Chemical and bacterial reduction of azo-probes: Monitoring a conformational change using fluorescence spectroscopy

Article abstract of DOI:10.1016/j.tet.2013.01.086

Sterically constrained probes 2,4-O-bisdansyl-6,7-diazabicyclo[3.2.1]oct-6- ene (8) and 2,4-O-bispyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-ene (9) exhibit specific dimer fluorescent characteristics (λ_max 555 nm and 511 nm, respectively), attributed to the 2,4-diaxial arrangement of the dansyl or pyrene groups. Reduction of the azo-conformational locking group in (8) and (9) yielded 1,3-bisdansyl-4,6-diaminocyclohexane (16) and 1,3-bispyrenoyl-4,6- diaminocyclohexane (17) in the tetra-equatorial chair conformation, thus minimising interaction of the bisdansyl or bispyrenoyl groups. This induces a change in fluorescence from a cooperative green emission dimer band to a blue-shifted, monomer type fluorescence, with λ_max 448 nm and 396 nm for the reduced forms (16) and (17), respectively. The azo-bond conformational lock can either be reduced under biomimetic conditions (using sodium dithionite) or with bacteria (Clostridium perfringens or Escherichia coli) utilising azo-reductase enzymes. These fluorescent probes have the potential to specifically detect azo-reductase expressing bacteria.

Full text of DOI:10.1016/j.tet.2013.01.086

Tetrahedron 69 (2013) 2758e2766
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Chemical and bacterial reduction of azo-probes: monitoring
a conformational change using fluorescence spectroscopy
a
a
a
a
Nicholas J.W. Rattray , Waleed A. Zalloum , David Mansell , Joe Latimer ,
b
a,
a,
*
*
Mohammed Jaffar , Elena V. Bichenkova , Sally Freeman
School of Pharmacy & Pharmaceutical Sciences, University of, Manchester, Oxford Road, Manchester M13 9PT, UK
Morvus Technology Limited, Ty Myddfai, Llanarthne, Carmarthen SA32 8HZ, UK
a
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Sterically constrained probes 2,4-O-bisdansyl-6,7-diazabicyclo[3.2.1]oct-6-ene (8) and 2,4-O-bispyr-
Received 30 October 2012
Received in revised form 12 January 2013
Accepted 28 January 2013
Available online 8 February 2013
enoyl-6,7-diazabicyclo[3.2.1]oct-6-ene (9) exhibit specific dimer fluorescent characteristics (lmax 555 nm
and 511 nm, respectively), attributed to the 2,4-diaxial arrangement of the dansyl or pyrene groups.
Reduction of the azo-conformational locking group in (8) and (9) yielded 1,3-bisdansyl-4,6-
diaminocyclohexane (16) and 1,3-bispyrenoyl-4,6-diaminocyclohexane (17) in the tetra-equatorial
chair conformation, thus minimising interaction of the bisdansyl or bispyrenoyl groups. This induces
a change in fluorescence from a cooperative green emission dimer band to a blue-shifted, monomer type
Keywords:
Fluorescent probe
Dimer
fluorescence, with lmax 448 nm and 396 nm for the reduced forms (16) and (17), respectively. The
azo-bond conformational lock can either be reduced under biomimetic conditions (using sodium
dithionite) or with bacteria (Clostridium perfringens or Escherichia coli) utilising azo-reductase enzymes.
These fluorescent probes have the potential to specifically detect azo-reductase expressing bacteria.
Ó 2013 Elsevier Ltd. All rights reserved.
Conformational lock
Azo-reductase
Escherichia coli
AMBER
1. Introduction
probe system, with the advantages of lower costs, flexibility and
ease of use. Through application of this type of probe to broth, plate
The ability of fluorescent probes to identify biochemical events
culturing media or cellular suspensions, it is possible to detect the
fluorogenic products of bacterial enzymatic action.
The detection of Escherichia coli using glycosidic enzymes as
a biomarker has been achieved using glycoside-coumarin probes.
8
,9
and abnormalities is an ever expanding area of research. From the
1,2
detection of DNA mismatches, through to the identification of
3
antigens expressed by specific cancer cells and hypoxic cellular
conditions,4 the range of practical applications spans from the
molecular to the multi-cellular level. The detection of bacteria has
traditionally relied upon direct culturing and subsequent micro-
scopic analysis of cellular detail or colonial growth patterns, which
is a comparatively timely process. Research has previously shown
how specific bacteria can be identified using different methods,
4-Methyl-umbelliferyl-
hydrolysis of the ester linkage by
fluorescent 4-methyl-coumarin (2) (Fig. 1). Altering the glycoside
attached to 4-methyl-coumarin can give rise to a range of probes
b
-
D
-glucuronide (1) is a substrate that upon
-glucuronidase, releases highly
b
10
that can detect other enzymes expressed by bacteria, such as b-
10e12
galactosidase or
b
-glucosidase.
However, the limitations of the
5
6
such as conjugation with DNA, PCR based hybridisation and ELISA
4-methyl-coumarin fluorophore are well documented. The narrow
7
assays. Although great sensitivity is achieved by these strategies,
they are expensive multistep procedures that require a high degree
of technical expertise.
Utilising bacterial enzyme expression as an indicative biomarker
is an alternative approach. A fluorescent probe can be designed as
a substrate for the relevant enzyme and emits a fluorescent re-
sponse upon contact. This confers a high degree of specificity to the
*
Corresponding authors. Tel.: þ44 (0) 161 275 2366; e-mail addresses: nicholas.
rattray@manchester.ac.uk (N.J.W. Rattray), elena.bichenkova@manchester.ac.uk
E.V. Bichenkova), sally.freeman@manchester.ac.uk (S. Freeman).
Fig. 1.
b-Glucuronidase catalysed hydrolysis of 4-methyl-coumarin-b-D-glucuronide
(
probe (1) yields fluorescent product 4-methyl-coumarin (2).
0
040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.01.086

N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
2759
working pH range due to the high pK
a
value of the fluorophore is
Here we outline the design and synthesis of two novel fluores-
cent molecular probes, 2,4-O-bisdansyl-6,7-diazabicyclo[3.2.1]oct-
6-ene (8) and 2,4-O-bispyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-ene
(9) (Fig. 4). Fluorescence spectroscopy is used to characterise the
probes and follow their subsequent chemical/biological reductions
that influence their conformation. This research is also supported
by molecular dynamic simulations that are implemented to de-
termine the most likely conformation (axial or equatorial) following
reduction of the azo trigger. Preliminary results on (8) with C.
known to compromise reproducibility. These probes often require
a base work up at the end of the assay in order to increase ionisa-
10,13
tion and release full fluorescence.
In this research we have designed a fluorescent probe that in-
teracts with a different bacterial enzyme system. E. coli and Clos-
tridium perfringens are known to express azo-reductase activity
and it has been proposed that the monitoring of this enzyme using
14e17
18
fluorescence could lead to rapid clinical detection of such bacteria.
21
Previous work has shown that the intramolecular heterodimer
fluorescence-quenched probe 4-O-dabsyl-2-O-dansyl-myo-inosi-
tol-1,3,5-orthoformate (3) can be reduced to (4) and decomposes to
perfringens have been reported in a communication.
