Elimination of Bisphenol A and Triclosan Using the Enzymatic System of Autochthonous Colombian Forest Fungi

Bisphenol A (BPA) and triclosan (TCS) are known or suspected potential endocrine disrupting chemicals (EDCs) which may pose a risk to human health and have an environmental impact. Enzyme preparations containing mainly laccases, obtained from Ganoderma stipitatum and Lentinus swartzii, two autochthonous Colombian forest white rot fungi (WRF), previously identified as high enzyme producers, were used to remove BPA and TCS from aqueous solutions. A Box-Behnken factorial design showed that pH, temperature, and duration of treatment were significant model terms for the elimination of BPA and TCS. Our results demonstrated that these EDCs were extensively removed from 5 mg L−1 solutions after a contact time of 6 hours. Ninety-four percent of TCS and 97.8% of BPA were removed with the enzyme solution from G. stipitatum; 83.2% of TCS and 88.2% of BPA were removed with the L. swartzii enzyme solution. After a 6-hour treatment with enzymes from G. stipitatum and L. swartzii, up to 90% of the estrogenic activity of BPA was lost, as shown by the yeast estrogen screen assay. 2,2-Azino-bis-(3-ethylthiazoline-6-sulfonate) (ABTS) was used as a mediator (laccase/mediator system) and significantly improved the laccase catalyzed elimination of BPA and TCS. The elimination of BPA in the absence of a mediator resulted in production of oligomers of molecular weights of 454, 680, and 906 amu as determined by mass spectra analysis. The elimination of TCS in the same conditions produced dimers, trimers, and tetramers of molecular weights of 574, 859, and 1146 amu. Ecotoxicological studies using Daphnia pulex to determine lethal concentration (LC50) showed an important reduction of the toxicity of BPA and TCS solutions after enzymatic treatments. Use of laccases emerges thus as a key alternative in the development of innovative wastewater treatment technologies. Moreover, the exploitation of local biodiversity appears as a potentially promising approach for identifying new efficient strains for biotechnological applications.


Introduction
Bisphenol A (BPA) and triclosan (TCS) are present in a great number of products of daily use. Residues of these products are found in many environmental matrices such as rivers and wastewater treatment plant effluents [1,2]. BPA is a monomer or plasticizer in various polymeric materials [3], and residues of this compound in water appear to be due to incomplete polymerization [4,5]. Triclosan (TCS) is used as a broad spectrum antibacterial agent and a preservative used in products of domestic use. It is found in cosmetics and personal care products such as shampoos, deodorants, and toothpastes and it also occurs in textile polymers and 2 ISRN Biotechnology �bers to give them antibacterial properties [6]. e TCS structure strongly resembles to that of estrogens and dioxins [7]. Both of these compounds are frequently encountered in aquatic environments [8] and can incite serious risks for the environment and public health [9].
BPA has been recognized as an endocrine disrupting chemical (EDC). is chemical has the ability to imitate the female estrogen hormones which disrupt the body's chemical messenger system [10] due to its direct interaction with steroid receptors [11]. It has been argued that endocrine disruptors may be responsible for decline in sperm counts, abnormalities in the female reproductive tract, slow development in infants, increases in the incidence of testicular and breast cancer, and other medical disorders [12,13]. e estrogenic effect of BPA is primarily linked to the presence of a core phenolic structure [14]. On the other hand, the chemical structure of TCS is related to many well-known xenobiotic compounds, such as halogenated diphenyl ethers and BPA. Little is known about the potential endocrine disrupting activities of TCS. Studies have shown that there are changes in �n length of the medaka �sh Oryzias latipes exposed to TCS [15] suggesting that its metabolites can act like estrogen receptor antagonists [7]. TCS alone can alter the thyroid hormone receptor transcript levels in the brain of premetamorphic tadpoles and induce a transient weight loss [16].
e objectives of this work were to study the removal of BPA and TCS from aqueous solutions by using laccases excreted by the ligninolytic fungi Ganoderma stipitatum and Lentinus swartzii collected from the Colombian forest [29]. ese WRF were selected as high laccase producers, aer a preliminary screening of 30 WRF collected from a tropical forest in Colombia. Tropical forest constitute an important reservoir of microbial diversity including white rot fungi which have been identi�ed as possible candidates for wastewaters treatment [21,30,31].
Speci�c aspects of this study included the utilization of a Box-Behnken-type factorial design to determine the effect of the experimental conditions (pH, temperature, and contact time) on their removal, the elimination of their estrogenic activity, and the potential improvement of the transformation of these chemicals in the presence of redox mediators. Finally, the products formed by the oxidative action of laccase on BPA and TCS were determined by mass spectra (MS) analysis and the acute toxicity of the liquid waste aer treatment was tested by using the freshwater crustacean species Daphnia pulex.

