Showdomycin as a versatile chemical tool for the detection of pathogenesis-associated enzymes in bacteria

Article abstract of DOI:10.1021/ja909150y

Showdomycin is a potent nucleoside antibiotic that displays a high structural similarity to uridine and pseudouridine. No detailed target analysis of this very unusual electrophilic natural product has been carried out so far. To unravel its biological function, we synthesized a showdomycin probe that can be appended with a fluorophor or a biotin marker via click chemistry and identified diverse enzymes which were important for either the viability or virulence of pathogenic bacteria. Our results indicate that the antibiotic effect of showdomycin against Staphylococcus aureus may be due to the inhibition of various essential enzymes, especially MurA1 and MurA2, which are required for cell wall biosynthesis. Although real-time polymerase chain reaction revealed that the MurA2 gene was expressed equally in four S. aureus strains, our probe studies showed that MurA2 was activated in only one multiresistant S. aureus strain, and only this strain was resistant to elevated concentrations of the MurA inhibitor fosfomycin, suggesting its potential role as an antibiotic bypass mechanism in the case of MurA1 inhibition. Moreover, we utilized this tool to compare enzyme profiles of different pathogenic strains, which provided unique insights in regulatory differences as well as strain-specific signatures.

Full text of DOI:10.1021/ja909150y

Published on Web 04/30/2010
Showdomycin as a Versatile Chemical Tool for the Detection
of Pathogenesis-Associated Enzymes in Bacteria
Thomas Bo¨ttcher and Stephan A. Sieber*
Center for Integrated Protein Science Munich (CIPSM), Department of Chemistry and
Biochemistry, Ludwig-Maximilians-UniVersita¨t Mu¨nchen, Butenandtstrasse 5-13,
81377 Munich, Germany
Received October 27, 2009; E-mail: stephan.sieber@cup.uni-muenchen.de
Abstract: Showdomycin is a potent nucleoside antibiotic that displays a high structural similarity to uridine
and pseudouridine. No detailed target analysis of this very unusual electrophilic natural product has been
carried out so far. To unravel its biological function, we synthesized a showdomycin probe that can be
appended with a fluorophor or a biotin marker via click chemistry and identified diverse enzymes which
were important for either the viability or virulence of pathogenic bacteria. Our results indicate that the antibiotic
effect of showdomycin against Staphylococcus aureus may be due to the inhibition of various essential
enzymes, especially MurA1 and MurA2, which are required for cell wall biosynthesis. Although real-time
polymerase chain reaction revealed that the MurA2 gene was expressed equally in four S. aureus strains,
our probe studies showed that MurA2 was activated in only one multiresistant S. aureus strain, and only
this strain was resistant to elevated concentrations of the MurA inhibitor fosfomycin, suggesting its potential
role as an antibiotic bypass mechanism in the case of MurA1 inhibition. Moreover, we utilized this tool to
compare enzyme profiles of different pathogenic strains, which provided unique insights in regulatory
differences as well as strain-specific signatures.
Introduction
enzymes. Many of those enzyme families have not been targeted
by ABPP probes before, which opens new avenues to study
Nature provides a great source of bioactive molecules that
are used in medicinal applications for the treatment of many
severe diseases, including bacterial infections. The repertoire
of small molecules that are produced by microorganisms and
sponges exerts a great diversity of biological activities that in
many cases is a result of evolutionary processes to develop
customized defense strategies against competitors in the fight
for limited nutritional resources. The urgent quest to discover
novel and biologically active natural product antibiotics for the
treatment of multiresistant bacterial pathogens has led to
extensive efforts in the fields of natural product isolation and
their corresponding total synthesis. However, in many cases the
application as a drug is limited since the precise mechanism of
action and the dedicated molecular targets remain unknown.
We have previously unraveled the molecular targets of bio-
mimetic ꢀ-lactones1-3 and ꢀ-lactams4,5 in bacteria by a chemical
proteomic strategy called activity-based protein profiling
(ABPP).6,7 In these studies, several targets with essential roles
in virulence and resistance could be detected. Here we extend
this strategy to the natural product showdomycin and explore
its full complement of biological targets in diverse pathogenic
bacteria. We show that a reporter-tagged showdomycin probe
labels and inhibits a great diversity of important bacterial
their precise function in various proteomes. We further show
that the showdomycin probe is an important tool for the
discovery of enzymes that display characteristic signature motifs
in closely related strains of pathogenic bacteria that help to
discriminate multiresistant from antibiotic-sensitive strains.
Finally, we identified molecular targets that are likely to be
responsible for the antibiotic effects of showdomycin.
Showdomycin is a potent C-glycosyl nucleoside antibiotic
that was isolated from Streptomyces showdoensis.8 It displays
a high structural similarity to uridine and pseudouridine, with
the exception that the bases are replaced by an electrophilic
maleimide moiety (Figure 1). It was therefore speculated that
nature designed this nucleoside analogue as a suicide inhibitor
to interfere with the uridine metabolism.9 So far, several studies
found evidence using cell-free in Vitro experiments with
eukaryotes that showdomycin targets uridine, UDP, and other
nucleotide-dependent enzymes like kinases, phosphorylases, and
dehydrogenases.9-11 However, no detailed target analysis of this
very unusual electrophilic natural product has been carried out
so far for bacteria.
Results and Discussion
To unravel its biological targets, we first applied a total
synthesis of showdomycin adapted from a strategy reported by
(1) Bo¨ttcher, T.; Sieber, S. A. J. Am. Chem. Soc. 2008, 130, 14400–14401.
(2) Bo¨ttcher, T.; Sieber, S. A. Angew. Chem., Int. Ed. 2008, 47, 4600–
4603.
(8) Nishimura, H.; Mayama, M.; Komatsu, Y.; Kato, H.; Shimaoka, N.;
Tanaka, Y. J. Antibiot. (Tokyo) 1964, 17, 148–155.
(3) Bo¨ttcher, T.; Sieber, S. A. ChemMedChem 2009, 4, 1260–1263.
(4) Staub, I.; Sieber, S. A. J. Am. Chem. Soc. 2008, 130, 13400–13409.
(5) Staub, I.; Sieber, S. A. J. Am. Chem. Soc. 2009, 131, 6271–6276.
(6) Evans, M. J.; Cravatt, B. F. Chem. ReV. 2006, 106, 3279–3301.
(7) Fonovic, M.; Bogyo, M. Exp. ReV. Proteom. 2008, 5, 721–730.
(9) Roy-Burman, S.; Roy-Burman, P.; Visser, D. W. Cancer Res. 1968,
28, 1605–1610.
(10) Tobin, T.; Akera, T. Biochim. Biophys. Acta 1975, 389, 126–136.
(11) Mu¨ller, T.; Bause, E.; Jaenicke, L. FEBS Lett. 1981, 128, 208–212.
9
6964 J. AM. CHEM. SOC. 2010, 132, 6964–6972
10.1021/ja909150y  2010 American Chemical Society

Showdomycin as a Chemical Probe for Pathogenic Bacteria
A R T I C L E S
Figure 1. Similarity of showdomycin to uridine and pseudouridine. Sulfhydryl groups in the enzyme’s binding site attack the reactive maleimide moiety
and covalently bind the probe. The alkyne linker of the probe allows the identification of labeled enzymes by the introduction of a fluorescent tag via click
chemistry after cell lysis and SDS gel analysis.
