Enzyme-mediated nutrient release: Glucose-precursor activation by β-galactosidase to induce bacterial growth

Article abstract of DOI:10.1039/c3ob27385g

Bacteria will gain an advantage if they are able to metabolize nutrients that are inaccessible for other bacteria. To demonstrate this principle, we developed a simple model system, which mimics how bacteria exploit natural carbon sources. A masked glucose precursor that is activated by β-galactosidase was used as a carbon source for bacterial growth in a glucose-deficient medium. No bacterial growth was observed in the presence of control substances in which β-galactosidase mediated cleavage did not lead to glucose release. This study represents a proof-of-principle example in which a bacterium can grow in a nutrient-free medium by inducible, enzyme-mediated nutrient release from a precursor.

Full text of DOI:10.1039/c3ob27385g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Enzyme-mediated nutrient release: glucose-precursor
activation by β-galactosidase to induce bacterial
growth
Cite this: Org. Biomol. Chem., 2013, 11,
2903
Naama Karton-Lifshin,a Uwe Vogel,b,c Eran Sella,a Peter H. Seeberger,b,c
Doron Shabat*a and Bernd Lepenies*b,c
Bacteria will gain an advantage if they are able to metabolize nutrients that are inaccessible for other
bacteria. To demonstrate this principle, we developed a simple model system, which mimics how bacteria
exploit natural carbon sources. A masked glucose precursor that is activated by β-galactosidase was used
as a carbon source for bacterial growth in a glucose-deficient medium. No bacterial growth was observed
in the presence of control substances in which β-galactosidase mediated cleavage did not lead to glucose
release. This study represents a proof-of-principle example in which a bacterium can grow in a nutrient-
free medium by inducible, enzyme-mediated nutrient release from a precursor.
Received 9th December 2012,
Accepted 24th February 2013
DOI: 10.1039/c3ob27385g
www.rsc.org/obc
A well-known gene cluster is the lac operon that regulates
the transport and metabolism of lactose in E. coli. It consists
Introduction
Bacteria will gain an advantage if they are able to adapt to of three genes encoding for the enzymes β-galactosidase
changing environmental conditions. For instance, an increase (β-gal), lactose permease and thiogalactoside transacetylase.
in the frequency of antibiotic resistance in bacteria due to The lac operon only becomes activated in the absence of
random mutations or exchange of genetic material between glucose and in the presence of lactose or non-metabolizable
different species of bacteria has been observed for most major lactose metabolites.
classes of antibiotics used to treat a wide variety of respiratory
Here, we disclose a system utilizing the process by which
illnesses, skin disorders, and sexually transmitted diseases.1 bacteria exploit natural carbon sources. A novel glucose
Resistance genes as well as genes encoding for enzymes release-system that is based on a masked glucose precursor
required for metabolizing nutrients are organized in operons. can be cleaved by the enzyme β-galactosidase. E. coli strains
This regulated, clustered expression ensures that these pro- expressing β-gal release glucose from the masked precursor
teins are not produced constitutively, but that expression is and utilize this non-natural precursor to grow. However, a
only induced if needed.
Bacteria preferentially exploit carbon sources that are most score the specificity of the system described here.
readily accessible and the expression of enzymes to metabolize The system described here may not just allow for release of
β-gal cleavage site was required for bacterial growth to under-
secondary carbon sources is usually suppressed.2 The presence nutrients, but also may be a basis for the design of novel anti-
of glucose in the culture medium prevents bacteria from the biotic release systems to combat bacterial pathogens. To
use of other carbon sources.3 However, if only secondary specifically target resistant bacteria, antibiotics will be released
carbon sources are available, the enzymes needed for metabo- from a masked precursor by the action of resistance proteins.
lizing these carbon sources become activated to allow for bac-
terial growth in the absence of glucose.
Results and discussion
aSchool of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences and
The inducible expression of operon-encoded resistance pro-
teins or catabolic enzymes can be exploited to release active
components from inactive precursors. In this proof-of-prin-
ciple study, we focused on the enzyme-catalyzed release of a
missing vital nutrient from an available precursor to provide
bacteria expressing the respective enzyme with an evolutionary
advantage. We employed an E. coli strain expressing the
Tel Aviv University, Tel Aviv, 69978, Israel. E-mail: chdoron@post.tau.ac.il;
Fax: +972-3-6409293; Tel: +972-3-6408340
bDepartment of Biomolecular Systems, Max Planck Institute of Colloids and
Interfaces, Am Mühlenberg 1, 14476 Potsdam, Germany.
