Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, April 2009, p. 1598-1609, Vol. 53, No. 4
0066-4804/09/$08.00+0 doi:10.1128/AAC.01329-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.
,
Department of Microbiology, Cornell University, Ithaca, New York 14853-8101
Received 3 October 2008/ Returned for modification 26 November 2008/ Accepted 6 January 2009
|
|
|---|
|
|
|---|
Antibiotics have also proven to be useful tools for microbial cell biology and have allowed the visualization of the subcellular location of cell envelope biosynthetic processes. For example, fluorescently labeled vancomycin and ramoplanin, antibiotics that bind specifically to the un-cross-linked PG precursors and/or lipid II, have served to confirm the helical arrangement of the lateral cell wall biosynthetic complexes (14, 58), which was also shown in studies using green fluorescent protein-tagged envelope proteins (53).
We have used Bacillus subtilis as a model system to investigate the genetic and physiological responses to both antibiotics and coculture with antibiotic-producing strains. To date, our studies have focused on cell envelope antibiotics (including vancomycin, bacitracin, and fosfomycin) and bacteriocins, such as nisin, duramycin, and sublancin (17, 43). Exposure to these compounds activates distinct cell envelope stress responses controlled by extracytoplasmic function (ECF)
factors and two-component regulatory systems (TCS) (34). In most cases, the activity of ECF
factors and TCS is controlled by transmembrane sensors (anti-
factors or membrane-bound histidine protein kinases, respectively) which thereby allow gene expression to be regulated in response to changes in the cell envelope. The identification of genes induced by a certain antibiotic stress provides insights into the nature of the antibiotic's target(s) and also aids in the identification of resistance functions, many of which are inducible by their cognate antibiotic (30).
Daptomycin, a cyclic lipopeptide antibiotic originally purified from Streptomyces roseosporus (46), is notable for its activity against methicillin-resistant Staphylococcus aureus and certain streptococci and enterococci. The mechanism of action of daptomycin has been controversial. Initial studies suggested that daptomycin inhibited lipoteichoic acid synthesis (6). However, these findings could not be verified (40). The current proposed mechanism of action involves the insertion of its decanoyl side chain into the cytoplasmic membrane in a Ca2+-dependent manner. Subsequent oligomerization, followed by depolarization of the membrane potential and efflux of potassium ions, leads to the arrest of protein, RNA, and DNA synthesis (56). It has been suggested that daptomycin approaches the bacterial membrane in the form of micelles composed of 14 to 16 daptomycin molecules and an equal number of Ca2+ ions, which are proposed to help mask the negative charge of daptomycin. After insertion into the membrane, daptomycin dissipates the membrane potential and leads to cessation of macromolecule synthesis (24, 54, 56). Daptomycin treatment does not result in cell lysis or in daptomycin entering the cytoplasm (6, 13).
Here, we have investigated the genetic and physiological responses of B. subtilis to daptomycin. Using transcriptional profiling, we demonstrate that daptomycin strongly activates the LiaRS TCS, which regulates the liaIHGFSR operon. Mutants defective for liaH, which encodes a phage-shock protein A (PspA)-like membrane stress protein (35), were threefold more susceptible to daptomycin. This susceptibility was further exacerbated in cells additionally lacking the paralogous gene pspA. Fluorescence microscopy studies using Bodipy FL-labeled daptomycin (daptomycin-BDP) together with strains having altered membrane lipid composition support a model in which the daptomycin-Ca2+ complex interacts preferentially with regions enriched in anionic lipids (primarily phosphatidylglycerol [PhG] in B. subtilis) and is localized at new cell division septa and in a helical pattern along the long axis of the cell.
(Part of this work was presented at the 47th Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL, September 2007.)
