α-Methylene-γ-butyrolactones attenuate Staphylococcus aureus virulence by inhibition of transcriptional regulation

Article abstract of DOI:10.1039/c3sc52228h

Bacterial pathogenesis is triggered by complex molecular mechanisms that sense bacterial density within an infected host and induce the expression of toxins for overriding the immune response. Virulence is controlled by a set of transcriptional regulators

Full text of DOI:10.1039/c3sc52228h

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
a-Methylene-g-butyrolactones attenuate
Staphylococcus aureus virulence by inhibition
of transcriptional regulation†
Cite this: Chem. Sci., 2014, 5, 1158
*
Martin H. Kunzmann, Nina C. Bach, Bianca Bauer and Stephan A. Sieber
Bacterial pathogenesis is triggered by complex molecular mechanisms that sense bacterial density within an
infected host and induce the expression of toxins for overriding the immune response. Virulence is
controlled by a set of transcriptional regulators that directly bind to DNA promoter regions of toxin-
encoding genes. Here, we identified an a-methylene-g-butyrolactone as
a potent inhibitor of
Staphylococcus aureus virulence. Treatment of bacteria not only resulted in a markedly decreased
expression of one of the most prominent virulence factors a-hemolysin (Hla) but also caused attenuated
invasion efficiency. Mass spectrometry (MS) based target identification revealed this biological effect
originating from the consolidated binding to three important transcriptional regulators SarA, SarR and
MgrA. MS investigation of the binding site uncovered a conserved cysteine in all three proteins which
gets covalently modified. Intriguingly, investigation of DNA binding demonstrated an impaired
DNA-protein interaction upon compound treatment. The functional correlation between target binding,
inhibition and the observed biological effect was proven by gene knockouts and confirmed the expected
mode of action.
Received 8th August 2013
Accepted 20th November 2013
DOI: 10.1039/c3sc52228h
www.rsc.org/chemicalscience
environmental conditions and transduce extracellular
signals inside the cell, thereby initiating selective molecular
Introduction
responses including virulence gene expression via tran-
scriptional regulators.8,9 In S. aureus virulence expression is
controlled by global regulatory elements such as staphylo-
coccal accessory regulators (Sar).7,10,11 The SarA protein binds
to several target gene promoters including a-hemolysin (Hla),
protein A and bronectin binding protein (FnBP).12–14 Anal-
ysis of multiple S. aureus genomes revealed the existence of
additional members of this regulatory family including SarR
and MgrA which share a high structural similarity to
SarA.7,15–17 Recent studies emphasized the existence of a
central Cys based redox sensor which is crucial for DNA
binding and thus regulatory activity.18 In addition, SarA
activity is controlled by phosphorylation of the same Cys
residue.19 Genetic deletion of sarA, sarR and mgrA regulators
revealed dramatic changes in the production of several
toxins, including Hla, that result in strains that cannot cause
infection within murine abscess models.20,21 The regulation
of toxin expression by transcriptional regulators, however, is
highly complex and inactivation of individual proteins such
as SarA can result in the up- or downregulation of Hla toxin
expression depending on the corresponding S. aureus
strain.22–24
Staphylococcus aureus represents an opportunistic pathogen
that infects a variety of human organs and causes severe
diseases such as endocarditis, pneumonia and toxic shock
syndrome.1 Tissue infection is promoted by
a variety
of bacterial virulence factors including hemolysins, leukoci-
dins and immune modulators that support bacterial
propagation within the host and attenuation of the immune
response.2
Due to the lack of novel and effective antibiotic drugs and
the rising concern about multidrug-resistant bacteria, the
targeting of virulence received increasing interest in the past
years as an alternative treatment strategy.3–6 Virulence factors
are not essential for bacterial viability and thus inhibition of
virulence reduces selective pressure, which is a major cause
of resistance development. Recent genomic and proteomic
studies have revealed insights into the complex pathways
by which bacteria control the expression of several virulence
factors.7 Multiple two component systems sense
Center for Integrated Protein Science CIPSM, Institute of Advanced Studies IAS,
¨
¨
Department Chemie, Lehrstuhl fur Organische Chemie II, Technische Universitat
¨
Munchen, Lichtenbergstrasse 4, 85747 Garching, Germany. E-mail: stephan.sieber@
In the search for novel therapies against multidrug-resis-
tant S. aureus (MRSA) strains, anti-virulence concepts become
more and more important. We previously described the
inhibition of ClpP, a central bacterial protease, which in
tum.de; Fax: +49 8928913210; Tel: +49 8928913302
† Electronic supplementary information (ESI) available: Synthesis, cloning,
mutagenesis, protein expression, purication and biochemical assays. See DOI:
10.1039/c3sc52228h
1158 | Chem. Sci., 2014, 5, 1158–1167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
addition to the transcriptional regulators, controls the
expression of crucial virulence factors.25,26 b-lactones were
found to be specic inhibitors which downregulate the
expression of several toxins including toxic shock syndrome
toxin, Hla and enterotoxin B.3 More recently, He et al.