E-mail: Bernd.Lepenies@mpikg.mpg.de; Fax: +49-30-838-59302;
Tel: +49-30-838-59572
cInstitute of Chemistry and Biochemistry, Department of Biology, Chemistry and
Pharmacy, Freie Universität Berlin, Arnimallee 22, 14195 Berlin, Germany
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2903–2910 | 2903

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 1 Masked precursor of glucose activated by β-gal to release free glucose.
lactose-metabolizing enzyme β-galactosidase (β-gal) that grows disassembled through elimination of imminium-methylene to
fast in the presence of the carbon source glucose. Instead of release a free glucose molecule.
glucose we nourished the E. coli with a masked precursor of
Compound 1 was synthesized commencing with a Curtius
glucose that can be activated by β-gal or, alternatively, with a rearrangement of acyl azide 2a to generate an isocyanate inter-
control compound that did not allow for β-gal mediated mediate that was reacted in situ with benzylalcohol derivative 2
glucose release.
to afford carbamate 2b (Fig. 3). Cleavage of the acetate esters
The β-gal activated masked precursor of glucose was using catalytic amounts of potassium carbonate in methanol
designed based on examples of β-gal activated prodrugs afforded compound 1.
(Fig. 1).4 A β-galactose triggering group is attached to glucose
As a control compound, we synthesized a masked glucose
through a self-immolative linker to form precursor type I. The precursor that cannot undergo activation with β-gal (Fig. 4).
precursor is designed to release the glucose moiety upon clea- Curtius rearrangement of acylazide 2a generated an isocyanate,
vage of the head β-galactose from the linker.
which was trapped in situ with benzylalcohol to give carbamate
Self-immolative linkers that “sacrifice” themselves in order 3a. The latter was subjected to mild basic conditions to afford
to implement their essential function are widely used for drug glucose precursor 3.
delivery systems5–8 and diagnostic applications.9,10 We have
Evaluation of β-gal-mediated glucose release
demonstrated the self-immolative disassembly in several mol-
ecular systems such as dendrimers11–15 and polymers16,17 and
To determine whether β-gal would release sufficient quantities
shown that such linkers can efficiently disassemble under
physiological conditions.
of glucose from the masked precursor 1 to promote bacterial
growth, the β-gal expressing E. coli strain BL21(DE3)19 was
grown in the presence of either 0.5 g l−1 or 1.0 g l−1 glucose or
in the presence of the glucose-releasing 1. The concentration
of 1 was normalized such that complete glucose release would
correspond to 0.5 g l−1 or 1.0 g l−1 glucose, respectively.
Design and synthesis of a β-gal activated precursor of glucose
The design of the glucose release system relies on a self-immo-
lative aminal-carbamate linkage we developed earlier (Fig. 2).18 Indeed, substantial bacterial growth was observed in the pres-
The β-gal activated precursor of 1 is composed of a glucose ence of 1 as the only rapidly metabolizable carbon source
(Fig. 5). Bacterial growth was reduced when compared to
molecule attached through a self-immolative linker to β-galac-
tose – a substrate for β-gal. Upon incubation of 1 with β-gal, growth in the presence of glucose indicating that glucose
phenol 1a is obtained and immediately undergoes 1,6-elimi- release from the precursor was incomplete. However, bacterial
growth in the presence of 1 was markedly increased compared
to control compound 3 that does not allow for β-gal-mediated
glucose release. This finding indicates that glucose release
from compound 1 is indeed specifically induced by β-gal clea-
vage. Interestingly, β-gal mediated cleavage of control com-
pound 3 releases one D-galactose residue that also represents a
carbon source, which can lead to E. coli growth. Yet, the
respective gal operon encoding the galactose-metabolizing
enzymes needs to be induced before galactose metabolization
can start.20 Thus, though our system is potentially leaky with
respect to D-galactose release from 3, we did not observe sub-
stantial bacterial growth at least during short incubation times
up to overnight. This finding indicates that bacterial growth
nation to release aminal intermediate 1b. The latter is further
Fig. 2 β-Gal activated glucose precursor to release free glucose.
Fig. 3 Chemical synthesis of β-gal activated glucose precursor 1.
2904 | Org. Biomol. Chem., 2013, 11, 2903–2910
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 4 Chemical synthesis of non-activated glucose precursor 3.
expression (XL-1 blue) in the presence of the masked glucose
precursor 1. As expected, no bacterial growth was observed
(data not shown). Though this E. coli strain might considerably
differ in its growth rate, the experiment provides further evi-
dence for the specificity of the glucose-release system.
Measurement of β-gal expression using a new NIR fluorescent
probe
To detect β-gal expression by the E. coli strain BL21(DE3), we
synthesized a new β-gal probe based on the recently developed
near infrared (NIR) fluorescent dye QCy7 (Fig. 6).21 Probe 4 is
composed of a β-galactose substrate attached to the NIR dye
through a short self-immolative linker. Cleavage of β-galactose
by β-gal generates intermediate 4a, which further disassembles
through 1,6-elimination to release phenolate 4b. The latter can
undergo internal charge transfer to form the NIR fluorescent
dye SulfoQCy7.
The synthesis of probe 4 commenced with the bromination
of benzylalcohol 2 by carbon tetrabromide and triphenyl phos-
phine generating benzylbromide derivative 4c (Fig. 7).