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Strains used in this study
|
|
View this table: [in a new window] |
TABLE 2. Oligonucleotides used in this study
|
|
View this table: [in a new window] |
TABLE 3. Daptomycin stimulona
|
|
View this table: [in a new window] |
TABLE 4. MIC of B. subtilis mutants and strains with altered membrane composition, or deletion of transcriptional regulators
|
![]() View larger version (19K): [in a new window] |
FIG. 1. Daptomycin stimulon in B. subtilis. The scatter plot represents the average expression levels of treated (+) versus untreated (–) cultures of B. subtilis W168 (daptomycin [DAP] at the MIC of 1 µg/ml; 20 min) from triplicate microarray analyses. The key lists highly expressed genes as grouped by their corresponding transcriptional regulators.
|
10-fold higher MIC than that induced by daptomycin treatment. Furthermore, daptomycin-BDP was able to induce cat-lacZ reporter fusions to liaI independent of the solvent. One hundred microliters of cells from exponential or stationary growth phases was incubated with 4 µl daptomycin-BDP (0.5 mg/ml) in 50% dimethyl sulfoxide for 10 min, with 2 µl vancomycin-BDP (0.1 mg/ml) mixed 1:1 with 0.1 mg/ml vancomycin, or with 1 µl FM 4-64 (0.5 µg/ml) for 20 min. After washing in Mueller-Hinton broth the cells were mounted in the antifade reagent Citifluor (Ted Pella, Inc.) on poly-L-lysine (Sigma-Aldrich)-treated slides. Nomarski optics differential interference contrast (DIC) or fluorescent images were taken with an Olympus BX61 epifluorescence microscope with a UPlanApo 100x (numerical aperture of 1.35) objective. The microscope is equipped with fluorescence filter cubes for viewing DAPI (4',6-diamidino-2-phenylindole), fluorescein isothiocyanate, and Cy3. Images were acquired using a Cooke SensiCam with a Sony interline chip. Image acquisition and postprocessing were performed using the SlideBook software package (Intelligent Imaging). Cluster analysis. Results of whole genome microarray analyses of B. subtilis with a set of antimicrobial compounds from our data and from a study by Hutter et al. (30) were compared by complete linkage clustering (arrangement based on treatment and genetic response similarity) using the Gene Cluster 3.0 software. The resulting cluster was visualized with Treeview 1.60, written by Michael Eisen (18).
Microarray data accession number. The complete set of raw and normalized data for each of the triplicate DNA microarray experiments involving B. subtilis treated with daptomycin is available at the Gene Expression Omnibus database (http://acbi.alm.nih.gov/geol/) under accession no. GSE13900.
|
|
|---|
83 genes at least twofold, and many of these are members of known antibiotic-responsive regulons (Fig. 1 and Table 3) (17, 44). The strongest response to daptomycin treatment was the induction of the autoregulated liaIHGFSR operon which encodes the antibiotic-responsive LiaRS TCS (Table 3). Lia was named in reference to its strong induction by lipid-II-interacting antibiotics (35, 44), although it is now clear that there is an imperfect correlation between lipid II binding and induction (61). The Staphylococcus aureus VraSR system is orthologous to the LiaRS TCS (35). It has been implicated in mediating antibiotic resistance (22, 38) and was also strongly induced by daptomycin, as deduced by microarray analyses (47).
In addition to the liaI operon itself, daptomycin induced several other genes in a LiaR-dependent manner. These included the yhcYZ operon, which was previously shown to be a direct target of LiaR activation (35), and the yvrI and yvrL genes. The yvrI gene has recently been shown to be a divergent member of the
70 family of transcriptional regulators (41) and activates expression of oxalate decarboxylase (OxdC), a major cell wall protein (1). It is not yet clear if yvrI is a direct or indirect target of LiaR activation.
In B. subtilis, a number of genes involved in antibiotic resistance and cell wall metabolism are regulated by the ECF
factors
W and
M. Daptomycin weakly induced six genes of the
W regulon (10, 12) and 19 genes of the
M regulon (17, 32) (Table 3). To monitor the transcriptional responses of the liaI, sigW, and sigM genes to daptomycin treatment, we used transcriptional (cat-lacZ) reporter fusions and found that all were induced by daptomycin (data not shown).
