Further Insights into the Mode of Action of the Lipoglycopeptide ...

Further Insights into the Mode of Action of the Lipoglycopeptide Telavancin through Global Gene Expression Studies Yang Song,a Christopher S. Lunde,b Bret M. Benton,b and Brian J. Wilkinsona Microbiology Group, School of Biological Sciences, Illinois State University, Normal, Illinois, USA,a and Theravance, Inc., South San Francisco, California, USAb

Telavancin is a novel semisynthetic lipoglycopeptide derivative of vancomycin with a decylaminoethyl side chain that is active against Gram-positive bacteria, including Staphylococcus aureus strains resistant to methicillin or vancomycin. A dual mechanism of action has been proposed for telavancin involving inhibition of peptidoglycan biosynthesis and membrane depolarization. Here we report the results of genome-wide transcriptional profiling of the response of S. aureus to telavancin using microarrays. Short (15-min) challenge of S. aureus with telavancin revealed strong expression of the cell wall stress stimulon, a characteristic response to inhibition of cell wall biosynthesis. In the transcriptome obtained after 60-min telavancin challenge, in addition to induction of the cell wall stress stimulon, there was induction of various genes, including lrgA and lrgB, lysine biosynthesis operon (dap) genes, vraD and vraE, and hlgC, that have been reported to be induced by known membrane-depolarizing and active agents, including carbonyl cyanide m-chlorophenylhydrazone, daptomycin, bacitracin, and other antimicrobial peptides These genes were either not induced or only weakly induced by the parent molecule vancomycin. We suggest that expression of these genes is a response of the cell to mitigate and detoxify such molecules and is diagnostic of a membrane-depolarizing or membrane-active molecule. The results indicate that telavancin causes early and significant induction of the cell wall stress stimulon due to strong inhibition of peptidoglycan biosynthesis, with evidence in support of membrane depolarization and membrane activity that is expressed after a longer duration of drug treatment.

S

taphylococcus aureus is a leading bacterial pathogen worldwide. When penicillin was first introduced in the 1940s, it effectively treated staphylococcal infections, although resistant strains producing penicillinase that rendered penicillin ineffective soon arose. Chambers and DeLeo (7) have described how several “epidemic waves” of antibiotic-resistant S. aureus have occurred since then. Methicillin-resistant S. aureus (MRSA) is endemic in most hospitals worldwide and is a leading cause of death in health care facilities (21). The rapid rise of serious community-associated MRSA infections over the past decade is alarming. In addition, strains showing reduced susceptibility to vancomycin (vancomycin-intermediate S. aureus [VISA]) have been encountered with increasing frequency worldwide (17), thereby compromising the therapeutic efficiency of this drug, which was the one to which MRSA remained uniformly susceptible. In addition, a few true vancomycin-resistant S. aureus (VRSA) isolates have been reported (17), which have a much higher vancomycin MIC than VISA. Against this background, the need for novel antistaphylococcal agents is obvious. Telavancin (Vibativ) is a novel semisynthetic lipoglycopeptide derivative of vancomycin that has recently been approved by the Food and Drug Administration for the treatment of complicated skin and skin structure infections. Telavancin exhibits more potent in vitro and in vivo activity than vancomycin against methicillin-susceptible S. aureus (MSSA), MRSA, VISA, and VRSA (10, 14, 23, 26, 30). The structure of telavancin includes a hydrophobic decylaminoethyl side chain attached to the vancosamine sugar of the molecule, which is therefore classified as a lipoglycopeptide (25). The mode of action of telavancin includes inhibition of peptidoglycan biosynthesis by hydrogen-bonding to D-Ala-D-Ala-containing lipid intermediates of peptidoglycan biosynthesis (15). In addition, Higgins et al. (15) provided evidence that telavancin disrupts membrane barrier function, causing increased cell permeability

June 2012 Volume 56 Number 6

resulting in depolarization. Lunde et al. (28) showed evidence that telavancin targets the lipid II intermediate of peptidoglycan biosynthesis. This targeting of lipid II results in potent inhibition of peptidoglycan biosynthesis and provides a model for disruption of membrane barrier function that includes positive membrane curvature. Recently, it was shown by using fluorescent conjugates that telavancin displays enhanced binding at the division septum, the site of cell wall biosynthesis, compared to vancomycin (29). A variety of antibiotics target lipid II (5, 49), and, for example, nisin is proposed to form pores in the membrane via nisin-lipid II complexes (6). In the past decade, global genome-wide studies of changes in expression patterns in response to existing and new antimicrobial agents have provided us with a deeper understanding of antimicrobial action (4, 18). Both transcriptomic and proteomic studies have been reported over the past 10 years, and the findings were drawn together in reviews by Wecke and Mascher (50) and Wenzel and Bandow (51) in 2011. Cell wall-active antibiotics typically induce what is termed the cell wall stress stimulon or cell envelope stress response in Gram-positive bacteria (19, 24, 43, 48). A key two-component system controlling part of the cell wall stress stimulon is VraSR in S. aureus, or its homolog LiaSR in Bacillus subtilis. Transcriptional profiling of the action of daptomycin, an antistaphylococcal lipopeptide, on S. aureus revealed gene expres-

Received 27 July 2011 Returned for modification 9 October 2011 Accepted 2 March 2012 Published ahead of print 12 March 2012 Address correspondence to Brian J. Wilkinson, [email protected]. Supplemental material for this article may be found at http://aac.asm.org/. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/AAC.05403-11

Antimicrobial Agents and Chemotherapy

p. 3157–3164

aac.asm.org

3157

Song et al.