Chemicals.
Commercial grade reagents of the highest purity available were used in this work. All chemicals were from Sigma-Aldrich (St. Louis, MO, USA) and all solvents were of HPLC grade.

Organisms and Cultivation
Conditions. e WRF strains G. stipitatum and L. swartzii were isolated from the Colombian tropical forest as previously reported [29]. e inocula were grown in Petri plates on malt extract agar 2% (w/v) at 25 ∘ C for about 10 days. ereaer, the strains were maintained at 4 ∘ C until used. For enzyme preparation, liquid fermentation was performed on a orbital shaker at 150 rpm and 30 ∘ C in 250 mL Erlenmeyer containing 75 mL of standard medium: 10 g L −1 glucose, 0.2 g L −1 ammonium tartrate, 0.5 g L −1 Tween 80, 2.5 mM veratryl alcohol, 3 g L −1 tartaric acid, 0.2 g L −1 yeast extract, 0.5 mM CuSO 4 , and 0.8 g L −1 KH 2 PO 4 . e initial pH was adjusted to 5.0 with NaOH 2 M prior to autoclaving. e fermentation medium was inoculated with four mycelia agar plugs of 5 mm diameter.
Aer 20 days of cultivation, the culture medium was �ltered through a 0.22 m membrane �lter (Millipore Corporation, Billerica, MA, USA). e �ltrate was dialyzed against distilled water using a 14 kDa membrane (Spectrum Laboratories, Rancho Dominguez, CA, USA) and then used as the source of lignin modifying enzymes. Respective laccase production of G. stipitatum and L. swartzii was 11,000 and 9,000 U/L �ltrate under these cultivation conditions [29].

Enzymatic
Treatment. e reaction mixture consisted of 5 mg L −1 of each of BPA (22 M) or TCS (18 M), 250 U L −1 of laccase activity, citric acid/disodium hydrogen phosphate buffer, and 1% v/v methanol. Prior to treatment, air was bubbled in the buffer solution overnight to saturate it with oxygen. e enzymatic treatments were performed in a 50 mL Erlenmeyer containing a 5 mL reaction media. e Erlenmeyer �asks were sealed with aluminum foil or para�n �lm and placed in an orbital shaker at 150 rpm.
To improve elimination of the two pollutants, some mediators were tested: ABTS, veratryl alcohol, 1hydroxybenzotriazole (1-HBT), and guaiacol. Pollutant elimination incubations using mediators were carried out in 50 mL �asks with a reaction volume of 5 mL at pH 4 and 30 ∘ C for 1 hour, using a �nal mediator concentration of 10 M and a laccase concentration of 10 U L −1 . e choice of a relatively low laccase concentration was made deliberately, ISRN Biotechnology 3 because a too high laccase concentration could provoke a rapid and complete elimination of the pollutants and mask the potential effect of the mediators on the elimination of the selected micropollutants. Control �asks were prepared to quantify elimination in the absence of enzyme, mediator, or both. e enzymatic treatments were stopped by adding 3 drops of concentrated HCl (37% w/v).