Barrett et al.12 and attached 5-hexynoic acid via the free primary
alcohol group of the ribose moiety (Supporting Information
Figure S1) to introduce an alkyne handle for subsequent target
analysis. The modification of the alkyne tag via 1,3-dipolar
Huisgen cycloaddition (click chemistry, CC)13,14 allows us to
append a fluorophor or biotin reporter group after proteome
labeling for target enzyme visualization via SDS-PAGE or
enrichment for target identification via mass spectrometry (MS),
respectively. Diastereomeric showdomycin and epi-showdomy-
cin probes were separated by preparative HPLC, and the
naturally occurring showdomycin isomer was used for all
subsequent studies.
(Supporting Information Figure S2A). Interestingly, while the
maleimide probe labeled a great deal of almost all proteins in
the lysate, the showdomycin probe revealed target selectivity
for only a limited number of proteins. A corresponding control
by which the proteome was heat-inactivated at 95 °C for 5 min
prior to probe addition further demonstrated that the labeling
of the majority of targets occurred in an activity-based manner,
as these proteins do not get labeled when they are unfolded
(Supporting Information Figure S2B). MS experiments revealed
that in aqueous solution showdomycin reacts only with thiols
but not with amines, indicating that cysteine residues are indeed
the preferred point of attachment (Supporting Information).
To test whether the showdomycin probe still displayed
antibacterial activity, we compared minimal inhibitory concen-
tration (MIC) values of S. aureus NCTC 8325 that was treated
with various concentrations of the two compounds in culture
medium. In fact, both molecules gave identical MIC values (500
µM for showdomycin and 500 µM for the corresponding probe),
indicating that the small alkyne tag modification does not hinder
the activity of the natural product. These MIC values are in the
order of magnitude determined previously by the agar-streak
dilution method for another S. aureus strain with a MIC around
220 µM.8 Maleimide is a strong electrophile that reacts
preferentially with cysteine residues of proteins, also including
those which are not of elevated nucleophilicity and are not part
of the active-site pocket. We therefore hypothesized that the
maleimide moiety, the reactive group of showdomycin, is fine-
tuned by a sterically restricted environment provided by the
ribose moiety to ensure specific target interactions, as previously
suggested.15 To evaluate the reactivity of showdomycin, we
incubated S. aureus proteome lysates with an N-labeled
fluorescent maleimide, unhindered at the site of attack, and the
showdomycin probe to directly compare the labeling pattern
These initial experiments provided evidence that the show-
domycin probe is in most cases a selective chemical tool to
label distinct enzymes in complex proteomes. We next tested
whether the compound is cell permeable and could label
enzymes in living pathogens such as S. aureus, multiresistant
S. aureus (MRSA), P. aeruginosa, and L. monocytogenes strains
under in situ conditions. After cell lysis, an azide-rhodamine
tag was appended to the labeled proteins of the cytosolic
proteome fraction via CC prior to fluorescent SDS-PAGE
analysis (Figure 2A). Gel bands of labeled proteins were
detected by fluorescence scanning and identified by MS.
Interestingly, several distinct labeling events were observed,
indicating that showdomycin is capable of entering the cell.
Some of the targeted proteins were of low abundance, as shown
by direct comparison of the relative intensities observed with
Coomassie staining vs fluorescence scanning (Figure 2A). The
optimal concentration for target labeling was determined by
variation of probe concentrations in several bacterial species
(Figure 2, Supporting Information Figure S4). It turned out that
50 µM was the optimal concentration for the full saturation of
the majority of targets. Selected enzymes displayed a very sensitive
labeling with the showdomycin probe down to 1.3 µM (Figure
2B). Preincubation of proteomes with 1 mM to 100 µM unmodified
showdomycin and subsequent addition of 10 µM probe resulted
in a total loss of fluorescent signals, indicating that showdomycin
and the corresponding probe compete for the same labeling sites
(Supporting Information Figure S3). Additionally, various
incubation times revealed that the probe very rapidly entered
(12) Barrett, A. G. M.; Broughton, H. B. J. Org. Chem. 1986, 51, 495–
503.
(13) Rostovtsev, V. V.; Green, J. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596–2599.
(14) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67,
3057–3064.
(15) Uehara, Y.; Fisher, J. M.; Rabinovitz, M. Biochem. Pharmacol. 1980,
29, 2199–2204.
9
J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010 6965

A R T I C L E S
Bo¨ttcher and Sieber
Figure 2. Labeling of the proteomes of bacterial pathogens by the showdomycin probe. (A) Comparison of the fluorescent gels (Fl) and Coomassie abundance
staining (Co) by in situ labeling of enzymes in three different bacterial strains. Enzyme identities are assigned to the corresponding gel band (for a list of
abbreviations, please refer to Table 1). (B) The probe exhibits remarkable sensitivity for selected targets, such as MurA1 that is labeled down to 1.3 µM. (C)
Examples of recombinantly expressed enzymes (-, before induction; +, after induction; P, native proteome).
the cell and achieved full saturation of targets within 7.5 min
(Supporting Information Figure S4).
Taking all bacterial pathogens together, we were able to
observe the labeling of about 13 different enzymes by MS (Table
1). MS results were confirmed by recombinant expression of
all major hits and subsequent labeling by the corresponding
probe (Figure 2C). The identified enzymes belong to two major
classes comprising oxidoreductases and transferases, and inter-
estingly several of them recognize substrates that contain
nucleosides, e.g., uridine (Table 1). Many of the corresponding
enzyme families require a nucleophilic cysteine residue in their
active site for catalysis, which is likely to attack the electrophilic
9
6966 J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010

Showdomycin as a Chemical Probe for Pathogenic Bacteria
A R T I C L E S
Table 1. Enzymes of Different Functions and Enzyme Classes (EC) That Were Identified in Pathogenic and Non-pathogenic Strains
function enzyme class
organismsa
oxidative-stress resistance alkyl hydroperoxide reductase subunit C (AhpC)
EC 1.11.1 S. aureus NCTC 8325, Mu50, P. putida, P.
aeruginosa
alkyl hydroperoxide reductase subunit F (AhpF)
thioredoxine reductase (TrxB)
phosphoenolpyruvate-protein phosphotransferase (PtsI)
acetolactate synthase (AlsS)
EC 1.6.4
EC 1.8.1
EC 2.7.3
EC 2.2.1
EC 1.1.1
S. aureus NCTC 8325, Mu50
S. aureus NCTC 8325, Mu50
S. aureus Mu50, L. monocytogenes EGD-e
S. aureus NCTC 8325
S. aureus NCTC 8325, Mu50, L.
monocytogenes EGD-e
virulence-associated
nucleotide biosynthesis
inosine-5′-monophosphate dehydrogenase (IMPDH)
ribonucleotide-diphosphate reductase subunit alpha (RNR)
phosphomethyl-pyrimidine kinase (ThiD)
formate acetyltransferase (FAT)
EC 1.17.4 L. monocytogenes EGD-e, P. aeruginosa
primary metabolism
EC 2.7.4
EC 2.3.1
EC 2.8.1
S. aureus NCTC 8325
S. aureus NCTC 8325, Mu50
P. aeruginosa
cysteine desulfurase (CDS)
resistance/cell-wall
biosynthesis
UDP-N-acetylglucosamine 1-carboxyvinyltransferase 1 (MurA1) EC 2.5.1
S. aureus NCTC 8325, Mu50
UDP-N-acetylglucosamine 1-carboxyvinyltransferase 2 (MurA2) EC 2.5.1
UDP-N-acetyl-glucosamine pyrophosphorylase, putative (GlmU) EC 2.7.7
S. aureus Mu50
S. aureus NCTC 8325
a As identified by mass spectrometry; those identified by labeling profile are not given.
medicinal interest (Supporting Information Table 2), including
enzymes that are crucial for oxidative stress resistance in S.
aureus, like alkylhydroperoxide reductase C (AhpC), alkylhy-
droperoxide reductase F (AhpF), and thioredoxin reductase
(TrxB). These enzymes have fundamental functions for colo-
nization and survival of the pathogen in the host.16 Inosine-5′-
monophosphate dehydrogenase (IMPDH) is required for the
biosynthesis of GMP and thus for the cells’ guanosine phosphate
pool.17 Besides its role for the synthesis of branched amino
acids, acetolactate synthase (AlsS) is important for the virulence
of S. aureus, as genetic knockouts resulted in reduced biofilm
formation.18 The UDP-N-acetylglucosamine 1-carboxyvinyl-
transferase 1 (MurA1) catalyzes the first step in cell wall
biosynthesis and is thus an essential bacterial enzyme with major
importance as an antibiotic drug target.19
One of the major challenges in the treatment of bacterial
infections is rapid detection of the disease-causing pathogen.
