生物富集

sf Sh nt, N ces, 2011 Accepted 20 September 2011 Available online 27 October 2011 ethe fish. In this study, the tissue distribution, excretion, depuration and biotransformation of 4,40-dibromod- aquatic organism, i.e. freshwater fishes (Luo et al., 2007). As one of the PBDE derivatives, hydroxylated (OH-) PBDEs are likely to be more potent in causing toxic effects than the parent PBDEs and they have been shown to possess antiestrogenic, antian- drogenic, antiprogestagenic potency, or to disrupt thyroid hor- mone homeostasis (Hamers et al., 2006; Li et al., 2010). OH- PBDEs have been detected in natural environment such as surface in rats, chickens and birds, etc. Results from exposed mammals and one study on marine flatfish demonstrated that OH-PBDEs were formed as metabolites from parent PBDEs in vivo (Hakk et al., 2002, 2006; Qiu et al., 2007; Munschy et al., 2010). The biotransfor- mation of PBDEs (such as BDE 99, 183, and 209) in freshwater fish had been investigated in several recent studies and they reported that metabolism primarily occurs by a reductive debromination pathway, resulting in the formation of lower brominated BDEs (Stapleton et al., 2004; Tomy et al., 2004; Nyholm et al., 2009; Noyes et al., 2010). No efforts having been made to evaluate the fate of lower brominated BDEs in freshwater fish and also no ⇑ Corresponding authors. Tel./fax: +86 25 89680360 (L. Mao), tel.: +86 25 89680359; fax: +86 25 89680360 (S. Gao). Chemosphere 86 (2012) 446–453 Contents lists available at s evi E-mail addresses: lmao@nju.edu.cn (L. Mao), ecsxg@nju.edu.cn (S. Gao). Polybrominated diphenyl ethers (PBDEs), a class of common brominated flame retardants, have aroused increasing concerns on account of their widespread use, ubiquitous environmental dis- tribution, great bioaccumulation potential, and possible toxicity (Xia et al., 2008). It has been found that, among all PBDE congeners, the lower brominated ones tend to be the most bioaccumulative in aquatic animals (Luo et al., 2007; Xu et al., 2009). Their relative high concentrations in fish, nearly three or ten times the average levels in meats or dairy products (Schecter et al., 2006), cause great concern. As a lower brominated PBDE, 4,40-dibromodiphenyl ether (BDE 15) was found to be the predominant congener present in 2009). Nevertheless, it remains difficult to trace the source of the OH-PBDEs in environmental samples from the monitoring data. In fish, OH-PBDEs may originate from cytochrome P450 (CYP450) enzyme-mediated metabolism of PBDEs or from natural sources (Marsh et al., 2004; Raff and Hites, 2006; Zhou et al., 2011). To ex- plore the occurrence of OH-PBDEs, earlier experiments in con- trolled laboratory settings had been conducted to investigate the biotransformation of 2,20,4,40-tetrabromodiphenyl ether (BDE 47), 2,20,4,40,5-pentabromodiphenyl ether (BDE 99) and decabromodi- phenyl ether (BDE 209) in different organisms (Hakk et al., 2002; Pirard and De Pauw, 2007; Van den Steen et al., 2007), for instance, Keywords: BDE 15 Freshwater fish Bioaccumulation Metabolism OH-PBDEs 1. Introduction 0045-6535/$ - see front matter � 2011 Elsevier Ltd. A doi:10.1016/j.chemosphere.2011.09.038 iphenyl ether (BDE 15) were investigated in crucian carp (Carassius auratus) which were exposed to spiked water solution at different concentrations for 50 d, followed by a 14-d depuration period. Bioac- cumulation parameters were calculated and the results showed that BDE 15 was mainly concentrated in the gill and liver. In particular, five biotransformation products of BDE 15 in carp were identified using GC–MS/MS. Besides two debrominated metabolites, three of the metabolites were mono-OH-BDE 15, diOH-BDE 15 and bromophenol. Our results unequivocally suggested that BDE 15 oxidation did occur via the formation of hydroxylated (OH-) metabolites in crucian carp exposed in vivo. These findings will be useful for determination of the metabolic pathways of PBDEs in freshwater