This document summarizes a study that assessed the phototoxicity of various xanthene derivatives against E. coli, S. aureus, and S. cerevisiae. Without light exposure, the compounds showed similar inherent toxicity to each organism, dependent on chemical structure. Upon illumination, the compounds demonstrated phototoxicity, with stronger effects on Gram-positive bacteria and yeast. Compounds with more halogen substituents generated higher levels of reactive oxygen and showed greater phototoxic activity. The results suggest that xanthene derivatives have potential as alternative antimicrobial agents.

1. The Phototoxicity of Xanthene Derivatives Against Escherichia coli,
Staphylococcus aureus, and Saccharomyces cerevisiae
Hong Wang, Lei Lu, Shiyun Zhu, Yahong Li, Weimin Cai
School of Environmental Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
Received: 8 March 2005 / Accepted: 11 September 2005
Abstract. We assessed the phototoxicity of a series of xanthene derivatives against E. coli, S. aureus,
and S. cerevisiae, measured the physicochemical properties of the photosensitizers, and found the
relationship between them. Without illumination, the dyes tested showed almost the same level of
inherent toxicity to the same organism, which showed the inherent toxicity of dyes was primarily
dependent on the structure of parent molecule. Upon illumination, the photosensitizers showed obvious
phototoxicity to all organisms. The dyes showed stronger phototoxicity to Gram-positive bacteria. With
the increasing number of halogen substituents, the singlet oxygen yields increased and the phototoxic
activity increased too. There was no obvious correlation between relative lipophilicity and activity in the
current study. Our results showed xanthenes had the potential to act as alternatives to conventional
antimicrobial compounds and also could be used for the decontamination of microbially polluted waters.
The research field of antimicrobial compounds is one
of the current constant challenges. Unfortunately, an
unpleasant cycle has recently appeared: as soon as a new
drug is introduced, the strains resistant to that drug
emerge [14]. There is an urgent need for some new
drugs with novel mechanisms of attack.
Extensive research had been carried out to screen
the photosensitizers with antimicrobial properties and
illuminate their photodynamic mechanisms [2, 3, 19,
20]. Photosensitizers can be electronically excited by
irradiation with light at the wavelength appropriate for
that photosensitizer after being localized in the target
organism [4]. Then the molecule will pass its excitation
energy onto other biomolecules by two mechanisms. In
the type I process, energy can be transferred between the
excited photosensitizer and nearby biomolecules,
yielding oxygenated free radicals; In the type II process,
energy can be transferred between the excited photo-
sensitizer and molecular oxygen, yielding singlet oxy-
gen, 1
O2 [6]. Such highly reactive products are able to
photomodify some biomolecules in cells such as lipids,
enzymes, and DNA [7, 11, 12], which will be lethal to
cells. The non-specific nature of these attacking modes
makes it unlikely for the bacteria to acquire the resis-
tance to the photosensitizers.
Many photosensitizers had been shown to possess
antimicrobial properties [20]. Xanthene derivatives were
a class of photosensitizing molecules and some of them
had been developed for commercial use as pesticides [5,
9]. In the present study, the phototoxicity of a series
of xanthene derivatives against E. coli, S. aureus, and
S. cerevisiae, was assessed.
Materials and Methods
Photosensitizers. All chemicals used in our experiments were of
analytical grade. Fluorescein (Fl)-derived photosensitizers (Table 1)
were purchased from Shanghai Chemical Regents Company (China).
Na2Fl, Na2FlBr4, Na2FlI4, and Na2FlBr4Cl4 were stored as aqueous
stock solutions (1 mM) at 4°C. FlBr2 and FlI2 were stored as 95%
ethanol stock solutions (1 mM) at 4°C.
Light source. 150-W metal halide lamp (Philips, The Netherlands),
giving a light fluence of 15 mW cm)2
, was used in toxicity tests. The
fluence of the polychromatic light was measured with a FZ-A light
meter (Handy, China).
