Phenothiazinium-fluoroquinolone drug conjugates

Article abstract of DOI:10.1016/j.ijantimicag.2009.11.022

Synthesis and antibacterial screening of a homologous series of 3-dialkylaminophenothiazinium-7-norfloxacin conjugates was carried out alongside a corresponding series of symmetrical methylene blue derivatives. The norfloxacin conjugates maintained typical methylene blue derivative photoproperties, such as long wavelength absorption, but produced no measurable singlet oxygen in the standard assay and provided no significant increase in the magnitude of photoantibacterial action, this being similar to the methylene blue homologues, although both the conjugates and homologues were considerably more active than methylene blue itself both against Staphylococcus aureus and Escherichia coli. DNA binding via intercalation was considerably greater for the series of norfloxacin conjugates than for the methylene blue homologues.

Full text of DOI:10.1016/j.ijantimicag.2009.11.022

International Journal of Antimicrobial Agents 35 (2010) 405–409
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents
journal homepage: http://www.elsevier.com/locate/ijantimicag
Short communication
Phenothiazinium–fluoroquinolone drug conjugates
Mark Wainwright∗, Ami Shah, Katie Meegan, Ciara Loughran, Andy Smith, Nasima Valli, Nicola Dempster
School of Pharmacy and Biomolecular Sciences, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 October 2009
Accepted 30 November 2009
Synthesis and antibacterial screening of a homologous series of 3-dialkylaminophenothiazinium-7-
norfloxacin conjugates was carried out alongside a corresponding series of symmetrical methylene blue
derivatives. The norfloxacin conjugates maintained typical methylene blue derivative photoproperties,
such as long wavelength absorption, but produced no measurable singlet oxygen in the standard assay
and provided no significant increase in the magnitude of photoantibacterial action, this being similar to
the methylene blue homologues, although both the conjugates and homologues were considerably more
active than methylene blue itself both against Staphylococcus aureus and Escherichia coli. DNA binding via
intercalation was considerably greater for the series of norfloxacin conjugates than for the methylene
blue homologues.
Keywords:
DNA
Fluoroquinolone
Methylene blue derivatives
Phenothiaziniums
Photodynamic antimicrobial chemotherapy
(PACT)
© 2009 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction
[3,4]. Anionic porphyrins have also been examined in conjunction
with the cationic peptide antibiotic polymyxin, with similar intent
Methylene blue and a number of structurally related phe-
nothiazinium derivatives have been reported to be efficacious
photoantimicrobials against a wide range of bacteria, fungi and
viruses [1]. Red light absorption by such compounds (620–670 nm)
is sufficient to circumvent endogenous absorption by natural chro-
mophores, e.g. haem and melanin, allowing the use of methylene
blue derivatives in body fluids and tissues.
Other classes of photosensitiser have been similarly investi-
gated. The porphyrin class is chief among these, mainly due to
its majority share in the clinical photodynamic therapy (PDT)
[5].
However, such chromophore attachment to peptides and poly-
mers suffers from the fact that there are multiple sites of
attachment, in this case between free carboxylic acid residues in
the chlorin ring and numerous amine residues in the polymers,
often chemically equivalent. Consequently, product mixtures are
formed rather than pure material and this may be disadvantageous
from the point of view both of structure–activity relationships and
of future clinical product registration.
Novel methylene blue derivatives are usually produced by
altering the auxochromic amine groups at positions 3 and 7
of the phenothiazinium ring, typically providing homologous
series, and both symmetrical and asymmetrical examples have
been reported in the literature [6,7]. However, owing to the
broad-spectrum activity associated with methylene blue deriva-
tives, the use of biomolecule attachment to the phenothiazinium
chromophore has received scant attention, although in the anti-
cancer application attachment of methylene blue via an extra
amino group at C-4 to oligonucleotides has been reported
[8].
of cancer, although this may not supply
a sound rationale
for antimicrobial discovery. Indeed, whereas the cationic phe-
nothiazinium class exhibits broad-spectrum antibacterial activity,
anionic porphyrins used in PDT, such as protoporphyrin IX or
haematoporphyrin derivative, are normally only effective against
Gram-positive bacteria [2].