19
(5) to detect enzyme-based redox systems (Fig. 2). The fluores-
cent dansyl and a quenching dabsyl group in (3) are located in close
proximity with electron transfer resulting in the static quenching of
the dansyl fluorescence. Reduction of the azo bond within the
2 2 4
dabsyl quenching group with either sodium dithionite (Na S O ) or
osteosarcoma cell extracts (containing bioreductive enzymes)
causes an increase in fluorescence, generating a measurable re-
sponse to the presence of reductive conditions. Earlier work has
also shown that the orthoformate trigger in (6) cleaves in the
presence of acid to give (7) with a ring flip of the myo-inositol ring
2
0
Fig. 4. Fluorescent probes (8) and (9) held in a strained tetra-axial conformation via
the azo locking group.
resulting in fluorescence (Fig. 3).
2
2
. Results and discussion
.1. Synthesis
Synthesis of probes (8) and (9) first required the synthesis of the
azo-linked, tetra-axial precursor, 6,7-diazabicyclo[3.2.1]oct-6-ene-
2
2e24
(
(
2,4)-diol (10), achieved by adapting literature protocols
Fig. 5). Diol (10) was prepared from the bis-epoxidation of 1,4-
cyclohexadiene (11). Ring-opening of the bis-epoxide (12) with
hydrazine gave (13), followed by oxidation with H yielded (10).
2 2
O
Fig. 2. Reduction of the azo bond in non-fluorescent 4-O-dabsyl-2-O-dansyl-myo-
inositol-1,3,5-orthoformate (3) gives a fluorescent signal from 2-O-dansyl-myo-inosi-
tol-1,3,5-orthoformate (5) via the decomposition of intermediate (4).
Fig. 5. Synthesis of key intermediate azo-diol (10). Reagents and conditions: (a)
m-chloroperbenzoic acid, DCM; 31% yield (b) hydrazine, 2-methoxyethanol; 82% yield
2 2 2
(c) H O , H O, 69% yield.
The syntheses of probes (8) and (9), together with mono-dansyl
and mono-pyrene standards (14) and (15), are shown in Fig. 6. Dual
labelled dansyl probe (8) was synthesised in a 55% yield via sulfo-
nation at the 2- and 4-positions of (10) using dansyl chloride in the
presence of sodium hydride. Bis-pyrenoyl labelled probe (9) was
also synthesised from precursor (10) by a Steglich type esterifica-
2
5
tion using pyrene-1-carboxylic acid with the DCC coupling agent
and DMAP catalyst, in a 21% yield.
Fig. 3. Acid induced activation of static quenched non-fluorescent probe (6) generat-
ing fluorescent probe (7).
2
0
Singly labelled dansyl standard (14) was synthesised in a 45%
yield from a regiospecific method using diol (10) and dansyl chloride
in the presence of silver oxide and potassium iodide. A coordination
complex formed at the 2- and 4-positions of the diol with the silver
ion, causes the deprotonation of one of the hydroxyl groups and thus
In this research we have combined these approaches to design
a probe system where reduction of the azo group leads to a con-
formational change of a cyclohexane ring. Cleavage of a reduction-
sensitive locking group by biomimetic chemical reduction or bac-
terial azo-reductase enzymes should confer a notable fluorescent
change within the reporter groups. Tracking of this change via
fluorescence-based spectroscopic methods could lead to the rapid
and sensitive identification of certain bacterial strains.
2
6,27
expediting the mono-sulfonyl esterification.
Singly labelled
pyrene standard (15) was adapted from the method used to syn-
thesise probe (9). A careful addition of reagents was carried out with
an equimolar equivalent of 1-pyrenecarboxylic acid being slowly

2760
N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
p
-stacking interactions between closely located pyrenoyl groups
0
0
within probe (9). For the dansyl probe (8) protons H-3 , H-5 and H-
0
6
experience up-field shifts of 0.20, 0.07 and 0.04 ppm, re-
spectively, upon comparison with the spectrum of the mono-dansyl
standard (14). There is greater evidence of stacking in-
pep
teractions in the bis-pyrene probe (9) when compared with the bis-
dansyl probe (8), which is rationalised by the closer proximity and
increase in size of the aromatic structures participating in the
stacking. Probe (8) is functionalised by tetrahedral sulfonyleester
bonds and a steric repulsion will occur between these large groups,
thus reducing the pep overlap. In the case of probe (9), the pyrene
groups are linked via planar carbonyl ester bonds. A combined ef-
fect of a smaller central ester atom and a planar structure means the
pyrene groups can be located closer and afford a stronger overlap,
when compared with the bis-dansyl derivative (8).
Reduction of the azo-locking group is expected to afford
bis-ammonium intermediates (16ax) and (17ax) (Fig. 7). With the
cyclohexane ring no longer retained in the tetra-axial conforma-
tions, relaxation to the tetra-equatorial conformations (16eq
)
and (17eq) is anticipated, which is next studied using molecular
dynamic simulations.
Fig. 6. Synthesis of azo-probes (8) and (9) and standards (14) and (15) from diol (10).
Reagents and conditions: (a) dansyl chloride, sodium hydride, dry DMF; 55% yield (b)
pyrene-1-carboxylic acid, DCC, DMAP, dry DCM; 21% yield for (9) and 28% yield for (15)
(
c) dansyl chloride, AgO, KI, dry DMF, 45% yield.
ꢀ
added over an hour to a mixture of diol (10), DCC and DMAP at 0 C,
to give (15) in a 28% yield.
It was also desirable to synthesis the reduced forms of (8) and
(9) as standards for fluorescence studies. However, due to technical
difficulties, this was only achieved for probe (8), where (16) was
Ò
generated using a H-Cube Continuous-flow Hydrogenation Re-
actor. Using a Pd/C catalyst accompanied with H
2
generation via
H
2
O electrolysis, (8) was converted into (16) in a 36% yield.
2
.2. Probe characterisation
Novel azo-probes (8) and (9) alongside standards (14), (15) and
1
13
(16) were characterised by a combination of H and C NMR
spectroscopy, atmospheric pressure chemical ionisation mass
spectrometry (APCI-MS), infrared spectroscopy and elemental
analysis. The assignments of proton and carbon signals were
assisted by 1H COSY and DEPT techniques. The numbering of the
1
atoms in each compound is shown in Fig. 6. Analysis of the H NMR
chemical shifts indicated that a proportion of the aromatic protons
from bisdansylated probe (8) and bispyrene probe (9) participate in
1
Fig. 7. Reduction of the azo-locking group within (8) and (9) results in the formation
pe
p
stacking interactions. Selected H NMR chemical shifts and
of the bis-ammonium tetra-axial conformations (16ax
A subsequent ring flip to the more stable tetra-equatorial conformations (16eq) and
17eq) then occurs.
) and (17ax), respectively.
coupling constants for compounds (8), (9), (14) and (15) are shown
in Table 1.
