Mutation Research 611 (2006) 34–41

Genotoxic evaluation of extracts from Aplysina fulva,a Brazilian marine sponge

Claudia Aiub a,b,∗, Ana Giannerini b, Flavia Ferreira b, Jose Mazzei b,Luiza Stankevicins b, Gisele Lobo-Hajdu c, Pedro Guimaraes d,

Eduardo Hajdu e, Israel Felzenszwalb b

a Department of Cellular Biology, Division of Genetics, Salzburg University, Salzburg A5020, Austriab Department of Biophysics and Biometry, University of Rio de Janeiro State, Rio de Janeiro 20551-030, Brazil

c Department of Cellular Biology and Genetics, Biology Institute Roberto Alcantara Gomes,University of Rio de Janeiro State, Rio de Janeiro 20551-030, Brazil

d Department of Organic Chemistry, Chemistry Institute, University of Rio de Janeiro State, Rio de Janeiro 20551-030, Brazile Department of Invertebrates, National Museum, University of Rio de Janeiro, Rio de Janeiro 20940-040, Brazil

Received 29 March 2006; received in revised form 13 June 2006; accepted 30 June 2006Available online 20 September 2006

Abstract

A range of biologically active secondary metabolites with pharmacological application has been reported to occur in marinesponges. The present study was undertaken to provide a set of data on the safety of a hydro-alcoholic extract (ALE) and an aqueousfraction (AQE) from Aplysina fulva Pallas, 1766 (Aplysinidae, Verongida, Porifera). Salmonella typhimurium strains TA97, TA98,TA100 and TA102, Escherichia coli strains PQ65, OG40, OG100, PQ35 and PQ37 and Balb/c 3T3 mouse fibroblasts were used todetect induction of DNA lesions by ALE and AQE. Assays used for these analyses were a bacterial (reverse) mutation assay (Amestest), the SOS-chromotest and the comet assay. Both extracts presented identical infrared 2-oxazolidone spectra. ALE treatmentinduced a higher frequency of type-4 comets, indicative of increasing DNA migration, in the alkaline comet assay. ALE also induceda weak genotoxic effect, as expressed by the induction factor (IF) values in the test with E. coli strain PQ35 (IF = 1.5) and by cytotoxiceffects in strains PQ35, PQ65 and PQ37. Positive SOS induction (IF = 1.7) was detected in strain PQ37 treated with diluted AQE. Nogenotoxic effects were observed in strains PQ35, PQ65, OG40 and OG 100 after treatment with AQE dilutions. Using the bacterial(reverse) mutation test and survival assays with or without S9 mix, after 60 min of pre-incubation, we observed for strain TA97treated with ALE a weak mutagenic response (MI = 2.2), while cytotoxic effects were seen for strains TA98, TA100 and TA102. AQEdid not show mutagenic activity in any of the strains tested, but a weak cytotoxic effect was noted in strain TA102. Our data suggest

that both ALE and AQE from A. fulva induce DNA breaks leading to cytotoxicity and mutagenicity under the conditions used.© 2006 Elsevier B.V. All rights reserved.

Keywords: Aplysina fulva; Sponge extracts; Cytotoxicity; Mutagenicity; Gen

∗ Corresponding author at: Department of Cellular Biology, Divi-sion of Genetics, Salzburg University, Hellbrunnerstrasse 34, SalzburgA5020, Austria. Tel.: +43 662 8715 44; fax: +43 662 80445776.

E-mail address: aiub@hotmail.com (C. Aiub).

1383-5718/$ – see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.mrgentox.2006.06.035

otoxicity; 2-Oxazolidone

1. Introduction

Marine sponges provide an array of biologicallyactive secondary metabolites with structural features dis-tinct from those of other natural sources [1]. Nineteenamong 21 sponge species screened inhibited growth

n Resea

ota

sossTmwdpatbeAtbrCbAhsepgr

sba

2

2Ae

ldacfivN1

u

C. Aiub et al. / Mutatio

f at least one species of microorganisms. Amonghose sponges, Aplysina fulva Pallas 1766 was stronglyntibacterial and slightly antifungal [2].

The different species of Aplysina (Porifera, Demo-pongiae, Verongida) can be distinguished by the lackf terpenes, the presence of carotenoids, and by aeries of brominated compounds derived from tyro-ine and of sterols in moderately high content [3–5].he bromometabolites, which are considered chemicalarkers of this order, have generally been extractedith low-polarity solvents and are soluble in eitherichloromethane or chloroform [6]. Some of these com-ounds present in Aplysina sp. possess in vitro cytotoxicctivities [7–9] and have shown activity against humanumor cells [10,7]. 2-Oxozoline derivatives have alsoeen found in A. fulva collected on the Virgin Islands,xtracts of which have presented cytotoxic activity [9].mong the 2-oxazoline derivatives were also bromo-

yrosine compounds known as fistularin-1 and -2, or ais-2-oxazoline derivative containing a dibromophenoling. These substances have also been isolated from otheraribbean sponges [5,11]. Other more polar and non-rominated secondary metabolites, such as aplysillin

[12] and some tryptophan-derived [13] substancesave been isolated from Aplysina species. Bromotyro-ine metabolites have represented one-fourth of the crudextract of the Mediterranean sponge, Aplysina aero-hoba. More recently, the isolation of a novel acidiclycogen from the Brazilian sponge A. fulva has beeneported [14].