sion patterns typical of inhibition of peptidoglycan biosynthesis and induction of membrane depolarization (35). In this paper we report the changes in gene expression that took place upon challenge of S. aureus with telavancin in comparison to other agents. The situation is such that it is now expected that this information be available for all antimicrobial agents. This is important in the case of telavancin in that the additional semisynthetic lipoglycopeptides dalbavancin and oritavancin are under development and the telavancin transcriptomic information is the first for this class of agents, which are all expected to have a dual mode of action (33). The results are consistent with a dual mode of action of telavancin involving inhibition of peptidoglycan biosynthesis and membrane depolarization and/or membrane activity. MATERIALS AND METHODS Strain, medium, and antimicrobial agents. MSSA strain ATCC 29213 was used in this study. This strain is a quality control strain used in antimicrobial susceptibility determinations (8) and has been used in modeof-action studies of telavancin (15, 28, 29) and other antimicrobials (35). The organism was grown in Mueller-Hinton broth supplemented with 25 mg of Ca2⫹ liter⫺1 and 12.5 mg of Mg2⫹ liter⫺1 (CA-MHB). Telavancin was supplied by Theravance, Inc. Vancomycin, carbonyl cyanide m-chlorophenylhydrazone (CCCP), and enduracidin were purchased from Sigma Chemical Company (St. Louis, MO). Growth inhibition and transcriptional profiling conditions. Overnight-grown cultures were used to inoculate 25 ml CA-MHB (1% vol/vol) in a 50-ml Erlenmeyer flask and were grown at 37°C with shaking at 210 rpm. Antimicrobial agents were added to the culture in log phase (optical density at 600 nm [OD600], ⬃0.35). Growth was measured at regular intervals using a Beckman DU65 spectrophotometer. Various concentrations of telavancin (1, 2, 4, 6, and 10 ␮g ml⫺1), vancomycin (2, 4, 6, 8, and 10 ␮g ml⫺1), CCCP (2 and 4 ␮g ml⫺1), and enduracidin (0.6, 0.8, 1, 1.5, 2.5, and 5 ␮g ml⫺1) were used to study growth-inhibitory effects (35). Based on the study of a range of concentrations of antimicrobial agents, concentrations of agents were chosen to give similar degrees of inhibition for transcriptional profiling studies and compatibility with previous studies (28, 35). The following concentrations of antimicrobial were added to mid-exponential-phase cultures for 15 and 60 min of challenge: telavancin, 8 ␮g ml⫺1; vancomycin, 10 ␮g ml⫺1; CCCP, 2 ␮g ml⫺1; and enduracidin, 1 ␮g ml⫺1. Telavancin and CCCP were dissolved in dimethyl sulfoxide and enduracidin in 50% methanol. Control cultures contained appropriate antimicrobial agent solvent controls and were also incubated for 15 min and 60 min. RNA extraction and purification. RNA extraction and purification were carried out as described by Muthaiyan et al. (35). Microarray hybridization and processing. The microarray hybridization and processing procedure was described by Muthaiyan et al. (35). Purified labeled cDNA was hybridized with S. aureus genome microarrays, version 6.0, provided by the Pathogen Functional Genomics Resource Center of the National Institutes of Allergy and Infectious Diseases (NIAID). The full genome array consists of 70-mer oligonucleotides representing open reading frames (ORFs) from S. aureus strains COL, MW2, MSSA476, Mu50, MRSA252, and N315. Each ORF is printed in quadruplicate on the array. Microarray data analysis. Hybridization signals were scanned using an Axon 4000B scanner (Molecular Devices, Sunnyvale, CA) with Acuity 4.0 software, and scans were saved as TIFF images. Scans were analyzed using TIGR-Spotfinder (www.tigr.org/software/) software, and the local background was subsequently subtracted. The data set was normalized by applying the LOWESS algorithm using TIGR-MIDAS (www.tigr.org /software/) software. The normalized log2 ratio of test to reference signal for each spot was recorded. Genes with fewer than three data points were considered unreliable, and their data points were discarded. The averaged

3158

aac.asm.org

TABLE 1 Primers used for RT-PCR Gene

Primer (5=–3=)

16S rRNA

Forward: CCACGCCGTAAACGATGAGT Reverse: CACATGCTCCACCGCTTGT

vraS

Forward: TTGCGGCAAGTATGATGCTT Reverse: TGCGAATCTTGAACCATTTTCTC

vraR

Forward: AAGCGACGACTCAGATTAAAAAAGA Reverse: ACTATCGACACCTGCATCTAATGC

lytM

Forward: AGGTCCAGACGCGAGCTATTAT Reverse: TCTTTCGCATGACCACTAGCTG

lrgB

Forward: TTGGCATCGTATCATCGGAG Reverse: GATACTGGTAACGCAATCGCTG

lrgA

Forward: AGACGCATCAAAACCAGCAC Reverse: GAGATACCGGCTGGTACGAA

log2 ratio for each remaining gene on the six replicate slides was ultimately calculated. Significant changes in gene expression were identified with SAM (significance analysis of microarrays; http://www-stat.stanford.edu /⬃tibs/SAM/) software using one class mode. SAM assigns a score to each gene on the basis of change in gene expression relative to the standard deviation of repeated measurements. For genes with scores greater than an adjustable threshold, SAM uses permutations of the repeated measurements to estimate the percentage of genes identified by chance, i.e., the false discovery rate. A cutoff of 2-fold for over- and underexpressed ORFs was used. To examine how genes with transcript level changes are distributed with regard to their function, we further classified these genes using our in-house software Gene Sorter according to the categories described in the comprehensive microbial resource of the J. Craig Venter Institute (http://cmr.tigr.org/tigr-scripts/CMR/shared/Genomes.cgi). Microarray validation by real-time reverse transcription-PCR (RTPCR). To confirm the validity of microarray data, gene-specific mRNAs were quantified by RT-PCR. Total RNA was prepared as described above for transcriptional profiling. Residual DNA was removed from the samples by performing an on-column DNase digestion step with RNase-free DNase (Qiagen). cDNA was synthesized using a high capacity RNA-tocDNA kit (Applied Biosystems, Foster City, CA). Specific primers for the genes tested and for 16S rRNA were designed using Primer Express software (Applied Biosystems) in order to design 100- to 200-bp amplicons (Table 1 shows the primer sequences). RT-PCR was performed with the ABI Prism 7300 sequence detection system (Applied Biosystems) and SYBR green technology. RT-PCRs were performed in a 25-␮l reaction volume containing 1 ␮l of template DNA (25 ng ml⫺1), 1 ␮l of genespecific primers (10 ␮M), 12.5 ␮l of Power SYBR green PCR Master Mix (Applied Biosystems, Foster, CA), and 9.5 ␮l of H2O. The expression levels of the tested genes were normalized using the 16S rRNA gene of S. aureus as an internal standard. Each assay was performed in triplicate. The data analysis was carried out using 7300 system software (Applied Biosystems). Microarray data accession number. The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE37272.

