Anthracene induces oxidative stress and activation of antioxidant and detoxification enzymes in Ulva lactuca (Chlorophyta)

In order to analyze whether the marine macroalga Ulva lactuca can absorb and metabolize anthracene (ANT), the alga was cultivated with 5 µM ANT for 0–72 h, and the level of ANT was detected in the culture medium, and in the alga. The level of ANT rapidly decreased in the culture medium reaching a minimal level at 6 h, and rapidly increased in the alga reaching a maximal level at 12 h and then decreased to reach a minimal level at 48 h of culture. In addition, ANT induced an increase in hydrogen peroxide that remained until 72 h and a higher increase in superoxide anions that reach a maximal level at 24 h and remained unchanged until 72 h, indicating that ANT induced an oxidative stress condition. ANT induced an increase in lipoperoxides that reached a maximal level at 24 h and decreased at 48 h indicating that oxidative stress caused membrane damage. The activity of antioxidant enzymes SOD, CAT, AP, GR and GP increased in the alga treated with ANT whereas DHAR remained unchanged. The level of transcripts encoding these antioxidant enzymes increased and those encoding DHAR did not change. Inhibitors of monooxygenases, dioxygenases, polyphenol oxidases, glutathione-S-transferases and sulfotransferases induced an increase in the level of ANT in the alga cultivated for 24 h. These results strongly suggest that ANT is rapidly absorbed and metabolized in U. lactuca and the latter involves Phase I and II metabolizing enzymes.


Results
Viability of U. lactuca exposed to increasing concentrations of ANT. In order to determine a sublethal concentration of ANT in U. lactuca, the alga was cultivated in artificial seawater without ANT (control) and with increasing concentrations of ANT corresponding to 1, 10, 50, 100 and 250 µM, for 7 days (Fig. 1). Cell viability was analyzed through visualization of chlorophyll fluorescence in chloroplasts using confocal microscopy. Chlorophyll fluorescence in the algae treated with 10 μM of ANT was similar to the control indicating full viability (Fig. 1A,B). Algae treated with 50-250 µM of ANT showed a progressive decrease in chlorophyll fluorescence compared to the control (Fig. 1C-F). Thus, the chosen sub-lethal concentration of ANT for further experiments was 5 µM.
ANT is rapidly incorporated and metabolized in U. lactuca. To determine whether U. lactuca can incorporate and metabolize ANT, the alga was cultivated in without ANT (control) and with 5 μM of ANT for 0-72 h, and the level of ANT was determined in the culture medium and in the alga. The level of ANT in seawater was 5 µM, which corresponds to 1.6 µmol in 300 mL of seawater, and it rapidly decreased reaching a minimal level at 6 h of culture corresponding to 0.05 µmol in 300 mL of seawater or 1,6 µM of ANT and almost completely disappeared at 48 h of culture ( Fig. 2A)  ANT increased antioxidant enzymes activities in U. lactuca. To analyze whether the oxidative stress condition induced the activation of antioxidant enzymes, the alga was cultivated without ANT (control) and with 5 µM of ANT for 0-72 h, and the activities of antioxidant enzymes superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (AP), dehydroascorbate reductase (DHAR), glutathione reductase (GR) and glutathione peroxidase (GP) were determined. The activity of SOD was 0.48 µmol of superoxide anions consumed per min −1 mg −1 of protein in control condition and it remained unchanged until 72 h of culture (Fig. 4A). SOD activity increased in treated alga from 0.48 to 1.5 µmol min −1 mg −1 of protein at 6 and12 h of culture and decreased to 0.7 µmol min −1 mg −1 of protein at 72 h (Fig. 4A). The activity of CAT was 15 µmol of hydrogen peroxide consumed per min −1 mg −1 of protein in control condition and it remained unchanged until 72 h (Fig. 4B). CAT activity increased in treated alga from 15 to 58 µmol min −1 mg −1 of protein 12 h of culture and decreased to 28 µmol min −1 mg −1 of protein at 24 h and increased again to 36 µmol min −1 mg −1 of protein at 72 h (Fig. 4B). The activity of AP was 69 µmol min −1 mg −1 of protein in control condition and it remained unchanged until 72 h (Fig. 4C). AP activity increased in