To remedy this lack of activity, various groups have investi-
gated the attachment of targeting or adjuvant moieties to anionic
porphyrin photosensitisers. Thus, polycationic molecules, such as
poly(lysine) and poly(ethyleneimine), have been attached to chlo-
rin e6 (Ce6), the resulting polycationic side chains having the effect
of permeabilising the outer membrane of Gram-negative bacte-
ria allowing photosensitiser ingress and thus photocytotoxicity
As part of an ongoing photosensitiser discovery programme,
it was considered that it might be possible to use small bioac-
tive molecules to aid in photoantibacterial targeting. Given the
requirement for single, pure, active species in drug development,
the reactive intermediate cation employed in asymmetrical
phenothiazinium synthesis was used as a starting point, since this
would guarantee a single active site for coupling with amines.
∗
Corresponding author. Tel.: +44 151 231 2039; fax: +44 151 231 2170.
E-mail address: mark wainwright@hotmail.com (M. Wainwright).
0924-8579/$ – see front matter © 2009 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
doi:10.1016/j.ijantimicag.2009.11.022

406
M. Wainwright et al. / International Journal of Antimicrobial Agents 35 (2010) 405–409
Fig. 1. Synthetic routes to the phenothiazinium homologues. Reagents: (i) bromine in acetic acid; (ii) dialkylamine in dichloromethane; (iii) iodine in dichloromethane; (iv)
dialkylamine in methanol; and (v) norfloxacin in methanol.
The fluoroquinolone class of antibacterial agents is one of the
most widely used in modern infection control, the presence of a flu-
orine atom at position 6 of the 4-quinolone ring coupled with amino
functionality at the neighbouring position (C-7) providing optimal
antibacterial activity [9]. Several front-line fluoroquinolones are
substituted in this latter position with a 4-piperazinyl moiety, thus
offering ideal potential reactive sites for the current proposition,
since this would furnish a dialkylamine equivalent auxochrome in
the resulting phenothiazinium derivative (Fig. 1), optimising light
absorption in the 660 nm region. A further reason for the use of flu-
oroquinolones here is their targeting of enzymes closely associated
with bacterial DNA [10].
2.2. Symmetrical photosensitisers
2.2.1. 3,7-Dibromophenothiazinium tribromide
One gram of phenothiazine (5 mmol) was dissolved in 70 mL
of glacial acetic acid, ensuring uniform mixing for approximately
30 min. To this solution, 5 mL of bromine in 40 mL of glacial acetic
acid was added rapidly. Then, 200 mL of cold water was quickly
added to this mixture with vigorous stirring, resulting in a dark
red solution containing a black precipitate. The solution was then
filtered through a sintered glass crucible and the resulting solid
was washed with water and then with diethyl ether to ensure all
bromine was removed. The yield of dark red–brown powder was
2.72 g (81%).
The present study covers the investigation of the antibacterial
activity of a series of methylene blue analogues, i.e. having increas-
ing size of dialkylamine group at position 3 and attached at position
7 to the fluoroquinolone norfloxacin via the 4-piperazinyl moiety.
2.2.2. 3,7-Bis(dialkylamino)phenothiazinium bromides (1b–1f)
The requisite dialkylamine (15 mmol) in 100 mL of DCM was
stirred in a flask followed by the addition of 1.00 g (1.7 mmol) of
solid 3,7-dibromophenothiazinium tribromide. The mixture was
stirred for approximately 4 h, washed with dilute hydrobromic
acid followed by water, the organic layer was separated and dried
(Na2SO4) and then concentrated under vacuum and the product
was precipitated using diethyl ether. Dissolution in DCM and ether
precipitation was repeated until spectrophotometric examination
of the resulting material gave an absorbance ratio (Aꢀ : A290) of
2.2.
2. Materials and methods
2.1. Materials
The following amines were purchased from Sigma-Aldrich
(Gillingham, UK) and were used without further purifi-
cation:
dimethylamine;
diethylamine;
di-n-propylamine;
di-n-butylamine; di-n-pentylamine; and di-n-hexylamine.
