Science against microbial pathogens: communicating current research and technological advances

_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

Antimicrobial activity of natural photosensitizing anthraquinones

S.C. Nez Montoya1, #, L.R. Comini1, # and J.L. Cabrera1.
1
Pharmacognosy, Department of Pharmacy, School of Chemical Sciences, National National University of Crdoba,
IMBIV-CONICET, Ciudad Universitaria 5016 Crdoba, Argentine.
#
S.C.N.M. and L.R.C. contributed equally to this publication

It was demonstrated that anthraquinone-rich extracts, obtained from the phototoxic vegetal species Heterophyllaea
pustulata Hook f. (Rubiaceae), exhibited bacteriostatic activity on Micrococcus luteus ATCC 9341, selectively inhibiting
both oxacillin-sensitive and resistant Staphylococcus aureus, and antifungal activity against important opportunist
microorganisms and against those involved in superficial mycosis, all from nosocomial origin. The acute in vitro
cytotoxicity evaluation of each anthraquinone (AQ) isolated from these bioactive extracts, on a mammalian eukaryotic cell
line (Vero cells), allowed us to establish the non-cytotoxic concentration range, which was used to evaluate the
antimicrobial effect. Four from nine AQs tested, soranjidiol, rubiadin, damnacanthal and (S)-5,5'-bisoranjidiol, showed in
vitro bacteriostatic/bactericide activity against S. aureus. The action mechanism seems to involve an increase in the levels
of superoxide anion and/or singlet oxygen molecular. Moreover, the effect of actinic irradiation as a boosting agent for the
production of both reactive oxygen species as well as its influence on antibacterial effect was assessed.

Keywords anthraquinones; antimicrobial activity

1. Introduction
Throughout history, natural products have been a rich source of compounds that have found many applications in the
field of medicine. In microbiology, particularly, several plant-derived compounds have been studied with this aim,
including alkaloids, flavonoids, tannins, quinones, essential oils and other secondary metabolites [1]. Among them,
anthraquinone derivatives (AQ) have aroused special interest since they have demonstrated potential therapeutic uses as
antibacterial, antiviral, antifungal agents and other biological activities [1-5]. Within this family of compounds, several
AQs have been thoroughly studied in relation to their photosensitizing properties in photodynamic reactions [6, 7]. On
the basis of this, some of them show good antibacterial and antiviral effects, by producing reactive oxygen species
(ROS) such as superoxide anion (O2), hydroxyl radical (OH.) and singlet molecular oxygen (1O2), in the presence of
light, with subsequent oxidative damage [8, 9]. Hypericin, a photosensitizing AQ found in certain members of the genus
Hypericum, is a clear case showing significant antimicrobial activity by means of a photodynamic photosensitization,
acting mainly through the generation of ROS, particularly 1O2 [10-12]. It is widely accepted that substances with
photosensitizing characteristics have become particularly relevant due to their potential applications in photodynamic
antimicrobial chemotherapy (PACT), which involves photosensitizers and visible or ultraviolet light. This therapeutic
(PACT) has been proposed in the treatment of local infections, especially those from caries, periodontal diseases, oral
candidiasis as well as wounds [9, 13].
In this context, bearing in mind the many potential applications of photosensitizers and the need for new chemical
structures with this particular feature, we started a series of chemical, physical and biological studies on
photosensitizing AQs isolated from a phototoxic plant species, Heterophyllaea pustulata Hook f. (Rubiaceae). This
vegetal species grows in the Andean mountain range in the northwest of Argentina [14] and the animals that ingest the
aerial parts of this plant experience a typical primary photosensitization reaction, clinically revealed by dermatitis and
blindness in severe cases [15, 16]. The chemical investigation of this plant revealed the presence of several constituents
(AQs, flavonoids and iridoids) with an evident predominance of aglicone-9,10-AQs. From leaves and stems, nine AQs
namely, soranjidiol, soranjidiol 1-methyl ether, rubiadin, rubiadin 1-methyl ether, damnacanthal, damnacanthol,
heterophylline, pustuline and (S)-5,5-bisoranjidiol (Fig. 1), were isolated and purified by using repeated combination of
several chromatographic techniques. The identification of each metabolite was made by applying different
spectroscopic/spectrometric techniques (UV-V, IR, 1H-RMN, 13C-RMN, HRMS) [17, 18]. In previous studies we
demonstrated that these AQs also exhibit photosensitizing properties by generation of O2 (Type I mechanism) and/or
1
O2 (Type II mechanism) [19, 20], which are directly involved in the phototoxic effects that H. pustulata produces on
cattle [21].
Following previous studies, we have focused our attention on studying the in vitro antibacterial and antifungal
activity of enriched AQ extracts obtained from H. pustulata. To analyze further the results obtained, we evaluated the
antibacterial activity of each isolated AQ, centering our interest on determining whether this effect was solely due to a
photodynamic process or to a concurrent combination of light-driven and dark processes.

FORMATEX 2011 3
Science against microbial pathogens: communicating current research and technological advances
______________________________________________________________________________
A. Mndez-Vilas (Ed.)

OH

O 1
2

O R1

8 9 1 O
R5 8a R2
7 9a 6
2 O
2 5
A C HO
6 3 5
4a 1
10a
R4 10 4 R3
5 HO OH
6
O
O

Anthraquinone nucleous (S)-5,5'-bisoranjidiol

AQs R1 R2 R3 R4 R5
soranjidiol OH CH3 H OH H
soranjidiol 1-methyl ether OCH3 CH3 H OH H
rubiadin OH CH3 OH H H
rubiadin 1-methyl ether OCH3 CH3 OH H H
damnacanthal OCH3 CHO OH H H
damnacanthol OCH3 CH2-OH OH H H
2-hydroxi-3-methyl AQ H OH CH3 H H
heterophylline OH CH3 H OH OCH3
pustuline H OH OCH3 H CH3
Fig. 1 Anthraquinones structures used in this study.