Extraction and Quantitative
Analysis. BPA and TCS were extracted using ethyl acetate in a 1/1 volumetric ratio (ethyl acetate/treated solution). e solutions were acidi�ed to pH 2 with HCl (37% w/v), shaken for 20 min, and then frozen overnight. e organic phase was separated and evaporated under a gentle stream of nitrogen. Each sample was dissolved in 100 L of methanol prior to quantitative chemical analysis.
e quantitative analysis of BPA and TCS was performed on an HPLC system consisting of a 600 controller, a 717 plus autosampler, and a 996 photodiode array detector (Waters, Milford, MA, USA). A Platinum EPS C 18 100A 5U 250 × 4.6 mm (Alltech, Deer�eld, IL, USA) was used for the chromatographic separation. A 1 mL min −1 isocratic elution by means of 50% (A) acetonitrile with 10% of bidistilled water and 50% (B) phosphate buffer solution (10 mM KH 2 PO 4 , pH 3.2) with 10% acetonitrile was used to quantify the micropollutants at a wavelength of 277 nm.

Determination of the Estrogenic Activity by the YES-Assay.
A recombinant yeast strain of Saccharomyces cerevisiae was used in a test (Yeast Estrogenic Screen, YES) designed to identify compounds interacting with the human estrogen receptor (hER) [33]. When the hER is bound to an estrogenlike compound, the receptor is coexpressed with the reporter gene lac-Z, which codes for the enzyme -galactosidase. is enzyme is secreted into the medium and catalyzes the transformation of the chromogenic substance chlorophenol red--D-galactopyranoside (CPRG), which is subsequently measured colorimetrically in the medium. e absorbance resulting from the change of color from yellow to red is a direct measurement of the estrogenic activity of the compound tested [34]. e test was adapted to microtiter plates whose absorbance was read at 540 nm (for the color) and at 620 nm (for turbidity). e TCS was not subjected to this test because its microbicidal activity made the yeast test impracticable [19].
e positive wells were identi�ed by a strong red coloring accompanied by the growth of the yeast. e estrogenic activity of the solution was correlated with the absorbance ( ) of the solution by (1) as corrected = 540 sample − 540 blank − 620 sample − 620 blank . (1)

2.�. �denti�cation of High �olecular �eight �eaction Products.
To identify high molecular weight reaction products, a 5 mL mixture consisting of 5 mg L −1 of BPA or TCS, citric acid/disodium hydrogen phosphate buffer and 1% v/v methanol was used. e temperature was set at 30 ∘ C and

Statistical
Analysis. e impacts of the operational conditions pH, temperature, and contact time on the EDCeliminations were determined by using a Box-Benhken-type factorial design [36]. e experimental conditions tested are presented in Table 1. e results presented in Table 2 were used for the determination of the regression coefficients of the second-order multiple regression model. e statistical analysis was performed using the Design Expert 6.0.11 soware (Stat-Ease Inc., Minneapolis, MN, USA). values of pH, temperature, and time were followed using the factorial experimental design of Box-Behnken [36] shown in Table 1. is design was used in order to evaluate the impact of the three variables and of their potential interactions with each other upon the degradation of BPA and TCS, thus reducing the number of experiments required to test each parameter individually. Previous studies revealed that these parameters are operational variables with statistically signi�cant in�uence on the performances of the biotreatment [19,25,26].

Results and Discussion
A multiregression model (see (2)) was developed to correlate the elimination of BPA and TCS by enzyme solutions produced by G. stipitatum or L. swartzii with the operational conditions by using a least squares method [37] as follows: where Y represents the predicted response (elimination of the EDC), x the kth variables (pH, temperature, and contact time), are constant regression coefficients of the model, and is the random error component of the system. e second order models for the respective elimination (Y) of BPA and TCS by the enzyme solutions from L. swartzii and G. stipitatum are presented through (3)-(5) considering only the signi�cant terms determined by the ANOVA analysis (Table 3) e 3D surface responses obtained from these quadratic expressions are presented in Figure 1 in order to show the effects of the independent variables contact time, temperature, and pH and the interactive effects of each variable on the removal of the compounds of interest.
ese quadratic models, based on a Box-Behnken experimental design, can be used to predict the optimal values of temperature and pH for the elimination of BPA and TCS among the range of values tested. Best temperatures were estimated to be between 40 ∘ C and 60 ∘ C for the two micropollutants and best pH was 5.0 except for a pH of 4.0 that was the best adapted for the elimination of BPA with the laccase from L. swartzii. e maximal extents of removal predicted by the models were obtained with laccase from L. swartzii for the 5 mg L −1 solution of BPA (97.5%, 60∘C, pH = 5) and with laccase from G. stipitatum for TCS (93.9%, 40 ∘ C, pH = 5) when these enzymes were used at a level of 250 U L −1 for a contact time of 6 hours. ese results are in agreement with a combination of stability produced by a higher pH and catalytic activity resulting from a higher temperature. An optimum pH of 3.0 for the initial rate of BPA removal by laccase from Coriolus versicolor was previously found [38], but the time course of the transformation of BPA under this pH condition was not reported. e optimum temperature and pH for the elimination of BPA using commercial laccase from T. versicolor were reported as 45 ∘ C and 5.0, respectively [25]. Finally, higher TCS removal by laccase of C. polyzona was observed at 50 ∘ C and pH of 5.0, whereas BPA removal culminated in a broad range between 40 and 50 ∘ C and pH of 5.0 [19].