RESULTS AND DISCUSSION

Growth response of S. aureus to telavancin and other antimicrobial agents chosen for comparison. In order to choose the conditions for determining telavancin transcriptomes, mid-exponential-phase cultures of strain ATCC 29213 were challenged with


Telavancin Transcriptome

FIG 1 Response of mid-exponential-phase S. aureus cultures to the concentrations of antimicrobial used in transcriptional profiling. (a) Telavancin; (b) vancomycin; (c) CCCP; (d) enduracidin.

various concentrations of telavancin, and the growth response of the culture was monitored. Based on these results, a concentration of 8 ␮g ml⫺1 was chosen for determining the telavancin transcriptomes. This concentration of telavancin has been shown to cause significant membrane depolarization (28). In order to provide perspective on the anticipated telavancin transcriptomes, we chose certain pertinent antimicrobial agents for determination of their transcriptomes. Vancomycin was chosen because it is the parent molecule of telavancin, but it lacks the lipophilic side chain present in telavancin. Enduracidin is a lipoglycodepsipeptide related to ramoplanin. Enduracidin preferentially inhibits the transglycosylation step of peptidoglycan biosynthesis through binding to lipid II (12). CCCP is a proton ionophore that collapses the pH and electrical gradients. The growth inhibition responses of the cultures to the chosen concentration of antimicrobial agents are shown in Fig. 1. Concentrations giving a similar degree of growth inhibition were employed. In addition, the transcriptomes of a


significant number of potentially relevant cell wall- or membraneactive antimicrobials have been published. The 15-min telavancin transcriptome shows strong induction of the cell wall stress stimulon. The telavancin and enduracidin transcriptomes were first determined after 15 min of challenge with 8 ␮g telavancin ml⫺1 and 1 ␮g enduracidin ml⫺1. Changes in expression of selected genes are shown in Table 2, and the full results are shown in Table S1 in the supplemental material. The cell wall stress stimulon (53) was clearly strongly induced by this short-term telavancin treatment. Challenge of S. aureus with cell wall-active antibiotics, including vancomycin (24, 32), oxacillin, bacitracin, and D-cycloserine (48), modulation of the transcription of cell wall synthesis gene murF (47), modulation of transcription of mevalonate biosynthesis pathway genes (3), and comparison of the transcriptomes of VISA strains versus vancomycin-susceptible S. aureus (31) have led to the identification of core genes of the cell wall stress stimulon. The cell wall stress

aac.asm.org 3159

Song et al.

TABLE 2 Overexpression of cell wall stress stimulon genes by telavancin and enduracidin Change (fold) in expression after 15 mina Locus ID

Gene

Protein

Subcategory

Tlv

End

Cell envelope SAV0212

No data

Hypothetical protein

8.6

0.0

SAV0276

lytM

Peptidoglycan hydrolase

5.3

2.2

SACOL1066

No data

Fmt

5.2

4.9

SAV1092 SAV2124

No data murZ

4.0 3.5

3.4 5.3

SAV1450

pbp2

Conserved hypothetical protein UDP-N-acetylglucosamine 1-carboxylvinyl transferase 2 PBP2

3.2

2.3

SACOL2354 SAV0470 SACOL1194

No data No data pbp1

Membrane protein, putative Conserved hypothetical protein Penicillin-binding protein 1

2.3 2.3 2.1

2.4 3.4 1.4

SAS1987

No data

D-Alanine-D-alanine

2.1

2.4

SAV1151

murI

Glutamate racemase

2.0

1.9

SAV0747

llm (tarO)

2.2

2.4

SACOL2385 SAS1654

No data No data

Lipophilic protein affecting bacterial lysis rate and methicillin resistance Heat shock protein, Hsp20 family Putative protease

Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis and degradation of murein sacculus and peptidoglycan Other Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis and degradation of murein sacculus and peptidoglycan Other Other Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis and degradation of murein sacculus and peptidoglycan Biosynthesis of teichoic acid

3.1 6.9

3.1 7.6

SAV2030 SACOL1637 SAV1153

groES dnaK No data

GroES protein DnaK protein Conserved hypothetical protein

3.3 2.1 2.9

2.3 2.6 2.7

SACOL1636 SAV0975

dnaJ clpB

DnaJ protein ClpB chaperone homologue

Protein folding and stabilization Degradation of proteins, peptides, and glycopeptides Protein folding and stabilization Protein folding and stabilization Degradation of proteins, peptides, and glycopeptides Protein folding and stabilization Role category not yet assigned

2.5 2.5

1.9 2.2

vraR vraS vraE vraD No data tcaA drp35

DNA-binding response regulator VraR Histidine kinase sensor VraE protein ABC transporter ATP-binding protein Conserved hypothetical protein TcaA protein Drp35

Protein interactions Protein interactions Similar to ABC transporter, permease Similar to ABC transporter Conserved Role category not yet assigned Role category not yet assigned