treated alga from 69 to 202 µmol of ascorbate consumed per min −1 mg −1 of   4C). The activity of DHAR was 16 µmol of dehydroascorbate consumed per min −1 mg −1 of protein and it did not change in control or treated conditions (Fig. 4D). The activity of GR was 115 µmol of glutathione consumed per min −1 mg −1 of protein in control condition and it remained unchanged until 72 h of culture (Fig. 4E). GR activity increased from 115 to 259 µmol min −1 mg −1 of protein at 12 h, decreased to 202 µmol min −1 mg −1 of protein at 24 h and increased again to 296 µmol min −1 mg −1 of protein at 72 h (Fig. 4E). The activity of GP was 22 µmol of glutathione consumed per min −1 mg −1 of protein in control condition and it did not change until 72 h of culture (Fig. 4F). GP activity increased from 22 to 40 µmol min −1 mg −1 of protein at 3 h of culture, decreased to 29 µmol min −1 mg −1 of protein at 6 h of culture, increased to 45 µmol min −1 mg −1 of protein at 12 h and remained unchanged until 72 h of culture (Fig. 4F). Thus, the oxidative stress condition induced by ANT is buffered by the activation of several antioxidant enzymes in U. lactuca. The maximal activity of antioxidant enzymes occurred at 12 h of culture, except for AP that reached a maximal level at 24 h of culture.
ANT increased the level of transcripts of antioxidant enzymes in U. lactuca. In order to analyze whether the increase in activities of antioxidant enzymes is due to an increase in their expression, the alga was cultivated without ANT (control) and with 5 μM of ANT for 0-72 h, and the relative levels of transcripts encoding antioxidant enzymes SOD, CAT, AP, DHAR, GR and GP were determined. The relative level of transcripts of SOD increased to reach a maximal level of 0.86 times of increase at 3 h of culture, decreased to 0.22 times at 12 h and slightly increased to 0.35 times at 72 h of culture (Fig. 5A). The relative level of transcripts encoding CAT increased to reach a maximal level of 0.91 times of increase at 12 h of culture and decreased to reach 0.42 times of increase at 72 h of culture (Fig. 5B). The relative level of transcripts encoding AP increased to reach a maximal level of 1.42 times of increase at 12 h of culture and decreased to 0.75 times at 24 h and to 0.44 times at 72 h of culture (Fig. 5C). In contrasts, the relative level of transcripts encoding DHAR did not change from 0 to 72 h of culture (Fig. 5D). The relative level of transcripts encoding GR increased to reach a maximal level of 13.3 times of increase at 12 h of culture and decreased to reach 6.3 times of increase at 72 h (Fig. 5E). The relative level of transcripts encoding GP increase to reach a maximal level of 4.4 times of increase at 12 h of culture and decreased to reach 2.5 times at 72 h (Fig. 5F). Thus, the increase in the level of transcripts encoding antioxidant enzymes occurred mainly at 12 h of culture, except those encoding SOD that reached a maximal level at 3 h of culture indicating that the increase in antioxidant enzyme activities is due, at least in part, to the increase in their expression. www.nature.com/scientificreports/ ANT induced an increase in activity and expression of GST metabolizing enzyme. In order to analyze the mechanisms involved in ANT detoxification, the alga was cultivated without ANT (control) and with 5 µM of ANT for 0-72 h, and the activity of the metabolizing enzyme GST and the relative level of transcripts encoding GST were determined. The activity of GST in control condition was 9.8 µmol min −1 mg −1 of protein and it remained unchanged until 72 h of culture (Fig. 6A). GST activity in treated algae increased from 9.8 to 53 µmol min −1 mg −1 of protein at 24 h and decreased to 41 µmol min −1 mg −1 of protein at 72 h of culture (Fig. 6A). The level of transcripts encoding GST increased in 4.2 times at 3 h of culture, decreased to 2.1 times at 6 h, increased to 3.6 times at 12 h and slightly decreased to 3 times at 72 h of culture (Fig. 6B). Thus, ANT induce an increase in the activity of the detoxification enzyme GST and this increase in activity is due, at least in part, to the increase in its expression. www.nature.com/scientificreports/

ANT induced an increase in Phase I and II metabolizing enzymes activities.