screened a library of compounds for dedicated MgrA inhibi-
tors and identied 5,5-methylenedisalicylic acid (MDSA) that
efficiently blocked DNA binding.27
Results and discussion
Synthesis of a-methylene-g-butyrolactones
Inspired by the diversity of a-methylene-g-butyrolactones in
nature we extended our existing collection of xanthatine
inspired molecules (ESI, Scheme 1†) by four additional mole-
cules that bear the core structure but are equipped with
different side chains.28 All molecular structures exhibit a
terminal alkyne handle which serves as a tag for downstream
target identication by activity based protein proling
(ABPP).23,30,31,36,37 In this strategy the small probe molecules
penetrate cells and bind to the dedicated target(s) within a
native environment (Fig. 1). Subsequent cell lysis, Huisgen–
Sharpless–Meldal cycloaddition (click chemistry) with func-
tionalized azides allow the binding, enrichment and visualiza-
tion of proteins of interest for subsequent identication by
SDS-gel electrophoresis and mass spectrometry, respectively
(Fig. 1).38–40
In our search for novel antibacterial compounds we
screened
a distinct collection of a-methylene-g-butyr-
olactones to reduce the expression of Hla (Fig. 1). a-Methy-
lene-g-butyrolactones represent a potent and privileged
structural motif that exhibits
a
huge diversity of
bioactivities and is present in about 3% of all known natural
products.28,29
We identied two compounds that revealed a signicant
reduction in toxin expression and subsequently started target
deconvolution by activity based protein proling (ABPP)
(Fig. 1).30–32 Proteomic proling and mass spectrometric
analysis showed that all active compounds exhibited binding
and reactivity preferences for the three transcriptional viru-
lence regulators SarA, SarR and MgrA via the conserved
Cys redox sensing residue. Moreover, subsequent EMSA
assays demonstrated that covalent binding to these partic-
ular regulatory proteins induced protein-DNA dissociation.
Concordantly with previous reports,13,33–35 the relevance of
these targets for virulence in S. aureus was veried by
genetic knockouts. Thus, these results not only expand the
scope of virulence inhibition but also emphasize the drugg-
ability of multiple response regulators for medical
applications.
The synthesis followed two different procedures in order to
account for aliphatic as well as aromatic substituted a-methy-
lene-g-butyrolactones (Scheme 1).
For aliphatic substituted lactones, compound 11 was synthe-
sized as described by Kamijo et al. and a subsequent reaction
with the indicated Grignard reagent lead to the precursors 12 and
13.41 A Reformatsky reaction with ethyl 2-(bromo-methyl)acrylate
yielded the nal aliphatic substituted lactone probes 1 and 2 as
racemic mixtures.42 In the case of phenyl substituted lactones, a
Wittig reaction of benzaldehyde with (carbethoxyethylidene) tri-
phenylphosphorane led to the formation of compound 14 which
was brominated at its allylic position by using NBS and AIBN to
yield lactone precursor 15.43 The nal phenyl substituted lactones
Fig. 1 An a-methylene-g-butyrolactone library was screened for hemolysis inhibition of S. aureus. The cellular targets of potent compounds
were identified by incubation with living S. aureus cells, subsequent lysis, click chemistry with rhodamine biotin azide and enrichment with biotin-
avidin beads followed by SDS page and mass spectrometric analysis (MS). In vitro assays and knockout studies were performed to verify the
identified proteins and the biological effect.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 1158–1167 | 1159

View Article Online
Chemical Science
Edge Article
Scheme 1 Synthesis of (A) aliphatic substituted and (B) phenyl substituted a-methylene-g-butyrolactone probes and the hydrated compound
10. For synthetic details please refer to the ESI† section.
3 and 4 (each racemic) were obtained by a Reformatsky reaction EC50 of > 100 mM. However, the presence of an electrophile
of 15 with 5-hexynal.42 Compound 3 was reduced by hydration seems mandatory since no reduction of Hla expression with
with Pd/C to yield probe 10.
compound 10 was observed (ESI, Fig. 1†). A better under-
standing of this SAR data would be achieved by the knowledge
of the corresponding cellular targets.
Test for anti-hemolytic and antibiotic activity
All compounds including the previously established a-methy-
lene-g-butyrolactones were tested for bioactivities against
Target deconvolution by ABPP
several S. aureus strains including antibiotic sensitive To unravel the molecular basis for the anti-hemolytic activity of
(NCTC8325) as well as MRSA (Mu50 and USA300) (Table S1†). 3 we utilized a proteomic approach by which the compounds
First, all compounds were investigated for their ability to inhibit were incubated with living bacteria of three different strains
S. aureus hemolysis of sheep derived erythrocytes. Compound 3 (antibiotic sensitive NCTC8325, and two MRSA strains: Mu50
exhibited the most potent IC50 value of 4 mM in NCTC8325 and and USA300) at various concentrations. The cells were lyzed and
2 mM in USA300 followed by compound 4 that was slightly less the alkyne was subsequently modied with a rhodamine azide
active (Fig. 2A and ESI, Fig. 1†). In order to verify if the anti- tag by click chemistry. The labeled cytosolic and cell envelope
hemolytic effect is based on the direct inhibition of virulence proteomes were individually separated by SDS-gel electropho-
rather than an indirect consequence of antibiotic activity, we resis and analyzed by uorescent scanning. Due to the reactive
additionally determined the minimal inhibitory (MIC) concen- Michael acceptor system, several uorescently labeled proteins
tration of all molecules (Fig. 2B). Interestingly, only weak anti- could be detected and a concentration of 5–10 mM seemed
biotic activity could be obtained for compound 3 (100 mM) optimal for saturated labeling (Fig. 3A, ESI, Fig. 3A and C†). The
which is 25–50 fold higher than the IC50 for hemolytic activity. target pattern can roughly be divided in to high molecular
All other compounds did not show any antibiotic effects below weight (hmw) proteins and low molecular weight proteins
100–500 mM. In addition, compounds 3 and 4 were added to (lmw). While some lmw protein bands were characteristic for
NCTC8325 at various concentrations close to their hemolysis the majority of compounds others were specic to 3 and 4, the
IC50. No adverse effect on the bacterial growth was observed that compounds that exhibited the best anti-hemolytic activity
could explain their anti-hemolytic effects (Fig. 2C) emphasizing (ESI, Fig. 3B†). This emphasizes that the corresponding
that the molecules directly target virulence gene expression.