21.1 13.8 37.4 7.2 30.6 7.8 15.3

9.3 11.2 40.5 32.9 13.0 7.1 13.4

Selected other genes SACOL1942 SAR1975 SAV2702 SAV2701 SACOL0625 SAV2356 SAV2688 a

ligase

Tlv, telavancin; End, enduracidin.

stimulon is characterized by induction of genes involved in cell wall synthesis and of genes encoding chaperones and proteases (52). There was a strong increase in expression of the two-component regulator that is the principal controller of the cell wall stress regulon, vraS (13.8-fold) and vraR (21.1-fold). Cell envelope-related genes that were upregulated include lytM (5.3-fold, glycylglycine endopeptidase), murZ (3.5-fold; UDP-N-acetylglucosamine-1-carboxyvinyl transferase), pbp2 (3.2-fold; penicillin-binding protein 2), SAS1987 (2.1-fold; D-alanylalanine synthetase), murI (2-fold; glutamate racemase), fmt (6.3-fold; autolysis and methicillin resistance-related protein), bacA (3-fold; undecaprenyl pyrophosphate phosphatase), and llm/tarO (2.2-fold; first committed enzyme in teichoic acid biosynthesis). Various protease and chaperone genes were upregulated, including grpE (3.9-fold), groES (3.3-fold), dnaJ (2.5-fold), groEL (2.4-

3160

aac.asm.org

fold), clpP (2.1-fold), dnaK (2.1-fold), and prsA (10.7-fold; peptidylprolyl cis-/trans-isomerase). In addition, two genes whose function is not completely understood but which are typically upregulated by cell wall-active antibiotics were strongly upregulated, and these were drp35 (15.3-fold), and tcaA (7.8-fold). The most highly upregulated gene was SACOL0625 (30.6-fold), which was recently designated cwrA (2). This gene is highly upregulated by a wide variety of cell wall-active antibiotics (2). The overexpression of these genes has been reported in many studies of the cell wall stress stimulon (2, 24, 31, 32, 47, 48). vraD and vraE were overexpressed 7.2- and 37.4-fold, respectively, by telavancin and 32.9- and 40.5-fold, respectively, by enduracidin. These genes encode part of the multidrug resistance transporter VraDE, which is induced strongly by various cationic antimicrobial peptides and more weakly by vancomycin (40, 44).



TABLE 3 Induction of genes indicative of membrane depolarization and/or membrane activity Change (fold) in expressiona Locus ID

Gene

Protein

Subcategory

Tlv

End

Van

CCCP

Dap

Amino acid biosynthesis SAV1328 SAV1395 SAV1396 SAV1397 SAV1394 SACOL1364 SAV1329

dhoM dapA dapB dapD asd thrB thrC

Homoserine dehydrogenase Dihydrodipicolinate synthase Dihydrodipicolinate reductase Tetrahydrodipicolinate acetyltransferase Aspartate Semialdehyde Dehydrogenase Homoserine kinase Threonine synthase

Aspartate family Aspartate family Aspartate family Aspartate family Aspartate family Aspartate family Aspartate family

12.0 11.6 8.7 8.6 7.3 6.2 5.2

4.1 2.5 NS NS 2.3 NS 2.8

NS NS 2.3 NS 2.8 NS NS

11.1 11.5 13.8 6.1 11.8 9.7 12.4

5.7 3.8 2.7 NS 3.2 2.5 3.1

Cell envelope SACOL0247

lrgA

Holin-like protein LrgA

5.2

NS

NS

14.3

2.7

SAR0260 SAV2420

lrgB hlgC

Holin-like protein Gamma-hemolysin component C

Biosynthesis and degradation of murein sacculus and peptidoglycan Other Toxin production and resistance

3.0 7.0

1.3 5.5

NS 1.7

8.1 32.4

3.6 3.5

vraE

VraE protein

82.0

32.4

16.1

NS

NS

vraD

ABC transporter ATP-binding protein

Similar to ABC transporter, permease Similar to ABC transporter

84.0

42.6

NS

NS

NS

Selected other genes SAV2702 SAV2701 a

Change in expression after 60 min (Tlv, End, and Van) or 15 min (CCCP and Dap). The data for Dap are from reference 32. The transcriptional profile for CCCP was generated using S. aureus genome microarrays version 8.0, provided by the Pathogen Functional Genomics Resource Center of the NIH. Tlv, telavancin; End, enduracidin; Van, vancomycin; Dap, daptomycin; CCCP, carbonyl cyanide m-chlorophenylhydrazone; NS, not significant.

VraDE is believed to be involved in antimicrobial peptide detoxification, including bacitracin and nisin resistance (16). A considerable number of genes were decreased in expression after 15 min of telavancin challenge. These included various genes of the cap (capsular polysaccharide) operon, various genes involved in fermentative and anaerobic metabolism, including pflB (⫺5.8-fold; formate acetyltransferase), ldh1 (⫺5.1-fold; L-lactate dehydrogenase), narG (⫺4.6-fold; respiratory nitrate reductase alpha chain), narK (⫺5.4-fold; nitrate extrusion protein), various pur genes of the purine biosynthesis operon, and pyr genes of the pyridine biosynthesis operon. Similar findings have been noted in previous studies of the cell wall stress response (24, 32, 39, 47, 48). It is well established that vancomycin induces the cell wall stress stimulon (24, 32), and many genes in the categories Cell Envelope and Protein Fate whose expression was increased by telavancin were also upregulated by vancomycin, although generally to a lesser extent than by telavancin (see Table S2 in the supplemental material). This is consistent with the finding that telavancin shows enhanced targeting of septal peptidoglycan synthesis compared to vancomycin (29). We are not aware of other reports of the S. aureus enduracidin transcriptome, although the response of B. subtilis to this agent has been reported (42). There was considerable similarity in the telavancin and enduracidin transcriptomes, with a significant portion of the telavancin transcriptome genes also being upregulated by enduracidin. Genes in the categories Cell Envelope and Protein Fate were strongly upregulated by enduracidin. This similarity of the telavancin and enduracidin transcriptomes is consistent with the proposed mode of action of this lipoglycodepsipeptide in preferentially inhibiting the transglycosylation step of peptidoglycan biosynthesis through binding to lipid II (12). Like telavancin, enduracidin has a lipophilic substituent as part of the molecule that likely promotes interaction with the membrane in a way that van-