In order to analyze the involvement of Phase I and II detoxification enzymes, the alga was cultivated with 5 μM of ANT (control) and 0.5 µM inhibitors of Phase I enzymes CYP450-dependent monooxygenases and dioxygenases, and phase II enzymes PPOs, GSTs and STs and with 5 µM ANT for 24 h, and the amount of ANT in the alga was determined (Fig. 7). The level of ANT in control algae was 0.4 µmol in 10 g of FT and in the alga treated with inhibitors of monooxygenases, dioxygenases, PPO, GST and sulfotransferases it was 0.9, 0.9, 1, 1.1 and 1.1 µmol in 10 g of FT, respectively, which represents increases of 36, 36, 43, 48 and 44% compare to the control (Fig. 7). The level of ANT in the alga increased with all inhibitors, mainly with those that inhibit PPOs, GSTs and STs (Fig. 7). Thus, Phase I enzymes, monooxygenases and dioxygenases, and Phase II enzymes, PPOs, GSTs and STs are involved in ANT detoxification in U. lactuca.

Discussion
In this work, we showed that ANT is rapidly incorporated and metabolized in the marine alga U. lactuca reaching 46% of metabolization at 6 h of culture and 98% at 48 h. These results are accord with those previously obtained in the green macroalgae U. intestinalis and C. glomerata that metabolized 42-49% of BaP in 6 h 22 . This contrasts with results obtained in brown macroalgae such as Fucus vesiculosus and Chorda filum that metabolized only 5%   24 . Thus, U. lactuca metabolized more efficiently the three-ring ANT than the five-ring BaP but, interestingly, the alga can metabolize both PAHs. ANT induced an oxidative stress condition in U. lactuca characterized by the accumulation of ROS, mainly superoxide anions. In addition, ANT increased the level of lipoperoxides with a maximal level of 1.25 µmol g −1 of DT at 24 h of culture which indicate that ROS induced membrane damage in the alga. These results are in accord with those obtained in U. lactuca cultivated with 5 µM of BaP that showed an increase in ROS, mainly superoxide anions, and an increase in lipoperoxides showing a maximal level of 0.68 µmol g −1 of DT at 6 and 12 h of culture 24 . Thus, it appeared that ANT induced a higher increase of oxidative stress than BaP reflected by membrane damage in U. lactuca. In this sense, it has been shown that ANT induced a higher inhibition of growth and a higher chlorosis than BaP in L. gibba 2,3 . Thus, it is possible that the higher toxicity of ANT compared to BaP is due to the induction of a higher membrane damage which may affect mainly chloroplasts inducing chlorosis. The aquatic liverwort R. communis cultivated with 0-10 μM of PHE for 72 h showed an increase in amino acid leakage from 2.5 to 10 µM and an increase in liperoxides and carbonylated proteins with 0.5 µM of PHE indicating that PHE also causes an oxidative stress condition and membrane damage 13 . Thus, PAHs cause oxidative stress and membrane damage in plants and algae and ANT is more toxic than BaP.  15 . It has been recently shown that U. lactuca cultivated with BaP for 72 h showed an oxidative stress condition and the activation of antioxidant enzymes SOD, CAT, AP, GR and GP, but not DHAR 24 . Thus, plants and marine alga exposed to PAHs displayed an oxidative stress condition that is partially buffered by the activation by antioxidant enzymes. As in the case of U. lactuca and BaP, the alga cultivated with ANT did not show and increase in DHAR activity indicating that the Halliwell-Asada-Foyer cycle is uncoupled 24,25 . DHAR activity uses GSH as substrate and produces oxidized GSH (GSSG) as product 24 . On the other hand, it was observed that GR and GP showed the higher increase in activities in U. lactuca cultivated with ANT and BaP and, it is important to note, that both enzymes consume GSH. Thus, it is possible that the lack of DHAR activity is due to the high consumption of GSH by enzymes such as GR and GP. Thus, an important molecule for PAHs tolerance in U. lactuca is the potential synthesis and the consumption of GSH.