10H-phenothiazine and norfloxacin were also purchased from
Sigma-Aldrich. The standard photosensitiser methylene blue
(Sigma-Aldrich) was purified by column chromatography on silica
gel [gradient elution with 0–10% methanol/dichloromethane
(DCM)]. Methanol (spectrophotometric grade), 1-octanol, DCM
and iodine were purchased from Fisher Scientific (Leicestershire,
UK) and were used without further purification.
2.3. Asymmetrical photosensitisers
Asymmetric analogues were prepared by the initial oxidation of
10H-phenothiazine with iodine, but with reaction of the resulting
cation with a lowered concentration of amine, furnishing a mono-
3-substituted derivative (2a–f). These preparative steps were
carried out as in previous work [11]. 3-Substituted intermediates

M. Wainwright et al. / International Journal of Antimicrobial Agents 35 (2010) 405–409
407
were further reacted with the second amine, i.e. the 4-piperazinyl
2.5. DNA intercalation
unit of norfloxacin, to obtain the asymmetrical product.
Following previous work [12], a DNA solution was made with
0.75 mg of DNA sodium salt type XIV from herring testes (Aldrich,
St Louis, MO) dissolved in 10 mL of distilled water. Using the extinc-
tion coefficient (ε) for DNA of 6600 M−1 cm−1, the concentration of
the DNA solution was found to be 100 ␮M.
Distilled water solutions of the candidate photosensitisers
(100 ␮M) were made up and ꢀmax values were measured spec-
trophotometrically.
For each of the candidates, 8 mL of 100 ␮M DNA solution was
mixed in a test tube with 1 mL of the 100 ␮M photosensitiser solu-
tion. The ꢀmax of the resulting mixture was recorded and compared
with that of the original photosensitiser solution. Both methylene
blue and norfloxacin were included as comparators.
2.3.1. Norfloxacin conjugates (3a–f)
A solution of the 3-dialkylamino- compound (2a–f, 0.75 mmol)
in methanol (20 mL) was stirred at room temperature and solid
norfloxacin (0.57 g, 1.8 mmol) was added in several portions over
10 min. Progress of the reaction was monitored by thin layer chro-
matography [11] as described above. The mixture was stirred for
approximately 5–6 h at room temperature. The solvent was then
removed from the reaction mixture in vacuo and the residual black
solid was extracted with DCM. The resulting solution was chro-
matographed on silica with a gradient MeOH/DCM eluent (0–10%
v/v) and the blue fraction was concentrated in vacuo before precip-
itation using diethyl ether.
Solutions (100 ␮M) of phenothiaziniums and conjugates were
made up in distilled water for use in the bacterial screen. All
spectrophotometric measurements were carried out on a Hewlett
Packard 8452A diode array spectrophotometer (Agilent Technolo-
gies, Waldbronn, Germany). The photosensitisers were found to
follow Beer’s law in the concentration range of 10−5–10−7 mol/L.
2.6. Singlet oxygen production and lipophilicity (log P)
These properties were measured spectrophotometrically, as
recounted in previous work [11].
3. Results
2.4. Antibacterial screening
The present study represents the first report of the synthesis of
phenothiazinium–drug conjugates. As expected, reaction between
the monosubstituted 3-dialkylaminophenothiazinium intermedi-
ate compounds and norfloxacin was straightforward owing to
the nucleophilic nature of the 4-piperazinyl moiety, providing
deep-blue-coloured reaction mixtures (Fig. 1). Products required
column chromatography for purification but yields were in line
with previous work involving the synthesis of asymmetrical phe-
nothiazinium derivatives [11]. In the same way, the absorption
maxima of the conjugates were similar to those of 3-dialkylamino-
7-heterocyclophenothiazinium derivatives reported by Gorman et
al. [7], since any electronic effects due to the fluoroquinolone
moiety are not transmissible to N-4 of the piperazinyl subunit
(Table 1).
Previously, N-4-piperazinyl derivatives of norfloxacin have been
synthesised as novel but conventional antibacterial agents using a
range of substituents, including heteroaromatics, several of which
exhibited increased activity against Gram-negative bacteria includ-
ing E. coli [13]. Simple 4-N-alkyl derivatives have also been shown
to have increased activity against conventional resistant bacteria,
again including E. coli and S. aureus [14].
