Phenothiazinium photosensitisers VII: Novel substituted asymmetric N-benzylphenothiaziniums as photoantimicrobial agents

Article abstract of DOI:10.1016/j.jphotobiol.2010.02.008

The synthesis of asymmetrical analogues of methylene blue, in which one of the dimethylamino groups is replaced by a diethylamino or di-n-propylamino group, and the other by benzylamino or 4-substituted benzylamino, is reported, the substituents being alkyl, alkoxyl or halogen. As expected, because of their longer alkyl chains these diethylamino- and di-n-propylamino derivatives proved to be considerably more lipophilic than the parent compound methylene blue, while maintaining suitable maximum absorption wavelengths and singlet oxygen efficiencies for photoantimicrobial use.Also as expected, in screening tests against Gram-positive and Gram-negative bacteria, the substituted benzylamino derivatives were highly active on illumination, presumably via singlet oxygen damage, and exhibited considerably increased activity against both classes relative to that of the standard, methylene blue. In addition, the more lipophilic derivatives exhibited greater activity against Escherichia coli. This may be due to increased interaction with the lipid-rich outer membrane of this Gram-negative bacterium. DNA binding of the derivatives was also increased, relative to methylene blue, showing large bathochromic shifts (>10. nm) on binding typical of strong intercalators.

Full text of DOI:10.1016/j.jphotobiol.2010.02.008

Journal of Photochemistry and Photobiology B: Biology 99 (2010) 74–77
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology B: Biology
journal homepage: www.elsevier.com/locate/jphotobiol
Phenothiazinium photosensitisers VII: Novel substituted asymmetric
N-benzylphenothiaziniums as photoantimicrobial agents
*
Mark Wainwright , Simon D. Brandt, Andrew Smith, Andrew Styles, Katie Meegan, Ciara Loughran
School of Pharmacy & Biomolecular Sciences, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis of asymmetrical analogues of methylene blue, in which one of the dimethylamino groups is
replaced by a diethylamino or di-n-propylamino group, and the other by benzylamino or 4-substituted
benzylamino, is reported, the substituents being alkyl, alkoxyl or halogen. As expected, because of their
longer alkyl chains these diethylamino- and di-n-propylamino derivatives proved to be considerably
more lipophilic than the parent compound methylene blue, while maintaining suitable maximum
absorption wavelengths and singlet oxygen efficiencies for photoantimicrobial use.
Also as expected, in screening tests against Gram-positive and Gram-negative bacteria, the substituted
benzylamino derivatives were highly active on illumination, presumably via singlet oxygen damage, and
exhibited considerably increased activity against both classes relative to that of the standard, methylene
blue. In addition, the more lipophilic derivatives exhibited greater activity against Escherichia coli. This
may be due to increased interaction with the lipid-rich outer membrane of this Gram-negative bacte-
rium. DNA binding of the derivatives was also increased, relative to methylene blue, showing large bath-
ochromic shifts (>10 nm) on binding typical of strong intercalators.
Received 22 September 2009
Received in revised form 9 February 2010
Accepted 11 February 2010
Available online 16 February 2010
Keywords:
Asymmetric phenothiazinium derivatives
Methylene blue
PACT
Photosensitiser
Photoantimicrobial
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
itself followed by amination have led to a greatly improved range
of structures [3].
Modern research into the biological application of photosensi-
tisers, was initially based on porphyrin derivatives from natural
sources, along with several synthetic dyes established as vital
stains [1,2]. The realisation that first generation porphyrin deriva-
tives had shortcomings in terms of non-optimal photoproperties
and side effects encouraged efforts in the synthesis of new, im-
proved derivatives from porphyrin and dyestuff leads.