Interestingly, the level of transcripts of antioxidant enzymes SOD, CAT, AP, GR and GP increased in U. lactuca exposed to ANT, but not those encoding DHAR. The increase in the level of transcript of the latter antioxidant enzymes showed maximal levels that are previous to maximal enzyme activities indicating that the increase in activities is due, at least in part, to their enhanced expression. In addition, the level of transcripts of  www.nature.com/scientificreports/ the metabolizing enzyme GST increased with a maximal level at 3 and 12 h and GST activity showed a maximal level at 24 h. Thus, the increase in GST metabolizing activity is also transcriptionally regulated. It is important to mention that GST also consumes GSH, as GR and GP, and this is another metabolic point of GSH consumption which may also explain the absence of DHAR activity (see above). It was previously shown that BaP also increase activity and expression of GST in U. lactuca 24  Interestingly, inhibitors of Phase I enzymes, CYP450-dependent monooxygenases and dioxygenases, and Phase II enzymes, PPOs, GSTs and STs are involved in ANT detoxification in U. lactuca mainly those that inhibit PPOs, GSTs and STs. In this sense, it has been shown in the green macroalgae E. intestinalis and C. glomerata metabolize BaP through activation of phenol oxidase activity and peroxidase 22 . In addition, GSTs are involved in PHE detoxification in willow trees (Salix viminalis) showing and increase in the level of GST transcripts and in those coding for enzymes involved in GSH synthesis 26 . GST activity was also increased in common beans (Phaseolus vulgaris) exposed to ANT and fluorene 27 . Regarding STs in plants, A. thaliana genome encode 21 ST genes and rice genome encode 29 functional genes and STs are involved in sulfation of hydroxyl groups present in flavonoids, phenolic acids, terpenes, gibberellic acids and glucosinolates, and in the detoxification of hydroxylated xenobiotics 28 . In genomes of red and brown micro-and macroalgae, several potential genes of STs have been detected but their substrate preference has not yet been identified 29 . Thus, plants green macroalgae exhibited Phase I and II enzymes activities that may be involved in PAHs detoxification.

Conclusion
The green macroalga U. lactuca rapidly incorporated and metabolized ANT reaching a complete degradation after 48 h of culture. ANT induced an oxidative stress condition characterized by the accumulation of ROS, mainly superoxide anions, and lipoperoxides. ANT induced the activation of antioxidant enzymes and the metabolizing enzyme GST and their activation was due, at least in part, by the increase in their expression. Using inhibitors, it was shown that monooxygenases, dioxygenases, GSTs, PPOs and STs participate in ANT metabolization in U. lactuca (see model in Fig. 8).  www.nature.com/scientificreports/ in González et al. 24 . Since ANT has low solubility in water, a 25 mM of ANT stock solution was prepared in DMSO with a purity ≥ 99% (Sigma-Aldrich, St Louis, CA, USA) and then an aliquot was dissolved in the culture medium adjusting the DMSO to a final concentration of 0.5% v/v in the medium, including the control cultures.

Materials and methods
These cultures were used to analyze the accumulation of ANT, for qPCR analyzes, the detection of enzyme activities, and only few fronds were used for cell viability analyzes. All these analyses were performed as independent triplicates. For assays performed with Phase I and Phase II inhibitors, 10 g of alga were cultivated in 300 mL of artificial seawater without or with the addition of 0.5 µM of mebendazole (MBZ), an inhibitor of CYP450-dependent monooxygenases; cyclohexanedione (CHD), an inhibitor of dioxygenases; 2-naphtoic acid (2-NA), an inhibitor of polyphenol oxidases; ellagic acid (ELA), an inhibitor of GST; and quercetin (QC), an inhibitor of sulfotransferases, and with 5 µM of ANT for 24 h, in triplicates.
Analysis of cell viability. Three algal laminae were cultivated without ANT and with increasing concentrations of ANT for 7 days and cell integrity and morphology were visualized using Axiovert 100 confocal microscope (Zeiss, Oberkochen, Germany) using an emission wavelength of 488 nm from an argon laser and a filter of 505-550 nm for detection of chlorophyll fluorescence. The images were analyzed using the software LSM510 (Zeiss, Oberkochen, Germany).