comycin, which lacks such a side chain, does not. In summary, the initial transcriptional response of S. aureus to telavancin is strongly reflective of inhibition of peptidoglycan biosynthesis. The 60-min telavancin transcriptome. In the studies of Higgins et al. (15) and Lunde et al. (28), membrane depolarization effects were seen with higher concentrations of telavancin for longer durations of treatment. Dengler et al. (9) studied the induction kinetics of various cell wall-active antibiotics and showed that the kinetics and maximal induction observed were dependent on the individual antibiotic and its concentration. Hence, the telavancin transcriptome was also determined after 60 min of challenge (Table 3; also, see Table S2 in the supplemental material). The results for the 15-min-CCCP transcriptome are shown for purposes of comparison. After 60 min of CCCP challenge, an excessively large number of genes were changed in expression, making the results more difficult to interpret. More genes (up- or downregulated 2-fold or more) were changed in their expression after 60 min than 15 min of telavancin challenge. A total of 752 genes were altered in expression after 60 min compared to 352 after 15 min. A total of 242 genes were common to both transcriptomes. Prominent categories of genes increased in expression included those corresponding to cell envelope (29 genes), energy metabolism (52 genes), protein fate (57 genes), regulatory function (28 genes), transport and binding proteins (57 genes), and hypothetical proteins (42 genes). As expected, the 60-min-telavancin transcriptome showed significant upregulation of the cell wall stress stimulon; vraS and vraR were overexpressed 11.1- and 7.3-fold, respectively (see Table S2 in the supplemental material). Upregulated genes in the Cell Envelope category included lytM (6.8-fold), murI (4.9-fold), murF (4.2-fold), pbp2 (2.8-fold), and murG (2.2-fold; N-acetylglucosaminyltransferase). A cell wall stress stimulon member gene, cwrA (SACOL0625), was overexpressed 58.7-fold. However, a

aac.asm.org 3161

Song et al.

number of the cell wall stress stimulon genes showed a lower magnitude of upregulation after 60 min than 15 min. Nevertheless, overall the transcriptome reflected strong upregulation of the cell wall stress stimulon. Genes in the category Protein Fate included grpE (13.2-fold), SAV1423, the methionine sulfoxide reductase B gene (10.2-fold), clpC (9.2-fold), SACOL2385 (Hsp20 family; 8.9-fold), SAV1424 (methionine sulfoxide reductase A; 8.0-fold), groES (5.9-fold), dnaK (4.4-fold), dnaJ (3.9-fold), clpP (3.1-fold), groEL (3.0-fold), and others. The upregulation of a significant number of genes in this category, some to quite a high degree of expression, may reflect considerable accumulation of damaged proteins as the time of exposure to the inhibitory agent is lengthened. SAV0524, which encodes a putative ATP:guanidophosphotransferase, was upregulated 19.1-fold. This is homologous to McsB in B. subtilis, which is part of a system targeting CtsR, the repressor of the clpC operon, for degradation by phosphorylation, which is part of the protein quality control system in Gram-positive bacteria (11). McsA is a redox-sensing protein that interacts with McsB and prevents it from binding and inactivating CtsR, the global heat shock repressor. However, when critical thiols of MscA are oxidized, MscB is freed, resulting in CtsR inactivation and induction of genes involved in protein quality control (11). Oxidative stress is a characteristic of the action of bactericidal antibiotics (22, 48). Among downregulated genes were various cap genes, encoding the capsular polysaccharide (see Table S2 in the supplemental material). Strongly downregulated genes were SAV1163 (alphahemolysin precursor; ⫺17.2-fold); sarR (staphylococcal accessory regulator R; ⫺10.2-fold); rot (repressor of toxins; ⫺10.1-fold); agrA (accessory gene regulator protein A; ⫺5.2-fold); genes involved in osmotolerance (betA, choline dehydrogenase, ⫺9.2fold, and gbsA, glycine betaine aldehyde dehydrogenase, ⫺2.7fold); and narK and narG (nitrite extrusion protein and respiratory nitrate reductase alpha subunit; ⫺20.7-fold and ⫺17.8-fold, respectively). Also, other genes involved in anaerobic metabolism were downregulated. There was a strong downregulation of a variety of genes involved in protein synthesis. Purine metabolism genes were also downregulated. Induction of genes by telavancin indicative of membrane depolarization. Higgins et al. (15) showed, using the membrane potential-sensitive fluorescent dye DiSC3, that telavancin dissipates the membrane potential in a time- and concentrationdependent manner. Lunde et al. (28) provide evidence that peptidoglycan lipid precursor II plays an essential role in telavancininduced membrane depolarization. Two key genes that were upregulated by telavancin were lrgA (5.2-fold; peptidoglycan hydrolase regulator) and lrgB (3.0-fold; antiholin protein) (Table 3). These genes were also upregulated by 15 min of CCCP treatment (Table 3). Patton et al. (37) have shown that CCCP increases the transcription of lrgA and lrgB and have attributed this mostly to a reduction in ⌬␾ (membrane potential) caused by CCCP. Muthaiyan et al. (35) have also shown that the known membrane-depolarizing agent daptomycin as well as CCCP induces lrgA and lrgB. These genes were not induced by vancomycin or enduracidin. Chitosan, which causes membrane depolarization, also induced lrgA and lrgB (41). Several genes of the lysine biosynthesis operon (dhoM, dapA, dapB, dapD, and asd) were strongly upregulated by telavancin (7.3-fold or more) (Table 3) and by CCCP and daptomycin (35),