Williamson etherification of 4c with commercially available
4-hydroxy isophthalaldehyde gave ether derivative 4d. Conden-
sation of dialdehyde 4d and indolium 5 followed by cleavage
of the acetate esters by a catalytic amount of potassium
carbonate afforded probe 4.
β-Gal expression can be up-regulated by lactose or lactose
metabolites such as isopropyl-β-D-1-thiogalactopyranoside
(IPTG).19 Since β-gal is only expressed at marginal levels in the
E. coli strain BL21(DE3), bacteria were grown in the presence
of lactose (Fig. 8) or IPTG (data not shown) to induce β-gal
expression. To analyze whether β-gal was present in the bac-
teria or secreted into the medium, bacterial lysates were pre-
pared and β-gal expression was measured in lysates as well as
the culture supernatant by adding probe 4. Indeed, fluor-
escence measurements indicated only a basal, but detectable
Fig. 5 Bacterial growth of E. coli strain BL21(DE3) in an MDG medium contain-
ing either glucose, or equivalent concentrations of 1 or the negative control
compound 3. The β-gal expressing E. coli strain BL21(DE3) was grown in the
presence of glucose, compound 1 or 3 as the only carbon source. While β-gal-
mediated glucose release from compound 1 led to substantial bacterial growth,
no bacterial growth was observed in the presence of control compound 3. Data
are presented as mean
experiments.
SD. Data are representative of two independent
was indeed specifically mediated by β-gal-mediated glucose
release from the masked glucose precursor 1. It should be
noted that in addition to the glucose precursor 1 or control
compound 3 the MDG minimal medium also contains vita-
mins and amino acids that could in theory be exploited as sec-
ondary carbon sources by E. coli. However, no bacterial growth
was observed in minimal medium in the absence of glucose
(data not shown) demonstrating that glucose supplementation
was essential for bacterial growth.
To further confirm the specificity of the system, we deter-
mined the growth rate of an E. coli strain lacking β-gal
Fig. 6 Mechanism β-gal activated probe 4 to release the NIR fluorescent dye SulfoQCy7.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2903–2910 | 2905

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 7 Chemical synthesis of a β-gal fluorescent probe.
fluorescence measurement assay. This finding further demon-
strates the utility of probe 4 to detect β-gal expression.
Conclusions
In summary, we have designed and synthesized a masked
glucose precursor that can be activated by β-galactosidase.
This glucose precursor is sufficient for bacterial growth in a
glucose-deficient medium. Glucose release was specifically
mediated by β-gal as demonstrated by a control compound
where glucose release was impossible. A new β-gal-activated
NIR fluorescent probe was employed to detect overexpression
of β-gal by confocal fluorescence microscopy. This proof-of-
principle study demonstrates that bacteria can accomplish an
evolutionary advantage. By adapting the teachings of this
study, we are currently designing an antibiotic release system
for activation by bacterial resistance proteins. Antibiotic
release from the precursor and “induced suicide” would only
occur in the presence of the resistance protein as a selective
tool to fight the bacterial defense mechanism.
Fig. 8 Fluorescence measurement of β-gal expression by E. coli strain BL21
(DE3) using probe 4. The β-gal expressing E. coli strain BL21(DE3) was grown in
an MDG medium with 0.5 g l−1 glucose, or with 0.5 g l−1 lactose to further
enhance β-gal expression. Bacteria were disrupted and β-gal expression was
measured either in bacterial lysate or supernatant by adding 50 μM of probe 4.
Samples were excited at 590 nm and conversion into SulfoQCy7 was measured
by recording the fluorescence spectrum between 600 nm and 850 nm. Only
marginal expression of β-gal was observed in bacterial lysate (blue curve, solid)
and supernatant (blue curve, dashed) when bacteria were grown in the pres-
ence of glucose. When bacteria were grown in lactose as the only carbon
source, only a slightly increased β-gal level was detected in the supernatant (red
curve, dashed). However,
a markedly increased β-gal expression could be
measured in the bacterial lysate (red curve, solid). The diagram is representative
of two independent experiments.
Materials and methods
Synthesis
expression of β-gal both in bacterial lysates and in the super-
natant when bacteria were grown in the presence of glucose as
General methods. All reactions requiring anhydrous con-
the only metabolizable carbon source (Fig. 8). Notably, β-gal ditions were performed under an argon atmosphere. All reac-
expression was sufficient to promote bacterial growth in the tions were carried out at room temperature unless stated
presence of compound 1 (ref. Fig. 5). Lactose led to an otherwise. Chemicals and solvents were either A.R. grade or
increased β-gal expression, which was mainly detected in the purified by standard techniques. Thin layer chromatography
bacterial lysate indicating that β-gal is only marginally released (TLC): silica gel plates Merck 60 F254: compounds were visual-
into the supernatant.