proteins could be involved in virulence. Another interesting
Although all structures share the same a-methylene-g- observation derived from the comparison of all three strains in
butyrolactone scaffold, the side chain decoration seems to the hmw segment of the gel. One additional protein band at a
signicantly alter and inuence the corresponding bioactivity. molecular weight of 60 kDa appears only in Mu50 and thus
The anti-hemolytic compounds 3 and 4 for instance represent might be resistance associated (ESI, Fig. 4†).
the only scaffolds with a substituent in 4-position. Interestingly,
In order to identify these particular cytosolic protein bands
the conversion of 3 from trans to cis 4 reduces the anti-hemolytic we next applied gel-based and gel-free MS identication
effect emphasizing a precise structure activity relationship methods. Labeled proteins were attached to a trifunctional tag
(SAR) for target interactions.
consisting of azide, rhodamine and biotin via click chemistry.44
All molecules were further tested for cytotoxicity in human Biotinylated proteins were enriched on avidin beads, washed
HeLa cells via the MTT assay (ESI, Fig. 2†). Compound 3 showed and released by heat denaturation. The proteins were then
a toxic effect with an EC50 of 12 mM, which is 3–6-fold above its either separated via SDS-PAGE, isolated, tryptically digested and
anti-hemolytic effect. Thus further optimization of the analyzed by LC-MS/MS or directly digested and analyzed by
compounds is desired in order to improve their toxicity prole LC-MS/MS. MS/MS fragmentation revealed peptide sequences
for pharmacological application. This could be achieved by a that were investigated by the SEQUEST search algorithm in
ne tuning of the electrophilic exocyclic double bond as a fully order to obtain protein identities. Analysis of the lmw protein
reduced a-methyl group in compound 10 revealed a toxicity IDs suggested thioredoxin A (TrxA), a staphylococcal disulde
1160 | Chem. Sci., 2014, 5, 1158–1167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
Fig. 2 (A) Inhibition of sheep blood hemolysis caused by S. aureus NCTC8325 and USA300 by probe 3. Data were fitted to the dose-response
function f(x) ¼ A1 + (A2 ꢀ A1)/(1 + 10^(log(x0 ꢀ x))p) with a variable Hill-slope given by parameter p. (B) MIC values for NCTC8325 strain of the
corresponding lactone probes in mM and mg mLꢀ1. (C) NCTC8325 growth curves based on OD600 with compounds 3 or 4. Each compound was
tested in at least three independent trials in triplicates; average values are shown and error bars display standard deviations from the mean.
reductase, as well as three transcriptional regulators, SarA, SarR a-methylene-g-butyrolactones could explain the observed
and MgrA as the most likely hits over several runs (ESI, Tables 2 antibiotic activity at higher concentrations. All other bands in
and 3†).
the 90 kDa region correspond to characteristic reactive proteins
This result is in line with the observed phenotype of in S. aureus that are unspecically labeled at concentrations
compounds 3 and 4 as SarA, SarR and MgrA are involved in above 20 mM.50 One predominant member has been assigned as
virulence regulation and hla expression.21–24,45,46 Interestingly, formate acetyltransferase before by several MS experiments.49,50
all other compounds did not or only weakly label SarA, SarR Since this study focuses on the anti-hemolytic activity of
and/or MgrA in the proteome which provides a link to the a-methylene-g-butyrolactones we commenced with target vali-
observed phenotype (ESI, Fig. 3 and 4†). In addition, the dation for proteins that are characteristic for the observed
distinct 60 kDa protein specic for Mu50 was identied as biological effects.
bifunctional AAC/APH, an enzyme that is involved in propa-
gating resistance to several antibiotics including gentamycin
and kanamycin (ESI, Fig. 4†).47 The protein band below at
Target validation
about 55 kDa was identied as MurA1 and MurA2 – two Before we started with the functional studies of the transcrip-
enzymes that are involved in cell wall biosynthesis and are tional regulators we rst analyzed MurA1 and MurA2 inhibition
crucial for bacterial viability.48,49 Their putative inhibition by by 3 in order to evaluate their potential contribution to the
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 1158–1167 | 1161

View Article Online
Chemical Science
Edge Article
Fig. 3 Target identification of a-methylene-g-butyrolactones. (A) In situ labeling of S. aureus NCTC8325 with probes 3 and 4 at indicated
concentrations. Identified proteins of the corresponding gel bands are shown on the left. For details about the protein abbreviations please refer
to Table S2†. (B) Target validation by labeling of recombinant proteins with the probes (p ¼ recombinant protein, DT ¼ heat denatured protein).
(C) MS/MS spectrum for the identification of the modified binding site (C* ¼ compound 3 modified cysteine residue).
observed anti-hemolytic effects at sub-MIC concentrations. Therefore, we focused our attention to the functional charac-
Incubation of recombinant MurA proteins49 with 3 and subse- terization of the lmw proteins with special emphasis on the
quent modication with the uorescent dye via click chemistry transcriptional regulators.
revealed a strong uorescent band on SDS-PAGE (Fig. 3B). No
TrxA, sarA, sarR and mgrA were cloned into strep-tag
labeling was observed if the proteins were heat denatured prior expression vector systems and expressed as recombinant
to compound addition emphasizing a specic interaction with proteins in E. coli BL21 DE3. Specic interactions of the probe
the folded and active enzyme. Enzyme activity of the cell wall molecules with the corresponding proteins were conrmed via
biogenesis regulating enzymes MurA1 and MurA2 was in situ labeling with living cells before and aer induction,
measured by monitoring the enolpyruvyl transfer reaction with in vitro labeling with heat denatured cell lysates containing the
phosphoenolpyruvate (PEP) and UDP-N-acetylglucosamine proteins as well as heat denaturation of the puried proteins
(UDPAG).49 Here, 100 mM compound 3 was required to signi- prior to labeling (Fig. 3B, ESI, Fig. 7†). The denatured proteins
cantly impair enzyme activity (ESI, Fig. 5†). This is comparable did not bind the compound emphasizing the importance of a
to the corresponding MIC value (100 mM) suggesting that inhi- properly folded binding pocket for recognition.
