A covalent reporter of β-lactamase activity for fluorescent imaging and rapid screening of antibiotic-resistant bacteria

Article abstract of DOI:10.1002/chem.201301654

Bacterial resistance to antibiotics poses a great clinical challenge in fighting serious infectious diseases due to complicated resistant mechanisms and time-consuming testing methods. Chemical reaction-directed covalent labeling of resistance-associated

Full text of DOI:10.1002/chem.201301654

FULL PAPER
DOI: 10.1002/chem.201301654
A Covalent Reporter of b-Lactamase Activity for Fluorescent Imaging and
Rapid Screening of Antibiotic-Resistant Bacteria
Qing Shao,[a] Yan Zheng,[b] Xueming Dong,[c] Kai Tang,[c] Xiaomei Yan,*[b] and
Bengang Xing*[a]
Abstract: Bacterial resistance to antibi-
otics poses a great clinical challenge in
fighting serious infectious diseases due
to complicated resistant mechanisms
and time-consuming testing methods.
Chemical reaction-directed covalent la-
beling of resistance-associated bacterial
proteins in the context of a complicat-
ed environment offers great opportuni-
ty for the in-depth understanding of
the biological basis conferring drug re-
sistance, and for the development of ef-
fective diagnostic approaches. In the
present study, three fluorogenic re-
agents LRBL1–3 for resistant bacteria
labeling have been designed and pre-
pared on the basis of fluorescence reso-
nance energy transfer (FRET). The hy-
drolyzed probes could act as reactive
electrophiles to attach the enzyme,
b-lactamase, and thus facilitated the co-
valent labeling of drug resistant bacte-
rial strains. SDS electrophoresis and
MALDI-TOF mass spectrometry char-
acterization confirmed that these
probes were sensitive and specific to
b-lactamase and could therefore serve
for covalent and localized fluorescence
labeling of the enzyme structure. More-
over, this b-lactamase-induced covalent
labeling provides quantitative analysis
of the resistant bacterial population
(down to 5%) by high resolution flow
cytometry, and allows single-cell detec-
tion and direct observation of bacterial
enzyme activity in resistant pathogenic
species. This approach offers great
promise for clinical investigations and
microbiological research.
Keywords: b-lactamase
tometry covalent labeling
FRET · fluorescent probes
· cell cy-
·
·
Introduction
observation of intrinsic resistance in individual living cells.[1]
Optical imaging techniques enable rapid, direct and sensi-
tive visualization of biological events at single-cell resolu-
tion, and thus have become powerful tools in monitoring
subcellular protein dynamics and analysis of pathogen–host
interactions.[2] However, imaging of antibiotic resistance has
been limited mainly by the lack of appropriate reagents to
report the resistance and to produce essential signal contrast
for reliable analysis. Incorporation of green fluorescent pro-
tein (GFP) or its color variants has allowed cellular detec-
tion of resistant genes[3] and whole-body imaging of anti-
biotic response to bacterial infections,[4] but the laboratory
strains expressing foreign genes are not identical to native
bacterial samples. The large size of such fluorescent tags
(~27 kDa) may potentially disturb protein function or
levels.[5] Approaches to imaging native bacterial strains and
studying their drug resistance usually involve staining with
fluorescently labeled affinity groups, such as metal com-
plexes, bacteria-binding peptides, antibodies and bacterio-
phages.[6] However, most of these modifications are associat-
ed with noncovalent interactions and have inadequate spe-
cificity to pathogens with potent resistance.[6f] Direct obser-
vation of resistant bacteria can be achieved by fluorescent
antibiotic derivatives on the basis of their different activities
towards antibiotic resistant and susceptible strains.[7] But the
intrinsic affinity between these fluorescent drug conjugates
and bacteria may present concerns for specificity. To avoid
nonspecific binding or adsorption thus requires more power-
ful designs to effectively report antibiotic resistance.
The remarkable increase of bacterial drug resistance has
been considered among the highest concerns towards
human healthcare, both within hospitals and in the commun-
ity. The growing prevalence of antibiotic resistance calls for
rapid detection of resistant strains and systematic under-
standing of the biological basis conferring bacterial drug re-
sistance. One most common method for the study of bacteri-
al susceptibility is the growth of cultures in antibiotic con-
taining media.[1] Although well-established, this method is
labor intensive and time-consuming, and cannot offer direct
[a] Q. Shao, Prof. B. Xing
Division of Chemistry and Biological Chemistry
School of Physical and Mathematical Sciences
Nanyang Technological University (Singapore)
E-mail: bengang@ntu.edu.sg
[b] Y. Zheng, Prof. X. Yan
The Key Laboratory of Analytical Sciences
The Key Laboratory for Chemical Biology of Fujian Province
Department of Chemical Biology
College of Chemistry and Chemical Engineering
Xiamen University, Xiamen, Fujian, 361005 (P.R. China)
E-mail: xmyan@xmu.edu.cn
[c] X. Dong, Prof. K. Tang
School of Biological Sciences
Nanyang Technological University (Singapore)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201301654.