The selective tumour staining activity of several cationic dyes,
such as Nile blue and methylene blue provided a sound rationale
for photosensitiser development, although the provision of large
series of compounds required for rapid drug discovery was not
straightforward due to synthetic chemistries based on obsolete
and unsuitable textile dye production. For example, the oxidative
nature of phenothiazinium dye production traditionally required
the use of dichromate (chromium(VII)), but the strength of this oxi-
dant is such that anilines employed as starting materials are invari-
ably converted to a variety of products, including those having
substituent alteration in the resulting chromophore. Routes to phe-
nothiaziniums via the oxidative coupling of anilinethiosulphonic
acids and anilines using a weaker oxidant such as silver(I) carbon-
ate and the halogen-mediated oxidation of 10H-phenothiazine
Recent investigations of the activity of methylene blue deriva-
tives – in the fields of both photodynamic therapy (PDT) and pho-
todynamic antimicrobial chemotherapy (PACT) – have mainly been
based on symmetrical auxochromic variation – i.e. the dimethyl-
amino groups at positions C-3 and C-7 of the phenothiazinium
chromophore being replaced by higher dialkylamino functionality
groups [4,5], although some asymmetric derivatives have also been
reported [6].
In a previous publication, it was demonstrated that 3-dialkyla-
minophenothiazinium derivatives having simple 7-anilino- and 7-
benzylamino moieties exhibited greatly improved photobacteri-
cidal activity, coupled with low dark toxicity, against Escherichia
coli and Staphylococcus aureus, compared to that of the lead com-
pound methylene blue [7]. Such activity in the arylamino deriva-
tives was unexpected, given the lack of associated singlet oxygen
production in the standard in vitro chemical screen. However, as
part of an ongoing programme of photosensitiser drug discovery,
the increased activity of these compounds has encouraged the syn-
thesis of peripherally-substituted analogues, the presence of the
pendant aryl groups allowing considerable potential for chemi-
cal/physicochemical variation without affecting the chromophore
and singlet oxygen production. The current paper covers the initial
chemical testing and antibacterial screening of a group of deriva-
tives as described, having standard aromatic substitution in the
7-benzyl pendant.
* Corresponding author.
E-mail address: mark_wainwright@hotmail.com (M. Wainwright).
1011-1344/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotobiol.2010.02.008

M. Wainwright et al. / Journal of Photochemistry and Photobiology B: Biology 99 (2010) 74–77
75
2.3. Singlet oxygen testing
2. Materials and methods
Singlet oxygen production by the photosensitisers was assayed
as in previous work [7], except that the decolourisation of 2,3,4,5-
tetraphenylcyclopentadienone (TPCPD) in dichloromethane was
employed rather than that of 1,3-diphenylbenzisofuran in metha-
nol. Thus the decrease in absorption at 500 nm was monitored
spectrophotometrically with time, using methylene blue as a stan-
dard photosensitiser. By assuming that the decrease in absorption
of TPCPD at 500 nm is directly proportional to its reaction with sin-
glet oxygen, the time for a 50% decrease in absorption caused by
each of the derivatives under identical conditions (t1/2MBD) thus
gives a measure of its photosensitising efficiency. Thus, if the time
for the DPIBF absorption to decrease by 50% due to MB photosen-
sitisation is t1/2MB, relative singlet oxygen yields for the deriva-
tives are given by:
2.1. Chemicals
10H-Phenothiazine, iodine, dialkylamines, benzylamines and
solvents were purchased from Sigma–Aldrich, UK, and used with-
out further purification. Both methylene blue and dimethyl meth-
ylene blue were also purchased from Sigma–Aldrich, but were
purified by column chromatography on silica gel (Fisher Scientific,
UK) using gradient elution in dichloromethane/methanol. Photo-
physical characterisation of the products was carried out using a
Hewlett Packard 8452A diode array spectrophotometer. This was
also used for the determination of lipophilicity. Accurate molecular
ion masses for the derivatives were obtained using a Micromass
LCT TOF mass spectrometer.
t1=2MB
Relative 1O2 yield ¼
t1=2MBD
2.2. Synthesis
i.e. the lower the t1/2 value for the derivative, the greater its 1O2
yield.
Both the precursors, phenothiazin-5-ium tetraiodide and the 3-
dialkylaminophenothiazinium triiodides were synthesised as de-
scribed previously [7].