bition of these essential enzymes is related to the observed
In order to identify the site of binding we incubated the
antibiotic activity. To evaluate the contribution of MurA1/2 puried proteins with 4-fold excess of 3 for 30 min, digested
inhibition to the reduction of hemolysis we utilized the well with trypsin and analyzed the resulting peptides via HPLC-MS/
characterized MurA1/2 inhibitor fosfomycin and applied it at MS. In case of SarA, SarR and MgrA a single cysteine residue
sub-MIC concentrations to S. aureus NCTC8325.48 No effect on (Cys9, Cys57 and Cys12) was identied as site of covalent
the hemolysis production was observed at a concentration of attachment (Fig. 3C and ESI, Fig. 8†). The identied cysteines
0.55 mM (MIC < 2.75 mM) emphasizing that these targets are not are highly conserved in all three structurally related proteins
responsible for the observed anti-hemolytic effects (ESI, Fig. 6†). and located at the dimerization interface.17,18,51 They form
1162 | Chem. Sci., 2014, 5, 1158–1167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
H-bonds with residues from the adjacent monomer thus likely
Several previous studies show that SarA, SarR and MgrA
explaining their elevated nucleophilicity and reactivity towards regulate hla expression.7,10 This regulation principle, however,
the electrophilic a-methylene-g-butyrolactones. These cysteine is highly complex and strain specic. For instance, a sarA
residues are important regulatory switches. For instance, Cys12 knockout in S. aureus strain 8325-4 downregulates hla
in MgrA is an important sensor for oxidative stress.18 Peroxides expression13,60,61 while the same knockout in strains V8 or
cause oxidation of the thiol to sulfenic acid and subsequent SH1000 upregulates Hla levels.23 Several reasons for this
dissociation from the DNA.18,52 In case of SarA, recent studies observation have been discussed and a strong link to the Sar
show phosphorylation of Cys9 as an additional regulation family protein amount (including SarS) has been sug-
principle reducing DNA binding affinity.19,53
gested.23,25 In order to further elucidate if the observed anti-
It is known that S. aureus adapts to changes of external hemolytic effect correlates indeed with reduced expression
oxygen concentrations by redox dependent processes regulating levels of Hla we utilized western blot analysis with anti-Hla
the expression of genes. Small thiol specic proteins such as antibodies. In agreement with the results of the blood hemo-
thioredoxin (TrxA) maintain the intracellular thiol-disulde lysis assay as well as the EMSA assay the expression of Hla was
balance.54,55 TrxA is an oxidoreductase enzyme featuring a signicantly diminished at a concentration of 2 mM 3 and
conserved active site Trp-Cys-Gly-Pro-Cys sequence that forms abolished at concentrations above 4 mM (Fig. 4C). This
or breaks disulde bonds by inter-protein exchange reactions.56 strongly suggests that at least the cumulative inhibition of all
As this regulation also indirectly triggers the redox state of three transcriptional regulators reduces hla expression in the
transcriptional regulators we were interested if 3 is able to S. aureus strains investigated here (NCTC8325 and USA3000).
inhibit the reductase activity of TrxA.54 The inhibition of TrxA by However, the individual role of each regulator has to be
3 was evaluated by an assay system which measures the investigated with the corresponding mgrA, sarA and sarR
reduction of disulde bonds within insulin.57 Reduced insulin knockout strains.
becomes insoluble and enzyme turnover is followed by
increasing turbidity. Concentration-dependent inhibition of
TrxA activity was observed and an IC50 value of 17 mM obtained
(ESI, Fig. 9†). Apart from its function in the cellular redox
pathway not much is known about the biological impact of TrxA
inhibition. A trxA deletion in Salmonella enterica exhibited a
reduction in intracellular replication and mouse virulence
emphasizing a possible role in virulence regulation that would
correlate with the above-mentioned function.58
These examples highlight the importance of a ne tuned
cellular redox mechanism that among other duties, maintains
cysteine oxidation states in transcriptional regulators for DNA
binding. In order to evaluate if covalent modication with a
small molecule induces DNA displacement effects we utilized
electrophoretic mobility shi assays (EMSA) and obtained cor-
responding IC50 values (Fig. 4 and ESI, Fig. 10†).
Based on previous studies we selected the agr promoter
DNA region as a conrmed binding site for SarA, SarR and
MgrA.45,59 A uorescent AlexaFluor488 dye at the 50site allowed
us to monitor protein-DNA interaction via uorescence scan-
ning of the corresponding native polyacrylamide gels. All three
proteins were incubated in the presence of the uorescent
DNA and various concentrations of compounds. Unspecic
binding was reduced by the addition of an excess of salmon
sperm DNA.
The corresponding mobility shi gels are shown in Fig. 4 and
ESI, Fig. 10†. Interestingly, SarA, SarR and MgrA showed
mobility shis at low molar excess of 3 (0.5-fold for SarR and
4-fold for SarA and MgrA). This emphasizes that the binding to
all three transcriptional regulators could be responsible for the
phenotypic effect. Importantly, compounds not exhibiting any
anti-virulence activity, except compound 5, did not show
mobility shis with SarA, SarR and MgrA, thereby demon-
strating a lack of interaction (Fig. 4 and ESI, Fig. 10,† right
Functional analysis by gene knockouts
In order to validate the relevance of transcriptional regulator
inhibition by a-methylene-g-butyrolactone
previously established knockout strains of sarA, sarR, mgrA as
well as a sarA/R double knockout. Phage mediated transduction
of the selective resistance cassettes into the corresponding
genes of the S. aureus NCTC8325 genome revealed corre-
3 we obtained
sponding
knockout
strains
for
further
functional
investigations.13,33–35
Subsequent hemolysis assays with the parental strain and
the knockouts revealed a signicant reduction in Hla produc-
tion in the case of DsarR and DmgrA (Fig. 5). Contrary, the DsarA
strain showed an increase in hemolysis as described for several
strains previously,23 thereby indicating SarA as being not a
responsible target for the anti-hemolytic effect of compound 3.