ized by irradiation with UV light. Flash chromatography (FC):
To further demonstrate that probe 4 is indeed valuable for silica gel Merck 60 (particle size 0.040–0.063 mm), eluent
measuring β-gal expression, the E. coli strain BL21(DE3) was given in parentheses. High pressure liquid chromatography
again grown in either glucose- or lactose-containing medium (HPLC): C18 5u, 250 × 4.6 mm, eluent given in parentheses.
and β-gal was visualized by confocal microscopy (Fig. 9). No Preparative HPLC: C18 5u, 250 × 21 mm, eluent given in par-
β-gal expression was detected in fixed bacteria that had been entheses. 1H-NMR spectra were measured using a Bruker
grown in the presence of glucose suggesting that the limit of Avance operated at 400 MHz as mentioned. 13C-NMR spectra
detection in the confocal microscopy measurement was higher were measured using a Bruker Avance operated at 100 MHz as
compared to spectrometric fluorescence measurement (ref. mentioned. The chemical shifts are expressed in δ relative to
Fig. 8). However, lactose again induced substantial β-gal TMS (δ = 0 ppm) and coupling constants J in Hz. The spectra
expression thus confirming the results obtained in the were recorded in CDCl3 as a solvent at room temperature
2906 | Org. Biomol. Chem., 2013, 11, 2903–2910
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 9 Confocal fluorescence microscopy measurement of β-gal expression by E. coli strain BL21(DE3) using probe 4. The β-gal expressing E. coli strain BL21(DE3)
was grown in an MDG medium with 1.0 g l−1 glucose, or with 1.0 g l−1 lactose to enhance β-galactosidase expression. Cells were pelleted and fixed before 200 μM
of probe 4 was added for β-gal detection. In addition, DAPI was used for staining bacterial DNA. The fluorescence of the cells was visualized by confocal microscopy.
While no β-gal expression was detected in fixed bacteria grown in the presence of glucose, lactose induced a substantial β-gal expression as measured with probe 4.
The experiment was repeated twice with similar results.
unless stated otherwise. All general reagents, including salts with MeOH–H2O–AcOH (1 : 1 : 0.1). The crude product was pur-
and solvents, were purchased from Sigma-Aldrich.
Compound 2b. Compound 222 (150 mg, 0.33 mmol) was 20 min) to give probe 1 (74 mg, 46%) as a white solid.
dissolved in 4 ml dry toluene. Compound 2a18 (240 mg, 1H NMR (400 MHz, MeOD): δ = 7.30 (2H, d, J = 8.4 Hz), 7.08
ified by preparative RP-HPLC (grad. 0–20% ACN in water,
0.55 mmol) was added, followed by a catalytic amount of (2H, d, J = 8.4 Hz), 5.05 (2H, bs), 4.95–4.78 (5H, m), 4.70 (1H,
DBTL. The reaction mixture was heated to 80 °C, stirred for d, J = 8.6 Hz), 3.90 (1H, d, J = 4.5 Hz), 3.81–3.57 (9H, m). 13C
30 min under an Ar atmosphere, and monitored by TLC dyed NMR (100 MHz, MeOD): δ = 162.94, 162.55, 162.48, 162.34,
with CAM (EtOAc–Hex 60 : 40). After completion, the solvent 162.16, 158.41, 158.28, 131.05, 129.96, 128.71, 117.03, 102.12,
was removed under reduced pressure. The crude product was 97.41, 77.39, 76.28, 74.29, 74.13, 73.34, 72.58, 71.56, 70.94,
purified by column chromatography on silica gel (THF–Hex 70.09, 69.55, 66.84, 61.82, 61.73. MS (ESI): m/z calc. for
40 : 60) to give compound 2b (293 mg, quan) as a viscous clear C21H31NO14: 521.2; found: 544.2 [M + Na]+, HRMS: m/z calc.
liquid.
544.1642; found 544.1643. HPLC grad. 0–20% ACN in water
1H NMR (400 MHz, CDCl3): δ = 7.43 (2H, d, J = 7.9 Hz), 7.12 20 min: 14.5 min, λ = 220 nm.