(
Table 1
Chemical shifts (d, ppm) and coupling constants (J, Hz) for selected aromatic protons
2.3. Molecular dynamic simulations
for compounds (8) and (9) alongside (14) and (15) to evaluate
pep stacking
interactions
Upon reduction of the azo bond in (8) and (9) to give (16) and
17), it is possible for three different amine protonation states to be
0
0
0
Dan H-6
Compound
Dan H-3
Dan H-5
(
Bis-dansyl (8)
Mono-dansyl (14)
d
d
8.02 (d), J¼8.0
8.22 (d), J¼8.6
d
d
7.51 (t), J¼8.5
d
d
8.26 (d), J¼8.5
8.30 (d), J¼7.5
formed: neutral, monocation and dication (Fig. 8). It is important to
consider these states since they influence the ability of the cyclo-
hexane ring to undergo the conformational ring flip from tetra-
axial to tetra-equatorial. It is proposed that if the resultant neu-
tral bis-amine was formed, the lone pairs of electrons would repel
each other in the diaxial conformation and aid in the conforma-
tional change to diequatorial. A similar situation is expected in the
bis-ammonium product (dication) where the positive charges
would repel and promote the cyclohexane ring to relax to the tetra-
equatorial conformation. This is the predicted protonation state
7.58 (m)
0
0
0
Compound
Pyr H-1
Pyr H-5
Pyr H-9
Bis-pyrene (9)
Mono-pyrene (15)
d
d
9.16 (d), J¼9.4
9.30 (d), J¼9.5
d
d
7.63 (d), J¼8.7
8.07 (d), J¼8.4
d
d
8.49 (d), J¼8.9
8.65 (d), J¼8.9
By comparison with mono-derivatised standards (14) and (15),
up-field proton shifts are seen in the H NMR spectra of probes (8)
and (9), attributed to the presence of stabilising interactions be-
tween the aromatic rings in each bis-substituted structure. When
1
0
0
0
comparing the pyrene protons H-1 , H-5 and H-9 , up-field shifts of
.14, 0.44 and 0.16 ppm, respectively, were observed, indicating
since the pK
a
of a primary amine group on a cyclohexyl ring is
2
8
0
w9.8. The third possibility involves the mono-protonation of an

N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
2761
in (9) are in orientations that promotes
pep
stacking, an essential
prerequisite for dimer/excimer fluorescence emission.
From Fig. 9 and Table 2 the neutral and dication forms of (16)
and (17) prefer to adopt the tetra-equatorial conformations. In
contrast, the lowest energy conformations of the monocation forms
of (16) and (17) are in the tetra-axial conformation. Viewing the
conformation (not shown), this is attributed to H-bonding between
the protonated and neutral amine groups across the ring. It must be
noted that energetically the dication and neutral forms of (16) and
(17) are in the tetra-equatorial forms and are significantly more
stable than the monocation in the tetra-axial forms (Table 1).
2.4. Fluorescence studies
A comparison of the fluorescence spectra of probe (8) and its
21
reduced form (16), also shown in our communication, indicates
that a change of fluorescence signal from dimer to monomer, re-
spectively, is observed. This is attributed to the strained ring
conformation in (8) no longer being present within the reduced
Fig. 8. Potential protonation states of the diamine product after reduction of the azo
groups to give (16) and (17): ring flip from tetra-axial to tetra-equatorial
eq) conformations. (R¼dansyl or pyrenoyl).
ax
( )
form as the N]N bridge has been cleaved. Subsequently, the
stacking interaction indicative of a dimer is minimised within
16) leading to the blue shifted signal. Probes (8) and (9), together
pep
(
(
amine group (monocation) within the probe. This is likely to create
a hydrogen bond bridge where the lone pairs on both amine groups
participate in proton sharing, locking the ring in the tetra-axial
conformation.
With the reduced amine products (16) and (17) proposed to adopt
the equatorial conformations, further molecular dynamic simula-
tions were carried out to determine fluorophore interactions (Fig. 8).
The products of reduction, 1,3-bisdansyl-4,6-diaminocyclohexane
with the monodansyl or monopyrenoyl standards (14) and (15),
were analysed for their fluorescence emission spectra, shown in
Fig. 10.
From the modelling data, it is proposed that cleavage of the azo-
locking functionality will cause a spatial relocation of each of the
ꢀ
20,29
fluorescent groups to equatorial positions (>10 A apart
thus preventing the group’s ability to form dimer complexes.
Spectra were recorded at a low concentration (5 M) to eliminate
) (Fig. 8),
m
(
16ax
)
and 1,3-bispyrenoyl-4,6-diaminocyclohexane (17ax), re-
any intermolecular aggregation effects that might alter the true
fluorescent signal of each compound. The spectra of mono-
substituted standards (14) and (15) showed locally excited state
emissions that are indicative of dansyl and pyrenoyl monomers
spectively, are expected to ring-flip to the tetra-equatorial confor-
mations, (16eq) and (17eq). Theoretically this will cause a spatial
relocation of each of the fluorescent moieties (>10 A apart
ꢀ
20,29
)
(Fig. 8), thus disabling the groups’ ability to form dimer complexes,
with emission lmax at 444 nm and 395 nm, respectively. In contrast,
with a concomitant fluorescence shift when compared with the non-
reduced probes (8) and (9). Insights into these stacking patterns and
most favourable geometries of the locked and reduced probes were
investigated theoretically. AMBER modelling software was used to
ascertain the theoretical lowest energy structural conformations of
azo-probes (8) and (9) and their reduced analogues (16) and (17),
under aqueous solvation conditions. As can be seen from the struc-
tures in Fig. 9, the fluorescent dansyl groups in (8) and pyrene groups
the emission spectra for probes (8) and (9) showed dimer and
typical green pyrene excimer emission bands at 555 nm and
505 nm, respectively. These large shifts in fluorescence emission
provide further evidence that the close proximity of the fluorescent
groups on the tetra-axial azo-locked cyclohexane ring conforma-
tion in (8) and (9) gives rise to associated
pep stacking of the ar-
omatic groups.
In order to test the response of probes (8) and (9) to act as
sensors for biological reductive conditions, initial reduction under
biomimetic conditions was carried out. The fluorescence spectra of
(
8) (5
biomimetic reducing agent Na
rescence spectra of (8)dS . Upon addition of Na
sequent tracking over 16 h (S eS16), a gradual disappearance of the
m
M) were recorded over a 16 h period after the addition of the
. Fig. 11 shows the initial fluo-
and sub-
2 2 4
S O
0
2 2 4
S O
0
dimer band at 555 nm is noted with the appearance of a blue
shifted structure-less band at 448 nm. A slight red-shift of 4 nm
(upon comparison to standard (14)) is noted, which is attributed to
differences in ionic strength from the Na
2 2 4
S O .
Probe (9) was also subjected to the same biomimetic conditions
used to reduce (8) and the individual time course fluorescence
spectra is shown in Fig. 12. Initial S
51 nm yields a strong green coloured excimer fluorescence band at
max 511 nm. Again, this value is shifted slightly in comparison to
the monomer standard due to environmental effects attributed to
Na . Upon comparison of the consecutive fluorescence spectra
eS12), the excimer band is shown to decrease in intensity whilst
a typical structured pyrene monomer band appeared at 396 nm.
This shift is attributed to the disappearance of stacking
0
excitation of probe (9) at
3
l
2 2 4
S O
(S
0
pep
interactions of the two pyrenoyl moieties in the ground state of (9),
initiated by the reduction and subsequent ring flip of the cyclo-
hexane ring from (17ax) to (17eq).
Fig. 9. Lowest energy conformations of (8) and (9) and their reduced amine analogues
(
16eq) and (17eq) as predicted by simulations on AMBER.