fish, especially about their oxidation metabolism. � 2011 Elsevier Ltd. All rights reserved. water and precipitation (Ueno et al., 2008), as well as in wild fishes, i.e., Baltic Sea Salmon (Marsh et al., 2004) and tuna (Wan et al., Received 17 April 2011 Received in revised form 16 September become ubiquitous environmental pollutants. Significant biotransformation of some PBDEs via reductive debromination has been observed. However, little is known about the fate of lower brominated BDEs in Bioaccumulation, depuration and biotran in crucian carp (Carassius auratus) Jie Cheng a, Liang Mao a,⇑, Zhigang Zhao a, Mengnan Shixiang Gao a,⇑ a State Key Laboratory of Pollution Control and Resource Reuse, School of the Environme bCollege of Agricultural and Environmental Sciences, Department of Crop and Soil Scien a r t i c l e i n f o Article history: a b s t r a c t Polybrominated diphenyl Chemo journal homepage: www.els ll rights reserved. ormation of 4,40-dibromodiphenyl ether en a, Shenghu Zhang a, Qingguo Huang b, anjing University, Nanjing 210046, PR China University of Georgia, Griffin, GA 30223, USA rs (PBDEs) are extensively used as a class of flame retardants and have SciVerse ScienceDirect phere er .com/locate /chemosphere LENOVO 高亮 LENOVO 高亮 LENOVO 高亮 LENOVO 高亮 LENOVO 高亮 fish were fed with pelleted food at a daily rate of 1% of their body pher weight, and checked daily for abnormal behavior, disease, and mortality. Triplicate water and fish were randomly sampled from each group on days 0, 1, 3, 6, 28, 50, 57, and 64. At each sampling point throughout the experiment, fish mass and length were recorded, and then fish were killed by a blow to the head. Each sampled fish was separated into liver, muscle and gill subsamples. The gut con- tent remaining in the intestinal track were removed and collected. All tissue samples were rinsed in deionized water and immediately stored at �20 �C before extraction and analysis. 2.3. Sample preparation and extraction definitive answer so far on whether the OH-PBDEs in freshwater fish could result from metabolism. More comprehensive under- standings of distribution and biotransformation behavior of lower brominated BDEs in freshwater fish are therefore desirable. The aim of this study was designed to simulate the exposure of BDE 15 to assess the in vivo biotransformation behavior of lower brominated BDEs in freshwater fish. We investigated the distribu- tion, bioaccumulation (uptake/depuration rate constants (k1/k2), half-life (t1/2), bioconcentration factor (BCF)), excretion and metab- olite formation of BDE 15 in crucian carp which is one of the daily foods and a mainstay of the commercial fishery in Yangtze River basin. These results are the first indications for the bioaccumula- tion and biotransformation of BDE 15 in freshwater fish. 2. Materials and methods 2.1. Materials The standards, solvents, and chemicals employed in this study are provided in Supplemental material (SM). Anhydrous sodium sulfate was baked at 450 �C for 4 h and then stored in sealed con- tainer. Prior to use, hydromatrix and neutral silica gel (60–100 mesh) were cleaned by sonication using dichloromethane (DCM) and methanol, respectively. Crucian carp (19.7 ± 3.6 g, 9.0 ± 1.0 cm) were obtained from Fisheries Research Institute of Nanjing. 2.2. Fish culture and exposure design Crucian carp were kept in de-chlorine water with dissolved oxygen, pH, temperature and hardness respectively at >5 mg L�1, 7.5 ± 0.3, 20 ± 1 �C and 118.5 ± 4.2 mg L�1 as CaCO3. Air stones were placed in each tank to maintain oxygen saturation in the water. A daily 12/12-h light/dark photoperiod cycle was used throughout the experiment. Fish were fed with pelleted food for at least 14 d prior to the experiment. Fish were randomly assigned to different BDE 15 exposure groups, each fish in a tank containing 2 L test medium, at nominal concentrations of 0 (control), 10 and 100 lg L�1. BDE 15 stocks were prepared in dimethylsulfoxide (DMSO) and