Chem. Eur. J. 2013, 19, 10903 – 10910
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10903

Recently, reporter enzyme fluorescence (REF) has been
successfully employed to non-invasively image resistant bac-
terial infections in vitro and in vivo with tunable emission
enhancement.[8] In these designs, fluorescent turn-on probes
can be efficiently activated by the resistance-associated
b-lactamase (Bla), a naturally occurring bacterial enzyme
that destroys penicillin and cephalosporin antibiotics. Tar-
geting Bla presents unique advantages, as these enzymes are
exclusively produced by antibiotic-resistant bacteria, which
ensures labeling specificity with minimum interference from
endogenous counterparts in mammalian cells. In addition, as
a key element to deal with bacterial resistance and treat-
ment of bacterial infections, the crystal structure and molec-
ular mechanism of one commonly encountered b-lactamase,
reduce the background by minimizing the diffusion of the
activated probes in living tissues, and thus shows the feasi-
bility to perform reliable and rapid observation and differ-
entiation of clinical resistance, especially for the systematic
investigation of individual cells in a mixed population.
Results and Discussion
Scheme 1 illustrates the rational design of LRBL reagents.
The labile p-hydroxybenzylic esters bearing FRET-quenched
fluorescent tags were connected to the Bla-sensitive cepha-
losporin structure, at which point the sulfide bond was oxi-
dized to sulfoxide for improved stability. Upon Bla hydroly-
sis, the released p-hydroxybenzylic derivatives underwent
1,6-elimination to generate fluorescent quinone–methide in-
termediates for the localized covalent labeling. Compared
with conventional activity-based probes (ABPs) that directly
interact with catalytic amino acid residues in the enzyme
structure,[12] formation of the quinone–methide nucleophilic
traps requires enzymatic cleavage of a precursor and succes-
sive spontaneous elimination.[13] In the process of these cas-
cade reactions for effective covalent fluorescence attach-
ment, the activated probes may not work as suicide inhibi-
tors, thus largely maintain the enzyme activities to enhance
Escherichia coli TEM-1,
a plasmid mediated bacterial
enzyme, has been well investigated.[9] These understandings
provide great opportunity for systematic profiling of enzyme
activation, and thus significantly facilitate the development
of new imaging substrates with improved enzymatic reactivi-
ties. Currently, several imaging probes have been designed
to readily identify the recombinant Bla reporters in mam-
malian cells or endogenous Bla in bacterial pathogens.[10]
Most of them employ sensitive enzyme substrates with well-
suited fluorescence resonance energy transfer (FRET)
pairs.[8,10a–c] Upon enzyme hydrolysis, the intense on–off fluo-
rescence switch or emission
shift can provide essential sig-
nals for investigating enzyme
activity or quantitatively imag-
ing antibiotic-resistant patho-
gens in living animals with
high sensitivity and specifici-
ty.[8a,b] Although all of these
prior studies have been signifi-
cant and successful in vitro
and in vivo, a covalent labeling
capability toward the straight-
forward and in situ recognition
of endogenous bacterial Bla
molecules that prevents probe
diffusion in living tissues will
still be greatly appreciated;[11]
relevant investigations have
not been fully exploited so far.
Inspired by previous success-
ful studies, we describe the ra-
tional design of novel optical
probes for detecting Bla and
covalent fluorescent labeling
of antibiotic-resistant bacteria.
More importantly, unlike the
traditionally used fluorescence
or culture methods to deter-
mine the susceptibility of mi-
crobes, this b-lactamase-re-
sponsive bacterial labeling
(LRBL) approach may greatly
Scheme 1. Reaction process of enzyme-responsive fluorescent probes (LRBL1–3) for covalent labeling of anti-
biotic-resistant bacteria.
10904
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 10903 – 10910

A Covalent Reporter of b-Lactamase Activity
FULL PAPER
the fluorescent signals for single cell observation. In this
study, three fluorescent molecules with different characteris-
tics were used, including fluorescein, water-soluble Cy3 and
near-infrared Cy5.5, to provide tunable emission properties,
which could greatly facilitate the microscopic imaging and
high-sensitivity flow cytometry (HSFCM) analysis. To maxi-
mize the signal-to-noise ratio, the selected fluorophores
were prequenched by DABCYL, BHQ2 and BHQ3 moie-
ACHTUNGTRENNUNG
Upon fluorescence enhancement activated by TEM-1 Bla,
the enzyme labeling profile was also studied by SDS-PAGE
analysis and matrix-assisted laser desorption ionization-time
of flight (MALDI-TOF) mass spectrometry. Typically, gel
electrophoresis was carried out by loading the incubation
mixture consisting of Bla (1 mg) and probe (2 mm) on a 10%
polyacrylamide/SDS gel and then subjected to 1 h of separa-
tion. Using a laser scanner, the labeled enzyme could be di-
rectly read out by the tagged fluorescence signals, clearly
suggesting the formation of the reactive quinone–
methide moiety for effective attachment to the enzyme
structure (the labeling mechanism is illustrated in Figure S1
in the Supporting Information). As a control, competition
assays pretreated with Bla inhibitor clavulanic acid (CA)
left the protein unlabeled, indicating the stability of the
probes under aqueous conditions and the labeling activity of
the specific enzyme reaction (Figure 2). MALDI-TOF mass
ACHTUNGTRENNUNGties, respectively, and the labeling process was accompanied
with recovered fluorescence that was employed for direct
observation and screening of antibiotic-resistant bacterial
strains. These probes were synthesized, purified by reverse
phase HPLC and further characterized by NMR and mass
spectrometry (see the Supporting Information).
