Environ. Sci. Technol. 2005, 39, 2599-2607
Accounting for Differences in Estrogenic Responses in Rainbow Trout (Oncorhynchus mykiss: Salmonidae) and Roach (Rutilus rutilus: Cyprinidae) Exposed to Effluents from Wastewater Treatment Works C. R. TYLER,† C. SPARY,† R. GIBSON,‡ E. M. SANTOS,† J. SHEARS,† AND E . M . H I L L * ,‡ Environmental and Molecular Fish Biology Group, Department of Biological Sciences, University of Exeter, Prince of Wales Road, Exeter, EX4 4PS, United Kingdom, and Centre for Environmental Research, School of Life Sciences, University of Sussex, Falmer, Sussex, BN1 9QJ, United Kingdom
Effluents from wastewater treatment works (WwTWs) contain estrogenic substances that induce feminizing effects in fish, including vitellogenin (VTG) synthesis and gonadal intersex. Fish vary in their responsiveness to estrogenic effluents, but the physiological basis for these differences are not known. In this study, uptake of estrogen from two WwTW effluents (measured in hydrolyzed bile) and estrogenic response (VTG induction) were compared in a salmonid (rainbow trout, Onchorhynchus mykiss) and a cyprinid fish (roach, Rutilus rutilus). Immature rainbow trout were more responsive than maturing roach to the estrogenic effluents. The more potent of the two estrogenic effluents (containing between 24.3 and 104.1 ng estradiol-17β equivalents/L [E2eq/L]) resulted in a 700fold and 240-fold induction of plasma VTG in male and female trout, respectively, but only a 4-fold induction in roach (and in males only). The less potent effluent (varying between 4.1 and 6.8 ng E2eq/L) induced VTG in the trout only, with a 4-fold and 18-fold induction in males and females, respectively. In fish exposed to tap water, the estrogenicity of the hydrolyzed bile was 0.03 ( 0.01 ng E2eq/µL (for both sexes in trout), 0.18 ( 0.04 ng E2eq/µL in male roach, and 0.88 ( 0.15 ng E2eq/µL in female roach. The higher bile content of estrogen in control roach reflected their more advanced sexual status (and thus higher endogenous estrogen) compared with the immature female trout. In trout maintained in effluents, the bile content of estrogen was 100-fold and 30-fold higher than controls at WwTW A and B, respectively. Bioconcentration factors (BCFs) for estrogenic activity in bile were between 16 344 and 46 134 in trout and between 3543 and 60 192 in roach (no gender differences were apparent). There were strong correlations between VTG induction and the estrogenic * Corresponding author phone: 44 1273 678382; fax: 44 1273 677196; e-mail:
[email protected]. † University of Exeter. ‡ University of Sussex. 10.1021/es0488939 CCC: $30.25 Published on Web 03/09/2005
2005 American Chemical Society
activity of bile extracts for both trout and roach. The results confirm that estrogenic contaminants bioconcentrate to a high degree in fish bile and that the level (and nature) of this accumulation may account for responsiveness to the endocrine disruptive effects of estrogenic effluents. Immature fish were the more appropriate life stage for quantifying estrogen exposure and uptake in bile, as they contain little circulating endogenous oestrogen compared with sexual maturing fish. The nature of the estrogenic contaminants is detailed in an accompanying paper.
Introduction Endocrine disruption has been established in a wide range of wildlife, and in some cases there are strong associations with chemicals that mimic or disrupt hormone function (1, 2). Induced effects range from alterations in reproductive function in mammals, birds, fish, and mollusks, disrupted development and limb deformities in amphibians, and altered reproductive behaviors. Laboratory based studies (both in vitro and in vivo) have identified a wide range of chemicals that act as disrupters/mimics of estrogens, thyroid hormones, and retinoic acids and, more recently, (anti-) androgens (3). Many (most) effects on reproduction appear to have resulted from exposure to estrogenic compounds (although in most cases this is not yet proven) and this has served to polarize endocrine disruption research on environmental estrogens. In vertebrates, estrogens play a fundamental role in gender determination, reproduction, and somatic cell function (47). In fish, exposure to xenoestrogens or to synthetic or natural estrogens has been reported to result in altered fecundity in females (8-10), reduced testicular development (11), reduced fertility in males, and in increased or decreased VTG production (the precursor of egg yolk protein) in both male and female fish (e.g., refs 12 and 13). Effluents from wastewater treatment works (WwTWs) are capable of disrupting reproduction in aquatic organisms (14), and some of these effects are mediated via estrogenic disruption (15-17). Induced feminizing effects of effluents in fish include vitellogenin synthesis (a precursor of yolk), altered titers of plasma sex hormones, and intersex (the simultaneous presence of both male and female characteristics in the same gonad; 18, 19). Intersex is widespread in populations of roach, Rutilus rutilus, living downstream of effluent discharges in U.K. rivers (18), and these fish produce gametes of poorer quality (20) with a reduced capacity for fertilization (21). Furthermore, there is a statistical association between altered sex cell development (oocytes in the testis) in wild roach populations and exposure to estrogenic WwTWs effluents (18). The compounds responsible for the estrogenic activity in WwTWs effluents are believed to be principally steroidal estrogens, both natural (estrone, E1, and estradiol, E2; 22-24) and synthetic (ethinyloestradiol, EE2, derived from the contraceptive pill (25) and, to a lesser extent, alkylphenol polyethoxylates (APEs; 22, 26, 27). Effluents from WwTWs, however, contain a complex mixture of chemicals, including other more weakly active estrogenic compounds (e.g., plasticizers, bisphenols herbicides, and pesticides) and their effects can be additive for VTG induction (28). Bile, a product synthesized in the liver and stored in the gall bladder, has been used in fish for some time for biomonitoring of aquatic pollutants. Pollutants monitored for in this way include hydrocarbon metabolites (ref 29 where levels of exposure were linked with toxicopathic lesions), chlorophenolics (30), and platinum group metals (31, 32). VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2599