2. In vitro antibacterial and antifungal activity of enriched AQ extracts obtained from

Heterophyllaea pustulata.
Since the AQ-rich extracts obtained from several plant species have shown significant antimicrobial activity [22], we
decided to evaluate the in vitro antibacterial and antifungal properties of enriched AQ extracts obtained from H.
pustulata [17].
In order to prepare extracts, the air-dried plant material was separated into stems and leaves. Both were ground
separately and extracted with benzene in a Soxhlet apparatus. Phytochemical studies have shown that benzenic stem
extract has damnacanthal, damnacanthol, rubiadin, soranjidiol and rubiadin 1-methyl ether. Benzenic leaf extract has a
similar chemical composition except for the absence of damnacanthal and damnacanthol, and the presence of
soranjidiol 1-methyl ether, heterophylline, pustuline and 5,5'-bisoranjidiol.
A screening with the benzenic stem and leaf extracts was performed in relation to their in vitro antibacterial activity
against the reference strain Micrococcus luteus ATCC 9341, oxacillin-sensitive and resistant Staphylococcus aureus,
coagulase-negative (c.n.) S. saprophyticus, Escherichia coli (two different strains), Proteus mirabilis, Pseudomonas
aeruginosa (two different strains) and the fungal species Candida albicans, C. glabrata, C. krusei, C. parapsilosis, C.
tropicalis, Cryptococcus neoformans, Aspergillus fumigatus, A. flavus, Trichophyton mentagrophytes, T. rubrum, and T.
floccosum [17]. The clinical importance of these pathogenic species (oxacillin-resistant and sensible S. aureus; the later,
a Beta-lactamase producer) should be stressed, since they participate in diverse infections such as abscesses,
endocarditis, pneumonia, meningitis and osteomyelitis. In addition, Candida is the fourth germ isolated from hospital
hemocultures, associated with 38% of mortality. C. albicans participates in cutaneous and deep candidiasis. C. krusei
offers natural resistance to fluconazole, an antibiotic commonly administered in numerous treatments. C. parapsilosis
represents the main cause of candidemia in children and catheter-bearing patients. Cryptococcosis and aspergillosis
produce mortality rates higher than 95% and are transmitted by the inhalatory route, frequently affecting AIDS patients.
T. mentagrophytes is associated with superficial skin, nail and hair mycosis [17].
The antimicrobial activity of benzenic stem and leaf extracts was determined using the agar-well diffusion method
modified according to the following experimental conditions, using nutrient agar for bacteria or Sabouraud agar for
fungi [17]. Except for M. luteus, all bacterial microorganisms were pathogenic strains isolated from patients with
different infections. All fungal microorganisms were isolated from different lesions of patients from the Hospital de
Clnicas (Buenos Aires), as described by Prez et al. [23]. Six-millimeter-diameter wells were punched into the agar
and filled with different extract dilutions of a stock solution (1 mg/ml in EtOH) and solvent blanks or standard antibiotic
solutions, according to the sensitivity of the microorganism tested [24]. The antimicrobial activity was evaluated by
measuring the inhibition-zone diameter observed.
In addition, the minimal bactericidal concentration (MBC) was determined [17]. Cylindrical pieces (4 mm diameter)
were extracted from the inhibition-zones of S. aureus produced by the highest concentration of the different H.
pustulata extracts. The pieces were transferred to sterile tubes containing tryptose broth. The tubes were incubated at 26
C for 24 h. A 100 ml aliquot of this broth was spread over Petri plates containing sterile nutrient agar. This was

4 FORMATEX 2011
 Science against microbial pathogens: communicating current research and technological advances
_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

incubated at 36 C for 24 h and the development of microorganisms was checked. The minimal extract concentration
that reduces the viable bacteria to 1:1000 or less is considered MBC.
The benzenic stem extract shows selective activity on Micrococcus luteus ATCC 9341, oxacillin-sensitive and
resistant Staphylococcus aureus, without acting against other Gram + as S. saprophyticus c.n. neither Gram as
Escherichia coli (two different strains), Proteus mirabilis nor Pseudomonas aeruginosa (two different strains). In this
aspect the leaf extract showed a similar activity to that of the stem [17]. A bacteriostatic effect of the different H.
pustulata extracts against oxacillin-sensible and resistant S. aureus is implied by the MBC/MIC (minimum inhibitory
concentration) ratio, whose results are clearly higher than 1 [17].
Moreover, the antimicrobial spectrum of the benzenic stem extract proves interesting because of its activity against
fungal species from nosocomial origin. Indeed, it was useful against fungal cultures isolated from different corporal
lesions such as several strains of C. albicans, C. krusei, C. parapsilosis, C. tropicalis, C. neoformans, A. fumigatus, A.
flavus and T. mentagrophytes (Table 1).

Table 1 Antifungal activity of Heterophyllaea pustulata stem extract (1mg/ml)

Microorganism Source material Inhibition I.A. (u)
Diameter
(mm)
Extract Mz

Opportunistic mycosis
Candida albicans I hemoculture 11 32 26
Candida albicans II urineculture MC - 22 6
Candida albicans III hemoculture 37203 18 46 11
Candida albicans IV hemoculture 37204 - 40 11
Candida albicans V hemoculture 15564 17 36 11
Candida albicans VI hemoculture 16655 22 60 11
Candida albicans VII hemoculture 1999 11 45 11
Candida glabrata I penis swab (candidemia by C. - - 12
glabrata) 35234
Candida glabrata II hemoculture 35202 - 38 11
Candida glabrata III hemoculture 36293 - 40 14
Candida krusei I laryngeal prothesis - - 11
Candida krusei II urineculture 38495 - - 11
Candida parapsilosis I hemoculture 35416 22 40 11

Candida parapsilosis II hemoculture 15760 25 30 11

Candida tropicalis hemoculture 16819 18 35 11
Cryptococcus neoformans CSF (HIV patient) 20772 15 25 11
Aspergillus fumigatus nasal polyps (fangal sinusitis) 13 40 13
Aspergillus flavus nasal swab (in leukemic patient) 21 40 12