The present study was undertaken to provide a basicet of data on the cytotoxicity, mutagenicity and DNA-reakage inducing capacity of a hydro-alcoholic extractnd its aqueous fraction of A. fulva.

. Material and methods

.1. Preparation, characterization and standardization of. fulva hydro-alcoholic (ALE) and aqueous (AQE)xtracts

Samples of A. fulva (about 120 g wet weight) were col-ected at Praia Joao Fernandinho, Armacao de Buzios (Rioe Janeiro, Brazil; 22◦44′49′′S/41◦52′54′′W), in May 2002t a depth of 2–7 m. They were washed with sea water,leaned of all visible surface debris, washed rapidly withreshwater and stored in a volume of 70% ethanol correspond-ng to 10 times its wet weight, for 3 months at −4 ◦C. A

oucher sample is deposited in the Porifera collection of Museuacional, Universidade Federal do Rio de Janeiro (MNRJ704).

A 10 mL of the hydro-alcoholic solution were evaporatednder vacuum at 55 ◦C, and the residue was lyophilized and

rch 611 (2006) 34–41 35

frozen at −20 ◦C until use. The residue was suspended in 2 mLof sterile distilled water and filtered through a 0.22-�m poremembrane.

Cytotoxicity and genotoxicity were evaluated with both theoriginal hydro-alcoholic extract and its aqueous fraction, afterdilution. Three hydro-alcoholic extract solutions were preparedby dilution-, 2-, 5- and 10-fold into 0.2 M sodium phosphatebuffer (pH 7.4). In the same way, three solutions of the aqueousfraction were prepared by 10-, 20- and 40-fold dilution. AQEsolutions were more diluted than the ALE solutions in viewof the higher concentrations in AQE after the preparative stepdescribed above.

Both extracts were identically fractionated following atypical method for bromotyrosines and other alkaloids inAplysina sponges [6,8,10]. The extracts (50 mL) were five-fold concentrated by removing the solvent under vacuumand dissolving the residue completely in 70% ethanol inwater (10 mL). The concentrated solutions were extractedconsecutively with hexane and chloroform at room temper-ature. The chloroform extracts were dried over anhydroussodium sulfate and the solvent removed under reduced pres-sure. The residues were passed through Sephadex LH-20(Pharmacia Biotech, USA) with chloroform-methanol 1:1.The solvent was removed from the eluted solution andthe residue was chromatographed through a column filledwith silica-gel 60(S) 70-230 mesh (Vetec-Brazil) eluted withchloroform–methanol 9:1. The solvent was evaporated andthe residues analyzed by ultraviolet and infrared spectrome-try.

The ultraviolet spectra of the final fractions (diluted inmethanol) were recorded on a Shimadzu UV-160A spec-trophotometer. After being dissolved in chloroform, the frac-tions were examined as a glassy film deposited over potas-sium bromide plate by Fourier-transform infrared spectrom-etry, recorded on a Perkin-Elmer Spectrum One instru-ment.

2.2. Bacterial strains

Escherichia coli SOS chromotest strains were obtainedfrom Dr. M. Hofnung, Institute Pasteur, Paris (France). Bac-terial strains PQ65, OG40 and OG100 are derived from E.coli K12. PQ35 and PQ37 are derived from GC4436 [15,16].The Salmonella typhimurium strains TA98, TA97, TA100 andTA102 were provided by Dr. B.N. Ames, University of Cali-fornia, Berkeley, CA, USA.

2.3. S9 fraction

An S9 fraction, prepared from livers of Sprague–Dawleyrats pre-treated with a polychlorinated biphenyl mixture

(Aroclor 1254), was purchased from Molecular ToxicologyInc. (MoltoxTM, USA). The S9 metabolic activation mixture(S9 mix) used for the SOS chromotest and for the bacterial(reverse) mutation test (Ames test) was used as previouslydescribed [17].

n Resea

36 C. Aiub et al. / Mutatio

2.4. SOS chromotest

The SOS chromotest was performed, for all ALE and AQEdilutions, according to Quillardet and Hofnung [16] with someadaptations [18]. Aflatoxin B1 (AFB-1, CAS # 1162-65-8)and 4-nitroquinoline 1-oxide (4-NQO, CAS # 56-57-5) bothat 0.5 �g/mL in DMSO from Sigma Chemical Co. (St. Louis,USA), were used as positive controls in assays with and with-out the presence of the S9 mix, respectively. Pre-incubationof ALE or AQE with the bacterial culture was performed at37 ◦C for 60 min under shaking, followed by centrifugationand suspension in 0.9% NaCl. The pre-incubation mixture con-sisted of 100 �L of each ALE or AQE concentration, 500 �Lof S9 mixture (4%), or the same volume of 0.2 M sodium-phosphate buffer (pH 7.4) for experiments without S9 mix, and100 �L of the bacterial suspension (2 × 109 cells/mL). Enzymeactivities were estimated in Miller units (Abs420 × 1000/pre-incubation time of the substrate, in minutes) [19]. The resultswere considered positive when the increase of the inductionfactor was higher than 1.5 combined with an increase in the�-galactosidase activity [16]. All experiments were done intriplicate and repeated twice.