Firstly, the enzymatic hydrolysis of LRBL1–3 was studied
by measuring the changes in their fluorescence emission
upon the addition of TEM-1 Bla in PBS buffer (0.1m,
pH 7.2). As shown in Figure 1, the probes were almost non-
Figure 2. Labeling of TEM-1 Bla by LRBL1–3. SDS-PAGE analysis of
Bla with probe incubation and imaged with: a) Coomassie blue staining,
and b) fluorescence of merged green, yellow and red channels. Lane 1:
molecular weight markers, in kDa; lane 2: Bla+LRBL1; lane 4: Bla+
LRBL2; lane 6: Bla+LRBL3; lanes 3, 5 and 7: clavulanic acid inhibited
Bla with LRBL1, -2 and -3, respectively.
Figure 1. a), c) Fluorescence emission of LRBL1–3 (10 mm) in the absence
or presence of Bla hydrolysis (PBS, pH 7.2) with excitation at 488, 525
and 650 nm, respectively. Insets: detection limit of TEM-1 Bla.
d) Enzyme kinetics of LRBL reagents to TEM-1 Bla.
fluorescent due to efficient FRET quenching. After incuba-
tion with TEM-1 Bla at 378C for 30 min, intense fluores-
cence enhancement (38-, 110- and 80-fold for LRBL1, -2
and -3, respectively) was observed at a maximum wave-
length of 518, 565 and 678 nm, respectively, corresponding
to the connected fluorophores. These results indicated that
enzyme hydrolysis would break FRET status by releasing
the individual fluorescence quenchers. Enzyme kinetics of
these probes to TEM-1 Bla hydrolysis were further studied
in PBS at 378C. The catalytic constants (kcat =4.92, 2.16,
1.49 minÀ1) and Michaelis constants (KM =3.08, 4.28,
5.24 mm) for LRBL1, -2 and -3 were determined, respective-
ly, by weighted least-squares fit of a double reciprocal plot
of the hydrolysis rate versus probes concentration (Fig-
ure 1d). The catalytic efficiencies (kcat/KM) were calculated
to be 2.66, 0.84 and 0.47ꢁ104 mÀ1 sÀ1, respectively, compar-
spectrometric analysis of the fluorescently labeled samples
provided further insight into the labeling profile of the
enzyme structure. As shown in Figure 3, besides the peak
from parent Bla enzyme (~28985, Figure S2 in the Support-
ing Information), the molecular weight peaks of 29572,
29800 and 30081 with additional probe fragments were also
observed in the MS spectrum, indicating the covalent attach-
ment of the quinone–methide intermediates (MW: 581, 813
and 1090 for LRBL1, LRBL2+Na and LRBL3, respective-
ly) into bacterial Bla. In the process of enzyme-responsive
fluorescent labeling, two or more modifications in the
enzyme sequence were also found in the MALDI-TOF anal-
ysis, implying the existence of more nucleophilic amino
acids that may potentially contribute to the capture of the
quinone–methide fragments after enzyme cleavage (see the
Supporting Information).
Chem. Eur. J. 2013, 19, 10903 – 10910
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10905

X. Yan, B. Xing et al.
cillin resistant E. coli JM109/pUC19 strain was selected as
our target due to the higher production of class A Bla—a
most prevalent plasmid-gene encoded bacterial enzyme
mainly responsible for b-lactam antibiotic resistance in both
clinic and community acquired infections.[9] Meanwhile, the
antibiotic susceptible E. coli JM109 strain that could not ex-
press Bla was used as a negative control. Cells were incubat-
ed with the probes (10 mm) and subjected to fluorescence
imaging. As shown in Figure 5a, after incubation, strong flu-
orescence emission of LRBL1 (green), -2 (yellow) and -3
(red) was observed in the resistant JM109/pUC19 cells
whereas there was no obvious fluorescence signal detected
in the cells pretreated with enzyme inhibitor CA (Figure S3
in the Supporting Information) or in the negative control of
E. coli JM109 strain. Similar bacterial imaging with LRBL1–
3 was also conducted by using confocal laser scanning mi-
croscopy. Fluorescent signals from the various depths of the
bacteria further proved that these LRBL probes could get
into the cells and produce fluorescence upon Bla hydrolysis
(Figure S4a–c in the Supporting Information). In addition,
the bacterial lysates, upon PopCulture/lysozyme treatment,
further indicated specific Bla hydrolysis and significant fluo-
rescence enhancement in the resistant bacteria with mini-
mum influence of other cellular proteins (Figure S5 in the
Supporting Information). These results clearly demonstrated
the native capability of the rationally designed enzyme-re-
sponsive probes to specifically detect endogenous Bla and
label the resistant bacterial strains.