Superficial mycosis
Trychophyton rubrum Tinea pedis (male patient) - - 2
Trychophyton mentagrophytes toe nail 12 30 10
Epidermophyton floccosum Tinea pedis (a year evolution in HIV patient) - - 2
Control: miconazole (MZ) (1 mg/ml). I.A.: inoculum absorbance at 580 nm. Inoculum dilution in the culture medium was 1/8 v/v.
Values are means of 2 to 4 data that differ in less than 10 % from the mean. - = no activity detected

On the other hand, acute toxicity studies were performed by administration schedules using unique doses on female
mice of CF1 strain, 3-month old and weight ranging between 31 and 40 g [17]. Because of the low solubility
characteristics of the benzenic stem and leaf extract, a mixture of dimethylsulfoxide (DMSO)/H2O (3:1) (v/v) was
selected among several solvents. The extract was administered subcutaneously in doses ranging from 60 to 4,000
mg/kg. The intravenous administration was performed through the tail veins with doses within 70 to 280 mg/kg range.
In both cases, the control mice were injected with the corresponding vehicle. The toxicity signs, including death, were
observed during 8 days following the extract administration. The LD50 was estimated through the up-and-down method
for small samples by Dixon [25].

FORMATEX 2011 5
Science against microbial pathogens: communicating current research and technological advances
______________________________________________________________________________
A. Mndez-Vilas (Ed.)

Upon subcutaneous administration, the mice tolerated doses of up to 4,000 mg/Kg body wt. without signs of toxicity,
indicating a LD50 equal or higher than that level. With the aim of determining LD50 values and taking into account the
difficulty of obtaining the extract, acute toxicity following intravenous administration was studied. We established a
LD50 value of 123 mg/Kg body wt. without higher signs of toxicity in the surviving mice. In both cases, the benzenic
extract of stem and leaf appears to exhibit similar patterns [17].
In conclusion, the AQ-rich extracts have proved antibacterial and antifungal effect, without a manifest toxicity in
experimental animals at the concentration assayed.
Some of these experiments were performed in the Department of Pharmacology, School of Dentistry, University of
Buenos Aires, under the direction of Dr. Cristina Prez.

3. Evaluation in vitro of biological effects of each AQ isolated from H. pustulata.

3.1 In vitro antibacterial activity of each AQ

Considering the above results of the antimicrobial activity of AQ-rich extracts, obtained from H. pustulata, we
evaluated the in vitro antibacterial activity of each AQ against a Gram (+) bacterium with the aim of establishing the
metabolites responsible for the effect shown by the extracts.
Routine susceptibility tests, including the determination of MIC, are an estimate of bacteriostatic action. In this case,
measurement of the MIC for each AQ was carried according to international standards from the Clinical and Laboratory
Standards Institute (CLSI) by means of broth macrodilution method in Mueller-Hinton medium (MH, Britania) [26].
The strain used for this purpose was S. aureus ATCC 29213 (1x105 colony-forming units (cfu)/ml). The working
solutions of each AQ were prepared by means of seriated dilutions from 256 g/ml up to 0.125 g/ml. MIC
determination was carried out after 24 h of incubation at 37 C by observing turbidity.
Our studies led to the conclusion that only soranjidiol, rubiadin, damnacanthal and 5,5'-bisoranjidiol showed
susceptibility to MICs within a 32-64 g/ml range.
Bearing in mind that for many substances having antibacterial activity there are mechanisms that elicit a
physiological response in bacteria by producing reactive oxygen species (ROS) [27-29], our objective is demonstrated if
the antibacterial effect shown by soranjidiol, rubiadin, damnacanthal and 5,5'-bisoranjidiol (through the determination
of the MIC) is directly linked to the increase of O2 and/or 1O2 levels. Nevertheless, it should be noted that some
photosensitizers also generate bacterial photoinactivation mainly by means of the generation of ROS, particularly
singlet molecular oxygen (1O2), but through a photodynamic photosensitization [13]. Therefore, as the AQs studied are
photosensitizing agents, we also assessed whether the irradiation causes photosensitization by these AQs, which
consequently increases the production of ROS and its antibacterial effect.
In order to reach this conclusion, the experimental model was carried out according to the following guidelines: the
AQs were dissolved in phosphate buffer solution (PBS) and the concentrations used throughout were below the minimal
inhibitory concentration (MIC) (32-64 g/ml), specifically was selected the subtoxic doses of 10 g/ml because higher
concentrations proved to be toxic for normal mammal cells (African green monkey kidney cells Vero) (see results
below) [30].
The four AQs that showed susceptibility to S aureus (soranjidiol, rubiadin, damnacanthal and 5,5'-bisoranjidiol) were
examined to ascertain whether this activity was associated with an increased generation of O2 with respect to the basal
production. The O2 determination was based on the NBT assay [26].
Two assays were simultaneously performed in darkness. One was carried out in the absence of AQs to measure the
basal level of the O2 production in the bacteria (control). The second preformed by adding AQs to the culture
medium, in order to establish whether these compounds induce an increase in the O2 generation through a
physiological response. Both assays were independently repeated under actinic irradiation to evaluate the
photosensitized response. The strain used was S. aureus ATCC 29213 (109 cfu/ml). In the NBT assay, the bacterial
suspension was incubated with each AQ solution and NBT at 37 C in darkness. Again, the control experiments were
carried out under the same conditions but without adding the AQ. Triplicate sets were incubated under actinic
irradiation. All these experiments were independently measured as a function of time [26].
Moreover, the same set of AQs used in the preceding paragraph was examined to evaluate an increase of 1O2
generation with respect to the basal production. The 1O2 production in the reaction medium was followed by
spectrophotometrical determination of the consumption of methionine (MET) at 236 nm as a function of time [31]. This
compound reacts chemically with the generated 1O2 with a rate constant (kr) of 2.1 x 107 L mol-1s-1 [32]. Here again, two
sets of experiments were done; one in darkness and the other under actinic irradiation. For each set, the procedure is
described below:
i) In order to determine the 1O2 basal production generated by bacteria (control), we incubated bacterial suspension
(109 cfu/ml), MET and PBS.