2.5. Bacterial (reverse) mutation test

Mutagenicity in both ALE and AQE dilutions was testedaccording to Maron and Ames [20] with some adaptations[18]. S. typhimurium strains TA97, TA98, TA100 and TA102were pre-incubated for 60 min with and without S9 mix.The pre-incubation mixture consisted of 100 �L of ALE orAQE dilutions, 500 �L of S9 mixture (4%), or the same vol-ume of 0.2 M sodium-phosphate buffer (pH 7.4) for experi-ments without S9 mix, and 100 �L of the bacterial suspension(2 × 109 cells/mL). The mixture was pre-incubated at 37 ◦Cunder shaking. After 60 min, 2 ml top-agar (0.6% agar, 0.6%NaCl, 50 �M l-histidine, 50 �M biotin, pH 7.4, 45 ◦C) wasadded in the test tube and the final mixture was poured ontoa Petri dish with minimal agar (1.5% agar, Vogel-Bonner Emedium, containing 2% glucose). This final mixture was incu-bated at 37 ◦C for 72 h, and the His+ revertant colonies werecounted.

Positive controls for assays in the absence of S9 mixwere: 4-NQO (1.0 �g/plate), for TA97; 2-nitrofluorene (2-NF, CAS#607-57-8) at 1.0 �g/plate for TA 98; sodium azide(SA, CAS#26628-22-8) at 0.5 �g/plate for TA100; mito-mycin C (MMC, CAS#50-07-7) at 0.5 �g/plate for TA102.Positive controls for assays in the presence of S9 mixwere: 2-aminofluorene (2-AF, CAS# 153-78-6) at 10 �g/platefor TA97; 2-amino-anthracene (2-AA, CAS#613-13-8) at1.0 �g/plate for TA98; benzo[a]pyrene (BaP, CAS#50-32-8)at 5.0 �g/plate for TA100; (2-AF) at 10 �g/plate for TA102.

All chemicals were from Sigma.

The substance or sample was considered positive for muta-genicity when: (a) the number of revertant colonies in the assaywas at least twice the number of spontaneous revertants (muta-tion induction, MI ≥ 2), (b) a significant response was found

rch 611 (2006) 34–41

with analysis of variance (ANOVA, p ≤ 0.05) and (c) a repro-ducible positive dose-response curve (p ≤ 0.01) was present.All experiments were done in triplicate and repeated twice.

2.6. Survival experiments

For determination of the cytotoxic effects of the ALE andAQE extracts, samples from the pre-incubation assay mixturein the bacterial (reverse) mutation test were diluted in 0.9%NaCl (w/v) to obtain a suspension containing 2 × 103 cells/mL.A suitable aliquot (100 �L) of this suspension was plated onnutrient agar (0.8% bacto nutrient broth (Difco), 0.5% NaCland 1.5% agar). The plates were then incubated at 37 ◦C for24 h and the colonies were counted [18]. All experiments weredone in triplicate and repeated twice.

2.7. Comet assay—single-cell gel electrophoresis (SCGE)

Balb/c 3T3 mouse fibroblasts (clone A31, ATCC CCL163) in exponential growth (3 × 104 cells) were incubated withALE and AQE extracts for 60 min. To verify DNA dam-age, cells were processed for the comet assay immediatelyafter treatment. To this end, adherent fibroblasts were har-vested by trypsin-EDTA treatment, centrifuged for 3 min at1000 × g, and resuspended in ice-cold Dulbecco’s minimalessential medium (DMEM) with 10% fetal bovine serum(FBS). To detect single- and double-stranded DNA breaks[21], 10 �L of the cell suspensions was mixed with 120 �L of0.5% low-melting agarose in phosphate-buffered saline (PBS)pH 7.4, and added onto microscope slides pre-coated with1.5% normal-melting agarose. Slides were recovered with amicroscope coverslip and refrigerated for 5 min, followed byimmersion in alkaline lysing solution (2.5 M NaCl, 10 mMTris, 100 mM EDTA, 10% DMSO, 1% Triton X-100, finalpH > 10.0) for 1 h. Slides were incubated for 20 min in an elec-trophoresis tray with an alkaline-EDTA solution (0.2 M NaOH,1 mM EDTA, pH 12.2) to allow the DNA to unwind. Elec-trophoresis was then conducted at room temperature for 25 minat 25 V/300 mA. The slides were rinsed with water, allowed todry at 37 ◦C and stained with 20 �g/mL ethidium bromide. TheDNA of individual cells was viewed by use of a fluorescencemicroscope (Olympus), with 516–560 nm excitation/emissionfrom a 50-W mercury light source. Semi-quantitative measure-ments of DNA breakage were made by visual scoring of 50randomly selected cells per slide, classifying them into fourcategories representing increasing degrees of damage, rangingfrom Comet 1, with a minimal detectable frequency of DNAlesions, to a maximum Comet length (type 4) [22]. The positivecontrol was 200 �M H2O2 (Fluka, CAS#7722-84-1). Statisti-cal analysis was conducted with Student’s or Welch t-test, usingInstat 2.01 software (GraphPad Software, CA, USA).