Interestingly, addition of compound 3 decreased hemolysis
production of DsarA bacteria, thus emphasizing SarR and MgrA
as the responsible anti-virulence-determining targets. In order
to investigate which of the two effects–activation of hemolysis
by DsarA or reduction of hemolysis by DsarR is dominating we
investigated a DsarADsarR double mutant. Interestingly, this
double mutation resulted in no detectable hemolysis demon-
strating that inactivation of at least these two transcriptional
regulators is sufficient to match the chemical phenotype of
a-methylene-g-butyrolactone 3. Thus, the results of ABPP, EMSA
studies as well as functional knockouts clearly support a
mechanism of action by which compound 3 covalently attaches
to a functional relevant cysteine in all three transcriptional
regulators, preventing DNA binding and downstream hemolysis
production.
Reduction of S. aureus invasion into human THP-1 cells
panel). This is also in line with labeling studies of the three Previous studies with sarA deletion strains in S. aureus indicated
recombinant proteins (ESI, Fig. 11†).
a reduced binding to bronectin, an important membrane
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 1158–1167 | 1163

View Article Online
Chemical Science
Edge Article
Fig. 4 Electrophoretic mobility shift assay (EMSA) with (A) SarA and (B) SarR (for MgrA please refer to ESI, Fig. 10†). The left part of the gel shows a
concentration dependent analysis with anti-virulence compound 3, the right part displays a single concentration analysis with all other
compounds. DNA ¼ unbound fluorescent promoter DNA; SarA/SarR-DNA ¼ protein bound DNA. (C) Luminescent western blot of S. aureus
NCTC8325 supernatant (grown with increasing concentrations of probe 3) incubated with a primary anti-Hla (a-hemolysin) and a secondary
(HRP) antibody.
associated protein in eukaryotic cells and a trigger for bacterial
invasion.62,63 Therefore, we nally investigated if binding of 3 to
the transcriptional regulators would reduce the phagocytosis
mediated invasion of S. aureus into human macrophages for
which we performed a compound-based assay.64 Bacterial cells
were grown in the presence of 20 mM 3 for 16 h, and incubated
with differentiated THP-1 cells in the presence of 2 mM 3 for 2 h.
In order to interrupt cellular uptake mechanisms cells were
washed with cold PBS and subsequently incubated in medium
supplemented with 50 mg mLꢀ1 gentamycin to kill all remaining
extracellular, non-internalized bacteria. The cells were lyzed,
lysates were plated and intracellular bacteria were counted as
colony forming units (CFU) aer one day of incubation at 37 ꢁC.
Interestingly, compound 3 signicantly reduced the invasive-
ness of S. aureus by 40% emphasizing that the pathogen is not
only disarmed in its Hla mediated virulence but also weakened
Fig. 5 Hemolysis of sheep red blood cells upon incubation with
supernatants of S. aureus USA300, NCTC8325 and the corresponding
in its general pathogenicity (ESI, Fig. 12†). This further high-
lights the relevance of these proteins as putative targets against
bacterial infections.
NCTC8325 knockout strains DmgrA, DsarA, DsarR and DsarADsarR
with and without 10 mM probe 3.
1164 | Chem. Sci., 2014, 5, 1158–1167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
(10 min, 13 000 g) and 100 mL of each supernatant (triplicates
for each culture) were pipetted into 96 well plates. Positive
controls (1 mL DMSO) and negative controls (1 mL sterile broth)
were included. Erythrocyte suspension (50 mL) was added to
each well and the plate was incubated for 10 min shaking at
37 C. The OD600, which correlates with the amount of intact
erythrocytes, was measured by a Tecan Innite® M200 PRO
plate reader.
Conclusion
Inhibition of virulence is a powerful approach to reduce the
pathogenicity of bacteria. The expression of toxins that
contribute to the devastating effects of bacterial infections is
regulated by a complex network of proteins that mediate a
precise timing in the initiation of virulence. This timing is
important, as bacteria need a certain density in order to overrun
the host. Several transcriptional regulators are activated by
quorum sensing and initiate toxin production. SarA, SarR and
MgrA are DNA binding proteins that attach to a promoter region
important for hla transcription. Once their central cysteine is
modulated by e.g. phosphorylation or oxidation these proteins
show reduced DNA affinity and therefore alter transcription of
the target genes. We were able to show that the covalent
modication of these particular central cysteine residues by
suitably designed a-methylene-g-butyrolactones not only
reduced DNA affinity but induced DNA dissociation that resul-
ted in a strong decrease of Hla expression. In addition, the
invasion of S. aureus into human THP-1 cells was signicantly
reduced demonstrating that bacteria are weakened in multiple
aspects of pathogenesis. This approach is novel and exhibits
several advantages compared to a previous MgrA specic
inhibitor (MDSA).27 a-Methylene-g-butyrolactone 3 abolishes
Hla production at 2–4 mM concentration and is thus more
potent compared to MSDA, which exhibits a hemolysis IC50 of
0.2 mM. Combined inhibition of several targets might be an
advantageous strategy as the effects are likely additive leading to
a more efficient depletion of Hla mediated virulence. Due to the
structural similarity of all three transcriptional regulators it is
not surprising that one compound with a suitable decoration
binds all proteins via the sensing cysteine. This irreversible
binding also exhibits the advantage of a stable link that is long
lasting and only reversed by protein turnover.