In these instances, suitable biomarkers are required that provide
a unique readout which is characteristic for the bacterial strain.
Our comparison of enzyme-labeling pattern between different
resistant and nonresistant strains and between pathogenic and
nonpathogenic bacteria (Listeria, Pseudomonas, and Staphylo-
coccus) with the showdomycin probe provided several of these
unique signatures (Figure 4 and Supporting Information Figure
S7). For instance, the comparative analysis of four different
strains of S. aureussthe reference strain S. aureus NCTC 8325,
the highly toxin-producing clinical isolate DSM 19041, and the
MRSA strains Mu50 and DSM 18827 (clinical isolate)srevealed
that phosphoenolpyruvate-protein phosphotransferase (PtsI) was
labeled only in the MRSA strains and the toxin-producing strain,
but not in the laboratory strain NCTC 8325 (Figure 4). PtsI
was found to be important for the virulence of S. aureus, S.
typhimurium, and H. influenzae, as shown by highly attenuated
pathogenicity of ∆ptsI mutants of these species in a mouse
model.20 This may indicate that the more pathogenic strains and
Figure 3. Identification of the binding sites of showdomycin by mass
spectrometry (MS). Peptide masses (MH+) and p-values are given for each
peptide containing a cystein residue. Only Cys39 and Cys168 were found
to be modified by showdomycin. A MS fragmentation pattern is given below
for the chymotryptic peptide of Cys39.
maleimide of showdomycin. Indeed, a substrate inhibition assay
with MurA1 demonstrated that showdomycin mediated inhibi-
tion of enzyme activity with an IC50 of 10 µM (Supporting
Information Figure S5). Moreover, we analyzed the site of
modification in the case of one representative reductase (AhpC,
see below) and demonstrated that showdomycin attached
selectively to two cysteine residues (Figure 3, Supporting
Information Figure S6). The active-site inhibition and great
coverage of mechanistically distinct enzyme families that have
not been addressed by ABPP before emphasizes the unprec-
edented utility of showdomycin as a new proteomic tool for
the identification of pathogenesis-associated enzyme targets.
(16) Cosgrove, K.; Coutts, G.; Jonsson, I. M.; Tarkowski, A.; Kokai-Kun,
J. F.; Mond, J. J.; Foster, S. J. J. Bacteriol. 2007, 189, 1025–1035.
(17) Shu, Q.; Nair, V. Med. Res. ReV. 2008, 28, 219–232.
(18) Cassat, J.; Dunman, P. M.; Murphy, E.; Projan, S. J.; Beenken, K. E.;
Palm, K. J.; Yang, S. J.; Rice, K. C.; Bayles, K. W.; Smeltzer, M. S.
Microbiology 2006, 152, 3075–3090.
(19) Du, W.; Brown, J. R.; Sylvester, D. R.; Huang, J.; Chalker, A. F.; So,
C. Y.; Holmes, D. J.; Payne, D. J.; Wallis, N. G. J. Bacteriol. 2000,
182, 4146–4152.
Interestingly, a large fraction of the identified enzymes are
involved in important cellular functions such as cell wall
biosynthesis, oxidative stress resistance, nucleotide biosynthesis,
and amino acid biosynthesis. Several targets are of special
(20) Kok, M.; Bron, G.; Erni, B.; Mukhija, S. Microbiology 2003, 149
(Pt. 9), 2645–2652.
9
J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010 6967

A R T I C L E S
Bo¨ttcher and Sieber
e.g., by the natural product antibiotic fosfomycin.19 Interestingly,
in our comparative experiment, only the MRSA strain Mu50
maintains two active MurA enzymes, although all S. aureus
strains share the genetic information for both enzymes. This
observation raised the question whether the lack of MurA2
labeling in the nonresistant strains is a consequence of expres-
sional regulation or posttranslational activation. In case of
differences in expressional regulation, quantitative real-time
polymerase chain reaction (RT PCR) should reveal the amount
of mRNA in different strains and allow a direct comparison.
We quantified the relative gene expression levels of murA1 and
murA2 in the four strains by RT PCR. The ratio of murA1/
murA2 was found to be 0.93 ( 0.01 for S. aureus NCTC 8325,
0.92 ( 0.01 for S. aureus Mu50, 0.91 ( 0.01 for S. aureus
DSM 18827, and 0.90 ( 0.02 for S. aureus DSM 19041
(Supporting Information Table 4). Thus, the expressional levels
of both murA1 and murA2 are comparable in all strains,
indicating that the difference in the in situ labeling is due to
posttranslational regulation of enzyme activity. In fact, regulation
of peptidoglycan biosynthesis by MurA inhibition with UDP-
MurNAc was already observed previously.22
Our chemical and genetic functional analysis of the MurA’s
therefore shows that two constitutively active copies of the
MurA enzyme in the MRSA strain Mu50 may contribute to
the antibiotic resistance of this strain. Indeed, only Mu50 was
resistant against fosfomycin treatment at concentrations up to
500 µM, while all other strains exhibited IC50 values between
1 and 5 µM of fosfomycin (Supporting Information Figure S9).
It should be noted that the S. aureus strains were indistinguish-
able on a proteome level by Coomassie staining, which again
illustrates the value of specific enzyme labeling by the activity-
based probe (Supporting Information Figure S10). Since show-
domycin targets both enzymes, its structure could represent a
starting point for customized selective MurA1 and 2 inhibitors.
Figure 4. Comparison of the in situ labeling profile of four different S.
aureus strains. The probe concentration was 50 µM each. Enzyme identities
are assigned to the corresponding gel band (for a list of abbreviations, please
refer to Table 1).
clinical isolates constitutively exhibit higher PtsI activity, even
under culture conditions, compared with the less infective
laboratory reference strain. Vice versa, the peroxidase AhpC
was found to be highly active in NCTC 8325 but not in the
other resistant or toxic strains. Interestingly, bioinformatic gene
analysis revealed that even species and strains in which less or
no labeling of a certain protein could be detected carried a copy
of the gene in their genome that was highly similar to or even
identical with the one that was labeled in the reference strain
or species (Supporting Information Table 3). This indicates that
regulatory processes at expressional or postranslational levels
are the reason for the observed differences in labeling. Further-
more, the comparison of S. aureus strains revealed a very distinct
labeling of MurA2 which occurred only at the MRSA strain
Mu50 and only in situ but not in Vitro (Supporting Information
Figure S8). It has been proposed that MurA2 arose from gene
duplication of murA1 in Gram-positive bacteria. The two
different copies of MurA are functionally identical and can
replace each other when one is inhibited.19 The two MurA
homologues share low sequence identity at the amino acid level.