(2H, d, J = 7.9 Hz), 5.86–5.84 (1H, m), 5.63–5.56 (2H, m), 5.40
Compound 3a. Compound 2a (80 mg, 0.18 mmol) was dis-
(1H, d, J = 3.8 Hz), 5.26–5.17 (5H, m), 5.15 (2H, s), 4.97 (1H, solved in 4 ml of dry toluene. Commercially available benzyl-
dd, J = 10.0, 3.4 Hz), 4.37–4.19 (6H, m), 2.39 (3H, s), 2.35 (3H, alcohol (25.9 mg, 0.24 mmol) was added, followed by a catalytic
s), 2.19 (6H, s), 2.17–2.14 (12H, m). 13C NMR (100 MHz, amount of DBTL. The reaction mixture was heated to 80 °C,
CDCl3): δ = 171.44, 171.14, 171.13, 171.03, 170.86, 170.36, stirred for 30 min under an Ar atmosphere and the progress
170.15, 158.92, 152.25, 136.49, 131.55, 130.81, 129.22, 128.98, was monitored by TLC (EtOAc–Hex 40 : 60). After completion,
126.26, 117.72, 100.23, 95.22, 71.74, 71.53, 71.23, 70.68, 69.32, the solvent was removed under reduced pressure. The crude
69.11, 68.40, 67.57, 67.46, 62.61, 62.02, 23.43, 21.92, 21.41, product was purified by column chromatography on silica gel
21.36. MS (ESI): m/z calc. for C37H47NO22: 857.3; found: 880.3 (EtOAc–Hex 30 : 70) to give compound 3a (67 mg, 70%) as a
[M + Na]+. HPLC grad. 10–90% ACN in water 20 min: 17.9 min, white solid.
λ = 220 nm.
1H NMR (400 MHz, CDCl3): δ = 7.34–7.33 (5H, m), 5.54 (1H,
Compound 1. Compound 2b (270 mg, 0.31 mmol) was dis- t, J = 9.8 Hz), 5.19 (2H, d, J = 3.1 Hz), 5.16 (2H, s), 5.04 (1H, t,
solved in 15 ml MeOH. K2CO3 (174 mg, 1.2 mmol) was added J = 9.8 Hz), 4.88 (1H, dd, J = 10.2, 3.1), 4.31–4.01 (4H, m), 2.05
to the suspension and the reaction mixture stirred at room (3H, s), 2.03 (3H, s), 2.01 (3H, s), 2.05 (3H, s). 13C NMR
temperature for 2 h and monitored by RP-HPLC (grad. 10% to (100 MHz, CDCl3): δ = 171.37, 170.77, 170.30, 156.83, 136.68,
90% ACN in water, 20 min). After completion, the reaction 129.24, 128.94, 128.56, 96.48, 71.59, 71.15, 70.64, 69.03, 68.27,
mixture was concentrated under reduced pressure and diluted 67.75, 62.48, 23.56, 21.29, 21.21, 21.17. MS (ESI): m/z calc. for
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2903–2910 | 2907

View Article Online
Paper
Organic & Biomolecular Chemistry
C23H29BNO12: 511.48; found: 534.2 [M + Na]+. HPLC grad. (59 mg, 0.21 mmol) was dissolved in 1 ml Ac2O. The reaction
10–90% ACN in water 20 min: 9.4 min, λ = 260 nm. mixture was stirred for 60 minutes at 80 °C under an Ar atmos-
Compound 3. Compound 3a (50 mg, 0.10 mmol) was dis- phere and the progress was monitored by RP-HPLC (grad.
solved in 5 ml MeOH. A catalytic amount of K2CO3 was added 10%–90 ACN in water, 20 min). After completion, the reaction
to the suspension and the reaction mixture stirred at room mixture was concentrated by evaporation under reduced
temperature for 60 min, and the progress was monitored by pressure and the crude product was taken to the next step
RP-HPLC (grad. 10% to 90% ACN in water, 20 min). After com- without further purification. HPLC grad. 10–90% ACN in water
pletion, the reaction mixture was diluted with 2.5 ml MeOH, 20 min: 15.1 min, λ = 426 nm.
2.5 ml water and 0.5 ml AcOH. The crude product mixture was
Compound 4. The acetate derivative 4e was dissolved in
purified by preparative RP-HPLC (grad. 10% to 90% ACN in 10 ml MeOH. K2CO3 (cat. amount) was added to the suspen-
water, 40 min) to give compound 3 (32 mg, 62%) as a white sion and the reaction mixture stirred at room temperature for
solid.
60 min and the progress was monitored by RP-HPLC (grad.
1H NMR (400 MHz, CDCl3+ drop of MeOD): δ = 7.30–7.25 10%–90 ACN in water, 20 min). After completion, the reaction
(5H, m), 5.05 (2H, s), 5.01 (2H, d, J = 3.1 Hz), 4.78 (1H, t, J = 8.6 mixture was diluted with 3 ml MeOH, 3 ml H2O, 600 μl AcOH,
Hz), 4.67 (1H, t, J = 6.3 Hz), 3.74–3.59 (4H, m), 3.51 (1H, dd, J = and purified by preparative RP-HPLC (grad. 10%–90 ACN in
10.2, 3.1), 3.43 (1H, t, J = 9.2). 13C NMR (100 MHz, MeOD): δ = water, 20 min) to give compound 4 (44 mg, 51%) as an orange
171.85, 158.13, 137.32, 128.91, 128.32, 128.25, 97.68, 74.35, solid.