Physicochemical properities. The lipophilicities of the photosen-
sitizers were calculated in terms of log P, the logarithm of their
partition coefficients between 1-octanol and deionized water. MaximalCorrespondence to: Hong Wang; email: hwang_cn@sjtu.edu.cn
CURRENT MICROBIOLOGY Vol. 52 (2006), pp. 1–5
DOI: 10.1007/s00284-005-0040-z Current
Microbiology
An International Journal
ª Springer Science+Business Media, Inc. 2005

2. dye absorption in the range of visible light was determined in aqueous
solutions by a Unico 2102 UV/Vis spectrophotometer (Japan). The
photosensitizing efficiency was measured as relative singlet oxygen
yield (F¢). According to the method described by Kraljic and Mohsni
[13], the system imidazole plus p-nitrosodimethylaniline (RNO) can be
used as a sensitive and selective test for the presence of 1
O2 in aqueous
solutions.The test is based on secondary bleaching of RNO as inducedby
the reaction of 1
O2 with imidazole. In the present study, the solutions of
different photosensitizers were irradiated in closed absorption cells of
10-mm light path for 5 min and the O.D. was determined before and after
irradiation at 440 nm, the maximal absorption of RNO. The
concentrations of photosensitizer, RNO, and imidazole were set to
30 lM, 50 lM and 8 mM, respectively. The experiments were carried out
at room temperature (25°C). We set F¢=1 for Na2FlBr4Cl4.
Bacterial strains and growth condition. E. coli (AS 1.129) and S.
aureus (AS 1.72) were grown aerobically in liquid LB medium for
approximately 18 h using an orbital incubator at 37°C, which
corresponded to the cell concentration of around 2 · 109
cfu ml)1
.
S. cerevisiae (AS 2.101) was grown in liquid PDA medium at 28°C for
approximately 24 h, which corresponded to around 2 · 108
cfu ml)1
.
For the phototoxicity tests, cells were diluted with fresh medium to a
final cell concentration of 105
cfu ml)1
.
Toxicity tests. Qualitative phototoxicity tests were carried out using
the methods modified from Daniels [1]. Photosensitizers were
dissolved in corresponding solvents to the concentration of 2 mg/
mL. Five microliters of each solution was applied to a piece of filter
paper disc (7-mm diameter, 10 lg/disc) and solvents were allowed to
evaporate in the dark. Forty microliters of cells at the concentration of
105
cfu ml)1
was spread onto a plate evenly with a sterile bent glass
rod. The discs were placed onto the freshly inoculated plates and all
plates were incubated in dark for 30 min. Then, after 30-min irradiation
period under the light fluence of 15 mW cm)2
, all plates were
incubated in dark again at the temperature mentioned above for another
48 h. The zones of inhibition were measured. For the dark control
plates, we omitted the 30-min irradiation period. All assays were
repeated 3 times and the results combined.
Toxicity tests were conducted as 10-fold replicates using 300-lL
flat-bottomed 96-well micro-titre plates (Costar). Fresh cell suspen-
sions were prepared as described above. These suspensions were then
mixed with the chemicals shown in Table 1 to obtain twofold dilution
series of each photosensitizer with the final concentrations varying
from 1 to 500 lM. Two hundred microliters of these mixtures were
then added to the wells of micro-titre plates and all plates were incu-
bated in dark for 30 min. Then, after a 30-min irradiation period under
the light fluence of 15 mW cm)2
, all plates were placed in an incubator
again in dark at the temperature mentioned above for 18 h. The
resulting 10-lL cultures were streaked onto agar medium plates and
incubated for another 18 h. After incubation, tested plates were
examined for bacterial growth and the lowest concentration at which
no colonies were observed was taken as the minimum lethal concen-
tration (MLC) of a given photosensitizer. The experiments were re-
peated until the same results appeared.