Differences in the active site have made it difficult, so far, to
develop MurA-specific antibiotics that effectively inhibit both
enzymes.21 It was recently shown that mutagenic disruption of
either murA1 or murA2 did not significantly alter cell growth,
while cells could not survive the removal of both genes. It can
therefore be speculated that MurA2 represents an essential key
as a resistance bypass strategy in the case of MurA1 inhibition,
Conclusion
In conclusion, we were able to establish an ABPP probe based
on the reactive nucleoside antibiotic showdomycin and validate
it as a tool for labeling and identification of enzymes in
pathogenic bacteria. Some of these enzymes indeed had uridine
or its derivatives as natural substrates, as the close structural
proximity to uridine suggested. Many of the identified enzymes
were important for either the viability or virulence of the
bacteria. The sum of all essential enzymes targeted by show-
domycin, among them especially MurA, which is one of the
most intense targets in S. aureus, is likely to explain the
antibiotic activity of showdomycin and emphasizes the utility
of profiling natural products to reveal their molecular targets.
Moreover, comparing enzyme activity profiles of different
pathogenic strains and between closely related pathogenic and
nonpathogenic species provides insights into regulatory differ-
ences of these strains as well as unique strain-specific signatures.
As demonstrated in this study, the showdomycin probe has a
great value as an analytical probe to increase the repertoire of
the ABPP toolbox and provides a simple means to study the
biological targets of a promising natural product.
Experimental Section
General. All chemicals were of reagent grade or better and used
without further purification. Chemicals and solvents were purchased
(21) Kedar, G. C.; Brown-Driver, V.; Reyes, D. R.; Hilgers, M. T.; Stidham,
M. A.; Shaw, K. J.; Finn, J.; Haselbeck, R. J. Antimicrob. Agents
Chemother. 2008, 52, 2009–2013.
(22) Mizyed, S.; Oddone, A.; Byczynski, B.; Hughes, D. W.; Berti, P. J.
Biochemistry 2005, 44, 4011–4017.
9
6968 J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010

Showdomycin as a Chemical Probe for Pathogenic Bacteria
A R T I C L E S
from Sigma Aldrich or Acros Organics. For all reactions, only
commercially available solvents of purissimum grade, dried over
molecular sieves and stored under an argon atmosphere, were used.
Solvents for chromatography and workup purposes were generally
of reagent grade and purified before use by distillation. In all
reactions, temperatures were measured externally. If not indicated
otherwise, column chromatography was performed on Merck silica
from pure ethyl actetate to ethyl acetate/(acetonitrile/methanol) 3:1.
Evaporation of the solvent yielded 1.20 g (35%) of 2 as a pale red
solid.
1H NMR (200 MHz, D2O): δ 6.59 (dt, J ) 8.4, 2.3 Hz, 1 H,
CdCH), 4.44 (dd, J ) 8.4, 4.4 Hz, 1 H, CHsCHdC), 3.69-3.60
(m, 2 H), 3.60-3.52 (m, 2 H), 3.36 (d, J ) 2.4 Hz, 2 H,
CH2sC(O)).
1
13C NMR (100 MHz, D2O): δ 178.9, 173.1, 134.4, 129.8, 73.5,
71.7, 69.6, 62.7, 33.2.
ESI-MS, positive mode (m/z): 232.0588 [M + H]+, calcd
232.0816; 249.1052 [M + NH4]+, calcd 249.1081. Negative mode
(m/z): 230.0678 [M - H]-, calcd 230.0670; 276.0739 [M + FA -
H]-, calcd 276.0725; 461.1446 [2M - H]-, calcd 461.1413.
Synthesis of Showdomycin and 1′-epi-Showdomycin (3a,b).
A stirred suspension of 2 (462 mg, 2.0 mmol) in 28 mL of dry
acetonitrile was heated to 65 °C, and phenylselenyl chloride (400
mg, 2.1 mmol) was added. The reaction was stirred for 29 h at 65
°C, allowed to cool to room temperature, and treated with 10 mL
of a 10% aqueous H2O2 solution. The reaction was monitored by
TLC (ethyl acetate/acetonitrile/methanol 15:4:1) and stirred until
all selenium-containing material was consumed, as visualized by
1% PdCl2 in 0.1 M HCl as TLC staining reagent. The solvent was
evaporated in Vacuo, and the products were purified by chroma-
tography on silica gel (ethyl actetate/ethanol 19:1, Rf ) 0.25). This
yielded 230 mg (50%) as a mixture of the diastereoisomers 3a and
3b. Purification by preparative HPLC and lyophilization gave the
single isomers as white solids. The diastereoisomers were identified
using two-dimensional NMR spectra.
gel (Acros Organics, 0.035-0.070 mm, mesh 60 Å). H NMR
spectra were recorded on a Varian Mercury 200 (200 MHz), a
Varian NMR-System 600 (600 MHz), or a Varian NMR-System
300 (300 MHz) spectrometer, and 13C NMR spectra were measured
with a Varian NMR-System 600 (600 MHz) or a Varian NMR-
System 300 (300 MHz) spectrometer, referenced to the residual
proton and carbon signal of the deuterated solvent, respectively.
For DEI measurements, samples were directly desorbed from
platinum wire (20-1600 °C, 120 °C min-1) and delivered to a
Finnigan MAT 95 mass spectrometer in EI mode (70 eV, 250 °C
source). ESI mass spectra were recorded with a Thermo Finnigan
LTQ FT spectrometer. HPLC analysis was accomplished with a
Waters 2695 separations module, for analytical runs with an
X-Bridge BEH130 C18 column (4.6 × 100 mm) or an X-Bridge
BEH130 PREP C18 column (10 × 150 nm) for preparative
applications, and a Waters 2996 PDA detector.
Strategy toward Synthesis of Showdomycin and Its ABPP
Probe. The synthesis of showdomycin was adapted from a strategy
developed by Barrett et al.23 Here, the maleimide moiety of
showdomycin is introduced via a Wittig reaction of the correspond-
ing triphenylphosphoranylidene succinimide with D-(-)-ribose.
Addition of phenylselenyl chloride triggers ring closure, and
subsequent selenoxide elimination restores the functional maleimide,
giving showdomycin and 1′-epi-showdomycin (Figure S1, Sup-
porting Information). These were coupled via the free primary
hydroxyl group of the ribose to DIC/HOBt-activated hex-5-ynoic
acid to give the ABPP probe 5′-O-hex-5-ynoylshowdomycin and
the corresponding 1′-epi-diastereoisomer. Separation with HPLC
of the last two reactions finally gave the pure products, showdo-
mycin and 5′-O-hex-5-ynoylshowdomycin, respectively.
Showdomycin (3a). HPLC analysis: mobile phase (HPLC grade)
A ) water, B ) acetonitrile. Gradient: T0, B ) 0%; T9, A ) 3%.
Retention time ) 5.1 min.
1H NMR (400 MHz, acetone-d6): δ 6.72 (d, J ) 1.7 Hz, 1 H,
CdCH), 4.71 (dd, J ) 4.8, 1.6 Hz, 1 H, 1′-H), 4.44 (br, 1 H, OH),
4.20 (dd, J ) 4.9, 4.9 Hz, 1 H, 2′-H), 4.16 (dd, J ) 5.1, 5.1 Hz, 1
H, 3′-H), 4.09 (br, 1 H, OH), 3.95 (td, J ) 5.1, 3.3 Hz, 1 H, 4′-H),
3.89 (app t, 1 H, 5′-OH), 3.81-3.76 (m, 1 H, CH(H)), 3.68-3.62
(m, 1 H, CH(H)).