6 FORMATEX 2011
 Science against microbial pathogens: communicating current research and technological advances
_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

ii) The production of 1O2 in the presence of AQs was measured incubating bacterial suspension (109 cfu/ml), MET, 1
ml of AQ and PBS.
By using sodium azide (NaN3) as a 1O2 physical quencher (kt = 5,8 108 L mol-1s-1) we can confirm whether the
consumption of MET is a direct consequence of the 1O2 generation in the medium [33]. Due to its large quenching
constant, NaN3 was added in a concentration high enough to suppress the consumption of 1O2 by MET as described
below:
iii) The bacterial suspension (109 cfu/ml) was incubated with MET, 1 ml of AQ, NaN3 and PBS.
Fig. 2 shows the increase in percentage of O2 in S. aureus with respect to basal situation when this strain was
treated with each AQ in darkness and under actinic radiation at 10, 20 and 40 min. As observed, there is no clear
indication of O2 at 10 min, neither for darkness nor for irradiation. At 20 and 40 min, all the AQs induce an increase
in O2 production with respect to the basal situation in darkness. Except for damnacanthal, this effect is more
noticeable at longer times. Under actinic irradiation, there is a further increase in O2 production with respect to basal
situation with the same exception shown previously.

80 b a b
Increase in percentage of superxide anion production with

a b a
AQ-Darkness AQ-Irradiation 67,7
70 66,1
64,5
a
60 56,3 a
51,5
respect to basal situation

50 b a
40,2
b 40
b
40 38 35,2

30,4
29,8
30

20 17,6
15,3
13,9
11,6

10 6,5
2 3
0 1 1 2 1 1
0
10 20 40 10 20 40 10 20 40 10 20 40

-10
Time (min)
Damnacanthal Rubiadin Soranjidiol 5,5'-bisoranjidiol

Results are given as mean SD, n=3.

a
p 0.05 measured with respect to darkness
b
p 0.05 measured with respect to 20 min time

Fig. 2 NBT assay. Increase in percentage of O2 in S. aureus ATCC 29213 with respect to basal situation for every AQ and time
measured.

Fig. 3 shows the MET consumption caused by the 1O2 generated in S. aureus under the conditions outlined in i, ii, iii
(evaluation of the 1O2 generation), both for darkness and under actinic irradiation for 5,5'-bisoranjidiol. As can be
observed, in the absence of AQ, small amounts of 1O2 are produced due to the normal breathing process. MET
consumption was higher when 5,5'-bisoranjidiol was present, which means that the 1O2 production increased. This
phenomenon is observed under both working conditions (darkness and actinic irradiation). The 1O2 generated was
counteracted by addition of NaN3. This physical quencher efficiently competes with MET for the deactivation of 1O2
and consequently, the trend line is similar to control experiments. These results confirm that MET consumption is a
direct consequence of the 1O2 generation. The preceding discussion is referred only to 5,5'-bisoranjidiol. The other AQs
tested also show a similar behavior, though with a smaller increase in 1O2 generation.

FORMATEX 2011 7
Science against microbial pathogens: communicating current research and technological advances
______________________________________________________________________________
A. Mndez-Vilas (Ed.)

1,02 Darkness 1,02 Actinic Irradiation

1,00 1,00

0,98 0,98

Abs/Abs0
Abs/Abs0

0,96 0,96

0,94 Control 0,94

In presence of 5,5'-bisoranjidiol Control

0,92 In presence of NaN3 0,92 In presence of 5,5'-bisoranjidiol
In presence of NaN3
0,90 0,90
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time (s) Time (s)

Fig. 3 Consumption of MET caused by 1O2 in S. aureus ATCC 29213 treated with 5,5'-bisoranjidiol (10 g/ml), in darkness and
under irradiation. Average of three experiments.

Fig. 4 shows the MET consumption caused by the 1O2 generation in bacteria treated with each AQ (see ii: evaluation
of the 1O2 generation), in darkness vs. irradiation. Its analysis (by comparison of trend lines) shows that the 1O2
production is higher in the presence of actinic radiation with respect to darkness because MET consumption increased,
except for damnacanthal.

1,02 1,02 Rubiadin

Rubiadin 1,02 1,02
Damnacanthal
Damnacanthal
1,00 1,00
1,00 1,00

0,98 0,98 0,98 0,98

Abs/Abs0
Abs/Abs0

Abs/Abs0
Abs/Abs0

0,96 0,96 0,96 0,96

0,94 0,94 0,94 0,94

In presence
In presence
of AQof+AQ
dark+ dark In presence
In presence
of AQof+AQ
dark+ dark
In presence
In presence
of AQof+AQ
light
+ light In presence
In presence
of AQof+AQ
light+ light
0,92 0,92 0,92 0,92

0,90 0,90 0,90 0,90

0 0500 500
1000 1000
1500 1500
2000 2000
2500 2500
3000 3000
3500 3500
4000 4000 0 0500 500
1000 1000
1500 1500
2000 2000
2500 2500
3000 3000
3500 3500
4000 4000

TimeTime
(s) (s) TimeTime
(s) (s)

1,02 1,02 5,5'-bisoranjidiol

5,5'-bisoranjidiol 1,02 1,02 Soranjidiol
Soranjidiol

1,00 1,00 1,00 1,00

0,98 0,98 0,98 0,98

Abs/Abs0
Abs/Abs0
Abs/Abs0
Abs/Abs0

0,96 0,96 0,96 0,96

0,94 0,94 0,94 0,94

In presence
In presence
of AQof+AQ
dark+ dark
In presence
In presence
of AQof+AQ
light
+ light
0,92 0,92 In presence
In presence
of AQof+AQ
dark
+ dark 0,92 0,92
In presence
In presence
of AQof+AQ
light
+ light
0,90 0,90 0,90 0,90
0 0 500 500
1000 1000
1500 1500
2000 2000
2500 2500
3000 3000
3500 3500
4000 4000 0 0 500 500
1000 1000
1500 1500
2000 2000
2500 2500
3000 3000
3500 3500
4000 4000

TimeTime
(s) (s) TimeTime
(s) (s)

Fig. 4 Consumption of MET caused by 1O2 in S. aureus ATCC 29213 treated with each AQ (10 g/ml), by comparing the process
occurring in darkness and under actinic irradiation. Average of three experiments.

