscopy strongly suggests that through the activity of hBD3, local lesions in the cell wall layer are created through which eventually membranes and cytoplasmic contents protrude as a result of osmotic pressure. These local lesions could thus represent the tesserae proposed in the “hydro-osmotic transtesseral extrusion and rupture (HOTTER)” mechanisms of AMP action (29).
Where in the cell could those local lesions occur? There is increasing evidence that bacterial cell walls are synthesized by highly organized multienzyme machineries which assemble at particular sites, i.e., the septum and septum initiation sites in coccoid cells and in rod-shaped bacteria and additionally at sites which follow the helically arranged cytoskeleton, to enable elongation growth (for a review, see, e.g., the work of Scheffers and Pinho ). These sites have been shown to be rich in negatively charged phospholipids such as phosphatidylglycerol and cardiolipin, an essential feature for the correct assembly of the division machinery for which a cationic helical stretch in the division protein MinD seems important (1). Necessarily, these sites are also rich in lipid II and therefore represent ideal binding sites for hBD3. It seems reasonable to argue that the presence of hBD3 at such an exposed site would interfere, like “sand in a gearbox,” with the coordinated assembly and function of the machinery, resulting in localized inhibition of cell wall biosynthesis and subsequent lesions as observed by electron microscopy in this study and in others (e.g., the work of Harder et al. ).
Transcriptional response data with hDB3, but also with other antimicrobial peptides (30, 34), clearly indicate that induction of the cell wall stress stimulon is the most prominent response of staphylococcal cells to treatment with HDPs. Nevertheless, membrane depolarization to various degrees (mild with hBD3 and rather strong with LL37) and the concomitant impact on energy-consuming cellular processes certainly contribute to killing as reported earlier (36). It remains to be studied whether the observed membrane depolarization results from destabilizing the lipid bilayer or from a direct impact of the peptides on energy generation processes. Most strikingly, a direct interference with electron transport and components of respiratory chains has been reported for a linear cationic peptide derived from the human bactericidal/permeability-increasing protein (BPI ) and for gramicidin S (25). If such activities can be identified more frequently, it appears that not only cell wall biosynthesis but also other multiprotein machineries organized in the cytoplasmic membrane, e.g., the electron transport chain, may be affected. Such pleiotropic effects of cationic HDPs may explain why antibiotic peptides remained active throughout evolution and why target microbes may have been able to develop mechanisms for reducing susceptibility (for a review, see, e.g., the work of Peschel and Sahl ) but scarcely developed high-level resistance mechanisms as observed with clinically used antibiotics.
The financial support by the German Federal Ministry of Education and Research (BMBF, SkinStaph project), the German Research Foundation (DFG, Sa292/13-1), the BONFOR program of the Medical Faculty, University of Bonn, and the European Commission (NAM project, 218191) is gratefully acknowledged.
Editor: F. C. Fang
Published ahead of print on 12 April 2010.