Moreover, the feasibility of labeling and screening the re-
sistant strains was also exploited with a laboratory-built
high-sensitivity flow cytometer (HSFCM). As a powerful
tool for performing rapid analysis of individual cellular func-
tions with excellent statistics in asynchronous cultures, flow
cytometry has found extensive application in combination
with effective detection reagents, including GFP, fluorescent
antibodies, recombinant phages or specific dye molecules
with good affinity to nucleic acids or bacterial mem-
branes.[15] With the covalent fluorescent labeling of bacterial
enzyme provided by the LRBL reagents, HSFCM could fa-
cilitate a clear insight into population heterogeneity of bac-
terial strains in terms of antibiotic-resistance properties.
As a proof-of-concept, E. coli cells were incubated with
LRBL1 (2 mm) after fixation and permeabilization. HSFCM
green fluorescence signals from individual bacterial cells
were collected at 520/35 nm under 488 nm excitation. Fig-
ure 5b and Figure S6 (in the Supporting Information) indi-
cated that a strong fluorescence enhancement (more than
20-fold) can be observed in the resistant bacteria as com-
pared to the negative control. The population of resistant
strain JM109/pUC19 was completely separated from that of
E. coli JM109 (Figure 5c), implying a high efficiency of
enzyme-responsive labeling for drug resistant bacterial
pathogens. The specificity of bacterial labeling was further
confirmed by CA inhibition with an expected result of low
fluorescence signal due to the potent inactivation of endoge-
nous Bla (Figure 5b–c and Figure S6-c2 in the Supporting
Information). The comparable scattering signals for E. coli
Figure 3. MALDI-TOF analysis of labeled TEM-1 Bla.
During the course of protein covalent labeling, the chemi-
cal modification may probably affect its biological functions
if the targeted specific amino acid residues locate at the pro-
tein active site. To this end, the enzymatic activities of the
labeled Bla were examined with the conventional hydrolysis
assay using benzylpenicillin as substrate.[14] Typically, TEM-1
Bla (0.5 mm) was incubated with LRBL probes (50 mm) for
30 min. Then the reaction mixture was added to benzylpeni-
cillin (0.5 mm in PBS) with 1:200 dilution and the enzymatic
hydrolysis profile of benzylpenicillin was monitored based
on the absorbance change. As shown in Figure 4, compared
to the hydrolysis activity with Bla alone, the similar absor-
ACHTUNGTRENNUNGbance change observed at 232 nm indicated that the enzyme
activities were almost identical for the TEM-1 Bla samples
that had been treated with different LRBL probes. As a
control, the enzyme hydrolysis of benzylpenicillin was also
performed in the presence of an efficient Bla inhibitor, clav-
ulanic acid (50 mm). The insignificant absorbance change
after incubation with Bla demonstrated that enzyme activity
was effectively suppressed. It is plausible that although mul-
tiple enzyme modifications induced by quinine–methide
fragments could occur, the covalent fluorescence labeling
did not take place predominantly at the active site of the
enzyme structure. Such unique behavior of catalytic forma-
tion and covalent labeling would ideally facilitate real-time
imaging and direct screening of antibiotic-resistant bacteria.
In order to evaluate the specific covalent labeling of re-
sistant bacteria, live cell imaging was performed with confo-
cal microscopy. In a typical study, the Gram negative peni-
Figure 4. Activity of TEM-1 Bla after incubation with LRBL1–3. Enzy-
matic hydrolysis of benzylpenicillin based on absorbance decrease at
232 nm after incubation with LRBL probes or enzyme inhibitor.
10906
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 10903 – 10910

A Covalent Reporter of b-Lactamase Activity
FULL PAPER
(Figure 6), whereas the preva-
lence of JM109/pUC19 in the
mixture was 5, 20, 50, 80 and
100%, correspondingly. Such
excellent linear correlation of
the ratios detected by FCM
proved the labeling specificity
to the antibiotic-resistant bacte-
ria and the non-detection of
LRBL probes of nonresistant
strains. Quantitative analysis of
antibiotic-resistant bacteria in a
mixture could be easily demon-
strated down to as low as 5%.