Synthesis of Triphenylphosphoranylidene Succinimide (1).
Triphenylphosphine (13.0 g, 50 mmol) was added to maleimide
(5.0 g, 51 mmol) in 100 mL of glacial acetic acid and refluxed at
100 °C. After 30 min the reaction was complete, as monitored by
TLC (isohexane/ethyl acetate 1:1). Diethyl ether was added to the
reaction mixture until a pale red precipitate was formed. This was
filtered off, washed with diethyl ether, and dried under vacuum to
yield 12.89 g (72%) of 1 as an amorphous colorless solid.
1H NMR (300 MHz, CDCl3): δ 7.64-7.61 (m, 9 H, Phe),
7.54-7.51 (m, 6 H, Phe), 3.03 (s, 2 H, CH2).
13C NMR (151 MHz, acetone-d6): δ 171.3, 170.6, 148.7, 129.1,
84.7, 78.0, 75.4, 71.4, 61.7.
ESI-MS, positive mode (m/z): 229.9854 [M + H]+, calcd
230.0659; 247.0896 [M + NH4]+, calcd 247.0924. Negative mode
(m/z): 228.0516 [M - H]-, calcd 228.0513; 264.0283 [M + Cl]-,
calcd 264.0280; 274.0572 [M + FA - H]-, calcd 274.0568;
457.1107 [2M - H]-, calcd 457.1099.
1′-epi-Showdomycin (3b). HPLC analysis: mobile phase (HPLC
grade) A ) water, B ) acetonitrile. Gradient: T0, B ) 0%; T9, A
) 3%. Retention time ) 4.5 min.
13C NMR (150 MHz, CDCl3): δ 177.9 (JC-P ) 16.8 Hz, CdO),
171.0 (JC-P ) 15.1 Hz, CdO), 133.6 (JC-P ) 10.5 Hz, Phe), 133.0
1H NMR (400 MHz, acetone-d6): δ 6.48 (d, J ) 2.1 Hz, 1 H,
CdCH), 4.94 (dd, J ) 3.7, 2.1 Hz, 1 H, 1′-H), 4.40-4.33 (m, 2 H,
2′-H and 3′-H), 4.24 (d, J ) 3.1 Hz, 1 H, OH), 4.12 (br, 1 H, OH),
3.94 (ddd, J ) 7.3, 4.1, 2.8 Hz, 1 H, 4′-H), 3.81 (ddd, J ) 11.5,
4.6, 2.9 Hz, 1 H, CH(H)), 3.73 (app t, 1 H, 5′-OH), 3.63 (ddd, J )
11.6, 6.5, 4.1 Hz, 1 H, CH(H)).
(JC-P ) 2.2 Hz, Phe), 129.4 (JC-P ) 12.6 Hz, Phe), 125.6 (JC-P
)
92.3 Hz, Phe), 38.5 (JC-P ) 10.2 Hz, CH2), 36.9 (JC-P ) 136.8
Hz, PdC).
DEI-MS (m/z): 258.0932 [M - H]-, calcd 358.0997.
Synthesis of (E)-3-(2,3,4,5-Tetrahydroxypentylidene)pyrrolidine-
2,5-dione (2). D-(-)-Ribose (2.3 g, 15 mmol) was added to a stirred
suspension of 1 (10.8 g, 30 mmol) in 75 mL of dry THF and
refluxed at 80 °C under nitrogen. The reaction was complete after
210 h, as monitored by TLC (chloroform/methanol 9:1). After the
solvent was evaporated in Vacuo, the crude product was partitioned
between 75 mL of water and 75 mL of dichloromethane. The
aqueous phase was collected and concentrated under reduced
pressure. Chromatography on a mixture of C18 reversed-phase
material and silica gel (100:1 w/w) with water as mobile phase
yielded a rapidly eluting crude product that was dried in Vacuo
and purified by chromatography on silica gel with ethyl acetate
and a gradient of an acetonitrile/methanol (4:1) mixture, reached
13C NMR (151 MHz, acetone-d6): δ 170.9, 148.7, 128.3, 82.8,
76.7, 72.8, 71.5, 61.8.
ESI-MS, positive mode (m/z): 229.9854 [M + H]+, calcd
230.0659; 247.0896 [M + NH4]+, calcd 247.0924. Negative mode
(m/z): 228.0516 [M - H]-, calcd 228.0513; 264.0283 [M + Cl]-,
calcd 264.0280; 274.0572 [M + FA - H]-, calcd 274.0568;
457.1107 [2M - H]-, calcd 457.1099.
Synthesis of 5′-O-Hex-5-ynoylshowdomycin and 5′-O-
Hex-5-ynoyl-1′-epi-showdomycin (4a,b). HOBT (40.5 mg, 0.30
mmol) was dissolved in 800 µL of DMF, and hex-5-ynoic acid
(33.1 µL, 0.30 mmol) and 1,3-diisopropylcarbodiimide (46,4 µL,
0.30 mmol) were added. The reaction mixture was stirred for 10
min at room temperature. The mixture of the diastereoisomers 3a
and 3b (40 mg, 0.17 mmol) was then added in 200 µL of DMF
and the reaction stirred for 12 h. The solvent was evaporated in
(23) Barrett, A. G. M.; Broughton, H. B.; Attwood, S. V.; Gunatilaka,
A. A. L. J. Org. Chem. 1986, 51, 495–503.
9
J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010 6969

A R T I C L E S
Bo¨ttcher and Sieber
Vacuo and the residue purified via HPLC, giving the showdomycin
probe and the 1′-epi-showdomycin probe.
addition of 50 µM probe and allowed to incubate for 30 min at
room temperature. Comparative experiments with maleimide were
carried out by incubation of the proteome with 50 µM Atto 550-
labeled maleimide (Sigma Aldrich) for 30 min at room temperature.
For heat controls the proteome was denatured with 2% SDS (4 µL
of 21.5% SDS) at 95 °C for 6 min and cooled to room temperature
before the probe was applied. Following incubation, reporter-tagged
azide reagents (13 µM rhodamine-azide (1 µL) for analytical or 20
µM rhodamine-biotin-azide for preparative scale) were added,
followed by 1 mM TCEP (1 µL) and 100 µM ligand (3 µL).
Samples were gently vortexed, and the cycloaddition was initiated
by the addition of 1 mM CuSO4 (1 µL). The reactions were
incubated at room temperature for 1 h.2,25 For analytical gel
electrophoresis, 50 µL of 2× SDS loading buffer was added and
50 µL applied on the gel. Fluorescence was recorded in a Fujifilm
Las-4000 luminescent image analyzer with a Fujinon VRF43LMD3
lens and a 575DF20 filter.