2762
N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
Table 2
Theoretical geometry based energy minimisation calculations for the reduction products (16) or (17) in the three possible protonation states indicated in Fig. 8 (EEL¼Energy of
Electrostatic interactions, EGB¼Energy of Generalised Born)
D
E (kcal/mol)
16 Neutral
16 Monocation
16 Dication
17 Neutral
17 Monocation
17 Dication
Bond stretching
Angular bending
Torsional
Van der Waals
EEL
3.4
11.2
45.0
ꢁ15.7
155.7
ꢁ202.0
22.9
ꢁ195.1
ꢁ174.6
Eq
3.5
16.4
84.6
3.3
7.5
43.5
ꢁ8.7
106.9
ꢁ184.8
24.1
ꢁ182.7
ꢁ190.9
Eq
5.7
9.5
5.6
8.9
5.9
10.1
15.6
ꢁ12.5
66.4
ꢁ170.5
38.7
ꢁ2.1
ꢁ48.7
Eq
14.3
ꢁ13.6
44.2
ꢁ148.8
38.3
10.1
ꢁ40.3
Eq
11.2
ꢁ16.2
ꢁ16.6
ꢁ69.0
37.9
13.8
ꢁ24.5
Ax
ꢁ13.6
23.5
EGB
ꢁ66.1
23.9
ꢁ178.0
ꢁ105.9
Ax
1
1
e4 Van der Waals
e4 EEL
Total
Axial or Equatorial
3
3
2
2
1
1
50
00
50
00
50
00
S
12
S
0
5
0
0
3
65
415
465
515
565
615
Wavelength (nm)
Fig. 12. Fluorescent time course spectra of the real time reduction of 2,4-O-bispyr-
enoyl-6,7-diazabicyclo[3.2.1]oct-6-ene (9) to generate its reduced form (17eq). At 5
concentration the probe was reacted with the biomimetic reducing agent sodium
dithionite (23 mM Na in phosphate buffer). Excitation at 351 nm initially showed
m
M
2 2 4
S O
an excimer type emission at ca. 511 nm. Subsequent tracking of fluorescence over
a 12 h period showed the appearance of a monomer band at ca. 396 nm. Spectra at 2 h
intervals are shown.
2.5. Fluorescence based microbiological assay
The initial biomimetic reduction studies of (8) and (9) showed
Fig. 10. A) Emission spectra of 2,4-O-bisdansyl-6,7-diazabicyclo[3.2.1]oct-6-ene
8, ———) and its monodansylated standard, 2-dansyl-6,7-diazabicyclo[3.2.1]oct-6-
ene-4-ol (14,▬). Excited at 335 nm. B) 2,4-O-Bispyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-
ene (9, ———) and its monopyrene standard 2-pyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-
that a dissociative state (monomer) type emission was observed
upon reduction of the azo locking group to give (16eq) and (17eq),
thus exhibiting a suitably large fluorescence shift, that is, distin-
guishable and quantifiable when compared to the original non-
reduced probe. The probes were then subsequently tested for
their suitability for detection of azo-reductase enzymes expressed
by E. coli and C. perfringens. Due to poor aqueous solubility, probe
(
ene-4-ol (15, ▬). Excited at 351 nm. Spectra were recorded at
5
mM
in DMSO/
ꢀ
100 mM pH 7.4 phosphate buffer (70/30 v/v) at 20 C.
S
16
3
3
2
2
1
1
50
00
50
00
50
00
(
9) was excluded from the microbiological studies. Initial bacterial
cell cultures were tested for evidence of azo-reductase activity
3
0
using a standard amaranth dye absorbance assay (Fig. 13).
S
0
C. perfringens
E. coli
0
0
0
.45
.4
.35
.3
.25
.2
0.15
.1
0.6
.5
0
0
0
0.4
0.3
0.2
0.1
0
5
0
0
0
3
60
410
460
510
560
610
660
0
Wavelength (nm)
0
.05
0
Fig. 11. Fluorescent time course spectra of the real time reduction of 2,4-O-bisdansyl-
,7-diazabicyclo[3.2.1]oct-6-ene (8) to generate its reduced form (16eq). The probe
M) was treated with the biomimetic reducing agent sodium dithionite (23 mM
in 100 mM phosphate buffer). Excitation at 335 nm initially showed a dimer
0
30
0
30
6
Time (min)
Time (min)
(
5 m
2 2 4
Na S O
Fig. 13. UV absorbance based amaranth enzyme viability assay indicating that the azo
bond within the amaranth dye has been reduced causing a decrease in absorbance at
25 nm. This indicates azo-reductase activity in the E. coli and C. perfringens cell culture
supernatants.
type emission at 555 nm. Subsequent tracking of fluorescence over a 16 h period
showed the appearance of a monomer band at 448 nm. Spectra at 2 h intervals are
shown.
5

N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
2763
Probe (8) was tested using E. coli cell culture supernatant and
the reduction was monitored over 12 h using fluorescence (Fig. 14,
A/B). These results showed a modest decrease in dimer fluores-
cence (18%), together with an increase in monomer type fluores-
cence, indicating that some reduction was taking place, attributed
to azoreductive enzymes contained within the E. coli supernatant.
Probe (8) was also tested with C. perfringens supernatant, which
exhibited a substantial shift from dimer to monomer fluorescence
over 12 h (Fig. 14, C/D). The dimer emission dropped by 89%
accompanied by a symmetrical increase in monomer fluorescence
broth dilutions were then carried out to deduce the lowest number
of colony forming units (C.F.U) capable of eliciting a detectable
fluorescent response from (8). Results indicated that after a 14 h
5
time period the cell extract from 1ꢂ10 C.F.U of C. perfringens was
capable of reducing 300 nM (8) and displaying the monomer
fluorescence signal indicative of its reduced form (16).
3. Conclusion
Fluorescence spectroscopy has been used to examine the bio-
mimetic and biological reduction of probes (8) and (9), which
contain an azo bond trigger in a strained tetra-axial cyclohexane
chair conformation. Molecular modelling studies predict the close
proximity of the fluorescent partners in probes (8) and (9), with
for 16eq
.
participation in
pep stacking interactions to account for the for-
mation of dimer/excimer type fluorescent emission. Upon re-
duction of the azo-bond trigger, changes in the fluorescence spectra
and NMR chemical shift data support that the tetra-axial confor-
mations of (8) and (9) undergo a ring flip to the tetra-equatorial
conformations to give (16) and (17). Both probes indicated
changes from dimer/excimer fluorescence to a monomer emission
upon biomimetic and bacterial azo-reductase reduction to (16eq
)
and (17eq) in the tetra-equatorial conformations, with molecular
modelling studies backing up these conformational changes. Hav-
ing tested azo-reductase expression in two bacterial strains we
have shown that, at a proof-of-principle level, how these types of
novel molecular probes may find application in the detection of
enzymatic biomarkers for the identification of bacteria.