Phage-shock protein A homologs LiaH and PspA protect against daptomycin. Of the liaIHGFSR genes, liaFSR are well conserved within gram-positive bacteria with a low G+C content. Only the bacilli also contain the phage-shock protein A homolog LiaH. S. aureus harbors the liaFSR orthologs SA1702 and vraSR (35). When testing for daptomycin susceptibility, Muthaiyan et al. observed an increased susceptibility of a vraSR mutant strain to daptomycin (0.78 µg/ml versus 1.0 µg/ml for the wild type) (47). Here, we have used growth inhibition studies to examine whether the daptomycin-induced liaI operon affects susceptibility to daptomycin. A deletion of the entire liaI operon significantly increased susceptibility from an MIC of 1.0 µg/ml for the wild type to an MIC of 0.3 µg/ml (Table 4). A null mutant of the negative regulator liaF, which overexpresses LiaIH (34, 35), did not appear to increase resistance.
To test for possible functional redundancy between liaH and its paralog pspA (which was not strongly induced by daptomycin treatment despite being part of the
W regulon) we tested single and double mutants of these loci. A pspA deletion produced almost no change in MIC compared to that of the wild type, but the liaH pspA double mutation further decreased the MIC by twofold relative to that of the liaH single mutant. These results indicate that the PspA (phage-shock protein A) homologs, LiaH and PspA, both contribute to decreased daptomycin susceptibility and that the induction of LiaH by daptomycin is adaptive.
The mechanisms by which PspA proteins protect cells against membrane disruption are unclear but are likely to involve direct interactions with the inner surface of the membrane (15). Escherichia coli PspA forms abundant, oligomeric ring-like structures that are speculated to coat the inner surface of the membrane and thereby prevent proton leakage (39). In vitro, E. coli PspA binds preferentially to liposomes containing anionic lipids and suppresses proton leakage (39). LiaH has also been observed to be an abundant oligomeric protein with a similar ultrastructure (T. Mascher, personal communication). Together, these findings are consistent with the notion that daptomycin toxicity results from disruption of the membrane integrity and that the two B. subtilis PspA paralogs can counteract this disruptive effect.
ECF
factors also contribute to decreased daptomycin susceptibility.
Since several
W and
M regulon members were upregulated upon daptomycin treatment, we tested null mutants of ECF
factor genes for daptomycin susceptibility. In the singly mutant strains, there was a slight decrease in MIC for sigM and sigW (0.8 µg/ml for both), whereas a sigX mutant was unaffected. Multiply mutant strains displayed even greater susceptibility, with the lowest MIC noted for the sigXM double (0.6 µg/ml) and the sigMWX triple mutants (0.6 µg/ml) (Table 4). Increased daptomycin susceptibility in the multiply mutant strains is consistent with the recent demonstration that these three ECF
factors have overlapping regulons and multiply mutant strains are often more susceptible to antibiotics than are single mutants (42).
In B. subtilis, several ECF
factors have been implicated in conferring resistance to cell envelope-active antibiotics. For example,
X regulates the dlt operon and the pssA ybfM psd operon, which reduce the net negative charge of the cell envelope by D alanylation of teichoic acids and insertion of phosphatidylethanolamine into the membrane, respectively. As a result, sigX mutants are more susceptible to cationic antimicrobial peptides (8). The
W regulon includes a large number of genes implicated in resistance against both small molecule inhibitors, such as fosfomycin, and peptide antibiotics, such as sublancin and SdpC (4). Finally, the
M regulon has recently been shown to include many genes known to be important for cell envelope synthesis, and sigM mutants are susceptible to some cell wall antibiotics, such as moenomycin and bacitracin (17, 42). The identity of the ECF
factor-dependent operons that confer daptomycin protection is not yet clear.
Mutants with altered membrane composition affect daptomycin susceptibility. Since the decanoyl side chain of daptomycin is predicted to insert into the membrane, we tested whether susceptibility is influenced in strains with altered membrane lipid composition. Daptomycin susceptibility was measured for B. subtilis strains lacking phosphatidylethanolamine (psd, pssA), lysyl-PhG (LPhG) (mprF), glycolipids (ugtP), or cardiolipin (ywnE). Of the null mutants, only the mprF mutant showed a significant difference compared to the wild type (MIC of 0.5 µg/ml versus 1.0 µg/ml [Table 4]). Moreover, overexpression of mprF led to slightly decreased susceptibility (MIC, 1.3 µg/ml). MprF catalyzes the tRNA-dependent modification of PhG with lysine to form the positively charged LPhG (57).