ꢁ
Electrophoretic Mobility Shi Assay (EMSA)
The Alexa Fluor 488 (AF488) labeled agr promoter fragment of
NCTC8325 (ref. 24) was generated by PCR and then puried by
an Omega Bio-Tek MicroElute Cycle-Pure Kit. MgrA, SarA and
SarR were puried as described in the ESI† and aliquots were
stored in 100 mM Tris–HCl pH 8.0, 150 mM NaCl and 1 mM ETDA
ꢁ
at ꢀ80 C. To 10.8 mL EMSA-Binding-Buffer (100 mM Tris–HCl
pH 8.0, 150 mM NaCl, 1 mM EDTA, 5% glycerol) containing MgrA
(1.75 mM), SarA (2.50 mM) or SarR (4.00 mM) were added 0.2 mL
probe of the desired concentration in DMSO and the mixture
was incubated 1.5 h at room temperature. Then 1 mL of agr
promoter fragment mixed with salmon sperm DNA (Carl Roth,
Karlsruhe) was added leading to nal concentrations of 5 nM
(7.5 ng) agr promoter fragment and 2 mg mLꢀ1 salmon sperm
DNA (Carl Roth, Karlsruhe). Aer incubation for 15 min at room
temperature, 2 mL EMSA-Loading-Buffer (100 mM Tris–HCl pH
8.0, 150 mM NaCl, 1 mM EDTA, 5% Glycerol, 0.1% bromphenol
blue) were added and the samples were applied on 5% TBE
polyacrylamide gels (150 V, 45 min). Fluorescence was recorded
in a Fujilm Las-4000 luminescent image analyzer with a Fuji-
non VRF43LMD3 lens and an Y515-Di lter. Forward primer:
50-CAATTTTACACCACTCTCCTC (50-AF488), Reverse primer:
50-CATCAACTATTTTCCATCACATCT (50-AF488).
Invasion assay for S. aureus infected human macrophages
Human THP-1 monocytes were seeded in a 6-well plate (1 ꢂ
106 cells per well in 2 mL RPMI1640 medium containing 10%
fetal calf serum and 5 mM L-glutamine) and PMA (phorbol 12-
myristate 13-acetate, 40 ng per well) was added. Cells were
Methods
Determination of MIC-values
S. aureus was grown in B broth shaking at 37 ꢁC until cell density
reached OD600 ¼ 0.6–0.8. Fresh B broth (Yeast extract (5.0 g),
tryptic peptone (10.0 g), NaCl (5.0 g), K2HPO4 (1.0 g), H2O
(1.0 L)) was inoculated with 2 ꢂ 106 bacteria per mL and 99 mL
of this culture were added to each well of a 96 well plate. Probe
in DMSO (1 mL, varying concentrations) was added and control
wells containing no probe (1 mL DMSO) or no bacteria (99 mL B
broth) were included. The plate was incubated overnight (16 h)
shaking at 37 ꢁC and the resulting optical density was measured
by a Tecan Innite® M200 PRO plate reader.
ꢁ
incubated for 2 days (37 C; 5% CO2), medium was removed,
cells were washed three times with PBS (RT) and fresh medium
was added. Bacteria (S. aureus NCTC8325) were inoculated in
fresh B-medium (1 : 100), incubated to an OD600 of 0.4–0.5 and
diluted to 3.4 ꢂ 107 bacteria per mL (formula y ¼ 4E +
07e1.0958x). Bacteria were supplemented with respective inhibi-
tory compounds (1000 ꢂ DMSO stocks) and incubated for 16 h
(37 ꢁC, 200 rpm). Bacterial culture was diluted in B-medium
(1 : 10) and OD600 was measured. CFU was calculated
(see formula) and 5 ꢂ 107 bacteria (MOI 50, conuent mono-
layer in 6-well plate 1 ꢂ 106 cells) were diluted in 2 mL
RPMI1640/10% FCS/5 mM L-glutamine. The respective
Hemolysis-assay
S. aureus was grown in B broth shaking at 37 ꢁC until cell density compound (non-lethal concentration, 2 mM, 1000 ꢂ DMSO
reached OD600 ¼ 0.6–0.8. Fresh B broth was inoculated with 2 ꢂ stocks) was added to bacterial suspensions and 2 mL of the
106 bacteria per mL and to 1 mL aliquots was added probe in bacterial suspension was added to the cells (2 mL without
DMSO (1 mL). The cultures were incubated overnight (16 h) bacteria as control). Infected cells were incubated for 2 h (37 ꢁC,
shaking at 37 ꢁC. The OD600 was measured to check the effect of 5% CO2), supernatant was removed and collected for CFU-
substances on bacterial growth. Cultures were centrifuged plating assay of the inoculum. Cells were washed four times
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 1158–1167 | 1165

View Article Online
Chemical Science
Edge Article
with ice cold PBS and 2 mL medium containing gentamycin 16 T. T. Luong, S. W. Newell and C. Y. Lee, J. Bacteriol., 2003,
(50 mg mLꢀ1) was added. Cells were incubated for 24 h with
185, 3703–3710.
gentamycin to kill all extracellular bacteria and proceed intra- 17 Y. Liu, A. C. Manna, C. H. Pan, I. A. Kriksunov, D. J. Thiel,
cellular replication of invaded bacteria. Aerwards cells were
washed once with PBS (1 mL) and 0.5% saponin (1 mL)/
A. L. Cheung and G. Zhang, Proc. Natl. Acad. Sci. U. S. A.,
2006, 103, 2392–2397.