Lastly, bearing in mind that, while those compounds that can reduce a minimum of 103 cfu/ml (3.0 Log10) are
considered bactericide agents, those below that range (which merely inhibit growth) are considered bacteriostatic agents
[27, 34, 35], we carried out assays with the purpose of establishing whether the AQs show any of such characteristics.
With this objective, S. aureus ATCC 29213 (108 cfu/ml) was incubated in triplicate with each AQ (damnacanthal,

8 FORMATEX 2011
 Science against microbial pathogens: communicating current research and technological advances
_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

rubiadin, soranjidiol and 5,5'-bisoranjidiol) at 37 C independently in darkness and under irradiation, replicating also the
whole set to be measured at 10, 40 and 90 min [26]. After incubation time, bacteria were serially 10-fold diluted with
PBS and each dilution was placed on plate count agar MH and incubated for 24 h at 37 C. Bactericidal activity was
determined by means of plate recount of colony-forming units per milliliter (cfu/ml).
A particular experiment was carried out to prove whether the suppression of 1O2 production by a chemical quencher
changes the reduction in the number of cfu/ml. A bacterial suspension (108 cfu/ml) with 5,5'-bisoranjidiol under
irradiation was treated with MET, after 40 min of incubation, allowing the system to proceed normally and measuring
the cfu/ml at 90 min.
Fig. 5 shows Log10 change in cfu/ml of S. aureus treated with each AQ in darkness, including the control. As noted, a
decrease, more pronounced for 5,5'-bisoranjidiol, is seen after an initial increase. This reduction could relate to the
capability of producing a physiological response in bacteria that generate O2 (Fig. 2) as well as 1O2 (Fig. 3) in the
absence of light. 5,5'-bisoranjidiol was the only AQ that showed bactericide effect on this microorganism with a
reduction of 3.0 Log10 of cfu/ml at 90 min. Rubiadin and soranjidiol produce a reduction of 2.2 and 2.0 Log10 of cfu/ml,
respectively. The increase in O2 production for rubiadin and sorianjidiol is similar to that of 5,5'-bisoranjidiol (Fig. 2).
Nevertheless, the increase in 1O2 generation is lower than the bianthraquinone. For damnacanthal, only an inhibitory
effect without producing death is noticed.

9 9

8 8

7 7
Log10 cfu/ml

Log10 cfu/ml
6 6

5 No AQ 5 NoAQ
Damnacanthal Damnacanthal
4 4
Rubiadin Rubiadin
3 Soranjidiol 3 Soranjidiol
5,5'-bisoranjidiol 5,5'-bisoranjidiol
2 2
5,5'-bisoranjidiol + MET (40 min.)
1 1
0 20 40 60 80 100 0 20 40 60 80 100

Time (min) Time (min.)

Error bars represent SD of the mean.

Fig. 5 Log10 change in colony-forming units per milliliter Fig. 6 Log10 change in colony-forming units per milliliter
(cfu/ml) of S. aureus ATCC 29213 treated with each AQ (cfu/ml) of S. aureus ATCC 29213 treated with each AQ
(10 g/ml) in darkness. (10 g/ml) under irradiation.

On the other hand, the actinic radiation and the consequent O2 and 1O2 increase by means of a photosensitization
phenomenon (Figs. 2 and 4), further increased the reduction of cfu/ml for all AQs, execept damnacanthal (Fig. 6).
These results clearly demonstrate that 5,5'-bisoranjidiol shows higher antibacterial activity than that of rubiadin and
soranjidiol, although the three increase O2 production at about the same level as a function of time (Fig. 2) regardless
that they are acting in darkness or under irradiation. In contrast, the 1O2 production for 5,5'-bisoranjidiol is high
irrespective of conditions (darkness or irradiation) [26]. This would suggest that, among the different species that
comprise ROS, 1O2 is the one mainly involved in the bactericidal effect, as already reported by Becerra et al. [31] using
precisely S. aureus. This becomes evident since, after 40 min of incubation, the addition of MET to a sample having
5,5'-bisoranjidiol shows that even under irradiation, the cfu/ml reduction is completely suppressed at 90 min (open
squares, Fig. 6).
In conclusion, these assays allowed us to identify the AQs: rubiadin, soranjidiol, damnacanthal and 5,5'-bisoranjidiol
as the compounds responsible for the antibacterial effects shown by AQ-rich extracts obtained from the phototoxic
vegetal species, H. pustulata. In addition, our results suggest that the antibacterial effect on S. aureus, found for these
AQs, is closely linked to the increase in O2 and/or 1O2 levels. This increase could result from the interaction between
bacteria and the AQs without needing light, i.e., without a photosensitizing process to produce a physiological response
eliciting O2 and 1O2. Actinic irradiation, on the other hand, generates photosensitization for rubiadin, soranjidiol and
5,5'-bisoranjidiol, consequently increasing their antibacterial effects (particularly bactericide). This effect is, in turn,
suppressed when a specific quencher of 1O2 is added, thus suggesting that the bactericidal activity derives mainly from
that particular ROS as already proposed in other reports [36].
Most importantly, the particular bactericidal activity shown under irradiation for rubiadin, soranjidiol and 5,5'-
bisoranjidiol at the concentration used, which does not affect normal mammal cells (subtoxic concentration - see results
above), led us to consider these AQs as potential agents for photodynamic antibacterial chemotherapy treatments.

FORMATEX 2011 9
Science against microbial pathogens: communicating current research and technological advances
______________________________________________________________________________
A. Mndez-Vilas (Ed.)

Some of these experiments were performed interdisciplinarily with Dr. Ins Albesa (Laboratory of Pharmaceutical
Microbiology, Department of Pharmacy) and Dr. Gustavo A. Argello (Department of Physical Chemistry), School of
Chemical Sciences, National University of Cordoba.