It has been shown earlier that a change in membrane charge due to mprF disruption affects the susceptibility to antimicrobial agents in S. aureus (49). Here, we speculate that the reduction of the net negative charge of the membrane upon increased production of LPhG functions to reduce the affinity of a positively charged daptomycin-Ca2+ complex due to electrostatic repulsion. Indeed, previous studies of S. aureus strains that were selected for increased daptomycin resistance found that point mutations in mprF frequently occurred as an early event during selection (21). However, the effect of these mutations alone was quite modest, and further selection led to additional mutations in the yycFG TCS and RNA polymerase subunit genes rpoBC (21). Since an mprF null mutant is more susceptible to daptomycin, we suggest that these mprF mutations may have been gain-of-function mutations. Independently, Jones et al. found that daptomycin resistance in S. aureus was correlated with the increased translocation of LPhG from the inner to the outer leaflet of the membrane without changing the overall concentration of LPhG (33). An increase of mprF gene expression was not seen upon daptomycin treatment in B. subtilis, but an increase in positive charge through LPhG translocation to the outer leaflet and an additional effect of reduction in the membrane net negative charge by
X could together affect the ability of daptomycin to insert into the membrane.
The physical properties of the membrane are determined by both the membrane head group composition and the length and desaturation of the fatty acyl side chains. In B. subtilis, the fluidity of the membrane is regulated, in large part, by the lipid desaturase Des, which introduces cis double bonds at the fifth position of the fatty acyl chains (
5) in response to reduction in temperature. The desaturase is under the control of the DesRK TCS (41a). Deletion of des resulted in increased susceptibility to daptomycin, and this effect was especially notable during growth at low temperatures (MIC at 25°C of 0.4 µg/ml versus 0.7 µg/ml for the wild type). The underlying mechanisms of the effect of des deletion are not entirely clear. The increased rigidity of the membrane in a des mutant might facilitate the membrane disruptive action of daptomycin and impair repair mechanisms by the cell, or the decrease in unsaturated fatty acyl moieties might affect interactions with the decanoyl side chain of daptomycin.
Depletion of PhG greatly decreases daptomycin susceptibility. The effect of reducing the net negative charge of the cell membrane on daptomycin susceptibility was especially apparent when we studied a strain in which PhG could be depleted from cells by using a conditionally expressed allele of pgsA. PgsA is required for the first step in PhG synthesis from phosphatidic acid (2). Depletion of this essential complex lipid, by transfer of cells to a medium lacking the inducer IPTG (isopropyl-β-D-thiogalactopyranoside), results in cells that lose the characteristic helical staining pattern associated with anionic-lipid-favoring membrane dyes (e.g., FM 4-64). This strain continues to grow for several hours even in the absence of IPTG, as cells gradually become depleted of PhG (2). When PhG-depleted cells were subcultured in a medium lacking IPTG, but containing daptomycin, they were able to grow in the presence of daptomycin concentrations significantly higher than that of the wild type. Conversely, when expression of pgsA was induced by IPTG addition, wild-type levels of daptomycin susceptibility were restored (Table 4). This increased resistance was specific for daptomycin; PhG-depleted cells were unaffected in susceptibility to vancomycin (which targets PG synthesis) and had increased susceptibility to duramycin, which interacts specifically with phosphatidylethanolamine (31). Presumably, in this case, depletion of PhG from the membrane led to an increase in the concentration of phosphatidylethanolamine.
The effects of membrane composition and charge on daptomycin insertion have also been studied by Jung et al. in artificial liposomes (37). By means of fluorescence spectroscopy, differential scanning calorimetry, and 31P nuclear magnetic resonance, they found that daptomycin (with Ca2+) binds to acidic and neutral lipids in different fashions and leads to a change of the structural organization of acidic membranes (induction of nonlamellar lipid phases and membrane fusion) (37). This again emphasizes the influence of membrane lipid composition on daptomycin susceptibility.