3162

aac.asm.org

bacitracin (48), and the cationic antimicrobial peptide indolicidin (27) and more modestly by enduracidin and to a limited extent by vancomycin (Table 3). Increasing production of the positively charged amino acid lysine may be a response by the cell to mitigate the membrane depolarization caused by telavancin, CCCP, and other antimicrobial peptides, such as daptomycin. Membrane depolarization results in dissipation of the pH gradient, and increased intracellular lysine concentrations of this basic amino acid may be a response by the cell to attempt to restore the pH gradient. Also, K⫹ leakage has been shown to be caused by telavancin (15) and daptomycin (46). K⫹ is the major cell cation, and increased production of lysine with its net positive charge may be a response to K⫹ depletion. Lysine is also a component of the positively charged phospholipid lysyl phosphatidylglycerol, and increases in its levels have been associated with increased surface positive charge and decreased susceptibility to multiple antimicrobial peptides (38) and with changes in membrane physical structure (1). hlgC, encoding gamma hemolysin component C, was induced by telavancin, enduracidin, daptomycin, and CCCP. HlgC forms bicomponent pores with HlgA or HlgB (45). Strong induction of vraD and vraE by telavancin. vraD and vraE were induced strongly by telavancin (84- and 82-fold, respectively) and enduracidin (42.6- and 32.4-fold), and more weakly (not significantly and 16.1-fold) by vancomycin (Table 3). VraDE comprise an ABC transporter involved in antimicrobial peptide detoxification (40, 43). The system is induced by various antimicrobial peptides, including bacitracin, nisin, mersacidin, indolicidin, ovisporin-1–NH2, temporin-1–NH2, dermaseptin, K4-S4 (1– 16)-NH2, vancomycin, and teicoplanin (13, 16, 27, 40, 43, 53). To this list we can add telavancin and enduracidin as strong inducers. Sass et al. (43, 44) have pointed out that strong inducers of vraDE disrupt bilayers by the carpet mechanism. Interestingly, vraD and vraE were not induced by CCCP, which is probably a pure uncoupling agent interacting with the membrane in a different way than the compounds above. Studies of the antimicrobial peptide susceptibilities of ⌬vraDE mutants indicate that VraDE is involved in bacitracin and nisin detoxification and resistance (16, 40). Recent work revealed an unexpected and complex mechanism of the control of vraDE expression (13, 16). Hiron et al. (16) showed that a novel two-component system associated with bacitracin resistance, BraSR, activates transcription of braDE and vraDE, encoding two ABC transporters. However, unexpectedly, the ABC transporter BraDE is proposed to be involved in bacitracin sensing and signaling through BraSR, rather than BraS doing the sensing. It is not yet known whether the other antimicrobial peptides that induce vraDE are sensed by BraDE. Vancomycin is comprised of a peptide with a backbone modified by the presence of carbohydrate residues that is a weak inducer of vraDE (16) (Table 3). It is noteworthy that telavancin, a semisynthetic derivative of vancomycin possessing a hydrophobic (decylaminoethyl) side chain appended to the vancosamine sugar, induces vraDE much more strongly than vancomycin. While we do not know how telavancin is sensed, it seems likely that its interaction with the bacterial membrane and/or lipid II, properties that are shared with many antimicrobial peptides, enable S. aureus to distinguish it from vancomycin. braE (SAV2623) was modestly induced by telavancin (2.3-fold), suggesting that BraDE may be involved in telavancin sensing (13, 16). Telavancin did not induce vraFG, mprF, or genes involved in the D-alanylation of teichoic acids, which are con-



TABLE 4 Confirmation by RT-PCR of gene expression changes in telavancin-treated S. aureus Change (fold) in expression after: 15 min

60 min

Gene

Microarray

RT-PCR

Microarray

RT-PCR

vraR vraS lrgA lrgB lytM

21.1 13.8 0.0 ⫺1.3 5.3

26.6 21.3 1.7 9.2 2.0

7.3 11.0 5.2 3.0 6.8

3.3 4.4 26.5 18.1 1.9

trolled by the two-component system GraSR, which controls resistance to cationic antimicrobial peptides (13, 16), suggesting little role for this system in telavancin sensing. Interestingly, LrgA and LrgB and HlgC are all potentially involved in pore formation and may, like VraDE, be involved in detoxification. This also seems to be a characteristic response to membrane depolarization and membrane activity of an antimicrobial agent. RT-PCR validation of gene expression changes. The results of measuring changes in expression of selected genes by RT-PCR compared to the microarray results are shown in Table 4. RT-PCR confirmed increased expression of vraR, vraS, lrgA, lrgB, and lytM due to telavancin as also indicated by microarray analysis. The values (fold) for upregulation of lrgA and lrgB as measured by RT-PCR were significantly higher than the values obtained by microarray analysis. Two additional semisynthetic lipoglycopeptides showing promise for clinical use are dalbavancin and oritavancin. Oritavancin carries a p-chloro-phenylbenzyl side chain that interacts with the cytoplasmic membrane disrupting membrane potential and increasing membrane permeability as well as inhibiting peptidoglycan biosynthesis (20, 33, 36). We are unaware of full reports of the transcriptomes of these agents, although Moisan et al. (34) reported in an abstract that oritavancin induced the cell wall stress stimulon. In conclusion, telavancin shows rapid and strong induction of the cell wall stress stimulon and later induction of genes that appear to be due to the membrane effects of telavancin. We believe that induction of lrgA, lrgB, lysine operon genes (dap), and vraD and vraE, among others, is diagnostic of a membrane-depolarizing effect and/or membrane interaction of certain compounds, as induction of vraS and vraR and other genes of the cell wall stress stimulon is diagnostic of the inhibition of cell wall synthesis. ACKNOWLEDGMENTS This work was funded by a grant from Astellas. The DNA microarrays were obtained through NIAID’s Pathogen Functional Genomics Resource Center, Division of Microbiology and Infectious Diseases, NIAID, NIH. We are grateful to Thorsten Mascher and Hans Georg-Sahl for discussions of gene expression in response to antimicrobial peptides.