4
4
. Experimental
.1. Materials and instrumentation
Fig. 14. Fluorescent spectra and kinetic plots of the reduction of 2,4-O-bisdansyl-6,7-
diazabicyclo[3.2.1]oct-6-ene (8) to generate its reduced form (16eq). At 10
concentration, the probe was treated with E. coli (A,B) or C. perfringens (C,D) cell
culture supernatants known to contain azoreductive enzymes. Excitation at 335 nm
initially showed a dimer type emission at ca. 552 nm (blue line). Subsequent tracking
of fluorescence over a 12 h period (S eS12) showed the appearance of a monomer band
0
at ca. 451 nm (red line). In both cases kinetic profiles (B) and (D) indicate a decrease in
dimer emission for (8) mirrored by an increase in monomer fluorescence for (16eq).
mM
Chemicals were purchased from Aldrich Chemical Co., Gilling-
ham, UK. Syntheses were monitored by thin layer chromatography
on pre-coated 60 F254 silica gel aluminium backed plates (Merck,
Darmstadt). Visualisation of spots for thin layer chromatography
was performed using a UV GL-58 Mineral-Light lamp. Flash column
grade 40e63 mm silica gel (Apollo Scientific, Stockport, UK) was
used in preparative scale column chromatography. Melting points
were determined using a Gallenkamp Melting Point apparatus
microscope (UK). IR spectra were recorded using a Jasco FT/IR 4100
2
.6. Limit of detection
The limit of bacterial detection by probe (8) was also deduced
ꢁ1
spectrometer, resolution 1 cm . NMR spectra were recorded using
(Fig. 15). An overall concentration value of 300 nM of probe (8) was
a Bruker Avance spectrometer equipped with a 5 mm single-axis Z-
determined as the lowest limit of detection by which a fluorescence
signal could be determined above baseline noise (being determined
as three times over this noise baseline level). A series of bacterial
1
gradient quattro nucleus probe, operating at 300 MHz for H and at
13
7
5 MHz for C unless otherwise stated. The spectrometers were
31
running TOPSPIN NMR system software (Version 2.1). Chemical
shifts ( ) are reported in parts per million (ppm), peak positions
relative to Me
d
1
13
4
Si (0.00 ppm) for H NMR and C NMR spectra.
Abbreviations used for splitting patterns are: s, singlet; d, doublet;
t, triplet; m, unresolved multiplet. Mass spectra were recorded at
the Swansea University EPRSC National Mass Spectrometry Service
Centre using a Thermo Scientific LTQ Orbitrap XL Hybrid Fourier
Transform Mass Spectrometer. Elemental analyses were recorded
at the School of Chemistry, University of Manchester.
4.2. Synthesis
4.2.1. 6,7-Diazabicyclo[3.2.1]oct-6-ene-2,4-diol (10). The syntheses
of (10) and its precursors were adapted from previous meth-
2
2e24
ods.
A solution of 6,7-diazabicyclo[3.2.1]octane-2,4-diol (13)
(
500 mg, 3.5 mM) was dissolved in water (5 mL) and 35% hydrogen
Fig. 15. Fluorescence spectra of 300 nM (8) upon exposure to cell supernatant from
peroxide (1.1 mL, 4 M equiv) was stirred in the dark for 72 h. After
completion of the reaction aqueous Raney Nickel slurry (0.1 mL)
was added to quench excess peroxide. The solution was filtered and
concentrated under reduced pressure to yield a crystalline solid.
5
1
ꢂ10 C. perfringens colony forming units containing azo-reductase enzymes. The limit
of detection for this probe system against azo-reductase expression by C. perfringens is
3
5
00 nM of (8) versus 1ꢂ10 colony forming units/mL after a 14 h time period. (Green
line (T
0
)¼0 h red line (T14)¼14 h).

2764
N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
Purification using flash column chromatography (acetone/CH
2
Cl
2
,
40 26 2 4
C H N O : C, 80.25; H, 4.38; N, 4.68. Found: C, 79.97; H, 4.14; N,
1
:2) was carried out to yield 339 mg (69%) of (10) as a colourless
4.42%.
ꢀ
ꢀ
24
ꢁ1
solid. Mp 142e145 C. Lit. 140e142 C. IR
n
max/cm : 1089 (CeO
str), 3312 (OH str).
str), 1298 (OH str), 1716 (N]N cis str), 2935 (CH
2
4.2.4. 2-Dansyl-6,7-diazabicyclo[3.2.1]oct-6-ene-4-ol (14). This pro-
cedure describing the mono-esterification of a diol has been
adapted from Bouzide et al. 6,7-Diazabicyclo[3.2.1]oct-6-ene-2,4-
diol (10) (78 mg, 0.55 mM) was dissolved in dry DMF (10 mL). Sil-
ver oxide (180 mg, 0.8 mM) was added and the mixture was stirred
for 10 min at room temperature. Dansyl chloride (162 mg, 0.6 mM)
and potassium iodide (10 mg, 0.06 mM) were added and the re-
action was stirred at room temperature for 90 min. The mixture was
then passed through a sintered filter and the filtrate was diluted
with ethyl acetate (50 mL). The organic layer was washed with
1
3
3
H NMR (CDCl
3
)
d
H
3
4.95 (2H, dd, J2e1/3eq 3.5 Hz, J4e3eq/5 5.4 Hz, H-
3
27
2
/4), 3.75 (2H, dt, J1e2/6eq 3.5 Hz, J5e4/6eq 5.5 Hz, H-1/5), 2.48 (1H,
2
3
2
d,
J
6axe6eq 12.3 Hz, H-6ax), 1.67 (1H, dt,
J
3eq-2/4 6.5 Hz,
J
3eqe3ax
3
2
15.4 Hz, H-3eq), 1.42 (1H, dt,
J
3axe2/4 3.5 Hz,
J
3axe3eq 15.4 Hz, H-
3
2
13
3
ax), 1.32 (1H, dt, J6eqe1/5 3.5 Hz, J6eqe6ax 12.3 Hz, H-6eq). C NMR
d
C
85.4 (C-2/4), 61.2 (C-1/5), 37.2 (C-3), 23.7 (C-6). MS (APCI) m/z
þ
[MþNH
4
]
160.1.
4.2.2. 2,4-O-Bisdansyl-6,7-diazabicyclo[3.2.1]oct-6-ene (8). Sodium
hydride (60%, 60 mg, 1.4 mM) was suspended in dry DMF (20 mL)
and the mixture was stirred at room temperature for 10 min under
a nitrogen atmosphere. Diol (10) (100 mg, 0.7 mM) was added and
the mixture was stirred for 10 min. Dansyl chloride (450 mg,
6ꢂ5 mL of water, dried over MgSO
4
and reduced in vacuo. The re-
sultant crude solid was crystallised from chloroform to yield 93 mg
ꢀ
ꢁ1
(45%) of (14). Mp 78e84 C. IR
n
max/cm 1171, 1351 (S]O str), 1572
1
3
(N]N str cis), 3388 (OH str). H NMR (CDCl
3
)
d
H
8.62 (1H, d, J
1
0
e2
e2
0
0
3
0
3
1
.7 mM) was added and the mixture was stirred under a nitrogen
atmosphere for 12 h. Completion of the reaction was monitored by
TLC (ethyl acetate/hexane, 2:1) R 0.38. The reaction was quenched
8.6 Hz, H-1 ), 8.30 (1H, d, J 7.5 Hz, H-6 ), 8.22 (1H, d, J
6 e5 3
0
0
0
0
0
0
0
3
4 e5
8.6 Hz, H-3 ), 7.57e7.48 (2H, m, H-2 /5 ), 7.20 (1H, d, J 7.5 Hz,
0 0
0
3
3
f
H-4 ), 5.16 (1H, t, J4e3eq/5 4.8 Hz, H-4), 4.95 (1H, t, J2e1/3eq 4.9 Hz,
by the addition of water (20 mL) and the mixture was diluted with
ethyl acetate (100 mL). The layers were separated and the organic
layer was washed with water (6ꢂ25 mL), followed by 10% aqueous
H-2), 4.61e4.53 (1H, m, H-5), 3.98e3.88 (1H, m, H-1), 2.90 (6H, s,
2
H-10), 2.51 (1H, d, J8axe8eq 12.8 Hz, H-8ax), 1.48e1.43 (2H, m, H-3ax
/
13
3eq), 1.28e1.16 (1H, m, H-8eq). C NMR
d
C
152.0, 132.1, 131.3, 130.7,
NaHCO
3
(2ꢂ20 mL). The organic layer was dried (MgSO
4
), filtered
129.9, 129.7, 129.0, 123.2, 119.0, 115.7 (naptheC), 84.2 (C-4), 81.4
and the filtrate concentrated under reduced pressure. The residue
was purified by flash column chromatography (ethyl acetate/hex-
(C-2), 70.7 (C-5), 59.9 (C-1), 45.4 (NMe
2
), 33.2 (C-3), 23.7 (C-6). MS
þ
(APCI) m/z [MþNH
4
]
376.13. Elemental analysis calcd for
ꢀ
ane 2:1) to yield 233 mg (55%) of a yellow solid (8). Mp 134e138 C.