Results and Discussion
The MLCs of Na2Fl, FlBr2, FlI2, Na2FlBr4, Na2FlI4, and
Na2FlBr4Cl4 (structures shown in Table 1) were deter-
mined when directly against three microorganisms with
or without illumination (Table 2). The results were
consistent with the qualitative phototoxicity assays
(Table 3). The spectral output of the light source coin-
Table 1. Structures of photosensitizers
Photosensitizer X Y Z
Fluorescein, disodium salt (Na2Fl) H H H
Dibromofluorescein (FlBr2) Br H H
Diiodofluorescein (FlI2) I H H
Tetrabromofluorescein. disodium salt (Na2FlBr4) Br Br H
Tetraiodofluorecein. disodium salt (Na2FlI4) I I H
Tetrachlorotetrabromofluorescein. disodium salt (Na2FlBr4Cl4) Br Br Cl
2 CURRENT MICROBIOLOGY Vol. 52 (2006)

3. cided with the max absorption wavelength of the
photosensitizers. Control experiments showed that in the
absence of photosensitizers, illumination alone or etha-
nol-added medium had no effect on all organisms (data
not shown).
Without illumination, Na2Fl showed no inherent
toxicity to three organisms under the concentration as
high as 500 lM. The inherent MLCs of the other
photosensitizers to E. coli, S. aureus, and S. cerevisiae
were 500 lM, 62.5 to 125 lM, and 125 to 250 lM,
respectively. In the present study, all photosensitizers
showed almost the same level of inherent toxicity to the
same organism. The present study showed that the
inherent toxicity was primarily dependent on the struc-
ture of parent molecule, and substituted groups had
slight effects on the inherent activity.
Upon illumination, the photosensitizers showed
obvious phototoxicity against either bacterium under
the conditions in our experiments. The xanthenes
showed stronger phototoxicity to Gram-positive bacte-
ria. S. cerevisiae was found to be much more sensitive
than bacteria to the same xanthene. Na2FlI4 and
Na2FlBr4Cl4 were the most phototoxic dyes tested with
the MLCs of 125, 1, and 1 lM when directly against
E. coli, S. aureus, and S. cerevisiae, respectively. The
physicochemical properities of the photosensitizers
used in the present study are given in Table 4. In vitro
chemical tests showed that each of the xanthene
derivatives was able to photosensitize the production of
singlet oxygen, in the order of Na2FlBr4Cl4 > Na2FlI4
> Na2FlBr4 > FlI2 > FlBr2 > Na2Fl. All dyes tested
except Na2Fl were phototoxic to three organisms and
generated significant levels of singlet oxygen. It is
suggested that type II mechanism of photosensitization
played an important role in such actions. With the
increasing number of halogen substituents, the singlet
oxygen yields increased and the phototoxic activity
increased too. Some research has shown that the
Table 2. Toxicity of photosensitizers against E. coil, S. aureus, and S. cerevisiae
MLC (lM)
E. coli S. aureus S. cerevisiae
Photosensitizers Dark Light Dark/Light Dark Light Dark/Light Dark Light Dark/Light
Na2Fl ) ) ) ) ) 15.6
FlBr2 500 250 2 125 31.3 4 250 3.9 64
FlI2 500 125 4 125 31.3 4 250 3.9 64
Na2FlBr4 500 125 4 125 2 62.5 250 2 125
Na2FlI4 500 125 4 62.5 1 62.5 250 1 250
Na2FlBr4Cl4 500 125 4 62.5 1 62.5 125 1 125
) = no obvious effects under the highest concentration.
Table 3. Qualitative phototoxicity assays
E. coli S. aureus S. cerevisiae
Photosensitizers Dark Light Dark Light Dark Light
Na2Fl ) ) ) ) ) +
FlBr2 ) + + + ) ++
FlI2 ) + + + ) ++
Na2FlBr4 ) + + ++ ) ++
Na2FlI4 ) + + ++ ) ++
Na2FlBr4Cl4 ) + + ++ ) ++
Clear zone diameter + = 1–5 cm; ++ = 5–10 cm; ) = no obvious effects.