REFERENCES 1. Andra J, Goldmann T, Ernst CM, Peschel A, Gutsmann T. 2011. Multiple peptide resistance factor (MprF)-mediated resistance of Staphylococcus aureus against antimicrobial peptides coincides with a modulated peptide interaction with artificial membranes comprising lysylphosphatidylglycerol. J. Biol. Chem. 286:18692–18700.


2. Balibar CJ, et al. 2010. cwrA, a gene that specifically responds to cell wall damage in Staphylococcus aureus. Microbiology 156:1372–1383. 3. Balibar CJ, Shen X, Tao J. 2009. The mevalonate pathway of Staphylococcus aureus. J. Bacteriol. 191:851– 861. 4. Brazas MD, Hancock REW. 2005. Using microarray gene signatures to elucidate mechanisms of antibiotic action and resistance. Drug Discov. Today 10:1245–1252. 5. Breukink E, de Kruijff B. 2006. Lipid II as a target for antibiotics. Nat. Rev. Drug Disc. 5:321–323. 6. Breukink E, et al. 2003. Lipid II is an intrinsic component of the pore induced by nisin in bacterial membranes. J. Biol. Chem. 278:19898 – 19903. 7. Chambers HF, Deleo FR. 2009. Waves of resistance: Staphylococcus aureus in the antibiotic era. Nat. Rev. Microbiol. 7:629 – 641. 8. Clinical and Laboratory Standards Institute. 2006. Methods for dilution antimicrobial susceptibility testing for bacteria that grow aerobically: approved standard, 7th ed. M7–A7. Clinical and Laboratory Standards Institute, Wayne, PA. 9. Dengler V, Stutzmann Meier P, Heusser R, Berger-Bachi B, McCallum N. 2011. Induction kinetics of the Staphylococcus aureus cell wall stress stimulon in response to different cell wall active antibiotics. BMC Microbiol. 11:16 –27. 10. Draghi DC, et al. 2008. Comparative surveillance study of telavancin activity against recently collected gram-positive clinical isolates from across the United States. Antimicrob. Agents Chemother. 52:2383–2388. 11. Elsholz AKW, et al. 2011. CtsR inactivation during thiol-specific stress in low GC, gram positive bacteria. Mol. Microbiol. 79:772–785. 12. Fang X, et al. 2006. The mechanism of action of ramoplanin and enduracidin. Mol. Biosyst. 2:69 –76. 13. Gebhard S, Mascher T. 2011. Antimicrobial peptide sensing and detoxification modules: unraveling the regulatory circuitry of Staphylococcus aureus. Mol. Microbiol. 81:581–587. 14. Hegde SS, Reyes N, Skinner R, Difuntorum S. 2008. Efficacy of telavancin in a murine model of pneumonia induced by methicillin-susceptible Staphylococcus aureus. J. Antimicrob. Chemother. 61:169 –172. 15. Higgins DL, et al. 2005. Telavancin, a multifunctional lipoglycopeptide, disrupts both cell wall synthesis and cell membrane integrity in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 49: 1127–1134. 16. Hiron A, Falord M, Valle J, Debarbonille M, Masadek T. 2011. Bacitracin and nisin resistance in Staphylococcus aureus: a novel pathway involving the BraS/BraR two-component system (SA2417/SA2418) and both the BraD/BraE and VraD/VraE ABC transporters. Mol. Microbiol. 81:602– 622. 17. Howden BP, Davies JK, Johnson PD, Sinear TP, Grayson ML. 2010. Reduced vancomycin susceptibility in Staphylococcus aureus, including vancomycin-intermediate and heterogeneous vancomycin-intermediate strains: resistance mechanism, laboratory detection, and clinical implication. Clin. Microbiol. Rev. 23:99 –139. 18. Hutter B, et al. 2004. Prediction of mechanisms of action of antibacterial compounds by gene expression profiling. Antimicrob. Agents Chemother. 48:2838 –2844. 19. Jordan S, Hutchings MI, Mascher T. 2008. Cell envelope stress response in Gram-positive bacteria. FEMS Microbiol. Rev. 32:107–146. 20. Kim SJ, et al. 2008. Oritavancin exhibits dual mode of action to inhibit cell-wall biosynthesis in Staphylococcus aureus. J. Mol. Biol. 377:281–293. 21. Klevens RM, et al. 2007. Invasive methicillin-resistant Staphylococcus aureus infections in the United States. JAMA 298:1763–1771. 22. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ. 2007. A common mechanism of cellular death induced by bactericidal antibiotics. Cell 130:797– 810. 23. Krause KM, et al. 2008. In vitro activity of telavancin against resistant Gram-positive bacteria. Antimicrob. Agents Chemother. 52:2647–2652. 24. Kuroda M, et al. 2003. Two-component system VraSR positively modulates the regulation of cell-wall biosynthesis pathway in Staphylococcus aureus. Mol. Microbiol. 49:807– 821. 25. Leadbetter MR, et al. 2004. Hydrophobic vancomycin derivatives with improved ADME properties: discovery of telavancin (TD-6424). J. Antibiot. (Tokyo) 57:326 –336. 26. Leuthner KD, Cheung CM, Rybak MJ. 2006. Comparative activity of the new lipoglycopeptide telavancin in the presence and absence of serum against 50 glycopeptide non-susceptible staphylococci and three vanco-

aac.asm.org 3163

Song et al.

27. 28. 29.

30.

31. 32. 33.

34.

35.