18 21 3 4
C H N O S: C, 57.58; H, 5.64; N, 11.19. Found: C, 57.19; H, 5.32, N,
ꢁ1
1
IR
n
max/cm : 1172, 1358 (S]O str), 1572 (N]N str cis). H NMR
10.98%.
3
0
0 0
8.0 Hz, H-1 ), 8.27e8.22 (4H,
1 e2
(
400 MHz, CDCl
3
)
d
H
8.64 (2H, d, J
0
0
0
0
3
m, H-3 , H-6 ), 7.61e7.52 (4H, m, H-2 , H-5 ), 7.26 (2H, d,
J
4
0
e5
0
4.2.5. 2-Pyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-ene-4-ol
(15). Diol
0
3
8
5
.5 Hz, H-4 ), 5.02 (2H, dt, H-2/4), 4.54 (2H, dt,
J
1e2/8eq, 5e4/8eq
(10) (50 mg, 0.35 mM), DCC (77 mg, 0.37 mM) and a catalytic
.0 Hz, H-1/5), 2.90 (12H, s, H-7 ), 2.48 (1H, d, 2J3axe3eq 13.8 Hz,
0
amount of DMAP (2 mg, 0.016 mM) was dissolved in dry DCM
2
3
ꢀ
H-3ax), 1.67 (1H, dt,
J
8axe8eq 17.3 Hz,
8eqe8ax 17.3 Hz, H-8eq), 1.20 (1H, dt,
J
8axe1/5 4.7 Hz, H-8ax), 1.36
3eqe3ax 13.8 Hz,
132.1, 131.3, 130.4, 129.8, 129.6,
(10 mL). This mixture was then cooled to 0 C. Separately,
2
2
(
1H, d,
J
J
1-pyrenecarboxylic acid (111 mg, 0.45 mM) was dissolved dry DCM
(10 mL). Both mixtures were kept under an inert nitrogen atmo-
sphere and left to stir for 10 min. 1 mL of the pyrene solution was
added every 5 min to the diol solution. After 45 min, the addition
was complete and the mixture was left to stir in the dark for 24 h.
Completion of the reaction was monitored by TLC (ethyl acetate/
3
13
J3eqe2/4 5.0 Hz, H-3eq). C NMR
d
C
0
1
29.3, 123.0, 119.1, 115.8, 81.6 (napth), 68.8 (C-2/4), 45.4 (C-10/10 ),
þ
41.0 (C-1/5), 30.6 (C-3), 24.0 (C-6). MS (APCI) m/z [MþNH
4
]
608.2.
Elemental analysis calcd for C30 : C, 59.19; H, 5.30; N,
32 4 6 2
H N O S
9
.20. Found: C, 58.94; H, 5.02; N, 9.14%.
f
hexane, 1:1) R 0.15. Filtration of the reaction mass was carried out
4.2.3. 2,4-O-Bispyrenoyl-6,7-diazabicyclo[3.2.1]oct-6-ene (9). Diol
to remove any precipitated dicyclohexylurea, then the filtrate was
(10) (230, 1.6 mM), 1-pyrenecarboxylic acid (1.009 g, 4.1 mM) and
evaporated in vacuo. The residue was dissolved in DCM, washed
DMAP (19.5 mg, 0.16 mM) were dissolved in dry DCM (20 mL)
under a nitrogen atmosphere. The mixture was then cooled to 0 C
twice with 0.5 N HCl and saturated NaHCO
3
, and then dried over
ꢀ
MgSO . Evaporation of solvent and flash column chromatography
4
before DCC (700 mg, 3.4 mM) was added, which was stirred for
(ethyl acetate/hexane, 1:1) yielded 32 mg (28%) of a brown solid
ꢀ
ꢀ
ꢀ
ꢁ1
5
min at 0 C and 24 h at 20 C. Completion of the reaction was
(15). Mp 102e106 C. IR
str), 3270 (OH str). H NMR (CDCl
n
max/cm : 1593 (N]N str cis), 1711 (C]O
1
3
monitored by TLC (ethyl acetate/hexane, 1:1) R 0.28. Filtration of
f
3
)
d
H
9.30 (1H, d,
J
1
0
e2
0
9.5 Hz,
0
3
0
0
0
0
the reaction removed any precipitated dicyclohexylurea, then the
filtrate was evaporated in vacuo. The residue was dissolved in DCM,
H-1 ), 8.65 (1H, d, J 8.9 Hz, H-9 ), 8.16e8.09 (6H, m, H-2 /3 /4 /
9 e8
0 0
3
0
0
0
0
0 0
8.4 Hz, H-5 ), 5.54e5.47 (2H, m, H-2/4),
5 e6
6 /7 /8 ), 8.07 (1H, d, J
5.35e5.31 (1H, m, H-1), 4.27e4.25 (1H, m, H-5), 2,76 (1H, d,
J3eqe3ax 12.55 Hz, H-3eq), 2.00 (1H, br s, HeOH), 1.93e1.83 (1H, m,
washed twice with 0.5 N HCl and saturated NaHCO
3
, then dried
2
over MgSO . The organic layer was subsequently concentrated
4
13
under vacuum and the residue was purified via flash column
chromatography (ethyl acetate/hexane 1:3) to yield (9) 202 mg
C
H-8ax/8eq), 1.57e1.51 (1H, m, H-3ax). C NMR d 166.8 (C]O), 134.7,
131.4, 131.0, 130.4, 129.9, 129.8, 128.4, 127.2, 126.5, 126.5, 126.4,
ꢀ
ꢁ1
(
21%) as a brown solid. Mp 178e180 C. IR
n
max/cm : 1249 (CeO
124.9, 124.7, 124.2, 124.1, 122.4 (PyreC), 84.4 (C-4), 81.3 (C-2), 63.8
1
þ
str), 1595 (N]N cis str), 1701 (C]O str). H NMR (CDCl
3
)
d
H
9.16
(C-1), 60.6 (C-5), 33.2 (C-3), 24.4 (C-6). MS (APCI) m/z [MþNH
371.23. Elemental analysis calcd for C23
N, 7.56. Found: C, 74.43; H, 4.61; N, 7.23%.