3. Results

The chloroform extracts of both the hydro-alcoholic and the aqueous fraction presented identi-

n Research 611 (2006) 34–41 37

ci1a1cg

abmIimh2coaofm

bouo

4auA

Fgfacdlew

Fig. 2. Comet-assay evaluation of DNA damage in exponentiallygrowing fibroblasts pre-treated with an aqueous extract (AQE) fromA. fulva, for 60 min. Quantification of DNA breakage was achievedby visual scoring of 50 randomly selected cells per slide, classifiedinto five categories each representing a different degree of damage,ranging from no comet (type 0, undamaged cells) to maximum length

C. Aiub et al. / Mutatio

al infrared spectra after elution from silica-gel, show-ng absorption maxima between 2950 and 2854 cm−1,500–1200 cm−1, and near 1740 cm−1 characteristic of2-oxazolidone spectrum. The strong absorption at

668 cm−1 and in the broad 3000–3600 cm−1 range indi-ates the possible presence of compounds with amideroups.

The lack of absorption near 1550 cm−1 indicated thebsence of the isoxazoline amide groups of the knownromotyrosines, which are commonly found in Aplysinaarine sponge extracts obtained with less polar solvents.

n addition, the lack of absorption near 2265 cm−1 isndicative of the absence of aeroplysin-I [9,8,10,5]. The

aximum absorption in ultraviolet spectra of both theydro-alcoholic extract and the aqueous fraction was at10 nm, not showing absorption near 280 nm, typical ofyclohexadiene derivatives, corroborating the absencef bromotyrosine derivatives as indicated by infrarednalyses. The spectral data were compared with thosef fistularin-3 and brominated dimetoxicetal isolatedrom Brazilian specimens of Aplysina caissara. Theseetabolites were not encountered in ALE and AQE.The comet assay is used to measure the extent of DNA

reakage and shows a large number of lesions in DNAf treated cells, especially with ALE. The comparison ofntreated cells with ALE- or AQE-treated cells can bebserved as a dose–response curve (Figs. 1 and 2).

ALE treatment induced a higher frequency of type-comet images with increasing DNA migration in the

lkaline comet assay (P < 0.01), when compared withntreated cells (Fig. 1) and when compared with theQE-treated cells (Fig. 2).

ig. 1. Comet-assay evaluation of DNA damage in exponentiallyrowing fibroblasts pre-treated with a hydro-alcoholic extract (ALE)rom Aplysina fulva, for 60 min. Quantification of DNA breakage waschieved by visual scoring of 50 randomly selected cells per slide,lassified into five categories each representing a different degree ofamage, ranging from no comet (type 0, undamaged cells) to maximumength comet (type 4, maximally damaged cells). Statistical values forach tested ALE dilution (*P < 0.05 or **P < 0.001) were comparedith the negative control (NaCl, 0.9%).

comet (type 4, maximally damaged cells). Statistical values for eachtested AQE dilution (*P < 0.05 or **P < 0.001) were compared with thenegative control (NaCl, 0.9%).

Tables 1 and 2 show genotoxicity induced by ALE andAQE, respectively, expressed as induction-factor val-ues (IF) either in the presence or absence of metabolicactivation by S9 mix in E. coli uvrA (PQ37), nth(OG40), oxyR (OG100) and PQ35 and PQ65 wild-typestrains.

A weak induction of sfiA expression (IF = 1.5) wasobserved for the wild-type strain PQ35 for the 1:5dilution of ALE in the presence of S9 mix (P > 0.05).Under metabolic activation conditions, cytotoxicity wasobserved for strain PQ35 only at the 1:2 ALE dilution(P < 0.05), and for strain PQ65 at the 1:5 (P < 0.05) and1:2 (P < 0.01) dilutions. All ALE test dilutions for strainPQ37, in the absence of S9 mix, caused a significantdecrease in alkaline phosphatase (P < 0.05), suggestingcytotoxicity. No genotoxic effects were observed forstrains OG40 and OG 100 treated with ALE dilutions(Table 1).

Positive SOS induction (IF = 1.7) was detected forstrain PQ37 treated with 1:20 and 1:10 AQE dilutions(P < 0.05) in the absence of S9 mix. For strains PQ35,PQ65, OG40 and OG 100 genotoxic effects were notobserved for the AQE dilutions tested (Table 2).