36. 37. 38. 39.

mycin-resistant Staphylococcus aureus. J. Antimicrob. Chemother. 58: 338 –343. Li M, et al. 2007. The antimicrobial peptide-sensing system aps of Staphylococcus aureus. Mol. Microbiol. 66:1136 –1147. Lunde CS, et al. 2009. Telavancin disrupts the functional integrity of the bacterial membrane through targeted interaction with the cell wall precursor lipid II. Antimicrob. Agents Chemother. 53:3375–3383. Lunde CS, Rexer CH, Hartouni SR, Axt S, Benton BM. 2010. Fluorescence microscopy demonstrates enhanced targeting of telavancin to the division septum of Staphylococcus aureus. Antimicrob. Agents Chemother. 54:2198 –2200. Madrigal AG, Basuino L, Chambers HF. 2005. Efficacy of telavancin in a rabbit model of aortic valve endocarditis due to methicillin-resistant Staphylococcus aureus or vancomycin-intermediate Staphylococcus aureus. Antimicrob. Agents Chemother. 49:3163–3165. McAleese F, et al. 2006. Overexpression of genes of the cell wall stimulon in clinical isolates of Staphylococcus aureus exhibiting vancomycinintermediate-type resistance to vancomycin. J. Bacteriol. 188:1120 –1133. McCallum N, Spehar G, Bischoff M, Berger-Bächi B. 2006. Strain dependence of the cell wall-damage induced stimulon in Staphylococcus aureus. Biochim. Biophys. Acta 1760:1475–1481. Mckay GA, et al. 2006. Oritavancin disrupts transmembrane potential and membrane integrity concomitantly with cell killing in Staphlococcus aureus and vancomycin-resistant enterococci, abstr C1-682. Abstr. 46th Intersci. Conf. Antimicrob. Agents Chemother., San Francisco, CA. Moisan H, Pruneau M, Gattuso M, Malouin F. 2008. Investigations on the mode of action of oritavancin against Staphylococcus aureus through transcriptional profiling, abstr A-061. Abstr. 108th Gen. Meet. Am. Soc. Microbiol. Muthaiyan A, Silverman JA, Jayaswal RK, Wilkinson BJ. 2008. Transcriptional profiling reveals that daptomycin induces the Staphylococcus aureus cell wall stress stimulon and genes responsive to membrane depolarization. Antimicrob. Agents Chemother. 52:980 –990. Patti GJ, et al. 2009. Vancomycin and oritavancin have different modes of action in Enterococcus faecium. J. Mol. Biol. 392:1178 –1191. Patton TG, Yang SJ, Bayles KW. 2006. The role of proton motive force in expression of the Staphylococcus aureus cid and lrg operons. Mol. Microbiol. 59:1395–1404. Peschel A. 2002. How do bacteria resist human antimicrobial peptides? Trends Microbiol. 10:179 –186. Petek M, et al. 2010. Revealing fosfomycin primary effect on Staphylococcus aureus transcriptome: modulation of cell envelope biosynthesis and phosphoenolpyruvate induced starvation. BMC Microbiol. 10:159 –171.

3164

aac.asm.org

40. Pietiainen M, et al. 2009. Transcriptome analysis of the response of Staphylococcus aureus to antimicrobial peptides and characterization of the roles of vraDE and vraSR in antimicrobial resistance. BMC Genomics 10:429 – 444. 41. Raafat D, von Bargen K, Haas A, Sahl H-G. 2008. Insights into the mode of action of chitosan as an antibacterial compound. Appl. Environ. Microbiol. 74:3764 –3773. 42. Rukmana A, Morimoto T, Takahashi H, Giyanto Ogasawara N. 2009. Assessment of transcriptional responses of Bacillus subtilis cells to the antibiotic enduracidin, which interferes with cell wall synthesis, using a highdensity tiling chip. Genes Genet. Syst. 84:253–267. 43. Sass P, et al. 2008. The lantibiotic mersacidin is a strong inducer of the cell wall stress response of Staphylococcus aureus. BMC Microbiol. 8:186 –197. 44. Sass V, Pag U, Tossi A, Bierbaum G, Sahl HG. 2008. Mode of action of human beta-defensin 3 against Staphylococcus aureus and transcriptional analysis of responses to defensin challenge. Int. J. Med. Microbiol. 298: 619 – 633. 45. Serra MD, et al. 2005. Staphylococcus aureus bicomponent ␥-hemolysin, HlgA, HlgB and HlgC, can form mixed pores containing all components. J. Chem. Infect. Models 45:1539 –1545. 46. Silverman JA, Perlmutter NG, Shapiro HM. 2003. Correlation of daptomycin bactericidal activity and membrane depolarization in Staphylococcus aureus. Antimicrob. Agents Chemother. 47:2538 –2544. 47. Sobral RG, et al. 2007. Extensive and genome-wide changes in the transcription profile of Staphylococcus aureus induced by modulating the transcription of the cell wall synthesis gene murF. J. Bacteriol. 189:2376 –2391. 48. Utaida S, et al. 2003. Genome-wide transcriptional profiling of the response of Staphylococcus aureus to cell-wall-active antibiotics reveals a cell-wall-stress stimulon. Microbiology 149:2719 –2732. 49. van Heijenoort J. 2007. Lipid intermediates in the biosynthesis of bacterial peptidoglycan. Microbiol. Mol. Biol. Rev. 71:620 – 635. 50. Wecke T, Mascher T. 2011. Antibiotic research in the age of omics: from expression profiles to interspecies communication. J. Antimicrob. Chemother. 66:2689 –2704. 51. Wenzel M, Bandow JE. 2011. Proteomic signatures in antibiotic research. Proteomics 11:3256 –3268. 52. Wilkinson BJ, Muthaiyan A, Jayaswal RK. 2005. The cell wall stress stimulon of Staphylococcus aureus and other gram-positive bacteria. Curr. Med. Chem. Anti-Infect. Agents 4:259 –276. 53. Yoshida Y, et al. 2011. Bacitracin sensing and resistance in Staphylococcus aureus. FEMS Microbiol. Lett. 320:33–39.