73.39, 72.65, 70.89, 70.31, 67.13, 61.85. MS (ESI): m/z calc. for
C15H21NO8: 343.13; found: 366.1 [M + Na]+. HPLC grad. 16.4 Hz), 8.55 (1H, d J = 16.2 Hz), 8.35 (1H, dd, J = 8.6, 2.2 Hz),
10–90% ACN in water 20 min: 5.0 min, λ = 260 nm. 8.13 (1H, d, J = 16.4 Hz), 8.04 (1H, d, J = 16.2 Hz), 7.93–7.49
1H NMR (400 MHz, MeOD): δ = 9.16 (1H, s), 8.68 (1H, d, J =
Compound 4c. Compound 2 (130 mg, 0.28 mmol) was dis- (11H, m), 7.23 (2H, d, J = 8.6 Hz), 5.37 (2H, s), 4.98–4.95 (4H,
solved in 4 ml DCM–THF 1 : 1. PPh3 (173 mg, 0.66 mmol) was m), 3.97–3.56 (7H, m), 3.14–2.86 (4H, m), 2.52–2.22 (4H, m),
added and after 3 minutes CBr4 (218 mg, 0.66 mmol) was 2.14 (6H, s), 1.91 (6H, s). 13C NMR (410 MHz, MeOD): δ =
added. The reaction mixture was stirred for 10 minutes at 183.47, 183.22, 163.58, 158.98, 154.54, 148.92, 144.70, 144.63,
room temperature under an Ar atmosphere and the progress 141.44, 138.98, 133.61, 130.53, 130.36, 130.27, 130.02, 128.96,
monitored by TLC (EtOAc–Hex 40 : 60). After completion, the 128.38, 124.81, 123.44, 117.43, 115.69, 115.40, 114.86, 114.50,
reaction mixture was evaporated under reduced pressure. The 102.05, 76.42, 74.19, 72.11, 71.64, 69.60, 61.87, 53.26, 46.35,
crude product was filtered through a small silica gel column 45.99, 30.04, 26.46, 26.22, 24.88, 23.56, 22.09. MS (ESI): m/z
(EtOAc–Hex 40 : 60) and taken for the next step without further calc. for C49H56N2O13S2−: 945.1; found: 967.3 [M + Na]+. HPLC
purification.
grad. 10–90% ACN in water 20 min: 10.6 min, λ = 426 nm.
Compound 4d. Crude 4c (82 mg, 0.16 mmol) was dissolved
in 1 ml dry DMF and commercially available 4-hydroxy-iso-
phthalaldehyde (23.8 mg, 0.16 mmol) was added and the sol-
Minimal medium
ution stirred at room temperature overnight under an Ar The medium used was based on a minimal medium for
atmosphere and the progress was monitored by TLC (EtOAc– growing freezer stocks and working cultures (MDG).23 The
Hex 30 : 70). After completion, the reaction mixture was diluted medium was prepared freshly by mixing autoclaved stocks of
with EtOAc, and was washed with saturated ammonium chlor- salts as 20× (0.5 M) Na2HPO4·7H2O (Sigma-Aldrich), 20×
ide solution. The organic layer was separated, dried over (0.5 M) KH2PO4 (VWR), 20× (1 M) NH4Cl (Serva), 20× (0.1 M)
MgSO4 and evaporated under reduced pressure. The crude Na2SO4 (Merck) and 500× (1 M) MgSO4·7H2O (Fluka). A 1000×
product was purified by column chromatography on silica gel trace metal solution was added, consisting of 50 μM
(EtOAc–Hex 30 : 70) to give compound 4d (54 mg, 63%) as a FeCl3·6H2O (Sigma), 20 μM CaCl2, 10 μM MnCl2·4H2O (Alfa-
white solid.
Aesar), 10 μM ZnSO4·7H2O (Santa Cruz Biotechnology), 2 μM
μM CuCl2·2H2O (Sigma
1H NMR (400 MHz, CDCl3): δ = 10.53 (1H, s), 9.97 (1H, s), CoCl2·6H2O (Santa Cruz),
2
8.37 (1H, d, J = 2.2 Hz), 8.13 (1H, dd, J = 8.7, 2.2 Hz), 7.40 (2H, Aldrich), 2 μM NiCl2·6H2O (Merck), 2 μM Na2MoO4·2H2O
d, J = 8.6 Hz), 7.22 (2H, d, J = 8.7 Hz), 7.07 (2H, d, J = 8.6 Hz), (Merck), 2 μM Na2SeO3·5H2O (AlfaAesar) and 2 μM H3BO3
5.50 (1H, dd, J = 10.5, 2.5 Hz), 5.48 (1H, d, J = 2.6 Hz), 5.26 (Santa Cruz). A 100× BME vitamin solution (Sigma-Aldrich)
(2H, s), 5.13 (1H, dd, J = 10.5, 3.4 Hz), 5.09 (1H, d, J = 7.9 Hz), was added, consisting of 0.1 mg ml−1 D-biotin, 0.1 mg ml−1
4.27–4.15 (2H, m), 4.10 (1H, t, J = 6.6 Hz), 2.21 (3H, s), 2.09 choline chloride, 0.1 mg ml−1 folic acid, 0.2 mg ml−1 myo-
(3H, s), 2.08 (3H, s), 2.04 (3H, s). 13C NMR (100 MHz, CDCl3): δ inositol, 0.1 mg ml−1 niacinamide, 0.1 mg ml−1 D-pantothenic
= 190.88, 189.22, 171.09, 170.98, 170.87, 170.13, 165.61, acid·1/2Ca, 0.1 mg ml−1 pyridoxal·HCl, 0.1 mg ml−1 pyri-
157.84, 136.32, 132.78, 130.59, 130.51, 129.93, 125.82, 117.97, doxine·HCl, 0.01 mg ml−1 riboflavin, 0.1 mg ml−1 thi-
114.28, 100.18, 71.81, 71.48, 71.36, 69.27, 67.51, 62.00, 21.49, amine·HCl.