3.2 Cytotoxic evaluation in vitro of each AQ

Prior to the antibacterial tests of each AQ, it was necessary to evaluate the cytotoxicity of these compounds on a
mammalian eukaryotic cell line (Vero cells) [37], with the aim of establishing the non-cytotoxic concentration range for
evaluating the antimicrobial effect of each AQ.
For this purpose, African green monkey kidney cells (Cercopithecus aethiops, Vero 76 ATCC CRL-587) were used.
The cytopathic effect produced by each AQ on the morphology of Vero cells was observed by optical microscopy [38].
From a stock solution of each AQ (1 mg/ml in PBS with 1% DMSO as co-solvent), 15 consecutive dilutions were
prepared with PBS, within a range of 1 to 30-50 g/ml according to AQ solubility. Each dilution was inoculated in
duplicate on a confluent cell monolayer (2.5 0.6 x 105 cells/ml, 48 h incubation). Cell controls (CC), containing only
maintenance medium (MM: Eagles minimum essential medium (MEM) with 2% Fetal calf serum (FCS), 1% L-
glutamine, gentamicine (50 g/ml) and 1% DMSO), were included. The cells were incubated at 37 C during 72 h, and
the development of cellular alterations such as rounding, membrane retraction, cell detachment and the presence of
granules in the cytoplasm was daily observed [38].
The cellular viability (CV) depending on the concentration of each AQ was measured by means of the uptake Neutral
Red (NR) assay [39]. Each dilution was inoculated in triplicate on a confluent monolayer of cells (2.5 0.6 x 105
cells/ml). The absorbance of the NR extracted after 48 h of incubation at 37 C was measured at 540 nm on a microplate
reader (BioTek ELx800). The percentage of cellular viability (CV%) was calculated by comparison with CC (100%
viability). The concentration of compound that reduces the viable cells to 50% (CC50) was determined by regression
(R2> 0.9) from the plot of cellular viability percentage vs. AQ concentration [39]. Maximum Non-Cytotoxic
Concentration (MNCC) was defined as the maximal sample concentration showing more than 90% viable cells and
exerting no cytotoxic effect as detected by microscopic monitoring [40].
Our results showed that, whereas rubiadin 1-methyl ether, damnacanthol and pustuline stand out by showing MNCC
values within a 16 and 22 g/ml range, this concentration was not higher than 10 g/ml for the remaining AQs (Table
2). Thus, the MNCC for soranjidiol 1-methyl ether, heterophylline and 5,5'-bisoranjidiol was near 10 g/ml, and
approximately 6 g/ml for soranjidiol, rubiadin and damnacanthal (Table 2).

Table 2 Cytotoxic Concentration to 50% (CC50), Maximum Non-Cytotoxic Concentration (MNCC) and Subtoxic
Concentration for each AQ tested on Vero cells.
AQs CC50 (g/ml) MCNC (g/ml) Subtoxic concentration (g/ml)
rubiadin 1-methyl ether 34.4 0.2 22.2 0.4 26.8 0.1
damnacanthol 33.7 0.2 19.6 0.2 a 23.9 0.1
pustuline nc 16.1 0.3 a 22.3 0.1
soranjidiol 1-methyl ether 27.1 0.2 10.5 0.3 a 18.4 0.1
heterophylline 23.69 0.04 9.7 0.2 b 15.64 0.04
(S)- 5,5'-bisoranjidiol 22.7 0.1 9.5 0.2 b 13.9 0.2
damnacanthal 20.1 0.1 6.7 0.2 c 11.2 0.1
soranjidiol 17.5 0.2 6.3 0.2 c 9.9 0.1
rubiadin 14.9 0.2 5.3 0.3 c 8.2 0.1
nc: not calculated
a
p <0.05 calculated with respect to b
b
p < 0.05 calculated with respect to c

Similar results were obtained by analyzing the CC50 determined for each AQ (Table 2), which established that
rubiadin 1-methyl ether was less cytotoxic, followed, in increasing order of cytotoxicity, by damnacanthol, pustuline,
soranjidiol 1-methyl ether, heterophylline, 5,5'-bisoranjidiol, damnacanthal, soranjidiol and rubiadin. The CC50 of
pustuline was not estimated because concentrations higher than 30 g/ml could not be tested due to solubility problems;
however, its MNCC was established. In addition, a subtoxic concentration was determined for all AQs tested (Table 2),
defined as the concentration that causes 10 - 20% cellular death [41] and produces slight morphologic changes observed
by microscopy.
In addition, in previous studies we demonstrated that some of the AQs tested had the ability to increase the O2
production in human leukocytes [19, 20]. Considering that this effect could generate the cytotoxic activity of AQs on
Vero cells, the ability of each AQ to produce this ROS and its relation to the cytotoxic effect were evaluated. To this
aim, the NBT reduction bioassay was performed and each AQ was evaluated at 10 g/ml, which represents a non-
cytotoxic or subtoxic concentration depending on the compound studied [42, 43].

10 FORMATEX 2011
 Science against microbial pathogens: communicating current research and technological advances
_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

Thus, rubiadin 1-methyl ether, damnacanthol and pustuline did not increase the production of O2 at 10 g/ml (Fig.
7) since it is a non-cytotoxic concentration for these AQs (approximately 95% CV) without evidencing any cytopathic
effect [30]. For the other AQs, this amount corresponds to a subtoxic value, increasing the generation of O2 (Fig. 7). It
was noted that those compounds producing a large increase in the O2 generation at 10 g/ml (soranjidiol and
damnacanthal, Fig. 7) exhibit a low CV% (about 80%) with increased cell damage [30]. However, those AQs that at the
same concentration showed a small increase in the O2 production (soranjidiol 1-methyl ether, heterophylline, 5,5'-
bisoranjidiol and rubiadin), reveal a high CV% (between 86 and 90%), except for rubiadin [30]. In general, we might
conclude that an increased production of O2 causes an s important cytopathic effect as observed by microscopy, with
a concomitant decrease in cellular viability. Rubiadin is excepted from this behavior, producing a significant cytopathic
effect at 10 g/ml, which results in a significant decrease in CV (30%), despite having a low production of O2 (Fig 7)
[30]. We may therefore assume the presence of another mechanism in the cytotoxicity of this compound. In addition,
when each AQ was tested at their CC50, the increase in O2 production was not the same in all AQs. Therefore, the
intracellular increase of this ROS would not be the sole cause for the loss of cellular viability at CC50.