Tables 3 and 4 show the mutation induction (MI)and cytotoxicity of ALE and AQE using the bacterial(reverse) mutation test and survival assays with or with-out S9 mix, after 60 min of pre-incubation.

For ALE, strain TA97 shows a mutagenic response,

in the absence of metabolic activation, at 1:2 dilution(MI = 2.2). For strain TA98, either in the presence or inthe absence of S9 mix, ALE was cytotoxic at 1:5 and1:2 dilutions. The cytotoxic effect was also observed for

38 C. Aiub et al. / Mutation Research 611 (2006) 34–41

Table 1SOS induction by hydro-alcoholic extract of Aplysina fulva (ALE)in different E. coli strains without (−S9) or with (+S9) metabolicactivation

Strain ALE −S9 +S9

�-Gala PAa IF �-Gala PAa IF

PQ35 0 4.15 0.58 1.0 3.29 7.81 1.01:10 4.55 0.47 1.4 3.11 7.33 1.01:5 4.93 0.53 1.3 4.11 6.54 1.51:2 3.90 0.55 1.0 3.67 4.66 1.9*

PQ37 0 1.90 0.17 1.0 0.21 3.37 1.01:10 1.88 0.07 2.5** 0.17 2.78 1.01:5 1.82 0.08 1.9* 0.13 2.34 1.01:2 1.50 0.08 1.6* 0.10 2.56 0.7

PQ65 0 4.30 0.88 1.0 0.82 6.12 1.01:10 4.05 0.88 1.0 0.60 3.62 1.31:5 4.27 0.90 1.0 0.54 1.96 2.1*

1:2 4.05 0.85 1.0 0.80 2.05 3.0**

OG40 0 1.03 0.78 1.0 1.14 6.62 1.01:10 0.80 0.88 0.7 1.28 6.26 1.11:5 0.90 0.95 0.7 0.95 5.88 1.01:2 1.35 1.10 0.9 1.16 5.18 1.3

OG100 0 0.85 0.75 1.0 0.23 7.14 1.01:10 0.80 1.01 0.7 0.24 7.62 1.01:5 0.96 1.28 0.7 0.27 7.49 1.31:2 1.08 1.01 0.9 0.12 7.81 0.7

a �-Galactosidase activity (�-Gal) and alkaline phosphatase activ-ity (PA) were measured in Miller units. The period of pre-incubationwith each substrate was 90 min. The induction factor (IF) corre-sponds to the ratio of the activities of �-Gal/PA divided by the controlgroup value. Statistical values for each tested ALE dilution (*P < 0.05or **P < 0.001) were compared with the negative control (NaCl,0.9%). Bold numbers indicate a positive toxicity and/or mutagenic-

Table 2SOS induction by aqueous extract of A. fulva (AQE) in E. coli without(−S9) or with (+S9) metabolic activation

Strain AQE −S9 +S9

�-Gala PAa IF �-Gala PAa IF

PQ35 0 3.23 7.70 1.0 0.24 3.56 1.01:40 2.80 6.43 1.0 0.21 3.60 0.91:20 3.57 7.40 1.2 0.16 2.78 0.81:10 3.43 7.57 1.0 0.18 2.60 1.0

PQ37 0 0.14 3.13 1.0 0.11 1.58 1.01:40 0.17 3.13 1.2 0.14 1.63 1.21:20 0.24 2.80 1.9* 0.08 1.53 0.71:10 0.24 3.13 1.7* 0.11 1.70 0.9

PQ65 0 0.30 0.63 1.0 0.33 0.22 1.01:40 0.40 0.60 1.4 0.33 0.27 0.81:20 0.20 0.60 0.7 0.29 0.33 0.61:10 0.20 0.67 0.6 0.34 0.26 0.9

OG40 0 2.43 3.47 1.0 0.29 1.47 1.01:40 2.53 3.50 1.0 0.31 1.84 1.11:20 2.63 4.23 0.9 0.31 1.37 1.11:10 2.13 4.33 0.7 0.42 1.89 1.1

OG100 0 0.11 1.97 1.0 0.10 1.01 1.01:40 0.07 1.77 0.8 0.11 0.90 1.21:20 0.06 1.83 0.6 0.11 0.86 1.31:10 0.04 1.77 0.4 0.08 0.90 0.9

a �-Galactosidase activity (�-Gal) and alkaline phosphatase activity(PA) were measured in Miller units. The period of pre-incubation witheach substrate was 90 min. The induction factor (IF) corresponds to theratio of the activities of �-Gal/PA divided by the control group value.Statistical values for each tested AQE dilution (*P < 0.05) were com-pared with the negative control (NaCl, 0.9%). Bold numbers indicate apositive toxicity and/or mutagenicity response. Positive controls for the

ity response. Positive controls for the assay without S9 mix: 4-NQO(1.0 �g/mL), for PQ37, IF = 5.8; with S9 mix: 2-amino-anthracene (2-AA) (1.0 �g/mL) for PQ37, IF = 6.4.

strain TA100 at 1:2 and for TA102 at 1:5 and 1:2 ALEdilutions, in the presence of S9 mix (Table 3).