The photobactericidal activities of the norfloxacin conjugates
and the methylene blue homologues in addition to those of the
known photosensitiser methylene blue and norfloxacin itself were
measured against a Gram-positive organism (Staphylococcus aureus
NCTC 6571) and a Gram-negative organism (Escherichia coli NCTC
10418). Both strains were grown in Mueller–Hinton broth and then
diluted to a concentration of 106 colony-forming units/mL. Aliquots
of the strains were treated in microtitre trays with various concen-
trations of photosensitiser ranging from 100 ␮M to 3 ␮M, with zero
photosensitiser concentration in each case used for control pur-
poses. The trays were then either illuminated for 20 min using an
array of 126 light-emitting diodes (660 nm) giving a light dose of
6 J/cm2 or alternatively foil-covered to provide dark controls. From
each well showing complete absence of growth of the microorgan-
ism (i.e. total bacterial kill), 1 ␮L was subcultured on nutrient agar
using the Miles–Misra method and incubated for 18 h at 37 ◦C.
Minimum bactericidal concentrations (MBCs) were thus deter-
mined as the lowest concentration of each photosensitiser giving
no bacterial growth. Each test was repeated (n = 6) to ensure an
absolute value for the cited MBC. Owing to the absolute nature of
the assay, i.e. the cut-off point was taken as a complete absence
of growth rather than fractional kill, no statistical treatment of the
resulting data was applied.
Series of symmetrical methylene blue derivatives (1a–f; Fig. 1),
i.e. having identical dialkylamine groups in positions 3 and 7
of the phenothiazinium ring, have been reported previously by
different groups [7,11] and exhibited the expected increase in
Table 1
Analytical data for the derivatives.
R3
R7
X−
m/za
% Yield
ꢀmax (nm)b
Log εmax
b
Derivative
Calculated
Observed
1a
1b
1c
1d
1e
1f
3a
3b
3c
3d
3e
3f
NMe2
NEt2
NMe2
NEt2
Cl−
Br−
Br−
Br−
Br−
Br−
I−
–
–
–
23
28
30
21
18
24
25
30
37
34
33
657
664
669
671
672
673
658
662
665
665
664
665
4.88
4.87
4.75
4.84
4.69
4.55
4.61
4.66
4.77
4.81
4.65
4.69
340.18
396.25
452.31
508.37
564.43
558.20
586.23
614.26
642.29
670.32
698.35
340.18
396.24
452.33
508.32
564.44
558.15
586.05
614.17
642.09
670.27
698.15
NPrn
NPrn
2
2
NBun
NBun
2
2
NPen
NPen
2
2
NHxn
NHxn
2
2
NMe2
NEt2
Norfloxacin
Norfloxacin
Norfloxacin
Norfloxacin
Norfloxacin
Norfloxacin
I−
NPrn
I−
2
NBun
I−
2
NPen
I−
2
NHxn
I−
2
a
Molecular ion mass measured by inductively coupled plasma mass spectrometry.
Measured in MeOH.
b

408
M. Wainwright et al. / International Journal of Antimicrobial Agents 35 (2010) 405–409
Table 2
Antibacterial and related data for the derivatives.
a
Compound
MBC (␮M)
Rel.1O2
Log P
DNA shift (nm)
Staphylococcus aureus
Escherichia coli
Light
Dark
Light
Dark
b
Norfloxacin
100
25
6.3
3.1
3.1
3.1
3.1
6.3
3.1
100
100
100
50
25
25
25
50
+0.1
−0.1
+0.8
+1.1
+1.3
+1.6
+1.8
−0.1
+0.2
+0.2
+0.4
+0.5
+0.4
0
+4
+3
+1
+2
1a
1b
1c
1d
1e
1f
3a
3b
3c
3d
3e
3f
1.00
0.55
0.59
0.61
0.26
6.3
6.3
3.1
3.1
3.1
6.3
3.1
3.1
12.5
12.5
12.5
100
100
100
100
100
100
100
100
100
100
50
50
100
100
100
25
100
100
100
50
+5
+6
0.21
b
+11
+10
+12
+11
+10
+4
b
b
b
b
b
3.1
12.5
12.5
12.5
MBC, minimum bactericidal concentration.
a
Yield of singlet oxygen relative to that of methylene blue.
b
1O2 measurement ≤2% that of methylene blue.
log P and ꢀmax with auxochrome size along the homologous
series.