Table 4. Physicochemical properties of photosensitizers
Photosensitizer kmax (nm)a
Log P F¢b
Na2Fl 490 )0.28 0.02
FlBr2 504 1.01 0.67
FlI2 507 1.21 0.69
Na2FlBr4 517 )0.25 0.81
Na2FlI4 527 )0.24 0.92
Na2FlBr4Cl4 539 )0.21 1
a
Measured in water.
b
Relative to the singlet oxygen yield of Na2FlBr4Cl4.
H. Wang et al.: The Phototoxicity of Xanthene Derivatives 3

4. presence of heavy bromine or iodine atoms enhanced
the yields of intersystem crossing to the reactive triplet
state of the xanthene dyes [11].
The Log P values measured in the experiments
showed that FlBr2 and FlI2 were lipophilic (Log P >
0), while Na2Fl, Na2FlBr4, Na2FlI4, and Na2FlBr4Cl4
were hydrophilic (Log P < 0). Variation in Log P was
expected to affect the uptake and localization of the
photosensitizers. In the current study, there was no
obvious correlation between relative lipophilicity and
activity. Due to the work of Pooler and Valenzeno
[18], we knew that xanthenes were typically localized
in cell membrane. At a molecular level, xanthenes
mostly photosensitized the cross-linking of proteins
and formed the hydroperoxides from unsaturated lip-
ids, thereby increasing the osmotic fragility of the
cells. It was not necessary for such photosensitizers to
pass through the membrane to attack intracellular
targets. However, Pimprikar and Heitz [17] observed
the toxicity ratio to Aedes mosquito larvae ranged up
to 2 orders of magnitude between the soluble and
insoluble forms of the same xanthene dye. With the
lipophilic xanthenes, mosquito larvae were able to
filter feed on dye particles and thereby received a
higher level of the dye. Thus, lipophilic xanthenes
showed higher activity against mosquito larvae than
organisms in present tests.
There was a clear tendency that xanthenes showed
higher activity against Gram-positive bacteria. The
additional layer of protection provided by the outer
membrane of Gram-negative bacteria could generally
hinder the binding of the photosensitizers and intercept
photo-generated reactive species. Researchers drew the
same conclusion in a previous study [16].
In the present study, photosensitizers showed rela-
tively higher toxicity upon illumination and lower tox-
icity without illumination against yeast S. cerevisiae
than bacteria. Yeast was shown to be more sensitive
to phototoxic reaction. Thus, it may be considered that
S. cerevisiae is better than E. coli or S. aureus in
screening useful photosensitizers.
Our results showed the dyes tested here could kill
bacteria and yeast at micromolar concentrations. In
relation to therapeutic use, although several photo-
sensitizers have been used successfully in photother-
apy, the dyes tested here still need some possible
structural modifications to make their kmax lie within
the window of 600–900 nm used for the treatment of
human conditions. All the tested dyes will not persist
or remain toxic for a very long time in the environ-
ment. Previous research showed that exposure of
Na2Fl or Na2FlBr4Cl4 to sunlight was expected to lead
to photodegradation with a half time of approximately
1 hour [10]. Thus, Suredye (69% Na2FlBr4Cl4 and
31% Na2Fl by weight of active components) becomes
the most successfully used photoinsecticide to control
the fruit fly [8]. In addition, Na2FlBr4Cl4 had been
approved by EPA (USA) for trial applications to
control corn root worms in 2000 and some more of
these dyes were in experiments for insect control. At
the same time, some researchers stated the photosen-
sitized inactivation of yeast and bacteria could be used
for the decontamination of microbially polluted waters
[15]. Thus, photoactivated dyes show potential as
antimicrobial agents; additional studies are needed to
define and explore their applications for biological
control and decontamination.