The results observed for strain TA100 treated withALE in the presence of metabolic activation was consid-ered a false-positive result, because at this concentrationtoxicity was observed (Table 3).

AQE revealed no mutagenicity in all tester strainsevaluated by the bacterial (reverse) mutation test, never-theless a weak cytotoxic effect was observed for TA102at 1:20 and 1:10 test dilutions, in the presence of S9 mix(Table 4).

4. Discussion

Poriferan species have been endowed with certainchemicals to compensate for their otherwise inadequatedefense mechanisms. Many of the compounds produced

assay without S9 mix: 4-NQO (1.0 �g/mL), for PQ37, IF = 5.8; withS9 mix: 2-amino-anthracene (2-AA) (1.0 �g/mL) for PQ37, IF = 6.4.

by marine sponges are toxic towards a diversity of organ-isms such as fish, bacteria, and fungi. By an intricateprocess, such chemicals appear in the surface tissues ofthe sponges or are secreted into the surrounding waters.These chemicals have helped to distinguish betweenspecies and clarify misconceptions within the taxonom-ical tree [2]. However, when the application of variouschemical compounds was investigated, the concept oftoxic chemical defense by sponges was developed. Inrecent years, sponges have been recognized for theirpharmacological purposes [12–14].

The apparent distinctiveness of the chemical profilesof Aplysina species collected in Brazil, when comparedwith profiles obtained from the same species elsewhere,

suggests the importance of assessing the geographicdimension of chemical diversity. Given the considerablemorphologic plasticity observed in A. fulva along theBrazilian coast [23], the possibility that A. fulva might

C. Aiub et al. / Mutation Research 611 (2006) 34–41 39

Table 3Induction of His+ revertants in Salmonella typhimurium strains TA97, TA98, TA100 and TA102 by hydro-alcoholic extracts ALE obtained from astored solution of sponge A. fulva, without (−S9) and with (+S9) metabolic activation

Strain ALE −S9 +S9

MIa His+b % Survivalc MIa His+b % Survivalc

TA97 0 1.0 115 100 1.0 92 1001:10 1.4 166 100 0.9 84 761:5 1.8 210 80 1.6 151 751:2 2.2 255 80 1.9 177 73PC 2.3 378 4.6 648

TA98 0 1.0 29 100 1.0 27 1001:10 1.2 35 100 1.4 38 841:5 1.6 46 64 1.5 40 311:2 1.7 49 58 1.7 46 23PC 15 488 4.1 763

TA100 0 1.0 134 100 1.0 71 1001:10 1.4 182 100 1.0 72 951:5 1.4 187 100 1.4 99 831:2 1.6 209 77 2.3 165 61PC 7.1 626 8.6 812

TA102 0 1.0 94 100 1.0 127 1001:10 1.1 108 83 0.9 116 881:5 0.9 86 82 0.9 119 441:2 0.9 86 80 0.8 106 28PC 3.6 591 6.9 1090

a Mutagenic index (MI): number of His+ revertants induced in the sample/number of spontaneous His+ revertants in the negative control.s, repeaThe do

7 sponse

cc

mefczchthrnl

tcebsld

b Number of His+ revertants/plate: mean values of three experimentc Survival proportion calculated in relation to the negative control.0%. Bold numbers indicate a positive toxicity and/or mutagenicity re

omprise a complex of sibling species cannot be dis-arded.

Our data have shown that in bacterial tests and inammalian cells, the ALE presents higher cytotoxic

ffects compared with AQE (Tables 3 and 4). The reasonor this could be based on the presence of the same polarompounds in different concentrations, such as oxa-olidone derivatives. Unfortunately, these compoundsould not be quantified, but it has been shown that N-exyl-2-oxazolidone causes a mutagenic response in S.yphimurium strain TA98 at 0.04 �g/mL [24]. Othersave shown that nitroso-5-methyloxazolidone given toats induced mainly forestomach neoplasms and a feweoplasms of the duodenum, suggesting that inducedesions, if not repaired, can lead to cancer [25].

Both ALE and AQE extracts show a significant cyto-oxicity for mammalian cells, when compared with theontrol group in the comet assay (Figs. 1 and 2). Thextent of spontaneous DNA diffusion was evaluated

y measuring the diameter of the cell nucleoids, con-idering that cells containing extremely low molecu-ar weight DNA associated with apoptosis or necrosisevelop spontaneous DNA diffusion into agarose gel and

ted twice.se was considered toxic when the percentage of survival was below. PC: positive control as described in Section 2.

consequently have larger diameters [26]. ALE treatmentinduced a higher frequency of type-4 comet images, sug-gesting necrotic and apoptotic processes by an increasein DNA migration in the alkaline comet assay (Fig. 1).These cytotoxic effects induced by ALE can be corrob-orated by the results shown in Tables 3 and 4. Thesefindings suggest that ALE and AQE induce DNA single-strand breaks that, if not repaired, may lead to cell death.