stored at room temperature in the dark. Final DMSO concentration in the exposure medium was 0.01% in all groups, including the control. The water was renewed daily during the exposure period in a semi-static manner by replacing the medium with the BDE 15 test solution. Water temperature and oxygen content were checked immediately before each renewal. After a 50 d uptake period, the fish were transferred to aquaria and allowed to depurate for 14 d with daily renewal of clean water. Throughout the experimental procedure, J. Cheng et al. / Chemos Sample preparation and extraction were performed following the procedures described by previous studies (Stapleton et al., 2006; Wurl et al., 2006; Xu et al., 2009) with slight modifications. Details are presented in SM. Briefly, water sample was extracted with n-hexane, and then cleaned up using silica gel. Fish tissue sample was extracted using DCM/acetone (1:1 v/v) by a Dionex Accelerated Solvent Extraction system (ASE 350, Sunnyvale, CA, USA). Lipid was removed from the extract with a gel permeation chromatography column (GPC; J2 Scientific, AccuPrep MPS), and further cleanup was carried out using an acidified silica gel column. Analysis of metabolites in liver and gut content samples was conducted following the protocol (Hovander et al., 2000; Ueno et al., 2008; Wan et al., 2009). In short, the ASE 350 was used for sample (about 1 g) extraction. Extraction solvents were n-hex- ane/DCM (1:1 v/v) and n-hexane/methyl tert-butyl ether (MTBE) (1:1 v/v) in sequence. Two cycles were performed for each solvent per sample. To prevent photodegradation, the tubes remained wrapped with aluminum foil during the entire process. After GPC, the eluate containing target compounds was reduced and re-dissolved in n-hexane (8 mL). Resulting extract was partitioned with KOH solution (0.5 M in 50% ethanol). The organic phase (neu- tral fraction) was transferred. The aqueous layer was extracted using n-hexane twice more. Then the aqueous phase was acidified by 2 M HCl, followed by extraction at least three times with n-hex- ane/MTBE (9:1; v/v) (phenolic fraction). The concentrated extract (phenolic fraction) was passed through a Florisil column which was wet-packed with deactivated Florisil (12 cm) and anhydrous sodium sulfate. The eluate was first concentrated to dryness and then dissolved in DCM for OH-PBDEs and bromophenols analysis. Once analyzed by GC-EI/MS and GC- EI/MS/MS, the extract was quantitatively transferred with n-hex- ane. A 100 lL aliquot of the derivating agent N, O-bis (trimethyl- silyl)-trifluoroacetamide (BSTFA) was added. Mixture was vortexed for 30 s and then heated at 60 �C for 60 min. The deriva- tized sample was nitrogen evaporated and reconstituted in 100 lL of n-hexane for analysis again. The neutral fraction was concentrated and sequentially purified by acidified silica gel column (SM). The resulting extract was used for analysis of PBDEs. 2.4. Instrumental analysis Quantification of BDE 15 was performed on an Agilent 7890 GC fitted with a micro-cell electron capture detector (lECD). Detailed chromatographic conditions are given in SM. Some selected sam- ples were analyzed by full-scan GC–MS for BDE 15 confirmation. Identification of metabolites was done using a Thermo Finnigan Trace gas chromatography interfaced with a Polaris Q ion trap tan- dem mass spectrometer (GC-ITMS/MS, Thermo, Finnigen, USA). Instrumental details can be found in SM. 2.5. Quality assurance/quality control (QA/QC) and data analysis Procedural blanks covering the entire procedure were operated in parallel with each batch of extractions. Spiked blanks, fish feed, control samples (fish collected before the beginning of the experi- ment), and matrix spiked samples were analyzed. The procedural blanks, fish feed, and control samples were all found to be free from contamination. Further information about QA/QC is pre- sented in SM. Instrumental QC included regular injections of sol- vent and standard