ACKNOWLEDGMENTS
We thank Dr. Xiaofan Zhang for his technical assistance.
Literature Cited
1. Daniels F (1965) A simple microbiological method for dem-
onstration phototoxic compounds. J Invest Dermatol 44:259–
263
2. Davies MJ, Truscott RJW (2001) Photo-oxidation of proteins and
its role in caractogenesis. J Photochem Photobiol B Biol 63:114–
125
3. DeRosa MC, Crutchley RJ (2002) Photosensitised singlet oxygen
and its applications. Coord Chem Rev 37:351–371
4. Dougherty TJ, Gomer CJ, Henderson BW, Jori G, Kessel D,
Korbelik M, et al. (1998) Photodynamic therapy. J Nat Cancer Inst
90:889–905
5. Dowell RV, Wilson W, Hennessy MK (1997) Toxicity of Sure-
dye to beneficial insects. Photochem Photobiol, 65S: Abstract
TMP D-6
6. Foote CS (1991) Definition of type I and type II photosensitised
oxidation. Photochem Photobiol 54:659–664
7. Girotti AW (2001) Photosensitised oxidation of membrane lipids:
reaction pathways, cytotoxic effects, and cytoprotective mecha-
nisms. J Photochem Photobiol B Biol 63:103–113
8. Grant TM, Roy TC, Steven LP (1999) Suppressing oriental fruit
fly populations with phloxine B-protein bait sprays. Pestic Sci
55:574–576
9. Heitz JR (1997) Historical perspectives of xanthene dyes as pes-
ticides. Photochem Photobiol, 65S: Abstract TPM D-1
10. Imamura M, Koizumi M (1955) Irreversible photobleaching of the
solution of fluorescent dyes: Kinetic studies on the primary. Bull
Chem Soc Jpn 28:117–124
11. Jori G, Reddi E (1991) Second generation photosensitizers for the
photodynamic therapy of tumours. In: Douglas RH, Moan J, Ronto
G (eds) Light in biology and medicine. London: Plenum Press, pp
253–266
12. Kawanishi S, Hiraku Y, Oikawa S (2001) Mechanism of guanine-
specific DNA damage by oxidative stress and its role in carcino-
genesis and ageing. Mutat Res 488:65–76
13. Kraljic I, Mohsni SE (1978) A new method of the detection of
singlet oxygen in aqueous solutions. Photochem Photobiol
28:577–581
14. Livermore DM (2001) Antibiotic resistance in Staphylococci. Int J
Antimicrob Agents 16:3–10
15. Merchat M, Spikes JD, Bertoloni G, Jori G (1996) Studies on the
mechanism of bacteria photosensitization by meso-substituted
4 CURRENT MICROBIOLOGY Vol. 52 (2006)

5. cationic porphyrins. J Photochem Photobiol B:Biol–35:149–
157
16. Phoenix DA, Sayed Z, Hussain S, Harris F, Wainwright M (2003)
The phototoxicity of phenothiazinium derivatives against Esc-
herichia coli and Staphylococcus aureus. FEMS Immunol Med
Mic 39:17–22
17. Pimprikar GD, Heitz JR (1984) Observations on unusually high
insecticidal activity of the free acid forms of xanthene dyes. J Miss
Acad Sci 78:30–34
18. Pooler JP, Valenzeno DP (1979) The role of singlet oxygen in the
photooidation of excitable cell membranes. Photochem Photobiol
30:581–589
19. Schmidt-Erfurth U, Hasan T (2002) Mechanisms of action of
photodynamic therapy with verteporfin for the treatment of
age-related macular degeneration. Surv Ophthalmol 45:195–
214
20. Wainwright M (1998) Photodynamic antimicrobial chemotherapy
(PACT). J Antimicrob Chemother 42:13–28
H. Wang et al.: The Phototoxicity of Xanthene Derivatives 5