Although in the presence of ALE a weak geno-toxic and frameshift-mutation response has been shownfor PQ65 and TA97, stronger cytotoxic effects can beobserved for most of the strains used in the presence ofS9 mix (Tables 1 and 3). These results indicate that polarcompounds present in ALE are weak, direct-acting geno-toxic and mutagenic agents in prokaryotic organisms.

The effect of metabolic activation suggests that muta-genic compounds present in the AQE dilution can bedetoxified, decreasing the induction factor. AQE polarchemicals that could induce DNA lesions repaired by

the nucleotide excision repair pathway lead to a higherresponse in the uvrA PQ37 strain. This mutagenic activ-ity is not detected in the presence of metabolic activation(Table 3).

40 C. Aiub et al. / Mutation Research 611 (2006) 34–41

Table 4Induction of His+ revertants in S. typhimurium strains TA97, TA98, TA100 and TA102 by aqueous fractions AQE obtained from a stored solutionof sponge A. fulva, without (−S9) and with (+S9) metabolic activation

Strain AQE −S9 +S9

MIa His+b % Survivalc MIa His+b % Survivalc

TA97 0 1.0 116 100 1.0 131 1001:40 0.9 105 100 1.0 129 1001:20 1.0 125 100 1.0 136 1001:10 0.9 107 97 1.0 134 100PC 2.3 378 4.6 648

TA98 0 1.0 27 100 1.0 34 1001:40 1.0 27 82 1.2 41 1001:20 1.0 28 80 1.0 35 1001:10 1.0 27 78 1.1 38 93PC 15 488 4.1 763

TA100 0 1.0 126 100 1.0 123 1001:40 0.9 108 99 1.0 120 1001:20 0.9 109 96 1.1 136 901:10 0.9 119 77 1.1 131 73PC 7.1 626 8.6 812

TA102 0 1.0 142 100 1.0 315 1001:40 1.1 151 72 0.9 271 841:20 1.1 153 71 0.7 213 611:10 1.3 180 70 0.8 239 60PC 3.6 591 6.9 1090

a Mutagenic index (MI): number of His+ revertants induced in the sample/number of spontaneous His+ revertants in the negative control.b + s, repea

The docontrol

Number of His revertants/plate: mean values of three experimentc Survival proportion calculated in relation to the negative control.

70%. Bold numbers indicate a positive toxicity response. PC: positive

In line with the assumption that marine sponges do notsuffer from neoplastic diseases it has been reported thatthey are able to inhibit the production of DNA lesionsinduced by mutagens or carcinogens, because they havean efficient bio-transformation and detoxification capac-ity [27,28].

Acknowledgements

The authors are thankful to Rita Maria P. de Sa,Departamento de Quımica Organica, UERJ, Brazil, forthe spectrometric analysis. We also thank Dr. RobertoGomes de Souza Berlinck, Chemistry Department andMolecular Physics, USP, Brazil for the spectral data ofsome bromoderivatives. This work was supported byFAPERJ, CNPq and SR2-UERJ.

References

[1] A.S. Sarma, T. Daum, W.E.G. Muller, Secondary Metabo-lites from Marine Sponges, Akademie Gemeinnutziger Wis-senschaften zu Erfurt, Ullstein-Mosby Verlag, Berlin, Germany,1993.

ted twice.se was considered toxic when the percentage of survival was belowas described in Section 2.

[2] G. Muricy, E. Hajdu, F.V. Araujo, A.N. Hagler, Antimicrobialactivity of southwestern Atlantic shallow-water marine sponge(Porifera), Sci. Mar. 57 (1993) 427–432.

[3] W.L. Lee, B.M. Gilchrist, Carotenoid patterns in twenty-nine species of sponges in the order Poecilosclerida(Porifera:Demospongiae): a possible tool for chemosys-tematics, Mar. Biol. 86 (1985) 21–35.

[4] A. Kelecom, G.J. Kannengiesser, Chemical constituents of Veron-gia Sponges I–A comparison between Brazilian and Mediter-ranean species, Ann. Acad. Bras. Ciencia 51 (1979) 633–637.

[5] P. Ciminiello, V. Costantino, E. Fattorusso, S. Magno, A. Man-goni, M. Pansini, Chemistry of Verongida sponges. II. Con-stituents of the Caribeean Sponge Aplysina fistularis forma fulva,J. Nat. Prod. 57 (1994) 705–712.

[6] Y. Gopichand, F.J. Schmitz, The microbial substances of sponges.IV. Structure of bromine-containing compound from a marinesponge, Tetrahedron Lett. 35 (1979) 2823–2826.

[7] A.D. Rodrıguez, I.C. Pina, The structures of aplysamisines I, IIand III: new bromotyrosine-derived alkaloids from the Caribbeansponge Aplysina cauliformis, J. Nat. Prod. 56 (1993) 907–914.

[8] M.R. Kernan, R.C. Cambie, P.R. Bergquist, Chem-istry of sponges, VII. 11,19-Dideoxyfistularin 3 and

11-hydroxyaerothionin, bromotyrosine derivatives fromPseudoceratina durissima, J. Nat. Prod. 53 (1990) 615–622.