B-medium was added. Incubation of cells proceeded for 10 min 18 P. R. Chen, T. Bae, W. A. Williams, E. M. Duguid, P. A. Rice,
ꢁ
at 37 C (5% CO2) and then the lysate was plated out by using
O. Schneewind and C. He, Nat. Chem. Biol., 2006, 2, 591–595.
different dilutions. Quantitative analysis of bacterial concen- 19 F. Sun, Y. Ding, Q. Ji, Z. Liang, X. Deng, C. C. Wong, C. Yi,
trations was performed by counting CFUs one day aer plating.
Dilutions for CFU-plating assay: inoculum: 10ꢀ2, 10ꢀ3, 10ꢀ4
supernatants: 100; lysed cells: 10ꢀ1, 10ꢀ2 in B-medium. Plates
L. Zhang, S. Xie, S. Alvarez, L. M. Hicks, C. Luo, H. Jiang,
L. Lan and C. He, Proc. Natl. Acad. Sci. U. S. A., 2012, 109,
15461–15466.
;
were incubated overnight (37 ꢁC) and colonies counted 20 I.-M. Jonsson, C. Lindholm, T. T. Luong, C. Y. Lee and
subsequently.64
A. Tarkowski, Microbes Infect., 2008, 10, 1229–1235.
21 A. K. Zielinska, K. E. Beenken, L. N. Mrak, H. J. Spencer,
G. R. Post, R. A. Skinner, A. J. Tackett, A. R. Horswill and
M. S. Smeltzer, Mol. Microbiol., 2012, 86, 1183–1196.
22 R. Babai, B. E. Stern, J. Hacker and E. Z. Ron, Infect. Immun.,
2000, 68, 5901–5907.
23 J. Oscarsson, A. Kanth, K. Tegmark-Wisell and S. Arvidson,
J. Bacteriol., 2006, 188, 8526–8533.
24 J. Morschhauser, G. Kohler, W. Ziebuhr, G. Blum-Oehler,
U. Dobrindt and J. Hacker, Philos. Trans. R. Soc. London,
Ser. B, 2000, 355, 695–704.
Acknowledgements
We thank Mona Wolff for excellent technical support, David
C. Hooper, Simon J. Foster and Erik Gustafsson for providing
S. aureus knockout strains and Matt Nodwell and Katrin Lorenz-
Baath for scientic discussions. S. A. Sieber was supported by
the Deutsche Forschungsgemeinscha SFB749, SFB1035,
FOR1406, an ERC starting grant (259024) and the Center for
Integrated Protein Science Munich CIPSM. M. H. Kunzmann
thanks the TUM Graduate School for project funding and the
ERC for nancial support.
25 D. Frees, J. H. Andersen, L. Hemmingsen, K. Koskenniemi,
K. T. Baek, M. K. Muhammed, D. D. Gudeta, T. A. Nyman,
A. Sukura, P. Varmanen and K. Savijoki, J. Proteome Res.,
2012, 11, 95–108.
26 D. Frees, K. Sorensen and H. Ingmer, Infect. Immun., 2005,
73, 8100–8108.
References
1 F. D. Lowy, N. Engl. J. Med., 1998, 339, 520–532.
2 W. W. Navarre and O. Schneewind, Microbiol. Mol. Biol. Rev.,
1999, 63, 174–229.
27 F. Sun, L. Zhou, B. C. Zhao, X. Deng, H. Cho, C. Yi, X. Jian,
C. X. Song, C. H. Luan, T. Bae, Z. Li and C. He, Chem.
Biol., 2011, 18, 1032–1041.
¨
3 T. Bottcher and S. A. Sieber, J. Am. Chem. Soc., 2008, 130, 28 M. H. Kunzmann and S. A. Sieber, Mol. BioSyst., 2012, 8,
14400–14401.
3061–3067.
4 L. Cegelski, G. R. Marshall, G. R. Eldridge and S. J. Hultgren, 29 R. R. Kitson, A. Millemaggi and R. J. Taylor, Angew. Chem.,
Nat. Rev. Microbiol., 2008, 6, 17–27.
Int. Ed., 2009, 48, 9426–9451.
5 A. E. Clatworthy, E. Pierson and D. T. Hung, Nat. Chem. Biol., 30 M. J. Evans and B. F. Cravatt, Chem. Rev., 2006, 106, 3279–
2007, 3, 541–548. 3301.
6 D. T. Hung, E. A. Shakhnovich, E. Pierson and 31 M. Fonovic and M. Bogyo, Expert Rev. Proteomics, 2008, 5,
J. J. Mekalanos, Science, 2005, 310, 670–674. 721–730.
7 A. L. Cheung, A. S. Bayer, G. Zhang, H. Gresham and 32 T. Bottcher, M. Pitscheider and S. A. Sieber, Angew. Chem.,
¨
Y. Q. Xiong, FEMS Immunol. Med. Microbiol., 2004, 40, 1–9.
8 R. P. Novick, Mol. Microbiol., 2003, 48, 1429–1449.
9 E. A. George and T. W. Muir, ChemBioChem, 2007, 8, 847–
855.
10 A. L. Cheung, K. A. Nishina, M. P. Trotonda and S. Tamber,
Int. J. Biochem. Cell Biol., 2008, 40, 355–361.
Int. Ed., 2010, 49, 2680–2698.
33 Q. C. Truong-Bolduc, Y. Ding and D. C. Hooper, J. Bacteriol.,
2008, 190, 7375–7381.
34 J. Oscarsson, K. Tegmark-Wisell and S. Arvidson, Int. J. Med.
Microbiol., 2006, 296, 365–380.
35 E. Gustafsson and J. Oscarsson, FEMS Microbiol. Lett., 2008,
284, 158–164.