In Situ Experiments. For analytical and preparative in situ
studies, bacteria were grown in BHB or LB medium and harvested
1 h after reaching stationary phase by centrifugation. Depending
on the cell density at stationary phase, 1 mL of L. monocytogenes
and P. aeruginosa strains and 0.5 mL of S. aureus strains were
harvested for analytical and 5 and 2.5 mL for preparative studies,
respectively. After washing with PBS, the cells were resuspended
in 100 and 500 µL of PBS for analytical and preparative experi-
ments. Unless indicated otherwise, bacteria were incubated for 30
min with varying concentrations of probe at room temperature for
analytical studies and with a fixed concentration of 100 µM probe
for preparative studies. Subsequently, the cells were washed twice
with PBS and lysed by sonication in 100 and 500 µL of PBS. The
proteomes were separated into cytosolic and membrane fractions,
followed by CC as described above. Reactions for enrichment were
carried out together with a control lacking the probe to compare
the results of the biotin-avidin-enriched samples with the back-
ground of unspecific protein binding on avidin-agarose beads. After
CC, proteins were precipitated using an equal volume of prechilled
acetone. Samples were stored on ice for 20 min and centrifuged at
13 000 rpm for 10 min. The supernatant was discarded, and the
pellet was washed two times with 200 µL of prechilled methanol
and resuspended by sonication. Subsequently, the pellet was
dissolved in 1 mL of PBS with 0.4% SDS by sonication and
incubated under gentle mixing with 50 µL of avidin-agarose beads
(Sigma-Aldrich) for 1 h at room temperature. The beads were
washed three times with 1 mL of PBS/0.4% SDS, twice with 1
mL of 6 M urea, and three times with 1 mL of PBS. Aliquots of
50 µL of 2× SDS loading buffer were added and the proteins
released for preparative SDS-PAGE by 6 min incubation at 95
°C. Gel bands were isolated, washed, and tryptically digested as
described previously.26
5′-O-Hex-5-ynoylshowdomycin (4a). HPLC analysis: mobile
phase (HPLC grade) A ) water, B ) acetonitrile. Gradient: T0, B
) 0%; T40, A) 40%. Retention time ) 20.8 min.
1H NMR (400 MHz, acetone-d6): δ 6.60 (d, J ) 1.8 Hz, 1 H,
CdCH), 4.75 (dd, J ) 3.7, 1.8 Hz, 1 H, 1′-H), 4.38-4.34 (dd, J )
11.9, 2.5 Hz, 1 H, 5′-H), 4.27-4.23 (m, 2 H, 2′-H and 5′-H),
4.13-4.08 (m, 2 H, 3′-H and 4′-H), 2.50 (t, J ) 7.4 Hz, 2 H,
C(O)sCH2), 2.36 (t, J ) 2.7 Hz, 1 H, CtCH), 2.26 (dt, J ) 7.1,
2.7 Hz, 2 H, CH2sCtCH), 1.82 (Ψ-quint, J ) 7.4 Hz, 2 H,
C(O)sCH2sCH2).
ESI-MS, negative mode (m/z): 322.0944 [M - H]-, calcd
322.0932; 358.0730 [M + Cl]-, calcd 358.0699; 645.2015 [2M -
H]-, calcd 645.1937.
5′-O-Hex-5-ynoyl-1′-epi-showdomycin (4b). HPLC analysis:
mobile phase (HPLC grade) A ) water, B ) acetonitrile. Gradient:
T0, B ) 0%; T40, A ) 40%. Retention time ) 21.9 min.
1H NMR (400 MHz, acetone-d6): δ 6.38 (d, J ) 2.11 Hz, 1 H,
CdCH), 4.88 (dd, J ) 3.6, 2.1 Hz, 1 H, 1′-H), 4.30-4.26 (m, 2 H,
2′-H and 5′-H), 4.19 (dd, J ) 8.2, 4.5 Hz, 1 H, 3′-H), 4.08 (dd, J
) 11.8, 5.5 Hz, 1 H, 5′-H), 4.01 (ddd, J ) 8.2, 5.5, 2.7 Hz, 1 H,
4′-H), 2.39 (t, J ) 7.3 Hz, 2 H, C(O)sCH2), 2.25 (t, J ) 2.7 Hz,
1 H, CtCH), 2.16 (dt, J ) 7.1, 2.7 Hz, 2 H, CH2sCtCH), 1.71
(Ψ-quint, J ) 7.8 Hz, 2 H, C(O)sCH2sCH2).
13C NMR (151 MHz, acetone-d6): δ 172.2, 170.8, 148.2, 128.4,
83.1, 79.6, 76.7, 73.4, 72.4, 69.6, 63.9, 32.3, 23.7, 17.2.
ESI-MS, positive mode (m/z): 664.2332 [2M + NH4]+, calcd
664.2348. Negative mode (m/z): 322.0942 [M - H]-, calcd
322.0932; 368.1009 [M + FA - H]-, calcd 368.0987; 645.2011
[2M - H]-, calcd 645.1937.
Bacterial Strains. Staphylococcus aureus strains NCTC 8325
(Institute Pasteur, France), Mu50/ATCC 700699 (Institute Pasteur,
France), DSM 19041 (DSMZ, Germany), and DSM 18827 (DSMZ,
Germany) were maintained in brain-heart broth (BHB) medium
at 37 °C. DSM 19041 is a highly toxin-producing isolate from a
human furuncle, and the MRSA strain DSM 18827 was isolated
from tracheal secretion of a clinical case.24 Listeria monocytogenes
strains EGD-e (Institute Pasteur, France) and F2365 (BCCM/LMG
Bacteria Collection, Belgium) as well as the nonpathogenic strain
Listeria welshimeri SLCC 5334 serovar 6b (DSMZ, Germany) were
maintained in BHB medium, while Pseudomonas aeruginosa PAO1
(Institute Pasteur, France) and Pseudomonas putida KT2440
(ATCC, USA) were maintained in Luria-Bertani (LB) medium.
All strains were grown at 37 °C.
MIC Measurements. Overnight cultures of S. aureus (NCTC
8325) were diluted in fresh BHB medium to OD600 ) 0.01, and
100 µL aliquots were incubated in Nunclon round-bottom 96-well
plates with 1 µL of the corresponding DMSO stock of showdomycin
and the showdomycin probe at varying concentrations. The samples
were incubated overnight at 37 °C and the optical densities obtained
for MIC calculation. All experiments were conducted at least in
triplicate, and DMSO served as control.
Preparation of Proteome for in Vitro Experiments. The
proteome of the bacterial strains Staphylococcus aureus NCTC 8325
was prepared from 1 L liquid cultures in BHB medium, harvested
1 h after transition in the stationary phase by centrifugation at 9000
rpm. The bacterial cell pellet was washed with PBS, resuspended
in PBS, and lysed by sonication with a Bandelin Sonopuls
instrument under ice cooling. Membrane and cytosol were separated
by centrifugation at 9000 rpm for 45 min.
Mass Spectrometry and Bioinformatics. Tryptic peptides were
loaded onto a Dionex C18 Nano Trap column (100 µm) and
subsequently eluted and separated by a Dionex C18 PepMap 100
(3 µm) column for analysis by tandem MS, followed by high-
resolution MS using a coupled Dionex Ultimate 3000 LC-
ThermoFinnegan LTQ-FT MS system. The mass spectrometry data
were searched using the SEQUEST algorithm against the corre-
sponding databases via the software “bioworks”. The search was
limited to only tryptic peptides, two missed cleavage sites,
monoisotopic precursor ions, and a peptide tolerance of <10 ppm.
Filters were set to further refine the search results. The Xcorr vs
charge state filter was set to Xcorr values of 1.5, 2.0, and 2.5 for
charge states 1+, 2+, and 3+, respectively. The number of different
peptides has to be g2, and the peptide probability filter was set to
<0.001. These filter values are similar to others previously reported
In Vitro Labeling Experiments. Proteome samples were
adjusted to a final concentration of 1 mg protein/mL by dilution in
PBS prior to probe labeling. Analytical experiments were carried
out in 43 µL total volume, such that once CC reagents were added,
the total reaction volume was 50 µL. Reactions were initiated by
(25) Speers, A. E.; Adam, G. C.; Cravatt, B. F. J. Am. Chem. Soc. 2003,
125, 4686–4687.
(26) Sieber, S. A.; Niessen, S.; Hoover, H. S.; Cravatt, B. F. Nat. Chem.
Biol. 2006, 2, 274–281.