Effects of Mediators
. e use of low-molecular weight oxidizable substances in the biocatalytic cycle of laccase expands the activity of this enzyme. is mediated oxidation involves two oxidative steps. In the �rst one, the laccase oxidizes a primary substrate, the mediator, and this substance acts as an electron transferring compound. In the second step, the mediator transfers the electron from the substance of interest [39]. ese mediators are known to increase the substrate range of laccase toward different substances [40]. Different mediators, namely, ABTS, veratryl alcohol, guaiacol, and 1-HBT, were tested for one hour at 30 ∘ C and pH 4 to degrade BPA and TCS, each present at initial concentration of 5 mg L −1 using laccases of G. stipitatum or L. swartzii at a �nal concentration of 10 U L −1 . As a control, the buffer with the enzyme alone was used. e presence of the mediator alone did not allow elimination of BPA or TCS (results not shown). Figure 2 presents the relative elimination of the BPA and TCS by the different laccase/mediator systems tested. A 100% relative efficiency refers to the removal of the EDC by the 7 best combination of laccase and mediator. For both of these micropollutants, better extents of removal were obtained with ABTS for the two laccase solutions tested. is mediator helped to eliminate a higher percentage of micropollutants than that obtained with laccase alone (one-hour treatment at pH 4 and 30 ∘ C). e G. stipitatum and L. swartzii enzyme systems lacking a mediator had respective efficacy of 22 and 38% for removal of BPA, compared to the situation where ABTS was present (Figure 2(a)). Values of 38 and 18% were observed for the removal of TCS (Figure 2(b)). For the other mediators, guaiacol, veratryl alcohol, and 1-HBT, the improvement of the EDC transformation appeared to be also dependent from the nature of the enzymatic solution used (Figure 2). For example, guaiacol increased the removal of BPA when using laccase from L. swartzii whereas this conversion decreased when using laccase from G. stipitatum (Figure 2(a)). In contrast, guaiacol increased TCS removal when using laccase from G. stipitatum ( Figure  2(b)). Veratryl alcohol (VA) strongly increased removal of BPA by G. stipitatum, but apparently had a slight decreasing effect on removal of this EDC by laccase of L. swartzii ( Figure  2(a)). VA had a moderate increasing effect on TCS removal by both enzyme solutions (Figure 2(b)). Respective relative removal efficiency around 40 and 60% were observed for BPA and TCS by both enzyme solutions in the presence of 1-HBT (Figures 2(a) and 2(b)). e different behavior of both enzymes should be related to their respective responsiveness towards mediators, which was not studied here.
In conclusion, based on these results, ABTS was identi�ed as the best mediator system for BPA and TCS elimination by laccases of all tested strains. is is consistent with the report by Cabana et al. [19] who found that the utilization of the laccase/ABTS system signi�cantly improved the elimination of the EDCs nonylphenol, BPA, and TCS under the same operational conditions when using enzyme preparation excreted by Coriolopsis polyzona [19]. On the other hand, Tsutsumi et al. [28] determined that 1-HBT (0.2 mM) as a mediator could enhance BPA removal aer one hour of treatment using 100 U L −1 of laccase from Trametes versicolor. In our own case, the same performance was obtained using a laccase/ABTS system but with only 10 U L −1 of laccase from any of the two new fungal strains described. Murugesan et al. [22] have observed that TCS was efficiently eliminated (up to 80%) by Ganoderma lucidum laccase in the presence of different mediators including, 1-HBT, ABTS, or syringaldehyde (SYD). Products of lower molecular weights than TCS including 2,4-dichlorophenol and dechlorinated forms were detected in the presence of 1-HBT or SYD. e radicals obtained from the oxidation of ABTS (2,2azino-bis(3-ethylbenzothiazoline-6-sulfonate), in particular the radical cation ABTS •+ , are quite stable [41] and acted as efficient mediators of laccase towards phenolics and some nonphenolic substrates, for example, Textile dyes [42] and polycyclic aromatic hydrocarbons [43]. ere are, however, very few studies regarding the reaction of ABTS derived radicals with phenolic compounds [44,45]. Yet, it seems that reaction schemes observed were not always compatible with a simple, phenol promoted, back reduction of ABTS derived radicals. Products obtained from triclosan in the presence of ABTS were not studied here and were apparently not reported in the literature [22,26].