These results made it clear that
the inevitable multiple labeling
of the quinone–methide probe
did not necessarily mean a seri-
ous problem in the cellular con-
text, as the large extent reten-
tion of the activated probes
around the site of enzyme activ-
ity would significantly prevent
the leakage from targeted cells
and therefore greatly facilitate
rapid cellular imaging and FCM
studies.
As
a
comparison, similar
enzyme activity and HSFCM
studies were also carried out by
using a commercial fluorogenic
Bla substrate (FluorocillinTM
Green 495/525 reagent, Invitro-
gen). After the enzyme hydrol-
ysis of this substrate in PBS
(0.1m, pH 7.2), a strong fluores-
cence enhancement could be
detected, which was similar to
the enzyme activities observed
with LRBL1–3 (Figure S7d in
the Supporting Information).
However, owing to the fact that
there was no covalent labeling
Figure 5. a) Fluorescent imaging of penicillin resistant E. coli JM109/pUC19 and penicillin susceptible E. coli
JM109 incubated with LRBL1, -2 and -3 (10 mm). Scale bar: 10 mm. b), c) HSFCM analysis of antibiotic-resist-
ant bacteria with LRBL1 (2 mm) labeling. Samples: E. coli JM109 (1), CA-treated E. coli JM109/pUC19 (2)
and E. coli JM109/pUC19 (3). Concentration of bacteria was approximately 108 cellsmLÀ1
.
JM109 and E. coli JM109/pUC19 regardless of whether or
not they were treated with CA inhibitor (Figure 5b) sug-
gests that the specific enzyme-responsive covalent fluores-
cence labeling occurred in the target strains and such effec-
tive fluorescent labeling would not induce bacterial aggrega-
tion.
The potential interference of coexisting nonresistant bac-
teria in the detection of target strains was also investigated.
A series of bacterial mixtures containing E. coli JM109 and
JM109/pUC19 with a total number of 108 cellsmLÀ1 but dif-
ferent percentages of JM109/pUC19 were prepared and in-
cubated with LRBL1 (2 mm). A clear separation of antibiotic
resistant and susceptible cells provided the percentage of de-
tected resistant bacteria at 6.6, 19.6, 40.5, 69.3 and 92.2%
with FluorocillinTM Green 495/525 reagent, upon incubation
with resistant E. coli JM109/pUC19 bacteria, the fluorescent
products could not be retained within the strains properly,
thus making FCM detection difficult (Figure S7 in the Sup-
porting Information). These experiments clearly demon-
strated the obvious advantage of the covalent attachment by
our designed probes.
The general applicability of detecting and labeling antibi-
otic-resistant bacteria was further examined with pathogenic
Gram positive Bacillus cereus and Staphylococcus aureus.
These species are well-known to be involved in various seri-
ous infections, like brain abscesses and pneumonia, and rep-
resent a therapeutic challenge in terms of penicillin and
cephalosporin resistance as a consequence of complicated
Chem. Eur. J. 2013, 19, 10903 – 10910
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10907

X. Yan, B. Xing et al.
Figure 6. Differentiation of resistant E. coli JM109/pUC19 cells in bacterial mixtures. a)–e) FCM analysis of bacterial mixtures stained by LRBL1 (2 mm).
The antibiotic-resistant and susceptible bacteria can be divided by the line. f) Linear correspondence between the theoretical and the FCM-measured
percentages of antibiotic-resistant bacteria. Error bars indicate standard deviation (n=3).
mechanisms.[16] In this work, three bacterial pathogens in-
cluding one clinically isolated methicillin-resistant S. aureus
(MRSA, ATCC BAA39), one penicillin resistant B. cereus
5/B (ATCC 13061), and one penicillin susceptible S. aureus
(ATCC 29213) were selected as our main targets owing to
their different extents of Bla production. Upon incubation
of the probes (10 mm) with different bacterial cells, the fluo-
rescent signals were monitored by confocal microscopy. As
shown in Figure 7, after incubation, strong green, yellow and
NIR fluorescence emissions could be clearly observed in the
penicillin resistant MRSA and B. cereus. As a control, the
signal detected for S. aureus was much weaker, which could
be attributed to the low level Bla production in this
strain.[17] Bacterial imaging by confocal laser scanning mea-
The same bacterial strain without treatment with probe was
used as the negative control. As expected, the FCM studies
confirmed the efficient covalent labeling of B. cereus cells,
and there was more than 40-fold fluorescence increase as
compared to the background autofluorescence in the un-
treated bacteria (Figure 8, Figure S8 in the Supporting Infor-
mation). Similar FCM analysis was also utilized to compare
the Bla activities in drug susceptible S. aureus and clinically
isolated MRSA. The effective bacterial labeling indicated
that there was almost tenfold fluorescence enhancement ob-
served in pathogenic MRSA over the control S. aureus
strain (Figure S9 in the Supporting Information); this sug-
gests the quantitative recognition of Bla activities between
these two naturally existing bacterial pathogens, and is in
ACHTUNGTRENNUNGsurements were also performed
in penicillin resistant MRSA
and B. cereus at various depths
(Figure S4d,e in the Supporting
Information); the results clearly
indicated the effective cellular
internalization
of
imaging
probes and their specific label-
ing of Bla reporter enzyme in
the antibiotic-resistant bacteria
with resultant significant fluo-
rescence enhancement.