Daptomycin preferentially localizes to the cell septum and in a helical pattern along the cell wall. Fluorescent imaging with daptomycin-BDP was used to study the localization of daptomycin in the cell envelope in B. subtilis. Strikingly, daptomycin-BDP was not distributed evenly throughout the cell membrane but, rather, in a complex, reproducible pattern. The highest concentration was found along the newly formed division septa and in a helical pattern along the long axis of the cell, whereas no fluorescence was detected at the cell poles (Fig. 2A and B). This helical pattern and localization to the septa are reminiscent of localization studies of both the cell wall biosynthetic machinery (58) and anionic phospholipids (including PhG) in B. subtilis (2). Sites of active cell wall biosynthesis have been visualized using vancomycin-BDP (58).
![]() View larger version (51K): [in a new window] |
FIG. 2. Daptomycin-BDP inserts preferentially at new division septa and in a spiral pattern. Fluorescent and DIC micrographs of B. subtilis stained with daptomycin-BDP (DAP-BDP) and vancomycin-BDP (VAN-BDP). (A) B. subtilis W168 treated with daptomycin-BDP at two times the MIC for 10 min (during exponential growth phase). (B) W168 treated with daptomycin-BDP at 10 times the MIC for 10 min. (C) W168 treated with equal amounts of vancomycin and vancomycin-BDP for 20 min. Panels A and B show a spiral localization of daptomycin-BDP and the preferential insertion at newly formed division septa, similar to that of vancomycin-BDP (C). The scale bar represents 2 µm.
|
![]() View larger version (27K): [in a new window] |
FIG. 3. Daptomycin-BDP (DAP-BDP) staining of stationary phase cells. Fluorescent and DIC micrographs of stationary growth phase culture of B. subtilis W168 treated with daptomycin-BDP at two times the MIC for 10 min. The staining shows a similar insertion pattern as that of exponential-growth phase cultures. The scale bar represents 2 µm.
|
![]() View larger version (22K): [in a new window] |
FIG. 4. Correlation between daptomycin-BDP staining and anionic phospholipid content and distribution. Fluorescent and DIC micrographs of wild-type W168 (A) and pgsA::pMutin grown in the presence (B) or absence (C) of 1 mM IPTG, stained with daptomycin-BDP at two times the MIC for 10 min. Daptomycin-BDP delocalizes when pgsA is not expressed from the IPTG-inducible promoter, suggesting preferential insertion of daptomycin in membrane lipid domains rich in anionic phosphatidylglycerol. W168 and pgsA::pMutin stained with the membrane lipid dye FM 4-64 in the absence of IPTG are shown in panels D and E, respectively. The scale bar represents 2 µm.
|
To determine whether the daptomycin stimulon of B. subtilis most closely resembled the stimulons for membrane-perturbing agents or cell wall synthesis inhibitors, we performed a hierarchical clustering analysis using data sets representing the transcriptional responses to daptomycin, vancomycin, moenomycin, and ramoplanin (our results) and 35 other antibiotics from a study by Hutter et al. (30). The daptomycin stimulon is most similar to a cluster of treatment conditions that includes compounds that inhibit PG synthesis (vancomycin, ristocetin, ramoplanin, and moenomycin) and that perturb membrane function (Triton X-114 and polymyxin B). All of these conditions induce the
M regulon, and several compounds are strong inducers of the LiaRS TCS (Fig. 5). These results lead us to suggest that insertion of daptomycin into membrane regions enriched in anionic lipids may have multiple effects, including both disruption of membrane function and perhaps interference with the assembly or function of cell wall biosynthetic complexes.
![]() View larger version (74K): [in a new window] |
FIG. 5. Cluster analysis of B. subtilis microarray studies with 40 different antimicrobial agents. The gene expression patterns of daptomycin-treated B. subtilis are most closely related to those of cells treated with the following cell membrane- and cell wall-active antibiotics: moenomycin, Triton X-114, polymyxin B, ramoplanin, ristocetin, and vancomycin. Cluster analysis was performed on whole genome data sets for each antibiotic (see Materials and Methods), and selected clusters enriched for daptomycin-induced genes are shown. Red indicates induction and green repression after treatment, whereas black corresponds to unchanged gene expression.