solutions. No sensible difference was observed on fish growth between the control and exposed groups. Neither mortality nor obvious alteration of fish health was noticed during the experiment in all groups. Levels of BDE 15 in the exposure water were relatively sta- e 86 (2012) 446–453 447 ble during the exposure period. Measured concentrations for the low-dose and high-dose groups were 0.126 ± 0.049 and 1.050 ± 0.509 lg L�1, respectively, which were in average only pher 448 J. Cheng et al. / Chemos 1.3% and 1.1% of the nominal concentrations. Such phenomenon was consistent with earlier studies (Sundt et al., 2006). And BDE 15 was not detected in fish from the control group on any collec- tion day throughout the exposure, indicating that the accumula- tion of BDE 15 observed in the exposed fish was a result of the uptake from the exposed solution. Expressed as means ± standard deviation, the data were ana- lyzed by using Origin 7.5. Details about the estimate method of bioaccumulation parameters are described elsewhere (Paterson and Metcalfe, 2008), and formulas are provided in SM. 3. Results and discussion 3.1. BDE 15 in fish tissues To examine the distribution, BDE 15 concentrations in different tissues were measured on days 0, 1, 3, 6, 28, 50, 57 and 64, and the results were shown in Fig. 1. It is evident that BDE 15 concentra- tions in the tissues of exposed fish increased rapidly at the begin- ning of the exposure for both dosage groups, and reached a maximum level after approximately 28 d. The concentrations in the gill tended to be the highest among all tested tissues and this result was consistent with earlier study which focused on the pref- erential accumulation of dichlorodiphenyltrichloroethane (DDT) in marine fish (Kwong et al., 2008). On day 28, concentrations in the gut content were approximately 3.23 and 1.97-fold higher than that in the muscle for the low-dose and high-dose groups, respec- tively. The decrease of BDE 15 concentrations after 28 d exposure may result from a lower uptake at the later stage of the exposure period and a more rapid metabolism (i.e., the discovery of metab- olites, Section 3.3) and excretion (i.e., high concentrations in the Fig. 1. Accumulation and depuration of BDE 15 (lg g�1 dry weight, dw) in the expo e 86 (2012) 446–453 gut content), which is consistent with previous studies (Baussant et al., 2001; Sun et al., 2006; Pirard and De Pauw, 2007). Fig. 1 suggests that the body burdens of BDE 15 depend on the exposure dosages, and the trends in different tissues were similar for both groups. Taking the day 28 as a reference exposure time, concentrations in the muscle and gill of the high-dose group were respectively about 8 and 9-fold greater than that in the low-dose group, while in the liver only 3-fold. This phenomenon might due to more rapid metabolism of BDE 15 in liver tissue. 3.2. Bioaccumulation parameters Measured values of k1, k2 and t1/2 for BDE 15 in carp are given in Table 1. Decrease of BDE 15 concentration in the gill (0.238 ± 0.023 d�1) was markedly rapid in contrast to that in the muscle (0.146 ± 0.008 d�1). Greater removal rate from the gill than other tissues was also reported in a previous study (Kwong et al., 2008). The average half-life of BDE 15 in carp was approximately 4.4 d, which was consistent with an early finding (Gustafsson et al., 1999) in which the half-lives of BDE 47, 99 and 153 in mus- sels were 7.7, 5.6 and 8.1 d, respectively. The observed rapid dep- uration rates and high concentrations in the gut content suggested that excretion is one of the major routes for BDE 15 elimination. BCF of BDE 15 was calculated by the ratio of the uptake and depuration rate constants (Table 1). These estimated BCF were 1.3 � 105, 1.7 � 105, and 7.5 � 104 for the liver, gill, and muscle in the low-dose group, respectively, and 8.2 � 104 for the liver in the high-dose group. Our result was comparable with the earlier studies, i.e., Gustafsson et al. (1999) reported that the BCF of BDE 47 and 99 in mussels were 1.3 � 106 and 1.4 � 106, respectively. Despite of the low levels of PBDEs in the environment, their high sed fish. Values represent the mean and standard deviation of three replicates. pher bioaccumulation potential for aquatic organisms calls for attention when assessing the risk of PBDE to human health. 