�.�. �denti�cation o� �i�� �olecular �ei��t �eaction �rod� ucts.
In the absence of a redox mediator, BPA and TCS were transformed by laccase of L. swartzii and G. stipitatum into high molecular weight reaction products.   Huang and Weber [46] have shown that horseradish peroxidase transformed phenol-like chemicals via the formation of phenoxyl radicals. ese radicals react with phenolic substances to form oligomers of the initial substrate. Cabana et al. [19] working with Coriolopsis polyzona laccase were the �rst to demonstrate that enzymatic treatment of nonylphenol (NP), bisphenol A (BPA), and triclosan (TCS) produced high molecular weight metabolites through a radical polymerization mechanism of NP, BPA, and TCS. is result was con�rmed for TCS by Murugesan et al. [22] with G. lucidum laccase. Support for this polymerization mechanism with L. swartzii and G. stipitatum laccases comes from the ESI(−) MS spectra of BPA and TCS solutions treated with the enzyme solutions from both strains. e high molecular weight reaction products formed resulted from the oligomerization of oxidized products generated by laccase. ESI(−) MS spectra showed that this oligomerization could occur at the level of C-C or C-O bond formation. ese types of bond could be between phenol moieties of BPA or TCS. is way of transformation has been proposed for BPA. e dimer produced from BPA by laccase of the WRF Trametes villosa was identi�ed by NMR as 5,5 � -bis-[1-(4-hydroxy-phenyl)-1-methyl-ethyl]-biphenyl-2,2 � -diol [11]. Huang and Weber [46] also proposed possible reaction pathways in which the polymerization mechanism of BPA occurred through C-O bonds. Recently, Murugesan et al. [22] have shown that the utilization of the laccase/mediator system for TCS transformation results in the release of chlorine into the reaction media, while the utilization of free laccase did not result in the formation of dechlorinated products. ese results suggested different reaction mechanisms for the transformation of TCS depending of the system used. In addition, Cabana et al. [47] identi�ed TCS dechlorinated products generated by the utilization of laccase/chitosan conjugates for the treatment of TCS that is present in solution. ese results indicated that the conditions used for the laccasemediated oxidation of triclosan in�uenced the extent or the mechanisms of the reaction, which have an impact on the transformation products detected. It should be also noted that the differences between the products detected by these works can be explained by the different analytical methods used (single quadrupole (this work), triple quadrupole [19,22], or quadrupole time-of-�ight (�TOF) [47]).
According to our results, the panel of reaction products associated with the elimination of BPA and TCS by the enzyme solution produced by the WRF tested appeared to be strain-and micropollutant-dependent. For example, laccase of L. swartzii was the most efficient for elimination of BPA,  whereas laccase of G. stipitatum was the most effective for the removal of TCS. is is most probably due to different kinetic behavior (e.g., variation of the Michaelis-Menten kinetic parameters) as a function of strain and substrate that may strongly affect the extent of reaction and the metabolites produced [48].