A 100× non-essential amino acid solution
21.42, 21.34. MS (ESI): m/z calc. for C29H30O12: 570.5; found: (Invitrogen) was added, consisting of 890 μg ml−1 alanine,
609.2 [M + K]+.
1500 μg ml−1 asparagine·H2O, 1330 μg ml−1 aspartic acid,
Compound 4e. A mixture of compound 4d (54 mg, 1470 μg ml−1 glutamic acid, 750 μg ml−1 glycine, 1150 μg ml−1
0.1 mmol), NaOAc (17.2 mg, 0.21 mmol) and compound 521 proline, and 1050 μg ml−1 serine.
2908 | Org. Biomol. Chem., 2013, 11, 2903–2910
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
As a metabolizable carbon source, the MDG medium was in 4% PFA (Merck) dissolved in PBS buffer, pH 7.2. Fixation
supplemented by adding a stock of 1 M glucose (Calbiochem) was carried out at RT for 30 min before washing the cells three
or stocks of the different compounds (10 mM) to the respective times and resuspending them in 5 mM HEPES buffer, pH 7.2.
final concentration. Since the compounds were dissolved in The following compounds were then added to the final con-
dimethyl sulfoxide (DMSO), equivalent amounts of DMSO centrations: DAPI (Invitrogen) to 1 μM, probe 4 to 200 μM,
were added to the medium as a control.
EDTA (pH 7.2) to 4 mM. Cells were incubated at RT for 30 min
and were washed two times in HEPES buffer. The fluorescence
of the cells was visualized by confocal microscopy vs. differen-
Bacterial growth
β-Gal expressing Escherichia coli strain BL21 (DE3) (Invitrogen) tial interference contrast (DIC) on an LSM 700 (Carl Zeiss) fol-
was maintained as glycerol stocks. Preliminary cultures were lowing the instructions for using the software Zen 2009.
grown in LB medium (Roth). Bacteria were grown in 50 ml
sterile glass flasks or 15 ml sterile plastic tubes (Greiner) in a
Statistical analysis
Multitron incubator (Infors) at 37 °C, shaking at 300 rpm. LB
Statistical analyses were performed with the unpaired Student
agar plates were inoculated overnight, single colonies were
t test. All statistical analyses were performed with the Prism
software (GraphPad).
then picked and grown to log phase in LB medium. An aliquot
of 500 μl of the LB liquid culture was taken and washed two
times in a minimal medium. To adapt the bacteria to the
minimal medium, the aliquot was transferred into a 20 ml
MDG medium with glucose added as a single carbon source at
Abbreviations
a final concentration of 1 g l−1. Bacteria were grown overnight
AcOH Acetic acid
and were then directly used for the experiments.
Ac2O
Acetic anhydride
For the bacterial growth assays, 2 ml of MDG medium were
supplemented with glucose at final concentrations of 0.5 g l−1
or 1.0 g l−1 or with the respective compounds (1 or control
compound 3). To assess glucose release-dependent bacterial
growth, concentrations of the compounds were normalized to
equivalent glucose concentrations as in a glucose-containing
medium. Subsequently, bacteria taken from the overnight pre-
culture were inoculated 1 : 200. Bacterial growth was detected
by measuring the absorbance at a wavelength of 600 nm with a
spectrophotometer (Ultrospec 6300pro, Amersham). Each
measurement was performed in triplicate.
K2CO3 Potassium carbonate
DBTL Dibutyltin dilaurate
DCM
DMF
Dichloromethane
N,N′-Dimethylformamide
EtOAc Ethylacetate
Hex n-Hexanes
MeOH Methanol
THF
CAM
Tetrahydrofuran
Ceric ammonium molybdate
Acknowledgements
Fluorescence measurements
We thank Nele Reeg and Georgi Tadeus for their assistance
with bacterial growth assays as part of their internships. D.S.
thanks the Israel Science Foundation (ISF) and the German-
Israeli Foundation (GIF) for financial support. P.H.S. and B.L.
thank the Max Planck Society for the very generous support.