It should be noted at the outset that testing of the photosen-
sitisers against bacteria carried out in this work was aimed at a
clinically realistic approach. Thus, the candidate photosensitisers
were mixed with the bacterial suspensions and illuminated almost
immediately with red light for 20 min. Antibacterial screening of
conventional agents usually entails a significant incubation time
in order for the drug to be taken up by the bacterial cells, typically
18 h. Consequently, under the same conditions as those used for the
phenothiazinium derivatives, norfloxacin was much less effective
than might normally be expected. In addition, although the fluo-
roquinolones are also known to be photosensitisers, the use of red
light for activation, rather than ultraviolet, elicited photoactivation
only of the phenothiazinium chromophores involved.
All of the conjugates and symmetrical analogues of methy-
lene blue exhibited greater efficacy in both organisms than either
methylene blue itself or norfloxacin (Table 2). The pattern of
increasing activity with alkyl group size among the symmetrical
derivatives was not repeated in the conjugates, the dibutyl-dihexyl
analogues (3d–f) among the latter exhibiting much lower effica-
cies. Whilst these activity profiles may correlate with lipophilicity
for the symmetrical derivatives, this was not the case with the
conjugates, since little variation in log P was observed across the
series, presumably due to the highly hydrophilic character of the
norfloxacin moiety.
Dark toxicity between both sets of derivatives was found to be
low, most derivatives being toxic at the highest concentrations
used (100 ␮M) or above, in line with methylene blue (Table 2).
This suggests a lack of essential targeting and does not support
a specific DNA-localising hypothesis. Bacterial DNA photodam-
age has been reported for methylene blue itself, but it has been
suggested that the initial activity occurs at the cell exterior [1].
In addition, the relatively amphiphilic nature of the norfloxacin
derivatives here (−0.1 < log P < +0.5) suggests some difficulty in
reaching the interior of mammalian cell nuclei, since such local-
isation has been not reported for the symmetrical derivatives
[7].
An increasing number of phenothiazinium derivatives is appear-
ing in the literature as potential biological photosensitisers. As far
as we are aware, the present study is the first time that conjugation
of the phenothiazinium chromophore to a standard drug has been
reported.
4. Discussion
Although it had been assumed that the presence of the
inert piperazinyl linker between the phenothiazinium and fluoro-
quinolone ring systems would prevent intramolecular interference
with singlet oxygen generation by the former under red light con-
ditions, this was not the case. In vitro singlet oxygen yields for the
conjugates were, in fact, too low to measure using the standard
spectrophotometric assay employed, unlike previous asymmetri-
cal examples (Table 2). However, it is again noted that such tests are
not necessarily of great utility as performance indicators for pho-
toantimicrobials owing to the potential influence of local electronic
effects in the cellular milieu, not replicated in photofading tests. It
should also be recalled that the cytotoxic agents produced during
photodynamic antimicrobial chemotherapy (PACT) are very short-
lived, e.g. singlet oxygen has a half-life of less than a microsecond
before decaying to ground state molecular oxygen [1].
The loss of in vitro singlet oxygen production in the conjugates
may be due to molecular aggregation at the concentrations used in
the test. The aqueous spectra of the conjugates in the 600–700 nm
region were similar in shape to those of the symmetrical deriva-
tives (1a–f), i.e. broadened with a small shoulder around 630 nm.
However, whilst this aggregation was not deleterious for the sym-
metrical species, it is possible that in the conjugates this would
allow the close proximity of neighbouring phenothiazinium and
quinolone chromophores, thus providing a deactivation route. The
photobactericidal effects observed (see below) are not thought to
develop from such aggregates.