Determination of Estrogenic Activity of the Treated BPA Solution by the YES Assay.
Although the elimination of BPA and TCS was studied using the Box-Behnken factorial design, it was of great concern to address the elimination of the estrogenic activity associated with these compounds.
To determine the estrogenic activity of the treated solutions, we used the well-established YES assay [33]. TCS was not subjected to this test because of its microbiocidal activity which makes this yeast test nonapplicable [19]. Figure 4 shows a clear estrogenic activity reduction observed aer the enzymatic treatment of a 5 mg L −1 BPA solution. Upon contacting this solution for 3.5 hours, with 250 U L −1 laccase from L. swartzii, there was a decrease of up to 70% in estrogenic activity and a 50% decrease with an equivalent treatment with laccase from G. stipitatum. Finally, aer a 6 h treatment, estrogenicity loss of up to 90% of the BPA solution was observed for the laccases from both strains. It had been previously observed that the estrogenic activity of a 0.88 mM solution of BPA was almost entirely removed by extending the laccase treatment to 12 h [28]. In our own experiments, the estrogenic activity of a 5 mg L −1 BPA solution was almost entirely removed aer a laccase treatment of only 6 hours. e loss of estrogenic activity of BPA solution is apparently due to structural modi�cation of the parent compound [19] as con�rmed by the MS analysis (see Figure 3). e laccase mediated oxidation results in the formation of phenoxy radicals involved in the formation of BPA oligomers [12,19,46]. ese products did not present the structural features necessary to bind to the hER [14]. Furthermore, the elimination of the estrogenic activity could be linked to the physical removal by precipitation of the polymers produced.
3.5. Elimination of the Acute Toxicity of the BPA and TCS by Enzymatic Treatments. e toxicities of the aqueous samples were measured using the microcrustacean D. pulex acute toxicity assay. Dilution factors used (0, 12.5, 25, 50, and 100%) to determine the acute toxicity represent the proportion of the BPA or TCS solution added to the culture media. e smaller the LC 50 value, the more toxic is the tested solution. Table 4 presents the LC 50 determined for the solutions containing the initial 5 mg L −1 concentration of the micropollutants and for the treated solutions using 250 U L −1 of laccase activity from the different WRF strains at a pH of 5 and a temperature of 60 ∘ C. e initial BPA solution presented a LC 50 value of 10% while the TCS solution had a LC 50 < 12.5%. For both contaminants, the best toxicity reductions were obtained with the enzyme solution from the WRF strain G. stipitatum. In the analysis performed on samples from the BPA solution treated with that enzymatic solution, there was no observation of D. pulex death aer the 48 hours incubation period at any of the dilutions tested ( Table 4) which indicates that the LC 50 is above the maximum concentration tested. On the other hand, the LC 50 value of the TCS solution treated with this enzyme solution was 93%. It can be concluded that there was an important elimination of the toxicity associated with these contaminants aer enzymatic conversion of BPA and TCS.

Conclusions
In this paper, it was shown that among two newly described WRF strains producing laccases [29], the most efficient elimination of BPA was reached using the laccase of L. swartzii whereas for the removal of TCS the laccase of G. stipitatum was the most effective. Nevertheless, both strains were able to remove the estrogenic activities associated with BPA within a 6 h treatment. e elimination of the estrogenic activity of the BPA was associated with the formation of higher molecular weight reaction products. e MS spectra con�rm the formation of oligomers (dimers through pentamers). It can also be concluded that the chemicals formed by the oxidative action of laccase presented a lower acute toxicity than the parent compounds.
e biotreatment of wastewater using laccase-containing solutions secreted by WRF seems therefore to be an attractive solution for the elimination of micropollutants, including the known and suspected endocrine disruptors BPA and TCS. Laccases, which are regarded as environmentally friendly and relatively low-cost catalysts, may emerge as a key alternative in the development of new tertiary or polishing wastewater treatment technologies. e present contribution shows also that the exploitation of local biodiversity appears as potentially promising for identifying new efficient strains for biotechnological applications.