(24) Gruner, B. M.; Han, S. R.; Meyer, H. G.; Wulf, U.; Bhakdi, S.; Siegel,
E. K. J. Clin. Microbiol. 2007, 45, 2684–2685.
9
6970 J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010

Showdomycin as a Chemical Probe for Pathogenic Bacteria
A R T I C L E S
for SEQUEST analysis.27 Minimum P values and maximum Xcorr
values of each run as well as the total number of obtained peptides
are reported in Table S1 (Supporting Information).
PCR products were identified on agarose gels, and gel bands
were isolated and extracted with an E.Z.N.A. MicroElute gel
extraction kit. Concentrations of DNA were measured by a
NanoDrop ND-1000 spectrophotometer. Samples of 100 fmol of
purified attB-PCR product and 50 fmol of attP-containing donor
vector pDONR201 in TE buffer were used for in Vitro BP
recombination reaction with BP Clonase II enzyme mix to yield
the appropriate attL-containing entry clone. After transformation
in chemically competent One Shot TOP10 E. coli (Invitrogen), cells
were plated on LB agar plates containing 25 µg mL-1 kanamycin.
Clones of transformed cells were selected and grown in LB medium
supplemented with kanamycin. Cells were harvested, and plasmids
were isolated using an E.Z.N.A. Plasmid mini kit. The correspond-
ing attB-containing expression clone was generated by in Vitro LR
recombination reaction of approximately 50 fmol of the attL-
containing entry clone and 50 fmol of the attR-containing destina-
tion vector pDest using LR Clonase II enzyme mix in TE buffer.
The expression clone was transformed in chemically competent
BL21 E. coli cells (Novagen) and selected on LB agar plates
containing 100 µg mL-1 carbenicillin. Validity of the clones was
confirmed by plasmid sequence analysis. Recombinant clones were
grown in carbenicillin LB medium, and target gene expression was
induced with anhydrotetracyclin.
Labeled Proteomes in Vitro. Cytosol and membrane fractions
of the proteome were labeled separately. Membrane fractions were
comparable with the cytosol in terms of intensive and specific
labeling pattern, with slight differences in some enzymes. Examples
for heat-denatured proteome controls are given in Figure S2
(Supporting Information). The majority of the labeled part of the
proteome appears to be heat sensitive and thus can be regarded as
the result of specific activity-based binding events. Labeling profiles
for both probes, O′-hex-5′-ynoylshowdomycin and O′-hex-5′-ynoyl-
1′-epi-showdomycin, are shown in Figure S10 (Supporting Informa-
tion). Surprisingly, there are no major differences in the labeling
profiles of the two isomers, although the showdomycin probe labels
more intensively than the 1′-epi-showdomycin probe.
MurA1 Inhibition Assay. The inhibition of MurA1 by show-
domycin was investigated by an assay system as described
previously.19,28 Activity was measured as the release of Pi from
the MurA1-catalyzed enolpyruvyl transfer reaction with phospho-
enolpyruvate (PEP) and UDP-N-acetylglucosamine (UDPAG) in a
malachite green-molybdate assay. MurA1 from S. aureus NCTC
8325 was recombinantly expressed with a Strep-Tag in E. coli BL21
cells and purified by affinity chromatography using a StrepTrap
HP column. The pure protein was eluted in elution buffer containing
100 mM Tris pH 8.0, 150 mM NaCl, 1 mM EDTA, 1 mM DTT,
and 2.5 mM desthiobiotin. Before use, MurA1 was conditioned
for at least 10 min with 1 mM UDPAG on ice. In a typical assay,
to 5 µL of 500 mM HEPES pH 7.5 were added sequentially 36 µL
of H2O, 2.5 µL of 40 mM UDPAG, 5 µL of MurA1 (17 µg), and
1 µL of showdomycin stock or DMSO. After 10 min incubation at
room temperature, 50 µL of 400 µM PEP in 50 mM HEPES pH
7.5 was added to start the reaction. After incubation at room
temperature for an appropriate time, 20 µL of the mixture was
withdrawn and mixed in a round-bottom 96-well plate with 80 µL
of a 3:1 mixture of 0.45% malachite green and 4.2% ammonium
heptamolybdate in 4 N HCl. The absorption was read with a
TECAN GENios Pro plate reader at 670 nm, and IC50 values were
calculated from curve fittings by Microcal Origin 6.0 (Figure S5,
Supporting Information).
Comparison of labeling with genetic properties and characteristics
of the different species and strains was conducted by BLAST search
for homologous sequences in the strains. These were used for a
more accurate alignment to identify sequence homology via amino
acid identities in the sequence by the SIM alignment tool of the
Expert Protein Analysis System (ExPASy, Swiss Institute of
Bioinformatics). The alignment was computed by the BLOSUM62
comparison matrix, with settings for gap open penalty at 12 and
gap extension penalty at 4.
Recombinant Expression. The major hits of MS analysis were
recombinantly expressed in E. coli as an internal control of the
MS results by using Invitrogen Gateway Technology. Target genes
were amplified from the corresponding genomes by PCR with an
AccuPrime Pfx DNA polymerase kit with 65 ng of genomic DNA,
prepared by standard protocols, or alternatively by colony PCR with
cells directly from the corresponding cryostocks. attB1 forward
primer and attB2 reverse primer were designed to yield the
following attB-PCR products needed for Gateway Technology.
Alkyl hydroperoxide reductase subunit C, ahpC, S. aureus NCTC
8325: forward primer, 5′-GGG GAC AAG TTT GTA CAA AAA
AGC AGG CTA CAT GTC ATT AAT TAA CAA AGA AAT;
reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA AGC
TGG GTG TTA GAT TTT ACC TAC TAA ATC.
Acetolactate synthase, alsS, S. aureus NCTC 8325: forward
primer, 5′-GGG GAC AAG TTT GTA CAA AAA AGC AGG
CTA CAT GAC TGA T AA AAA GTA CAC T; reverse primer,
5′-GGG GAC CAC TTT GTA CAA GAA AGC TGG GTG TTA
GTT AAA TAC GTC AGG TAA.
Inosine monophosphate dehydrogenase, (IMPDH), L. monocy-
togenes EGD-e: forward primer, 5′-GGG GAC AAG TTT GTA
CAA AAA AGC AGG CTA CAT GTG GGA AAC AAA ATT
TGC A; reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA
AGC TGG GTG TTA AGA AAT ACT ATA ATT TGG CG.
Peroxiredoxin (POR), L. monocytogenes EGD-e: forward primer,
5′-GGG GAC AAG TTT GTA CAA AAA AGC AGG CTA CGT
GGC AGA ACG TTT AGT A; reverse primer, 5′-GGG GAC CAC
TTT GTA CAA GAA AGC TGG GTG TTA AAC AAT TGT TTT
TTC ACC G.
Phosphoenolpyruvate-protein phosphotransferase, ptsI, S. aureus
Mu50: forward primer, 5′-GGG GAC AAG TTT GTA CAA AAA
AGC AGG CTA CAT GTC TAA ATT AAT TAA AGG TAT T;
reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA AGC
TGG GTG TTA TTT TAC GTA GTT GTT AAC T.
Ribonucleotide-diphosphate reductase subunit alpha (RNR), P.
aeruginosa PAO1: forward primer, 5′-GGG GAC AAG TTT GTA
CAA AAA AGC AGG CTA CAT GCA TAC CGA CAC CAC
ACG C; reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA
AGC TGG GTG TCA CTG GCA GGC TTC GCA GT.