3.3. Identification of biotransformed products One focus of this study was to screen and identify possible metabolites in BDE 15 exposed fish. By partitioning with 0.5 M KOH solution, the phenolic compounds in aquatic layer could be separated from the neutrals (organic phase) (Hovander et al., 2000). In the phenolic fraction, one bromophenol and two hydrox- ylated dibrominated diphenyl ether (OH-BDE 15, diOH-BDE 15) were found in both liver and gut content extracts. In the neutral fraction, the debrominated metabolites 4-bromodiphenyl ether (BDE 3) and diphenyl ether (DE) were observed only in the gut con- tent. These brominated products in the exposed fish were not pres- ent in fish feed or the control fish, suggesting that BDE 15 biotransformation did occur in crucian carp. The retention time and spectra characteristics of metabolites (before derivatization) are shown in Table SM-1. 3.3.1. Phenolic compounds GC-EI/MS generates positive ions and provides molecular ion (M+) with m/z value corresponding to molecular weight of analyte molecule (M). The identification of the three phenolic metabolites was based on i) their mass spectra; ii) comparison of their isotopic ratio with their theoretical ratio; iii) the comparison of ion frag- mentation patterns with authentic congener standards. In addition, the possible products were derivatized using BSTFA and then ana- lyzed by GC–MS, which further supported the formation of pheno- lic metabolites. Table 1 Estimated bioaccumulation parameters of 4,40-dibromodiphenyl ether (BDE 15) from waterborne exposure using crucian carp. Tissuea Uptake rate constant k1 (L d�1 g�1 dw) Depuration rate constant k2 (d�1)b Half-life t1/2 (d) Bioconcentration factor BCF (�104, dw) Muscle (LD) 11.0 ± 4.5 0.146 ± 0.014 (0.942) 4.8 ± 0.3 7.5 Gill (LD) 40.2 ± 24.1 0.238 ± 0.028 (0.914) 2.9 ± 0.3 16.9 Liver (LD) 19.2 ± 10.7 0.150 ± 0.022 (0.864) 4.6 ± 1.1 12.8 Liver (HD) 11.3 ± 9.8 0.137 ± 0.025 (0.805) 5.2 ± 1.0 8.2 a LD = the low-dose exposed group; HD = the high-dose exposed group. b Coefficient of determination for the model is shown in parentheses. J. Cheng et al. / Chemos 3.3.1.1. Hydroxylation. Chromatographic patterns of the metabolic products in exact were obtained using GC-EI/MS operated in both full-scan mode and SRM mode. In the EI full-scan spectra, the molecular ion cluster as brominated compound observed in parent BDE 15 wasm/z 326, 328, 330, the fragment ion atm/z 168 was [M- 2Br]+ ion, and another ion fragment cluster at m/z 219, 221 was corresponding to [M-(BrCO)]+. Fig. 2a illustrates a typical mass spectrum with clear ion frag- mentation patterns to provide structural information. In the EI mode, the molecular ion cluster (m/z 342, 344, 346) was the dom- inant fragment as expected for brominated compounds containing two of bromine at equal proportions of the 79 and 81 bromine iso- topes. This unknown compound yielded an intense fragment ion at m/z 184 as a base peak which indicated the loss of two bromine atoms ([M-2Br]+) from the molecular ion. Furthermore, another ion fragment cluster at m/z 235, 237 was assumed to be an ion of [M-(BrCO)]+. The molecular ion cluster and other fragment ions observed in this unknown compound (metabolite I) were all differ by 16 m/z units with that observed in parent BDE 15. All these re- sults suggested that a hydroxy group (m/z 17) was placed at the substituted position. The retention time of this metabolite in the GC chromatogram was later than that of BDE 15 (Table SM-1). In addition, the mass spectrometric fragmentation pattern was similar to that of OH-BDE 47 standard, which had the molecula