4
]
3
0
3
0
(
2H, d, J
1
0
e2
0
9.4 Hz, H-1 ), 8.49 (2H, d, J
9
0
e8
0
8.9 Hz, H-9 ), 8.19 (2H,
9.4 Hz, H-2 ), 8.10 (2H, d, J 8.9 Hz, H-8 ), 7.95e7.88 (6H,
8.7 Hz, H-6 ), 7.63 (2H, d,
4e3eq/5 5.5 Hz, H-2/4), 5.64 (2H, t,
1e2/8eq, 5e4/8eq 6.0 Hz, H-1/5), 2.89 (1H, d, J8axe8eq 12.6 Hz, H-8ax),
18 2 3
H N O : C, 74.58; H, 4.90;
3
0
3
0
d, J
m, H-3 , 4 , 7), 7.82 (2H, d,
2 e1 8 e9
0
0
0
0
0
0
3
0
3
J
6
0
e5
0
J
5
0
e6
0
0
3
8
.7 Hz, H-5 ), 5.73 (2H, t, J2e1/3eq
,
4.2.6. 1,3-Bisdansyl-4,6-diaminocyclohexane (16). Azo-linked probe
(8) (50 mg) was dissolved in dry DMF and cycled within a Tha-
3
2
J
.31 (1H, d, 2
J
3eqe3ax 17.1 Hz, H-3eq), 2.01 (1H, dt,
3axe3eq 17.1 Hz, H-3ax), 1.67 (1H, dt,
3
J
3axe2/4 5.5 Hz,
lesNano H-Cube continuous flow hydrogenation reactor for 30 min
Ò
2
2
3
8eqe1/5 6.0 Hz, 2J8eqe8ax
J
J
whilst passing through a Pd/C catalyst cartridge. The internal
hydrogen source was obtained from H O electrolysis and the pres-
2
13
1
1
1
2.6 Hz, H-8eq). C NMR
29.9, 129.8, 128.4, 126.9, 126.5, 126.3, 126.3, 124.7, 124.4, 123.9,
23.8, 122.6 (PyreC), 81.6 (C-2/4), 63.4 (C-1/5), 33.9 (C-3), 24.9
d
C
166.4 (C]O), 134.4, 131.3, 130.9, 130.1,
sure within the system was maintained at 60 bar alongside a system
ꢀ
temperature of 60 C. Upon completion, concentration under
þ
(
C-6). MS (APCI) m/z [MþNH
4
]
598.2. Elemental analysis calcd for
reduced pressure yielded a brown residue that when purified by

N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
2765
flash column chromatography (ethyl acetate/hexane, 2:1) gave
4.5. Molecular dynamic simulations
1
1
9 mg (36%) of a yellow solid (16). H NMR for reduced probe (16)
3
0
(
400 MHz, DMSO-d
J
6
)
d
H
8.65 (2H, d, J
1
0
e2
0
8.2 Hz, H-1 ), 8.04 (2H, d,
The structures of the molecular probes (8, 9), and their reduced
forms (16, 17) in the neutral, monocation and dication states were
built using SYBYL software 8.0. Atomic point charges were derived
from the AM1-BCC charge model with ANTECHAMBER in AMBER10.
Molecular dynamics (MD) simulations for all molecular probes
3
0
3
0
0
0
7.9 Hz, H-6 ), 7.86 (2H, d, J
, H-2 ), 7.64e7.55 (2H, m, J
0
0
8.0 Hz, H-3 ), 7.70e7.60 (2H, m,
6
e5
3
e2
3J
0
3
, H-5 ), 7.33 (2H, d, J
0
3
0
0
0
0
0
0
0
0
2
e1
/3
5
e4
/6
4
e5
0
8
2
.0 Hz, H-4 ), 4.55e4.51 (2H, m, H-2/4), 3.94e3.88 (2H, m, H-1/5),
0
2
.80 (12H, s, H-7 ), 2.31 (1H, d, J3axe3eq 14.2 Hz, H-3ax), 2.20e2.12
3
2
(
1H, m, H-8ax), 1.70e1.62 (1H, m, H-8eq), 1.58e1.50 (1H, m, H-3eq
)
were carried out using AMBER10 with a GAFF force field under an
implicit water system applying the Generalised Born continuum
þ
4
] 612.2.
m/z [MþNH
3
3
solvent mode. The system was minimised using the steepest
descent and then conjugate gradient algorithms for 500 cycles to
remove any strains. Then, the system was heated to 300 K for 15 ps
with a gradual removal of the restraints on (8) and (9). Finally, the
10 ns molecular dynamics simulation was performed without any
restraints at a constant temperature of 300 K and a constant
pressure of 1 atm. The SHAKE algorithm of the hydrogen atoms
bond length was used. Cut-off distance for the calculation of non-
bonded forces was set to 16 Ǻ. The integration time step was set
at 2 fs and the energies and atomic coordinates were saved every
1 ps.
4
.3. UV and fluorescence studies
UVevisible spectra were recorded in 1.4 mL Hellma QG Far-UV
quartz cuvettes using a Cary 4000 UV-visible spectrophotometer
equipped with a Peltier-thermostated cuvette holder. pH was
measured using a Hanna-instruments HI 9321 microprocessor pH
ꢀ
meter, calibrated with standard buffers (Sigma) at 20 C. Fluores-
cence emission and excitation spectra were recorded in 1.4 mL
Hellma QS Fluorescence quartz thermostated cuvettes using
a temperature controlled Cary-Eclipse fluorescence spectropho-
tometer. Emission and excitation spectra were recorded in 0.1 M
phosphate buffer pH 7.4. The excitation wavelength used for com-
pound (8) was 335 nm and for compound (9) was 351 nm (both
values were determined by the absorption bands on UV analysis).
Slit-widths were 10 nm. Photo bleaching of the sample was mini-
mised by using the automatic shutter on function.
4.6. Molecular dynamics and energy minimisation studies
The 15 ns of each MD simulation was clustered by Kclust script,
where a fixed cluster radius (based on Cartesian coordinates RMSD)
was enabled. Structures (8, 9, 16, 17 and charge study variants)
obtained from the implicit water MD simulation were clustered
with a fixed radius of 3.0 Ǻ and in the case of (9) a 2 Ǻ distance was
used. These returned many clusters with only one highly popu-
lated. The structure with the least RMSD to the ‘centroid’ of the most
populated cluster was selected as a representative structure of each
molecule (Fig. 9). These representative structures were again
energy-minimised using SYBYL to determine the energy of lowest
energy conformation (Table 2).
To monitor the biomimetic reduction of (8) and (9) in real-time
(
with subsequent decomposition to (16) and (17)), 4.2 mg of
sodium dithionite (23.0 mM) was added to the quartz cuvette
containing (8) or (9) (at 5
mM) dissolved in 1 mL of DMSO/pH 7.4
ꢀ
P.B.S mixture (50/50 v/v). The reaction mixture was stirred at 20 C
with the first fluorescence spectrum being recorded immediately
after the addition of sodium dithionite.
The UV amaranth dye assay confirmed the ability of E. coli and C.
perfringens to produce azo-reductase enzymes. An aliquot (1 mL) of
cultured cell media was placed in a cuvette alongside 10
mg of
Acknowledgements
amaranth dye (10 L of a 1 mg/mL stock). The absorbance maxima
m
at 525 nm was then noted. Readings were then taken after 30 min
to deduce the extent of amaranth dye reduction.