Enzyme-responsive covalent
labeling and bacterial screening
in Gram positive strains were
also monitored by HSFCM
analysis. First, the penicillin re-
sistant B. cereus was chosen to
Figure 7. Fluorescence imaging of Gram positive B. cereus, MRSA and S. aureus after incubation with LRBL1,
incubate with LRBL1 (2 mm). -2 and -3 (10 mm). Scale bar: 10 mm.
10908
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 10903 – 10910

A Covalent Reporter of b-Lactamase Activity
FULL PAPER
observation of biological pathways in bacterial growth and
multiplication, real-time monitoring of bacterial infections
and assessment of therapeutic efficacy in vitro and in vivo.
Experimental Section
Flow-cytometric analysis of antibiotic-resistant bacterial labeling: Live
bacterial cells (108 cellsmLÀ1) were incubated with LRBL1 (10 mm) in the
dark for 60 min and washed with PBS. The fluorescence of the cells was
detected by a laboratory-built high-sensitivity flow cytometer described
previously.[19] To study the labeling specificity, the bacterial cells were
preincubated with Bla inhibitor clavulanic acid for 1 h, then stained with
LRBL1 and subjected to FCM analysis. For the strains that were fixed
and permeabilized, the cells were first fixed with PFA (1%) for 5 min
and then incubated with permeating solution (25 mm Tris-HCl, 1.8% glu-
cose, 10 mm EDTA, pH 7.4) for 20 min. After being washed with PBS,
the treated cells were incubated with LRBL1 (2 mm) in the dark for
20 min, washed with PBS and subjected to FCM analysis. For compari-
son, the commercial FluorocillinTM Green 495/525 b-lactamase substrate
(20 mm) was incubated with the permeabilized E. coli JM109/pUC19 cells
for 20 min. After being washed with PBS, the bacterial cells were subject-
ed to FCM analysis.
Figure 8. FCM analysis of resistance in Gram positive B. cereus by
LRBL1 labeling. B. cereus (unstained, 1), 0.5 mm CA-treated B. cereus
(2 mm LRBL1 stained, 2), and B. cereus (2 mm LRBL1 stained, 3). Con-
centration of B. cereus was approximately 108 cellsmLÀ1
.
Live cell imaging of labeled bacteria: An overnight bacterial suspension
culture was diluted to 108 cellsmLÀ1 and incubated with LRBL probes
(10 mm). After being washed with PBS, the bacteria were spotted on
poly-l-lysine pretreated glass slides and covered with coverslips. Cell
imaging tests were conducted with a Nikon Eclipse TE2000 Confocal Mi-
croscope.
accord with the data obtained in previous fluorescence
imaging measurements. In addition, FCM analysis showed
that treatment of drug resistant bacterial strain (e.g.,
B. cereus) with Bla inhibitor could also result in an obvious
decrease in fluorescence intensity (Figure 8). Considering
the fact that the efficacy of bacterial enzyme inhibitors
varies with different enzyme classes and different strains,[18]
fast reading of inhibitory efficiency through the enzyme-re-
sponsive bacterial covalent labeling may facilitate a simple
and overall inspection of chemical–microbe interactions.
This approach of enzyme-responsive bacterial labeling is
therefore expected to find potential applications in clinical
tests and new drug development to circumvent microbial
resistance.
Acknowledgements
This work was supported by Start-Up Grant (SUG), MOE, Tier 1 (RG
64/10) in Nanyang Technological University, Singapore, and the NSF of
P.R. China (21225523, 21027010, 20975087, and 90913015).
[1] S. B. Levy, B. Marshall, Nat. Med. 2004, 10, S122.
[2] a) D. Kentner, V. Sourjik, Annu. Rev. Microbiol. 2010, 64, 373; b) L.
Shapiro, H. H. McAdams, R. Losick, Science 2009, 326, 1225; c) J. F.
Lovell, T. Liu, J. Chen, G. Zheng, Chem. Rev. 2010, 110, 2839.
[3] a) P. Meyer, J. Dworkin, Res. Microbiol. 2007, 158, 187; b) I. Gold-
ing, J. Paulsson, S. M. Zawilski, E. C. Cox, Cell 2005, 123, 1025.
[4] a) M. Zhao, M. Yang, E. Baranov, X. Wang, S. Penman, A. R.
Moossa, R. M. Hoffman, Proc. Natl. Acad. Sci. USA 2001, 98, 9814;
b) L. E. Mansson, K. Melican, J. Boekel, R. M. Sandoval, I. Haute-
fort, G. A. Tanner, B. A. Molitoris, A. Richter-Dahlfors, Cell. Micro-
biol. 2007, 9, 413.