Financial support by the German Federal Ministry of Edu-
cation and Research (to B.L.) is also gratefully acknowledged.
A 2 ml bacterial culture taken from a pre-culture was incubated
overnight in an MDG medium with 0.5 g l−1 D(+)-glucose (Cal-
biochem) or with 0.5 g l−1 D(+)-lactose (Fluka), respectively.
Bacteria were disrupted by treating them on ice with a
Branson-sonifier (B12) for 5 s and were then incubated with
50 μM of probe 4 (stock at 50 mM in DMSO). To prevent degra-
dation of the disrupted bacteria, 1 mM PMSF (Roth) was
added. Bacteria were incubated for 2 h following the bacterial
growth protocol, pelleted afterwards at 10 000 g in an Eppen-
dorf centrifuge and the supernatant was sterile filtered.
To analyze the β-galactosidase-mediated conversion of 4
into SulfoQCy7, samples of bacterial lysate or the supernatant
were excited at 590 nm and the fluorescence spectrum
between 600 nm and 850 nm was recorded on a Fluorescence
spectrometer (LS 55, Perkin Elmer). A serial dilution made
from a 140 U ml−1 stock solution of β-galactosidase (Sigma-
Aldrich) in an MDG medium was used as a positive control.
References
1 D. P. Calfee, Curr. Opin. Infect. Dis., 2012, 25, 385–394.
2 B. Gorke and J. Stulke, Nat. Rev. Microbiol., 2008, 6,
613–624.
3 J. Deutscher, Curr. Opin. Microbiol., 2008, 11, 87–93.
4 T. Legigan, J. Clarhaut, I. Tranoy-Opalinski, A. Monvoisin,
B. Renoux, M. Thomas, A. Le Pape, S. Lerondel and
S. Papot, Angew. Chem., Int. Ed., 2012, 51, 11606–11610.
5 R. J. Amir, M. Popkov, R. A. Lerner, C. F. Barbas 3rd and
D. Shabat, Angew. Chem., Int. Ed., 2005, 44, 4378–4381.
6 A. Gopin, S. Ebner, B. Attali and D. Shabat, Bioconjugate
Chem., 2006, 17, 1432–1440.
Confocal laser scanning microscopy (LSM)
An overnight E. coli BL21 (DE3) culture was incubated in an
MDG medium containing 1 g l−1 glucose or 1 g l−1 lactose,
respectively, as described. Cells were pelleted and resuspended
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2903–2910 | 2909

View Article Online
Paper
Organic & Biomolecular Chemistry
7 K. Haba, M. Popkov, M. Shamis, R. A. Lerner, C. F. Barbas 16 A. Sagi, R. Weinstain, N. Karton and D. Shabat, J. Am.
3rd and D. Shabat, Angew. Chem., Int. Ed., 2005, 44,
716–720.
8 M. Shamis, H. N. Lode and D. Shabat, J. Am. Chem. Soc.,
2004, 126, 1726–1731.
Chem. Soc., 2008, 130, 5434–5435.
17 R. Weinstain, A. Sagi, N. Karton and D. Shabat, Chem. Eur.
J., 2008, 14, 6857–6861.
18 N. Karton-Lifshin and D. Shabat, New J. Chem., 2012, 36,
386–393.
9 E. Sella and D. Shabat, Chem. Commun., 2008, 5701–5703.
10 E. Sella and D. Shabat, J. Am. Chem. Soc., 2009, 131, 19 F. W. Studier and B. A. Moffatt, J. Mol. Biol., 1986, 189, 113–
9934–9936. 130.
11 L. Adler-Abramovich, R. Perry, A. Sagi, E. Gazit and 20 M. J. Weickert and S. Adhya, Mol. Microbiol., 1993, 10, 245–
D. Shabat, ChemBioChem, 2007, 8, 859–862. 251.
12 R. J. Amir, E. Danieli and D. Shabat, Chemistry, 2007, 13, 21 N. Karton-Lifshin, E. Segal, L. Omer, M. Portnoy, R. Satchi-
812–821.
Fainaro and D. Shabat, J. Am. Chem. Soc., 2011, 133, 10960–
13 R. J. Amir, N. Pessah, M. Shamis and D. Shabat, Angew.
Chem., Int. Ed., 2003, 42, 4494–4499.
14 R. J. Amir and D. Shabat, Chem. Commun., 2004, 1614–1615.
10965.
22 N. H. Ho, R. Weissleder and C. H. Tung, ChemBioChem,
2007, 8, 560–566.
15 M. Shamis and D. Shabat, Chem. Eur. J., 2007, 13, 4523–4528. 23 F. W. Studier, Protein Expr. Purif., 2005, 41, 207–234.
2910 | Org. Biomol. Chem., 2013, 11, 2903–2910
This journal is © The Royal Society of Chemistry 2013