Extraction of anthracene from algae and seawater. The extraction of ANT from algal tissue was performed as described by Sadowska-Rociek et al. 30 , with modifications. One gram of algal dried tissue (DT), which corresponds to 7 g of algal fresh tissue (FT), were pulverized in a mortar with liquid nitrogen and homogenized in 15 mL of cyclohexane having HPLC grade purity (Merck, Darmstadt, Germany). Cellular debris were removed using Quechers AOAC method kit (Agilent Technologies, Santa Clara, CA, USA) and centrifugation at 10,000×g for 10 min 31 . Pigments were removed using Quechers SPE dispersive kit (Agilent Technologies, Santa Clara, CA, USA) (Gratz et al. 31 ). The extracts recovered (7 mL) were filtered using PVDF filters with 0.2 µm pore size (Finetech, Taichung, Taiwan) and stored in amber glass vials at 4 °C.
Extraction of ANT from artificial seawater was performed as described by Colombo et al. 32 , with modifications. The culture medium (300 mL) was extracted twice with 200 mL of cyclohexane using an extraction funnel, and the extracts were concentrated to 2 mL using a rotatory evaporator HeiVap (Heidolph Industries, Schwabach, Germany). The extracts were filtered using PVDF filters with 0.2 µm pore size (Finetech, Taichung, Taiwan) and stored in amber glass vials at 4 °C.
Extraction of ANT from cultures with inhibitors was performed according to Warshawski et al. 17 , with modifications. Fresh tissue (10 g) was mixed with 250 mL of ethyl acetate HPLC grade purity (Merck, Darmstadt, Germany) and subjected to sonication for 30 min at room temperature. The solvent was collected and concentrated to 5 mL using a rotatory evaporator HeiVap. For the elimination of chlorophyll and pigments, a reaction mixture of 1 mL was prepared with an aliquot of 100 µL of the extract, 150 mM tert-butyl hydroperoxide and 50 mM sodium hydroxide in acetonitrile. The preparation was agitated in vortex for 60 min and then centrifuged at 10,000×g for 10 min. The supernatant was recovered and filtered using PVDF filters with 0.2 µm pore size and stored at 4 °C.
Quantification of anthracene by HPLC. ANT in extracts was analyzed according to Gratz et al. 31 using an Infinity 1260 series HPLC (Agilent Technologies, Santa Clara, CA, USA) having a reverse phase column Zorbax Eclipse XDB-C18 (4.5 mm × 15 mm and particle size of 5 µm, Agilent Technologies, Santa Clara, CA, USA) and a fluorescence detector. The elution program consisted in a mobile phase of H 2 O/CH 3 CN with a flux of 0.8 mL min −1 at 25 °C and a step from 0 to 1.5 min of 40% H 2 O and 60% CH 3 CN, a step from 1.5 to 7 min of 10% H 2 O and 90% CH 3 CN, and a step from 7 to 13 min of 0% H 2 O and 100% CH 3 CN. ANT was detected using excitation wavelength of 260 nm and emission wavelength of 520 nm a at retention time of 6.03 min, representative chromatograms for the standard of ANT and for ANT in tissue and in the culture medium are presented in Supplementary Figure S1.
Quantification of hydrogen peroxide, superoxide anions and lipoperoxides. Hydrogen peroxide was determined as described in González et al. 33 . Algae (1.5 g of FT) were collected from cultures every 30 min and immediately incubated in 100 mM phosphate buffer pH 7.0 containing 10 µM of 2′,7′-dichlorodihydrofluorescein-diacetate (DCF-DA, Invitrogen, Carlsbad, CA, USA) for 30 min. Algae were rinsed with fresh artificial seawater and dried with tissue paper. The samples were pulverized in a mortar with liquid nitrogen, homogenized in 5 mL Tris-HCl pH 7.0 and centrifuged at 10,000×g for 15 min. DCF fluorescence was determined using an excitation wavelength of 480 nm and an emission wavelength of 590 nm using a spectrofluorometer PerkinElmer model LS-5 (PerkinElmer, Waltham, MA, USA) The concentration of hydrogen peroxide was calculated using a calibration curve prepared with 0-1 µM of DCF.
Superoxide ions were determined as described in González et al. 33 . Algae (1.5 g of FT) were collected from cultures every 2 h and immediately incubated in 100 mM phosphate buffer pH 7.0 containing 100 µM hydroethidine (Invitrogen, Carlsbad, CA, USA) for 30 min. Algae were pulverized in a mortar with liquid nitrogen, homogenized in 5 mL Tris-HCl pH 7,0 and centrifuged at 10,000×g for 15 min. 2-Hydroxy ethidium (2-HE) fluorescence was detected using an excitation wavelength of 488 nm and an emission wavelength of 525 nm using a spectrofluorometer PerkinElmer model LS-5. The concentration of superoxide ions was calculated using the extinction coefficient of 2-HE (ε = 9.4 mM −1 cm −1 ).