We thank Dr. R. Bryce for the use of computational facilities and
Dr. A. McBain for helpful discussions regarding the microbiology
work. We thank the EPSRC National Mass Spectrometry Centre in
Swansea for MS analysis. NR thanks Morvus Technologies and the
BBSRC for a CASE studentship.
8
The reduction of probe (8) using E. coli (4.6ꢂ10 colony forming
7
units per mL) and C. perfringens (1.8ꢂ10 colony forming units per
mL) cell culture supernatants was also tracked by fluorescence. A
10 mM concentration of the probe was observed over a 12 h period
in a bacterial thioglycollate broth supernatanteDMSO mix (50/50
v/v at pH 7.4). Spectra were recorded every hour and corrected
against background interference.
References and notes
1. Bichenkova, E. V.; Gbaj, A.; Walsh, L.; Savage, H. E.; Rogert, C.; Sardarian, A. R.;
Etchells, L. L.; Douglas, K. T. Org. Biomol. Chem. 2007, 7, 1039e1051.
. Bichenkova, E. V.; Yu, X.; Bhadra, P.; Heissigerova, H.; Pope, S. J. A.; Coe, B. J.;
Faulkner, S.; Douglas, K. T. Inorg. Chem. 2005, 44, 4112e4114.
. Liu, T.; Wu, L. Y.; Hopkins, M. R.; Choi, J. K.; Berkman, C. E. Bioorg. Med. Chem.
Lett. 2010, 20, 7124e7126.
2
3
4
.4. Microbiology
For the initial reduction profile of (8) going to (16), stock cultures
8
4. Komatsu, H.; Harada, H.; Tanabe, K.; Hiraoka, M.; Nishimoto, S. Med. Chem.
Commun. 2010, 1, 50e53.
of E. coli (4.6ꢂ10 colony forming units per mL) and C. perfringens
7
(
1.8ꢂ10 colony forming units per mL) were maintained on brain
5
. Wang, Z. P.; Xu, H. A.; Wu, J.; Ye, J.; Yang, Z. Food Chem. 2010, 125, 779e784.
6. Yu, G. F.; Niu, J. J.; Shen, M. S.; Shao, H. J.; Chen, L. Clin. Chim. Acta 2006, 366,
81e286.
. Paffard, S. M.; Miles, R. J.; Clark, C. R.; Price, R. G. J. Immunol. Methods 1996, 192,
33e136.
heart infusion agar (Oxoid, UK). Single colonies were used to
inoculate 5 mL brain heart infusion/thioglycollate broth (Oxoid,
UK), which was incubated for 24 h. Cultures were then incubated
for a further 30 min. All of the above steps were carried out under
2
7
1
8. Zaytsev, A. V.; Anderson, R. J.; Bedernjak, A.; Groundwater, P. W.; Huang, Y.;
Perry, J. D.; Orenga, S.; Roger-Dalbert, C.; James, A. Org. Biomol. Chem. 2008, 6,
682e692.
. Feng, P. C.; Hartman, P. A. Appl. Environ. Microbiol. 1982, 43, 1320e1329.
10. Manafi, M.; Kneifel, W.; Bascomb, S. Microbiol. Rev. 1991, 55, 335e348.
11. Dealler, S. F. Rev. Med. Microbiol. 1993, 4, 198e206.
2. Perry, J. D.; James, A. L.; Morris, K. A.; Oliver, M.; Chilvers, K. F.; Reed, R. H.;
Gould, F. K. J. Appl. Microbiol. 2006, 101, 977e985.
3. Gee, K. R.; Sun, W. C.; Bhalgat, M. K.; Upson, R. H.; Klaubert, D. H.; Latham, K. A.;
Haugland, R. P. Anal. Biochem. 1999, 273, 41e48.
III
strictly anaerobic conditions in a Mk workstation (Don Whitley,
UK) at 37 C. Cell culture aliquots (1 mL) were then centrifuged at
ꢀ
9
1
3,000 rpm for 4 min and the culture supernatant was passed
through a 0.22 m syringe filter. For the limit of detection study, C.
perfringens was grown in a thioglycollate broth and a cell count of
m
1
7
1
.8ꢂ10 colony forming units per mL was noted. Initially, a DMSO/
1
6
broth supernatant mixture (50/50 v/vd9ꢂ10 colony units/mL)
1
1
1
4. Chen, H.; Hopper, S. L.; Cerniglia, C. E. Microbiology 2005, 151, 1433e1441.
5. McBain, A. J.; Macfarlane, G. T. J. Med. Microbiol. 1998, 47, 407e416.
6. Stingley, R. L.; Zou, W.; Heinze, T. M.; Chen, H. Z.; Cerniglia, C. E. J. Med.
Microbiol. 2010, 59, 108e114.
was applied to a 300 nM sample over 15 h. This gave the initial high
concentration value which was reduced by serial dilution to give
5
a reduced bacterial cell unit count of 3 ꢂ 10 /mL.

2766
N.J.W. Rattray et al. / Tetrahedron 69 (2013) 2758e2766
1
1
7. McBain, A. J.; MacFarlane, G. T. J. Med. Microbiol. 2001, 50, 833e842.
8. James, A. L.; Perry, J. D.; Jay, C.; Monget, D.; Rasburn, J. W.; Gould, F. K. Lett. Appl.
Microbiol. 2001, 33, 403e408.
25. Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17, 522e524.
26. Sureshan, K. M.; Shashidhar, M. S.; Praveen, T.; Gonnade, R. G.; Bhadbhade, M.
M. Carbohydr. Res. 2002, 337, 2399e2410.
1
9. Kadirvel, M.; Rattray, N. J. W.; Rajendran, R.; Zalloum, W. A.; Gbaj, A.;
Demonacos, C.; Bichenkova, E. V.; Freeman, S. New. J. Chem. 2011, 35, 701e708.
0. Kadirvel, M.; Gbaj, A.; Mansell, D.; Miles, S. M.; Arsic, B.; Bichenkova, E. V.;
27. Bouzide, A.; Sauve, G. Org. Lett. 2002, 4, 2329e2332.
28. Brown, H. C. Determination of Organic Structures by Physical Methods; Academic:
New York, NY, 1971.
2
Freeman, S. Tetrahedron 2008, 64, 5598e5603.
1. Rattray, N. J. W.; Zalloum, W. A.; Mansell, D.; Latimer, J.; Schwalbe, C. H.; Blake,
29. Turro, N. J. Modern Molecular Photochemistry of Organic Molecules; University
Science Books: 2010.
2
A. J.; Bichenkova, E. V.; Freeman, S. Chem. Commun. 2012, 6393e6395.
30. Rau, J.; Knackmuss, H.-J.; Stolz, A. Environ. Sci. Technol. 2002, 36, 1497e1504.
31. Crummett, W. Anal. Chem. 1980, 52, 2242e2249.
32. Tsui, V.; Case, D. A. Biopolymers 2000, 56, 275e291.
33. Cieplak, P.; Caldwell, J.; Kollman, P. J. Comput. Chem. 2001, 22, 1048e1057.
2
2
2
2. Chenevert, R.; Jacques, F. Tetrahedron: Asymmetry 2006, 17, 1017e1021.
3. Kavadias, G.; Velkof, S.; Belleau, B. Can. J. Chem. 1978, 56, 404e409.
4. Saumi, T. J. J. Org. Chem. 1975, 40, 456e461.