2.4. Lipophilicity (Log P)
2.2.1. 3-Dialkylamino-7-benzylaminophenothiazinium iodides (1a–e,
2a–e)
The lipophilicities of the photosensitisers were calculated in
terms of Log P, the logarithm of their partition coefficients between
phosphate-buffered saline and 1-octanol. The data were calculated
using the standard spectrophotometric method [8] based on the
relationship:
3-Diethylamino- or 3-di-n-propylaminophenothiazinium triio-
dide (0.75 mmol) was suspended in methanol (10 ml) and the
requisite benzylamine (1.8 mmol) in 10 ml methanol was added
dropwise, the reaction being monitored by TLC (SiO2, 3% aqueous
NH4OAc/CH3OH 1:17). Reaction times to the exhaustion of the
triiodide salt were in the region of 1.5–2 h. Products were iso-
lated by evaporation of the methanol in vacuo, redissolution in
dichloromethane, extraction with 5% v/v hydroiodic acid then
water, drying of the organic layer over anhydrous sodium sul-
phate, evaporation to a small volume and repeat precipitations
in dry diethyl ether. Compounds impure by thin-layer chroma-
tography at this stage were chromatographed on silica gel (Fish-
er Scientific, UK) using gradient elution in dichloromethane/
methanol.
(
)
A ꢀ A1 Vw
Log P ¼ Log
ꢁ
A1
Vo
where A and A1 are the absorption intensities before and after par-
titioning respectively and Vw and Vo are the respective volumes of
the aqueous and 1-octanol phases. Determinations were repeated
three times.
2.5. Antibacterial screening
Synthetic yields and analytical data for the derivatives are given
in Table 1.
The photobactericidal efficacies of the derivatives in addition to
that of the known photosensitiser methylene blue were measured
Table 1
Analytical data for the derivatives.
+
2
-
.
b
c
R
X
m/za
% Yield
kmax (nm)b
Log emax
Relative 1O2
Log P
Calc.
Found
MB
1a
1b
1c
1d
1e
2a
2b
2c
2d
2e
–
Et
Et
Et
Et
Et
n-Pr
n-Pr
n-Pr
n-Pr
n-Pr
–
H
Cl
F
MeO
Me
H
Cl
F
–
–
–
656
644
642
644
646
648
646
648
650
650
654
4.88
4.62
4.66
4.71
4.78
4.63
4.65
4.63
4.77
4.69
4.67
1.00
0.69
0.71
0.53
0.54
0.57
0.47
1.23
0.46
0.90
0.38
ꢀ0.1
+0.8
>2.0
>2.0
>2.0
+1.9
>2.0
>2.0
>2.0
>2.0
>2.0
374.17
408.13
392.16
404.18
388.18
402.20
436.16
420.19
432.21
416.22
374.18
408.12
392.10
404.18
388.18
402.22
436.16
420.11
432.21
416.22
29
35
26
34
36
32
37
28
30
33
MeO
Me
a
b
c
By ICP-MS.
Measured in MeOH.
Yield of singlet oxygen relative to that of MB.

76
M. Wainwright et al. / Journal of Photochemistry and Photobiology B: Biology 99 (2010) 74–77
against a Gram-positive and a Gram-negative bacterium, S. aureus
(National Culture Tissue Collection, NCTC 6571) and E. coli (NCTC
10418) respectively. Both strains were grown overnight in Muel-
ler–Hinton Broth and then diluted to a concentration of 106 col-
ony-forming units/ml. Aliquots of the strains in the growth phase
were then incubated for 1 h at 37 °C in microtitre trays with vari-
100
90
80
70
60
50
40
30
20
10
0
ous concentrations of photosensitiser ranging from 100 to 3 lM,
Light
Dark
with zero photosensitiser concentrations in each case for control
purposes. 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 cmꢀ2 or alternatively foil-covered to provide dark controls.
From each well apparently showing a complete absence of growth
of the micro-organism (i.e. total growth inhibition), 1 ll was sub-
cultured on nutrient agar, using the Miles-Misra method [9], and
incubated for 18 h at 37 °C. The minimum bactericidal concentra-
tions (MBCs) were thus determined as the lowest concentration
for each photosensitiser giving no bacterial growth. Each test was
repeated to ensure an absolute value for the cited MBC with
n = 6. Due to the absolute nature of the assay, i.e. complete absence
of growth, rather than fractional kill, no statistical treatment of the
resulting data was applied.