Fig. 7 NBT assay. Increase in percentage of O2 in Vero cells with respect to basal situation, produced by two different
concentrations of every AQ.

Finally we carried out an AQ incorporation assay in Vero cells [19], which constitutes a spectrophotometric
determination of intracellular content of AQs needed to stimulate O2 production and other biologic effects. Soranjidiol
was chosen to study the incorporation of AQs in Vero cells because this AQ is the predominant compound in the aerial
parts of H. pustulata. Thus, by means of this assay, we have established that soranjidiol enters Vero cells, 29 3% with
respect to the initial concentration after 30 min incubation. Although only a single AQ was tested, no significant
differences are expected in the rate of incorporation for the other AQs since they all have similar partition coefficients
[44].
In conclusion, this work allowed us to establish the concentration range where each AQ exhibits low or no cytotoxic
effect and therefore, these concentrations may be used in order to test their potential antimicrobial effects. From the
nine AQs tested, we were able to identify three derivatives: rubiadin 1-methyl ether, damnacanthol and pustuline, with
low or no cytotoxicity (95 5% VC) in a concentration range limited by the MNCC (Table 2). The estimation of the
subtoxic concentration for the other AQs (soranjidiol, soranjidiol 1-methyl ether, rubiadin, damnacanthal,
heterophylline and 5,5'-bisoranjidiol) allowed us to consider that a concentration of about 10 g/ml could be used to test
different biological activities, since this concentration ensured in our experiments more than 80% CV (Table 2).

Acknowledgements The support by Agencia Nacional de Promocin Cientfica y Tecnolgica (FONCYT), Consejo Nacional de
Investigaciones Cientficas y Tcnicas (CONICET), Ministerio de Ciencia y Tecnologa de la Provincia de Crdoba (Argentina), and
Secretara de Ciencia y Tecnologa de la Universidad Nacional de Crdoba (SeCyT-UNC) is gratefully acknowledged.

References
[1] Cowan MM. Plant Products as antimicrobial agents. Clin. Microbiol. Rev. 1999;12(4):564-582.
[2] Wijnsma R, Verpoorte R. Anthraquinones in the Rubiaceae. In: Hill RA, Krebs HCh, Verpoorte R, Wijnsma R, Herz W,
Grisebach H, Kirby GW, Tamm Ch, eds. Progress in the Chemistry of Organic Natural Products. Wien: Springer-Verlag;
1986,49:79-149.
[3] Rath G, Ndonzao M, Hostettmann K. Antifungal anthraquinones from Morinda lucida. Int. J. Pharmacognosy. 1995;33(2):107-
114.

FORMATEX 2011 11
Science against microbial pathogens: communicating current research and technological advances
______________________________________________________________________________
A. Mndez-Vilas (Ed.)