[9] Y. Gopichand, F.J. Schmitz, Marine natural products: fistularin-1, -2 and -3 from the sponge Aiplysina fistularis forma fulva,Tetrahedron Lett. 41 (1979) 3921–3924.

n Resea

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

C. Aiub et al. / Mutatio

10] A.I. Acosta, A.D. Rodrıguez, 11-Oxoaerothionin: a cytotoxicantitumor bromotyrosine-derived alkaloid from the Caribbeanmarine sponge Aplysina lacunose, J. Nat. Prod. 55 (1992)1007–1012.

11] P. Ciminiello, E. Fanttorusso, S. Magno, M. Pansini, Chemistry ofVerongida sponges. VI. Comparison of the secondary metaboliccomposition of Aplysina insularis and Aplysina fulva, Biochem.Syst. Ecol. 24 (1996) 105–113.

12] N.K. Gulavita, S.A. Pomponi, A.E. Wright, M. Garay, M.A. Sills,Aplysillin A, a thrombin receptor antagonist from the marinesponge Aplysina fistularis forma fulva, J. Nat. Prod. 58 (1995)954–957.

13] K. Kondo, J. Nishi, M. Ishibashi, J. Kobayashi, Two newtryptophan-derived alkaloids from the Okinawan marine spongeAplysina sp., J. Nat. Prod. 57 (1994) 1008–1011.

14] M.S. Zierer, R.P. Vieira, B. Mulloy, P.A.S. Mourao, A novelacidic glycogen extracted from the marine sponge Aplysina fulva(Porifera–Demospongiae), Carboh. Res. 274 (1995) 233–244.

15] O. Huisman, R. dıAri, An inducible DNA replication—cell divi-sion coupling mechanism in E. coli, Nature 290 (1981) 797–799.

16] P. Quillardet, M. Hofnung, The SOS chromotest, a colorimetricbacterial assay for genotoxins procedures, Mutat. Res. 147 (1985)65–78.

17] L. Da Costa Lopes, F. Albano, G. Laranja, L. Alves, L. Martins eSilva, G. Souza, I. Araujo, J. Nogueira-Neto, I. Felzenszwalb, K.Kovary, Toxicological evaluation by in vitro and in vivo assays

of an aqueous extract prepared from Echinodorus macroophyllusleaves, Toxicol. Lett. 116 (2000) 189–198.

18] C. Aiub, L.F. Ribeiro Pinto, I. Felzenszwalb, N-Nitrosodiethylamine mutagenicity at low concentrations,Toxicol. Lett. 145 (2003) 36–45.

[

rch 611 (2006) 34–41 41

19] J. Miller, Experiments in Molecular Genetics, Cold Spring HarborLaboratory Press, Cold Spring Harbor, NY, 1972.

20] D. Maron, B. Ames, Revised methods for the Salmonella muta-genicity test, Mutat. Res. 113 (1983) 173–215.

21] D. Fairbairn, P. Olive, K. OıNeill, The comet assay: a compre-hensive review, Mutat. Res. 18 (1995) 1054–1056.

22] K. Kovary, T. Louvain, M. Costa e Silva, F. Albano, B. Pires,G. Laranja, C. Lage, I. Felzenszwalb, Biochemical behavior ofnorbixin during in vitro DNA damage induced by reactive oxygenspecies, Br. J. Nutr. 85 (2001) 431–440.

23] B.M. Saeki, A.C. Granato, R.G.S. Berlinck, A. Magalhaes,A.B. Schefer, A.G. Ferreira, U.S. Pinheiro, E. Hajdu, Twounprecedent dibromotyrosine-derived alkaloids from the Brazil-ian endemic marine sponge Aplysina caissara, J. Nat. Prod. 65(2002) 796–799.

24] J.T. Fruin, L.J. Sauers, F.R. Pulliam, The mutagenic potencial ofN-hexyl-2-oxazolidone. Letterman Army Inst. Res. San FranciscoCA (25) (1981) A831301.

25] W. Lijinsky, M.D. Reuber, Carcinogenicity of hydroxylated alkyl-nitrosoureas and of nitrosooxazolidones by mouse skin paintingand by gavage in rats, Cancer Res. 43 (1983) 214–221.

26] M. Vasques, R.R. Tice, Comparative analysis of apoptosis versusnecrosis using the single cell gel (SCG), Environ. Mol. Mutagen.29 (1997) 28.

27] B. Kurelec, S. Britvic, S. Krca, W.E. Muller, R.K. Zahn,Metabolism of some carcinogenic aromatic amines in four species

of marine sponges, Comp. Biochem. Physiol. C 86 (1987)17–22.

28] S. De Flora, M. Bagnasco, C. Bennicelli, A. Camoirano, A. Bojne-mirski, B. Kf genotoxic agents in marine sponges. Mechanismsand modulation, Mutagenesis 10 (1995) 357–364.