Conclusion
In summary, this work presents a set of fast and sensitive re-
agents for specific labeling and screening of antibiotic-resist-
ant bacteria. By utilizing the endogenous production of
b-lactamases in the resistant strains, three fluorescent qui-
none–methide intermediates could be activated; these serve
as reactive electrophiles for the covalent labeling of drug-re-
sistant bacteria. This specific b-lactamase responsive bacteri-
al labeling, with significant fluorescence enhancement and
tunable emission properties, allows single-cell detection and
quantitative observation of resistant bacteria in a mixed
sample, and thus provides a promising tool for clinical inves-
tigation and microbiological research. The Bla specificity
and fluorescence activation can be applied to high-through-
put enzyme inhibitor screening, either in new drug develop-
ment or clinical diagnosis. Moreover, such covalent fluores-
cent labeling will provide a great opportunity for the direct
[5] R. Y. Tsien, Annu. Rev. Biochem. 1998, 67, 509.
[6] a) W. M. Leevy, S. T. Gammon, H. Jiang, J. R. Johnson, D. J. Max-
well, E. N. Jackson, M. Marquez, D. Piwnica-Worms, B. D. Smith, J.
Am. Chem. Soc. 2006, 128, 16476; b) V. van Dam, N. Olrichs, E.
Breukink, ChemBioChem 2009, 10, 617; c) F. Liu, A. S. Y. Ni, Y.
Lim, H. Mohanram, S. Bhattacharjya, B. G. Xing, Bioconjugate
Chem. 2012, 23, 1639; d) B. G. Xing, C. W. Yu, P. L. Ho, K. H. Chow,
T. Cheung, H. W. Gu, Z. W. Cai, B. Xu, J. Med. Chem. 2003, 46,
4904; e) A. M. Hujer, K. S. Keslar, N. J. Dietenberger, C. R. Bethel,
A. Endimiani, R. A. Bonomo, BMC Microbiol. 2009, 9, 46; f) A. E.
Smartt, S. Ripp, Anal. Bioanal. Chem. 2011, 400, 991; g) R. Edgar,
M. McKinstry, J. Hwang, A. B. Oppenheim, R. A. Fekete, G. Giuli-
an, C. Merril, K. Nagashima, S. Adhya, Proc. Natl. Acad. Sci. USA
2006, 103, 4841.
[7] a) I. Staub, S. A. Sieber, J. Am. Chem. Soc. 2008, 130, 13400;
b) B. G. Xing, T. T. Jiang, W. G. Bi, Y. M. Yang, L. H. Li, M. L. Ma,
C. K. Chang, B. Xu, E. K. L. Yeow, Chem. Commun. 2011, 47, 1601;
Chem. Eur. J. 2013, 19, 10903 – 10910
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10909

X. Yan, B. Xing et al.
c) P. M. Pereira, S. R. Filipe, A. Tomasz, M. G. Pinho, Antimicrob.
Agents Chemother. 2007, 51, 3627.
Suarez, S. Puthenveetil, A. Y. Ting, Proc. Natl. Acad. Sci. USA 2010,
107, 10914; e) S. Tsukiji, M. Miyagawa, Y. Takaoka, T. Tamura, I.
Hamachi, Nat. Chem. Biol. 2009, 5, 341; f) A. Keppler, S. Gendrei-
zig, T. Gronemeyer, H. Pick, H. Vogel, K. Johnsson, Nat. Biotechnol.
2002, 21, 86; g) R. M. Yusop, A. Unciti-Broceta, E. M. V. Johansson,
R. M. Sꢂnchez-Martꢃn, M. Bradley, Nat. Chem. 2011, 3, 241; h) Y. A.
Lin, J. M. Chalker, B. G. Davis, J. Am. Chem. Soc. 2010, 132, 16805;
i) T. Komatsu, K. Johnsson, H. Okuno, H. Bito, T. Inoue, T. Nagano,
Y. Urano, J. Am. Chem. Soc. 2011, 133, 6745; j) C. C. Hughes, Y.-L.
Yang, W.-T. Liu, P. C. Dorrestein, J. J. La Clair, W. Fenical, J. Am.
Chem. Soc. 2009, 131, 12094.
[8] a) H. Xie, J. Mire, Y. Kong, M. Chang, H. A. Hassounah, C. N.
Thornton, J. C. Sacchettini, J. D. Cirillo, J. Rao, Nat. Chem. 2012, 4,
802; b) Y. Kong, H. Q. Yao, H. J. Ren, S. Subbian, S. L. G. Cirillo,
J. C. Sacchettini, J. Rao, J. D. Cirillo, Proc. Natl. Acad. Sci. USA
2010, 107, 12239; c) J. X. Zhang, Y. Shen, S. L. May, D. C. Nelson,
S. W. Li, Angew. Chem. 2012, 124, 1901; Angew. Chem. Int. Ed.
2012, 51, 1865; d) Q. Shao, B. G. Xing, Chem. Commun. 2012, 48,
1739.