Lipoperoxides were determined as described in Ratkevicius et al. 34 . Algae (1 g of FT) were pulverized in a mortar with liquid nitrogen and homogenized with 5 mL of 0.1% v/v of trichloroacetic acid (TCA). The mixture was centrifuged at 10,000×g for 20 min, a sample of 200 µL was collected. The sample was mixed with 800 µL of Detection of antioxidant enzyme activities. Antioxidant enzyme activities were detected as described in González et al. 24 . Superoxide dismutase (SOD) activity was determined in 1 mL of reaction mixture containing 30 mM Tris-HCl pH 7.0, 0.1 mM EDTA, 20 mM riboflavin, 0.6 mM NBT and without protein extract or with 50 µg of protein extract. The reaction mixture was incubated under white light for 15 min and absorbance of formazan, the reduced form of NBT, was determined at 560 nm. The specific activity of SOD was calculated considering that 1 U of enzyme is the amount of the enzyme that reduces 50% of total NBT which is equal to the amount of dismutated superoxide anions. Catalase (CAT) activity and other antioxidant enzyme activities were determined according to Ratkevicius et al. 34 in 1 mL of reaction mix containing 100 mM sodium phosphate buffer pH 7.0, 1 mM hydrogen peroxide and without protein extract or with 30 µg of protein extract. The decrease in absorbance due to consumption of hydrogen peroxide was detected at 240 nm for 2 min. The specific activity of CAT was calculated using a calibration curve prepared with hydrogen peroxide from 1 nM to 1 mM.
Ascorbate peroxidase (AP) activity was determined in 1 mL of reaction mixture containing 100 mM sodium phosphate buffer pH 7.0, 0.4 mM ascorbate, 1 mM hydrogen peroxide and without protein extract or with 20 µg of protein extract. The decrease in absorbance due to the consumption of ascorbate was detected at 290 nm for 1 min. The specific activity was calculated using the molar extinction coefficient of ascorbate (ε = 2.8 mM −1 cm −1 ).
Dehydroascorbate reductase (DHAR) activity was determined in 1 mL of reaction mixture containing 100 mM sodium phosphate buffer pH 7.0, 1 mM glutathione, 0.5 mM of dehydroascorbate, and 50 µg of protein extract. The increase in absorbance due to production of ascorbate was detected at 290 was for 2 min. The specific activity was calculated using the molar extinction coefficient of ascorbate (ε = 2.8 mM −1 cm −1 ).
Glutathione reductase (GR) activity was determined in 1 mL of reaction mix containing 100 mM sodium phosphate buffer pH 7.0, 0.5 mM GSSG, 0.15 mM NADPH and 20 µg of protein extract. The decrease in absorbance due to consumption of NADPH was detected at 340 nm for 1 min. The specific activity was calculated using the NADPH molar extinction coefficient of NADPH (ε = 6.2 mM −1 cm −1 ).
Glutathione peroxidase (GP) was determined in 1 mL of reaction mix containing 100 mM sodium phosphate buffer pH 7.0, 0.5 mM glutathione, 0.15 mM NADPH, 1 U glutathione reductase, 1 mM hydrogen peroxide and without protein extract or with 20 µg of protein extract. The decrease in absorbance due to consumption of NADPH was detected at 340 nm for 2 min. The specific activity was calculated using the NADPH molar extinction coefficient of NADPH (ε = 6.2 mM −1 cm −1 ).
Detection of detoxification enzyme. Detection of detoxification enzymes was performed as described in González et al. 24 . Glutathione-S-transferase (GST) activity was determined in 1 mL of reaction mix containing 100 mM sodium phosphate buffer pH 7.0, 0.5 mM GSH, 1 mM CDNB and 50 µg of protein extract. The increase in absorbance due the formation of the adduct GSH-CDNB was detected at 340 nm 2 min. The specific activity was calculated using molar extinction coefficient of the adduct (ε = 9.6 mM −1 cm −1 ).