MB 1a 1b 1c 1d 1e 2a 2b 2c 2d 2e
Fig. 2. Antibacterial activities of the compounds against Staphylococcus aureus. MBC
– minimum bactericidal concentration.
100
90
80
70
60
2.6. DNA interaction
A solution of DNA sodium salt Type XIV (Sigma–Aldrich) was
made up in pH 7.4 (PBS) buffer. This solution was diluted to
Light
Dark
50
40
30
20
10
0
100
100
l
M, according to the method of Huang et al. [10]. Similarly,
lM solutions of methylene blue and compounds 1a–e and
2a–e were made up in pH 7.4 (PBS) buffer. A 100 lM solution of
1,9-dimethyl methylene blue (Sigma–Aldrich) was also made up
for comparison purposes. The photosensitiser solutions were either
diluted 1:8 v/v with buffer, or mixed in this ratio with DNA solu-
tion. The spectra of the resulting solutions were then measured
for kmax comparison.
MB 1a 1b 1c 1d 1e 2a 2b 2c 2d 2e
Fig. 3. Antibacterial activities of the compounds against Escherichia coli. MBC –
minimum bactericidal concentration.
3. Results
range as in the previous study [7], although the dipropylamino/4-
chlorobenzylamino derivative (compound 2b) produced more sin-
glet oxygen than did methylene blue.
All of the benzylamino derivatives, whether ring-substituted or
not, were more effective photobactericidal agents than the lead
compound, methylene blue, being up to eight times as active (Figs.
2 and 3). Interestingly, the activities of the derivatives were gener-
ally higher against the Gram-negative bacterium, E. coli, which was
unexpected.
As with previous work [7], the synthesis of the asymmetrical
derivatives (Fig. 1) was straightforward, with product yields typi-
cally around 30% (Table 1). The 4-phenyl substituents in the benzyl
moieties employed did not cause significant shifts in the absorp-
tion spectra of the series, kmax values remaining within 10 nm of
that of the original benzyl derivative in each case. Several of the
3-dipropylaminophenothiazinium series exhibited kmax values
close to that of the parent photosensitiser, methylene blue
(656 nm). Similarly, singlet oxygen yields in vitro were in the same
Fig. 1. Synthesis of asymmetrical phenothiazinium derivatives.

M. Wainwright et al. / Journal of Photochemistry and Photobiology B: Biology 99 (2010) 74–77
77
Table 2
reportedly due to a considerably stronger binding interaction with
the nucleic acid [12]. Increased interaction with bacterial DNA
would support the increased photobactericidal activities exhibited
by the derivatives in the current work, compared with that of
methylene blue. Increased lipophilicity in the derivatives might
also explain the higher relative activity against E. coli, given the li-
pid-rich outer membrane associated with Gram-negative bacteria,
a multiple site of action hypothesis being currently favoured for
photoantimicrobials, as mentioned above.
While the number of compounds in the present study is not
great enough to produce robust structure–activity data, it may be
seen from the results of the antibacterial screen that greater pho-
tobactericidal activity was associated with derivatives where the
pendant aryl group contained a methoxyl group or halogen atom.
In addition, the noticeably increased activity of the 3-di-n-propyl
derivatives against E. coli relative to that against S. aureus supports
the increased outer membrane interaction of more hydrophobic
compounds (Figs. 2 and 3). It is intended that future work will ex-
plore greater functional and positional variation in benzyl ana-
logues, for example including those of a hydrophilic nature.
Bathochromic shifts for the derivatives caused by DNA intercalation.
kmax (nm) alone kmax (nm) + DNA Change
Methylene blue
665
652
650
653
654
645
658
656
656
658
658
669
666
662
665
665
659
670
669
666
669
668
660
+4
+14
+12
+14
+13
+14
+12
+13
+10
+11
+10
+12
1a
1b
1c
1d
1e
2a
2b
2c
2d
2e
1,9-Dimethyl methylene blue 648
As part of the current study, the binding of the photosensitisers
to DNA was measured to investigate possible binding changes
caused by molecular alteration in the phenothiazinium derivative.