|
Analysis of strains with altered membrane composition suggests that daptomycin interacts preferentially with regions of the membrane enriched in anionic lipids, including PhG, the dominant anionic lipid in B. subtilis. For example, a strain in which PhG is depleted becomes less susceptible to daptomycin, although it retains normal (or even increased) susceptibility to other cell wall- and membrane-active antibiotics. Conversely, strains lacking MprF, which synthesizes the cationic lipid LPhG, are more susceptible to daptomycin. Mutations in mprF have been previously associated with daptomycin resistance (21), although in light of the results here, it seems likely that these are gain-of-function mutations that increase MprF levels or activity. Indeed, overexpression of mprF in B. subtilis decreases daptomycin susceptibility.
Cells depleted of PhG not only display decreased daptomycin susceptibility, they also lose the characteristic helical staining pattern seen in wild-type cells treated with daptomycin-BDP. This may result directly from the reduction in levels of negatively charged membrane lipids, which would thereby decrease the affinity of the positively charged daptomycin-Ca2+ complex for the membrane. Alternatively, or in addition, PhG depletion might result in altered composition or localization of membrane proteins or cell envelope biosynthetic complexes. For instance, Campo et al. reported that reduced PhG led to delocalization of the translocation ATPase SecA in the B. subtilis membrane (5), and Barák et al. observed enrichment of the cell division protein MinD in anionic phospholipid spirals in the membrane (2).
Daptomycin is used clinically as a reserve antibiotic against complicated skin and skin structure infections (23) as well as against S. aureus-induced bacteremia and infective endocarditis (20). To date, reports about resistance to daptomycin in clinical settings have been relatively rare (25, 55). To better understand the evolution of daptomycin resistance, S. aureus strains with increased daptomycin resistance (either selected in the laboratory or arising during clinical treatment) were chosen for DNA sequence analysis. These studies indicate that the evolution of resistance is a multigenic phenomenon. Often, mutations in mprF emerge early, and other contributing mutations occur in the essential yycFG TCS and rpoBC genes (21). To date, there are no documented examples of high-level daptomycin resistance emerging due to a single gene mutation. In light of these findings, it is interesting that cells depleted of PhG display such a dramatic decrease of susceptibility. However, null mutations that confer daptomycin resistance are unlikely to arise in pgsA since, at least in B. subtilis, it is an essential gene.
This work was supported by grant GM-047446 from NIH.
Published ahead of print on 21 January 2009. ![]()
Supplemental material for this article may be found at http://aac.asm.org/. ![]()
|
|
|---|
W, an extracytoplasmic-function
factor in Bacillus subtilis. J. Bacteriol. 183:2380-2383.
X factor regulates modification of the cell envelope and resistance to cationic antimicrobial peptides. J. Bacteriol. 186:1136-1146.
factors. J. Bacteriol. 184:6123-6129.
W regulon: a comparative analysis of promoter consensus search, run-off transcription/macroarray analysis (ROMA), and transcriptional profiling approaches. J. Mol. Biol. 316:443-457.[CrossRef][Medline]
M, an extracytoplasmic function
factor, and is conferred by YqjL and BcrC. J. Bacteriol. 187:2948-2956.
W and
M regulons. Mol. Microbiol. 45:1267-1276.[CrossRef][Medline]
M regulon and its contribution to cell envelope stress responses. Mol. Microbiol. 67:830-848.[Medline]
M, an ECF RNA polymerase sigma factor of Bacillus subtilis 168, is essential for growth and survival in high concentrations of salt. Mol. Microbiol. 32:41-50.[CrossRef][Medline]
X protein is an extracytoplasmic function
factor contributing to survival at high temperature. J. Bacteriol. 179:2915-2921.
W: autoregulation and partial overlap with the
X regulon. J. Bacteriol. 180:3765-3770.
W. Mol. Microbiol. 31:361-371.[CrossRef][Medline]
M-responsive genes of Bacillus subtilis and their promoters. J. Bacteriol. 189:4534-4538.
factor and two accessory proteins regulate oxalate decarboxylase expression in Bacillus subtilis. Mol. Microbiol. 69:954-967.[CrossRef][Medline]
factors. J. Bacteriol. 189:6919-6927.This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»