11 P. M. Dunman, E. Murphy, S. Haney, D. Palacios, G. Tucker-
Kellogg, S. Wu, E. L. Brown, R. J. Zagursky, D. Shlaes and 36 J. Krysiak and R. Breinbauer, Top. Curr. Chem., 2012, 324,
S. J. Projan, J. Bacteriol., 2001, 183, 7341–7353.
43–84.
12 Y. Chien, A. C. Manna, S. J. Projan and A. L. Cheung, J. Biol. 37 W. P. Heal and E. W. Tate, Top. Curr. Chem., 2012, 324, 115–
Chem., 1999, 274, 37169–37176.
135.
13 P. F. Chan and S. J. Foster, J. Bacteriol., 1998, 180, 6232–6241. 38 V. V. Rostovtsev, J. G. Green, V. V. Fokin and K. B. Sharpless,
14 A. C. Manna, M. G. Bayer and A. L. Cheung, J. Bacteriol.,
1998, 180, 3828–3836.
15 T. T. Luong, P. M. Dunman, E. Murphy, S. J. Projan and
C. Y. Lee, J. Bacteriol., 2006, 188, 1899–1910.
Angew. Chem., Int. Ed., 2002, 41, 2596–2599.
39 C. W. Tornøe, C. Christensen and M. Meldal, J. Org. Chem.,
2002, 67, 3057–3064.
40 R. Huisgen, Proc. Chem. Soc., 1961, 357–396.
1166 | Chem. Sci., 2014, 5, 1158–1167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Edge Article
Chemical Science
41 S. Kamijo and G. B. Dudley, J. Am. Chem. Soc., 2006, 128, 53 D. F. Fujimoto, R. H. Higginbotham, K. M. Sterba,
6499–6507.
S. J. Maleki, A. M. Segall, M. S. Smeltzer and
B. K. Hurlburt, Mol. Microbiol., 2009, 74, 1445–1458.
54 A. Ballal and A. C. Manna, J. Bacteriol., 2010, 192, 336–345.
42 M. S. Sawant, R. Katoch, G. K. Trivedi and U. R. Desai,
J. Chem. Soc., Perkin Trans. 1, 1998, 0, 843–846.
43 L.-W. Ye, X. Han, X.-L. Sun and Y. Tang, Tetrahedron, 2008, 55 O. Uziel, I. Borovok, R. Schreiber, G. Cohen and
64, 1487–1493. Y. Aharonowitz, J. Bacteriol., 2004, 186, 326–334.
44 J. Eirich, J. L. Burkhart, A. Ullrich, G. C. Rudolf, A. Vollmar, 56 J. Messens, I. Van Molle, P. Vanhaesebrouck, M. Limbourg,
S. Zahler, U. Kazmaier and S. A. Sieber, Mol. BioSyst., 2012, 8,
2067–2075.
K. Van Belle, K. Wahni, J. C. Martins, R. Loris and
L. Wyns, J. Mol. Biol., 2004, 339, 527–537.
45 S. Ingavale, W. van Wamel, T. T. Luong, C. Y. Lee and 57 A. Holmgren, J. Biol. Chem., 1979, 254, 9627–9632.
A. L. Cheung, Infect. Immun., 2005, 73, 1423–1431.
46 A. K. Zielinska, K. E. Beenken, H. S. Joo, L. N. Mrak,
58 E. Bjur, S. Eriksson-Ygberg, F. Aslund and M. Rhen, Infect.
Immun., 2006, 74, 5140–5151.
L. M. Griffin, T. T. Luong, C. Y. Lee, M. Otto, L. N. Shaw 59 A. C. Manna and A. L. Cheung, Mol. Microbiol., 2006, 60,
and M. S. Smeltzer, J. Bacteriol., 2011, 193, 2948–2958. 1289–1301.
47 D. D. Boehr, W. S. Lane and G. D. Wright, Chem. Biol., 2001, 60 A. S. Khan, I. Muhldorfer, V. Demuth, U. Wallner,
8, 791–800.
T. K. Korhonen and J. Hacker, Mol. Gen. Genet., 2000, 263,
96–105.
48 W. Du, J. R. Brown, D. R. Sylvester, J. Huang, A. F. Chalker,
C. Y. So, D. J. Holmes, D. J. Payne and N. G. Wallis, 61 M. Steinert, U. Hentschel and J. Hacker, Naturwissenschaen,
J. Bacteriol., 2000, 182, 4146–4152. 2000, 87, 1–11.
49 T. Bottcher and S. A. Sieber, J. Am. Chem. Soc., 2010, 132, 62 Y. Q. Xiong, A. S. Bayer, M. R. Yeaman, W. Van Wamel,
¨
6964–6972.
A. C. Manna and A. L. Cheung, Infect. Immun., 2004, 72,
1832–1836.
63 A. Karlsson, P. Saravia-Otten, K. Tegmark, E. Morfeldt and
S. Arvidson, Infect. Immun., 2001, 69, 4742–4748.
50 M. B. Nodwell, H. Menz, S. F. Kirsch and S. A. Sieber,
ChemBioChem, 2012, 13, 1439–1446.
51 Y. Liu, A. Manna, R. Li, W. E. Martin, R. C. Murphy,
A. L. Cheung and G. Zhang, Proc. Natl. Acad. Sci. U. S. A., 64 M. P. Horn, S. M. Knecht, F. L. Rushing, J. Birdsong,
2001, 98, 6877–6882.
52 G. A. Somerville and R. A. Proctor, Microbiol. Mol. Biol. Rev.,
2009, 73, 233–248.
C. P. Siddall, C. M. Johnson, T. N. Abraham, A. Brown,
C. B. Volk, K. Gammon, D. L. Bishop, J. L. McKillip and
S. A. McDowell, J. Pharmacol. Exp. Ther., 2008, 326, 135–143.
This journal is © The Royal Society of Chemistry 2014
Chem. Sci., 2014, 5, 1158–1167 | 1167