UDP-N-acetylglucosamine 1-carboxyvinyltransferase murA1, S.
aureus NCTC 8325: forward primer, 5′-GGG GAC AAG TTT GTA
CAA AAA AGC AGG CTA CAT GGA TAA AAT AGT AAT
CAA AGG; reverse primer, 5′-GGG GAC CAC TTT GTA CAA
GAA AGC TGG GTG TTA ATC GTT AAT ACG TTC AAT GT.
UDP-N-acetylglucosamine 1-carboxyvinyltransferase murA2, S.
aureus Mu50: forward primer, 5′-GGG GAC AAG TTT GTA CAA
AAA AGC AGG CTA CAT GGC TCA AGA GGT AAT AAA
AAT; reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA
AGC TGG GTG CTA TAC AGT TTC CGT CCA AA.
UDP-N-acetylglucosamine pyrophosphorylase glmU, S. aureus
NCTC 8325: forward primer, 5′-GGG GAC AAG TTT GTA CAA
AAA AGC AGG CTA CAT GCT AGA TAA AAA TCA ATT
AGC; reverse primer, 5′-GGG GAC CAC TTT GTA CAA GAA
AGC TGG GTG TTA AAT TAA ACC CAT TCT ACG TA.
Reactivity of Showdomycin. To study the reactivity of show-
domycin toward sulfhydryl and primary amino groups, 17 µL of
10 mM DTT and isopropylamine were incubated with 1 µL of 10
mM showdomycin for 1 h. Next, 80 µL of 25 mM NH4HCO3 was
added, and after incubation for 12 h, the samples were analyzed
(28) Dunsmore, C. J.; Miller, K.; Blake, K. L.; Patching, S. G.; Henderson,
P. J.; Garnett, J. A.; Stubbings, W. J.; Phillips, S. E.; Palestrant, D. J.;
de Angeles, L. J.; Leeds, J. A.; Chopra, I.; Fishwick, C. W. Bioorg.
Med. Chem. Lett. 2008, 18, 1730–1734.
(27) Mirza, S. P.; Halligan, B. D.; Greene, A. S.; Olivier, M. Physiol.
Genom. 2007, 30, 89–94.
9
J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010 6971

A R T I C L E S
Bo¨ttcher and Sieber
by ESI-MS. For the DTT treatment, showdomycin was identified
with one DTT molecule attached at the maleimide double bond
and with a hydrolyzed imide ring (M ) 401.08142).
MurA2: forward primer, 5′-CAA ATA TTG AAG TTG ACG
AAG GCA-3′; reverse primer, 5′-ACT TAA AAG CTT TAG GTG
CAG ATA TTT-3′.
ESI-MS, positive mode (m/z): 142.0297 [M + 2H + Na]3+, calcd
142.0284. Negative mode (m/z): 422.0536 [M + Na - 2H]-, calcd
422.0561.
RT PCR was conducted in a two-step thermal protocol using a
SsoFast EvaGreen Supermix kit (Bio-Rad Laboratories) from an
optimized concentration of the cDNA in Thermo-Fast 96 semi-
skirted plates (ABgene) using an Eppendorf realplex4 epgradient
S Mastercycler instrument. The Sso7d-fusion polymerase was
activated in an initial denaturation step for 30 s at 95.0 °C. Forty
cycles were then run, consisting each of 5 s at 95.0 °C, followed
by 20 s annealing and elongation at 57.6 °C. A melting curve was
measured to ensure specific replication of the target sequences only.
For all strains the experiments were run twice in triplicate each,
and the MurA1/MurA2 ratios were calculated from Ct values of
each strain.
Fosfomycin Growth Inhibition Assay. The resistance of the
different S. aureus strains against fosfomycin was assessed by
studying their growth in BHB medium at various concentrations
of the antibiotic. Overnight cultures of S. aureus strains NCTC 8325,
Mu50, DSM 18827, and DSM 19041 were diluted accordingly to
an OD600 of 0.01 in fresh BHB medium. In a 96-well plate, 100
µL of the bacteria in BHB was added to 1 µL of the corresponding
fosfomycin stock dilutions in sterile deionized water. After incuba-
tion for 20 h at 37 °C under shaking at 250 rpm, the OD600 of
every well (dilutions in BHB medium) was determined using a
spectrophotometer.
Isopropylamine treatment, however, did not produce any adducts,
neither by attack of the maleimide double bond nor of the carbonyl
groups. Only showdomycin with hydrolyzed ring (due to long
exposure to NH4HCO3)29 could be detected (M ) 247.06920).
ESI-MS, negative mode (m/z): 246.0622 [M - H]-, calcd
246.0619.
Binding Sites of AhpC. To discover the binding sites of
showdomycin in AhpC, purified enzyme was labeled by showdo-
mycin, and after chymotryptic digestion the peptide fragments were
analyzed by MS. First, 50 µg of AhpC (2 pmol) in 17 µL of PBS
was incubated for 30 min at room temperature with 1 µL of 1 mM
showdomycin. The buffer was exchanged to wash away unreacted
probe and adjusted to 25 mM NH4HCO3, giving a total volume of
100 µL. Then 1 mM CaCl2 was added and the digest started with
chymotrypsin (sequencing grade, Roche Diagnostics, Germany) for
12 h at room temperature. The sample was brought to 0.1% formic
acid and analyzed by LC-MS with MS-MS and high-resolution mass
spectra. The spectra of the modified peptides are shown in Figure
3 and Supporting Information Figure S6.
Gene Expression Analysis of MurA Enzymes. S. aureus strains
NCTC 8325, Mu50, DSM 18827, and DSM 19041 were inoculated
from overnight cultures at 1:100 dilution in 20 mL of BHB medium,
grown to stationary phase, and kept for an additional 1 h in culture
(total of 8 h). Total RNA of the bacteria was isolated using an
Aurum Total RNA mini kit (Bio-Rad Laboratories) and transcribed
in cDNA by reverse transcriptase and a random primer set using
an iScript Select cDNA synthesis kit (Bio-Rad Laboratories). For
RT PCR, specific primers were designed for MurA1 and MurA2
to amplify 186 and 210 bp sequences close to the 3′ end of the
genes, respectively. Because they had identical gene sequences in
these regions, the primers were applicable for all strains. Special
care was taken in the design to avoid any secondary structure
elements in the annealing sequence by using the MFOLD 2.0
(Mobyle@pasteur) package.
Acknowledgment. We thank Prof. Dr. Thomas Carell and his
group for their generous support and the excellent working
environment as well as for many fruitful scientific discussions. We
gratefully acknowledge funding by the Emmy Noether Program of
the Deutsche Forschungsgemeinschaft (DFG), a DFG grant (SFB
749), a stipend by the Ro¨mer-Stiftung, and support by the Fonds
der Chemischen Industrie and by the Center for Integrated Protein
Science Munich (CiPSM). T.B. acknowledges funding by the
Studienstiftung des Deutschen Volkes. We thank Lukas Lercher
and Mathias Leidl for contributions in synthesis and Kerstin Kurz
for excellent technical assistance.
Supporting Information Available: Additional figures con-
taining chemical structures, fluorescent gels, mass spectra, and
enzyme inhibition data. This material is available free of charge
via the Internet at http://pubs.acs.org.
MurA1: forward primer, 5′-GCT AAT ATC AAT GTA GAA
GGT CGT-3′; reverse primer, 5′-GGC TAT GTT GAC TTA CAC
GGT-3′.
(29) Arendes, J.; Zahn, R. K.; Hantke, E.; Muller, W. E. Cancer Lett. 1978,
4, 259–264.
JA909150Y
9
6972 J. AM. CHEM. SOC. VOL. 132, NO. 20, 2010