All of the derivatives 1a–e and 2a–e demonstrated considerable
bathochromic shifts compared to the known intercalator, methy-
lene blue. The shifts observed for the derivatives were in the range
10–14 nm, compared to the 4 nm seen for the lead compound
(Table 2).
References
[1] A. Policard, Etude sur les aspects offers par des tumeurs experimentales
examinees a la lumiere de wood, Compt. Rend. Soc. Biol. 91 (1924) 1423–1424.
[2] M. Wainwright, H. Mohr, W. Walker, Phenothiazinium derivatives for
pathogen inactivation in blood products, J. Photochem. Photobiol. B: Biol. 86
(2007) 45–58.
4. Discussion
The increases in absorption wavelength observed for the benzyl
derivatives were, in fact, due to the increased alkyl chain length in-
cluded in the dialkylamine auxochrome (i.e. diethylamino or dipro-
pylamino, compared to dimethylamino), since – as a comparison –
the mono-demethylated analogue of methylene blue, azure B, has a
kmax value of 645 nm in methanol – i.e. the aromatic residue in the
benzyl group is too far removed to have an electronic effect on the
phenothiazinium chromophore. Such increases in kmax have been
reported previously for the homologous series of methylene blue
derivatives having bis(diethylamino-) up to bis(dihexylamino-)
auxochromes [4].
Although each of the benzyl derivatives 1a–e and 2a–e pro-
duced singlet oxygen in vitro, this data did not correlate with pho-
tobactericidal efficacy. Such data should, of course, be used
carefully, since cellular environments are quite different to those
encountered in a test tube.
[3] M. Wainwright, R.M. Giddens, Phenothiazinium photosensitisers: choices in
synthesis and application, Dyes Pigments 57 (2003) 245–257.
[4] K.J. Mellish, R.D. Cox, D.I. Vernon, J. Griffiths, S.B. Brown, In vitro photodynamic
activity of a series of Methylene Blue analogues, Photochem. Photobiol. 75
(2002) 392–397.
[5] N. Leventis, M. Chen, C. Sotiriou-Leventis, Synthesis of phenothiazines
analogous to methylene blue by electrophilic and nucleophilic substitutions
in tandem, A mechanistic perspective, Tetrahedron 53 (1997) 10083–10092.
[6] S.A. Gorman, A.L. Bell, J. Griffiths, D. Roberts, S.B. Brown, The synthesis and
properties of unsymmetrical 3,7-diaminophenothiazin-5-ium iodide salts:
potential photosensitisers for photodynamic therapy, Dyes Pigments 71 (2006)
153–160.
[7] M. Wainwright, K. Meegan, C. Loughran, R.M. Giddens, Phenothiazinium
photosensitisers VI. Photobactericidal asymmetric derivatives, Dyes Pigments
82 (2009) 387–391.
[8] J. Pooler, D.P. Valenzo, Physicochemical determinants of the sensitizing
effectiveness for photooxidation of nerve membranes by fluorescein
derivatives, Photochem. Photobiol. 30 (1979) 491–498.
[9] A.A. Miles, S.S. Misra, The estimation of the bactericidal power of the blood, J.
Hyg. 38 (1938) 732–749.
[10] C.Z. Huang, Y.F. Li, D.J. Zhang, X.P. Ao, Spectrophotometric study on the
supramolecular interactions of Nile blue sulphate with nucleic acids, Talanta
49 (1998) 495–503.
[11] M. Wainwright, Photodynamic antimicrobial chemotherapy (PACT), J.
Antimicrob. Chemother. 42 (1998) 13–28.
[12] S.J. Wagner, A. Skripchenko, D. Robinette, J.W. Foley, L. Cincotta, Factors
That the mode of action of cationic photosensitisers against bac-
teria is multifactorial is attested to by the various reports of post-
treatment damage to nucleic acid, ribosomes and cell wall in the
same organism [11].
In work reported by Wagner et al., the intercalation of DNA by
1,9-dimethyl methylene blue caused a 12 nm bathochromic shift,
affecting virus photoinactivation by
a series of phenothiazine dyes,
Photochem. Photobiol. 67 (1998) 343–349.