[9] a) C. Jelsch, L. Mourey, J. M. Masson, J. P. Samama, Proteins Struct.
Funct. Genet. 1993, 16, 364; b) G. Minasov, X. J. Wang, B. K. Shoi-
chet, J. Am. Chem. Soc. 2002, 124, 5333; c) C. Damblon, X. Raquet,
L. Y. Lian, J. Lamotte, R. Brasseur, E. Fonze, P. Charlier, G. C. K.
Roberts, J. M. Frere, Proc. Natl. Acad. Sci. USA 1996, 93, 1747.
[10] a) G. Zlokarnik, P. A. Negulescu, T. E. Knapp, L. Mere, N. Burres,
L. Feng, M. Whitney, K. Roemer, R. Y. Tsien, Science 1998, 279, 84;
b) B. G. Xing, A. Khanamiryan, J. Rao, J. Am. Chem. Soc. 2005, 127,
4158; c) S. Mizukami, S. Watanabe, Y. Akimoto, K. Kikuchi, J. Am.
Chem. Soc. 2012, 134, 1623; d) W. Z. Gao, B. G. Xing, R. Y. Tsien, J.
Rao, J. Am. Chem. Soc. 2003, 125, 11146; e) S. Mizukami, T. Yama-
moto, A. Yoshimura, S. Watanabe, K. Kikuchi, Angew. Chem. 2011,
123, 8909; Angew. Chem. Int. Ed. 2011, 50, 8750; f) H. Yao, M. K.
So, J. Rao, Angew. Chem. 2007, 119, 7161; Angew. Chem. Int. Ed.
2007, 46, 7031; g) B. G. Xing, J. Rao, R. R. Liu, Mini-Rev. Med.
Chem. 2008, 8, 455; h) T. Jiang, B. G. Xing, J. Rao, Biotechnol.
Genet. Eng. Rev. 2008, 25, 41; i) Z. M. Yang, P. L. Ho, G. L. Liang,
K. H. Chow, Q. G. Wang, Y. Cao, Z. H. Guo, B. Xu, J. Am. Chem.
Soc. 2007, 129, 266.
[12] a) B. F. Cravatt, A. T. Wright, J. W. Kozarich, Annu. Rev. Biochem.
2008, 77, 383; b) A. M. Sadaghiani, S. H. L. Verhelst, M. Bogyo,
Curr. Opin. Chem. Biol. 2007, 11, 20.
[13] a) M. Kurogochi, S. I. Nishimura, Y. C. Lee, J. Biol. Chem. 2004,
279, 44704; b) T. Komatsu, K. Kikuchi, H. Takakusa, K. Hanaoka, T.
Ueno, M. Kamiya, Y. Urano, T. Nagano, J. Am. Chem. Soc. 2006,
128, 15946.
[14] S. B. Boggio, O. A. Roveri, Microbiology 2003, 149, 445.
[15] a) L. N. Wu, T. T. Huang, L. L. Yang, J. B. Pan, S. B. Zhu, X. M.
Yan, Angew. Chem. 2011, 123, 5995; Angew. Chem. Int. Ed. 2011,
50, 5873; b) D. Cohen, J. A. Dickerson, C. D. Whitmore, E. H.
Turner, M. M. Palcic, O. Hindsgaul, N. J. Dovichi, Annu. Rev. Anal.
Chem. 2008, 1, 165.
[16] a) E. J. Bottone, Clin. Microbiol. Rev. 2010, 23, 382; b) H. F. Cham-
bers, F. R. Deleo, Nat. Rev. Microbiol. 2009, 7, 629.
[17] E. C. Nannini, K. V. Singh, B. E. Murray, Clin. Infect. Dis. 2003, 37,
1194.
[18] S. M. Drawz, R. A. Bonomo, Clin. Microbiol. Rev. 2010, 23, 160.
[19] L. L. Yang, Y. X. Zhou, S. B. Zhu, T. X. Huang, L. N. Wu, X. M.
Yan, Anal. Chem. 2012, 84, 1526.
[11] a) J. Rayo, N. Amara, P. Krief, M. M. Meijler, J. Am. Chem. Soc.
2011, 133, 7469; b) W. Song, Y. Wang, J. Qu, Q. Lin, J. Am. Chem.
Soc. 2008, 130, 9654; c) E. M. Sletten, C. R. Bertozzi, Angew. Chem.
2009, 121, 7108; Angew. Chem. Int. Ed. 2009, 48, 6974; d) C. Utta-
mapinant, K. A. White, H. Brauah, S. Thompson, M. Fernꢂnde-
Received: April 30, 2013
Published online: July 8, 2013
10910
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 10903 – 10910