In terms of nucleic acid interaction, it was noticeable that, whilst
the symmetrical series of methylene blue derivatives exhibited
similar bathochromic shifts to that of methylene blue itself on in
vitro mixing with DNA [15], the conjugates exhibited much greater
shifts (typically >10 nm) (Table 2). The magnitude of these shifts
was more akin to that reported for Taylor’s blue (1,9-dimethyl
methylene blue), understood to be a much stronger intercalator
[15]. The bathochromic shifts observed in the red region of the
spectrum underline the hypothesis that intercalation involved the
phenothiazinium chromophore, rather than the fluoroquinolone
moiety, as no shifting of the norfloxacin ꢀmax (324 nm) was
observed.
Funding: No funding sources.
Competing interests: None declared.
Ethical approval: Not required.

M. Wainwright et al. / International Journal of Antimicrobial Agents 35 (2010) 405–409
409
References
[8] Moller U, Schubert F, Cech D. Versatile procedure of multiple introduction of
8-aminomethylene blue into oligonucleotides. Bioconjug Chem 1995;6:174–8.
[9] Anderson MI, MacGowan AP. Development of the quinolones. J Antimicrob
Chemother 2003;51:1–11.
[10] Blondeau JM. Fluoroquinolones: mechanism of action, classification, and devel-
opment of resistance. Surv Ophthalmol 2004;49(Suppl. 2):S73–8.
[11] Wainwright M, Meegan K, Loughran C, Giddens RM. Phenothiazinium pho-
tosensitisers. Part VI: photobactericidal asymmetric derivatives. Dyes Pigm
2009;82:387–91.
[12] Huang CZ, Li YF, Zhang DJ, Ao XP. Spectrophotometric study on the
supramolecular interactions of Nile blue sulphate with nucleic acids. Talanta
1998;49:495–503.
[13] El-Din G, Abuo-Rahma AA, Sarhan HA, Gad GFM. Design, synthesis, antibacterial
activity and physicochemical parameters of novel N-4-piperazinyl derivatives
of norfloxacin. Bioorg Med Chem 2009;17:3879–86.
[14] Vila J, Sanchez-Cespedes J, Sierra JM, Piqueras M, Nicolas E, Freixas J, et al.
Antibacterial evaluation of a collection of norfloxacin and ciprofloxacin deriva-
tives against multiresistant bacteria. Int J Antimicrob Agents 2006;28:19–
24.
[15] Wagner SJ, Skripchenko A, Robinette D, Foley JW, Cincotta L. Factors affecting
virus photoinactivation by a series of phenothiazine dyes. Photochem Photobiol
1998;67:343–9.
[1] Wainwright M. Photodynamic antimicrobial chemotherapy (PACT). J Antimi-
crob Chemother 1998;42:13–28.
[2] Nitzan Y, Dror R, Ladan H, Malik Z, Kimel S, Gottfried V. Structure–activity rela-
tionship of porphines for photoinactivation of bacteria. Photochem Photobiol
1995;62:342–7.
[3] Soukos NS, Ximenez-Fyvie LA, Hamblin MR, Socransky SS, Hasan T. Tar-
geted antimicrobial photochemotherapy. Antimicrob Agents Chemother
1998;42:2595–601.
[4] Tegos GP, Anbe M, Yang C, Demidova TN, Satti M, Mroz P, et al. Protease-
stable polycationic photosensitizer conjugates between polyethyleneimine
and chlorin(e6) for broad-spectrum antimicrobial photoinactivation. Antimi-
crob Agents Chemother 2006;50:1402–10.
[5] Nitzan Y, Gutterman M, Malik Z, Ehrenberg B. Inactivation of Gram-negative
bacteria by photosensitized porphyrins. Photochem Photobiol 1991;55:89–96.
[6] Mellish KJ, Cox RD, Vernon DI, Griffiths J, Brown SB. In vitro photody-
namic activity of a series of methylene blue analogues. Photochem Photobiol
2002;75:392–7.
[7] Gorman SA, Bell AL, Griffiths J, Roberts D, Brown SB. The synthesis and proper-
ties of unsymmetrical 3,7-diaminophenothiazin-5-ium iodine salts: potential
photosensitisers for photodynamic therapy. Dyes Pigm 2006;71:153–60.