[4] Ali AM, Ismail NH, Mackeen MM, Yazan LS, Mohamed SM, Ho ASH, Lajis NH. Antiviral, cytotoxic and antimicrobial
activities of anthraquinones isolated from the roots of Morinda elliptica. Pharm. Biol. 2000;38(4): 298-301.
[5] Agarwal SK, Singh SS, Verma S, Kumar S. Antifungal activity of anthraquinone derivatives from Rheum emodi. J.
Ethnophamacol. 2000;72:43-46.
[6] Gutirrez I, Bertolotti SG, Biasutti MA, Soltermann AT, Garca NA. Quinones and hydroxyquinones as generators and
quenchers of singlet molecular oxygen. Can. J. Chem. 1997;75:423-428.
[7] Gollnick K, Held S, Mrtire DO, Braslavsky SE. Hydroxyanthraquinones as sensitizers of singlet oxygen reactions: quantum
yields of triplet formation and singlet oxygen generation in acetonitrile, J. Photochem. Photobiol. A: Chem. 1992;69:155-165.
[8] Zerdin K, Horsham MA, Durham R, Wormell P, Scully AD. Photodynamic inactivation of bacterial spores on the surface of a
photoactive polymer. Reac. Funct. Polymers.2009;69:821-827.
[9] Luksien Z. New approach to inactivation of harmful and pathogenic microorganisms by photosensitization. Food Technol.
Biotechnol. 2005;43(4):411-418.
[10] Lopez-Bazzocchi I, Hudson JB, Towers GHN. Antiviral activity of the photoactive plant pigment hypericin. Photochem.
Photobiol. 1991;54(1):95-98.
[11] Carpenter S, Kraus GA. Photosensitization is required for inactivation of equine infectious anemia virus by hypericin.
Photochem. Photobiol. 1991;53(2):169-174.
[12] Lenard J, Rabson A, Vanderoef R. Photodynamic inactivation of infectivity of human immunodeficiency virus and other
enveloped viruses using hypericin and rose Bengal: Inhibition of fusion and syncytia formation. Proc. Natl. Acad. Sci. USA.
1993;90:158-162.
[13] Pellieux C, Dewilde A, Pierlot Ch, Aubry J-M. Bactericidal and virucidal activities of singlet oxygen generated by thermolysis
of naphthalene endoperoxides. Methods Enzymol. 2000;319:197-207.
[14] Bacigalupo NM. Rubiaceae. In: Cabrera AL, ed. Coleccin Cientfica INTA. Buenos Aires: INTA; 1993, tomo XIII, parte
IX:375-380.
[15] Hansen EW, Martiarena CA. Contribucin al estudio de la toxicidad de Heterophyllaea pustulata Hook cegadera en el
ganado. Rev. Inv. Agropec. Patol. Animal (INTA). 1967;4(7):81-113.
[16] Aguirre DH, Neumann RA. Intoxicacin por cegadera (Heterophyllaea pustulata) en caprinos del noroeste argentino. Med.
Vet. 2001;18(7-8):487-490.
[17] Nez Montoya SC, Agnese AM, Prez C, Tiraboschi IN, Cabrera JL. Pharmacological and toxicological activity of
Heterophyllaea pustulata anthraquinones extracts. Phytomedicine. 2003;10:569-574.
[18] Nez Montoya SC, Agnese AM, Cabrera JL. Anthraquinones derivatives from Heterophyllaea pustulata. J. Nat. Prod.
2006;69:801-803.
[19] Nez Montoya SC, Comini LR, Sarmiento M, Becerra MC, Albesa I, Argello Gustavo A, Cabrera JL. Natural
anthraquinones probed as Type I and Type II photosensitizers: singlet oxygen and superoxide anion production. J. Photochem.
Photobiol. B: Biol. 2005;78(1):77-83.
[20] Comini LR, Nez Montoya SC, Sarmiento M, Cabrera JL, Argello Gustavo A. Characterizing some photophysical,
photochemical and photobiological properties of photosensitizing anthraquinones. J. Photochem. Photobiol. A: Chem.
2007;188:185-191.
[21] Nuez Montoya SC, Comini LR, Rumie Vittar NB, Fernandez IM, Rivarola VA, Cabrera JL. Phototoxic effects of
Heterophyllaea pustulata (Rubiaceae). Toxicon. 2008;51:1409-1415.
[22] Sydiskis RJ, Owen DG, Lohr JL, Rosler K-A, Blomster RN. Inactivation of enveloped viruses by anthraquinones extracted
from plants. Antimicrob. Agents Chemother. 1991;35(12):2463-2466.
[23] Prez C, Tiraboschi IN, Agnese AM, Cabrera JL. Inhibition of different nosocomial microorganisms by the essential oil of
Senecio graveolens. In: Singh S, Singh VK, Govil JN, eds. Recent progress in medicinal plants. Houston, Texas, USA:
Research Periodicals and Books Publishing House; 2002:331-338.
[24] Edwards D. Antimicrobial Drug Action. London: Macmillan; 1980:259.
[25] Dixon WJ. The up-and-down method for small samples. Amer. Statistical Assoc. Journal. 1965;12:967-971.
[26] Comini LR, Nez Montoya SC, Paz PL, Argello Gustavo A, Albesa I, Cabrera JL. Antibacterial Activity of Anthraquinone
Derivatives from Heterophyllaea pustulata (Rubiaceae). J. Photochem. Photobiol. B: Biol. 2011;102:108-114.
[27] Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA. A common mechanism of cellular death induced by bactericidal
antibiotics. Cell. 2007;130:797-810.
[28] Becerra MC, Albesa I. Oxidative stress induced by criprofloxacin in Staphylococcus aureus. Biochem. Biophys. Res. Comm.
2002;297:1003-1007.
[29] Albesa I, Becerra MC, Battn PC, Pez PL. Oxidative stress involved in the antibacterial action of different antibiotics,
Biochem. Biophys. Res. Comm. 2004;317:605-609.
[30] Nez Montoya SC, Aguilar JJ, Konigheim BS, Contigiani MS, Cabrera JL. Cytotoxic activity in vitro of natural
anthraquinones. Biocell 2007; 31:95.
[31] Becerra MC, Sarmiento M, Pez PL, Arguello Gustavo A, Albesa I. Light effect and reactive oxygen species in the action of
ciprofloxacin on Staphylococcus aureus. J. Photochem. Photobiol. B: Biol. 2004;76:13-18.
[32] Miskoski S, Garca NA. Influence of the peptide bond on the singlet molecular oxygen mediated (O2 [1 delta g]
photooxidation of histidine and methionine dipeptides. A kinetic study. Photochem. Photobiol. 1993;57(3):447-452.
[33] Haag WR, Mill T. Rate constants for interaction of (1o21g) with azide ion in water. Photochem. Photobiol. 1987;45(3):317-
321.
[34] Barry AL, Craig WA, Nadler H, Reller LB, Sanders CC, Swenson JM. Methods for determining bactericidal activity of
antimicrobial agents; Approved guideline. In: Clinical and Laboratory Standards Institute (CLSI), Document M26-A.
Pennsylvania, USA: Wayne; 1999.
[35] Murray PR, Rosenthal KS, Pfaer MA. Microbiologa Mdica, 5th ed. Espaa: Elsevier; 2007.

12 FORMATEX 2011
 Science against microbial pathogens: communicating current research and technological advances
_______________________________________________________________________________
A. Mndez-Vilas (Ed.)

[36] Muller-Breitkreutz K, Mohr H, Briviba K, Sies H. Inactivation of viruses by chemically and photochemically generated singlet
molecular oxygen. J. Photochem. Photobiol. B: Biol. 1995;30:63-70.
[37] Rates SMK. Plants as source of drugs. Toxicon 2001;39:603-613.
[38] Ooi LSM, Wang H, Luk Ch-W, Ooi VEC. Anticancer and antiviral activities of Youngia japonica (L.) DC (Asteraceae,
Compositae). J Ethnopharmacol. 2004;94:117-122.
[39] Borenfreund E, Puerner JA. Toxicity determined in vitro by morphological alterations and neutral red absorption. Toxicol.
Lett. 1985;24:119-24.
[40] Liu A-L, Shu S-H, Qin H-L, Lee SMY, Wang Y-T, Du G-H. In vitro Anti-Influenza Viral Activities of Constituents from
Caesalpinia sappan. Planta Med. 2009;75:337-39.
[41] Cholewa M, Legge GJ, Weigold H, Holan G, Birch CJ. The use of a scanning proton microprobe to observe anti-HIV drugs
within cells. Life Sci. 1994;54:1607-1612.
[42] Rook GAW, Steele J, Umar S, Dockrell HM. A simple method for the solubilisation of reduced NBT, and its use as a
colorimetric assay for activation of human macrophages by -interferon. J. Inmonolog. Met. 1985;82:161-67.
[43] Choi HS, Kim JW, Cha Y-N, Kim Ch. A Quantitative Nitroblue Tetrazolium Assay for Determining Intracellular Superoxide
Anion Production in Phagocytic Cells. J. Immunoassay Immunochem. 2006;27:31-44.
[44] Comini LR, Nez Montoya SC, Argello Gustavo A, Cabrera JL. Determinacin del coeficiente de particin (log P) para
derivados antraquinnicos aislados de Heterophyllaea pustulata Hook. f. (Rubiceas). Lat. Ame. J. Pharm. 2006;25:252-255.

FORMATEX 2011 13