Studies of this nature have been undertaken in wild fish in both freshwater (e.g., perch, Perca fluviatalis, and roach living downstream of a bleached kraft mill effluent; 30) and in marine environments (e.g., English sole; Parophrys vetulus; 33). Measurement of chemicals in the bile of caged fish has also been employed to assess dilution of effluents in rivers downstream from point source discharges (e.g., whitefish Coregonus lavaretus; 34). Metabolites of natural and synthetic steroid estrogens and alkylphenolic chemicals such as 4-nonylphenol (35) and 4-octylphenol (36) are all excreted via bile, principally as glucuronides (37). Larsson et al. (38) measured environmental estrogens in bile from juvenile rainbow trout exposed to effluent from WwTWs (for 4 weeks) at concentrations between 10 000 and 1 000 000-fold higher than the surrounding water. A bioconcentration factor for 4-tert-octylphenol residues in bile in fish of 69 000 has been reported after a 10-day exposure to the alkylphenol (36). Thus, bile offers a tissue reservoir for monitoring exposure to EDCs. Very few studies have compared the effects of estrogens and their mixtures in different vertebrates. One recent study indicated that rainbow trout (Oncorhynchus mykiss) responded more rapidly to an estrogenic WwTWs effluent compared with carp (Cyprinus carpio; 39). A study on wild flounder and eels (fish with similar ecological niches) living in a contaminated environment found very different bile contents of hydrophobic pollutants such as PAHs (40), suggesting differences in uptake or processing of chemicals from the aquatic environment in those species. Furthermore, although phase I and phase II biotransformation pathways are involved in the metabolism of most EDCs, different fish can metabolize the same product in different ways. For instance, a major difference in the metabolism of 4-tertoctylphenol between rainbow trout and rudd (Scardinius erythrophthakamus) has been reported (41). In the rudd, the most common metabolites of the alkylphenol were products of phase I biotransformation, namely, aromatic and aliphatic hydroxylation reactions, together with conjugation of 50% of the metabolites to glucuronide and 25% of the metabolites to glucoside residues. In contrast, the products of 4-tertoctylphenol metabolism were simpler in rainbow trout where the major metabolite was the glucuronide conjugate of the parent alkylphenol and only one metabolite was a product of hydroxylation. Several studies on the metabolism of nonylphenol in salmonid fish have shown that glucuronide conjugates and hydroxylates are the principle metabolites in bile and some of these have indicated that tissue specific metabolism can occur (42-44). In this study, the estrogenic response (VTG induction) of a salmonid fish (rainbow trout) was compared with that of a cyprinid fish (roach) indigenous to U.K. rivers on exposure to WwTWs effluents, and the estrogenic content of both the effluents and the bile of the fish was measured using an estrogen responsive recombinant yeast screen. Measurement of plasma VTG has been established as an effective measure of the net effect of exposure (and response) to estrogenic contamination in effluent and river systems, and it allows for transformation processes within the whole organism, such as degradation or bioaccumulation of estrogenic chemicals. In parallel, analytical chemistry was employed to identify and measure the concentrations of estrogenic chemicals in the effluents and in the bile of the exposed fish and this is reported in an accompanying paper (Gibson et al.) (45).
Materials and Methods Fish. Rainbow trout (Oncorhynchus mykiss) and roach (Rutilus rutilus) were chosen as representative species of salmonid and cyprinid fish, respectively, for this comparative exposure analysis. Rainbow trout are sensitive to estrogen exposure (15, 27) and the dynamics of the vitellogenic response on exposure to estrogens has been well-character2600
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
ized (46). Sexual development in wild roach populations is affected by exposure to estrogenic effluents (18), and a series of controlled effluent exposures have studied VTG induction and altered gonad duct development in this species (24, 47). Maturing roach (Rutilus rutilus) of mixed sex and 3+ in age were obtained from a wild fishery from Framlingham Fisheries, Suffolk. Rainbow trout (Oncorhynchus mykiss) were obtained from two commercial fish farms: one in Cumbria (Hawkshead Trout Farm) and the other in Dorset (Houghton Springs Fish Farm). The Study Effluents. Both study WwTWs were estrogenic to fish, and the effluent at WwTWs A had been shown to induce VTG synthesis in all life stages and cause gonad duct disruption for exposures during early life (47). Treatment works A had a population equivalent of 138 000, with approximately 6% of the influent derived from trade sources. WwTWs A effluent received biological treatment via both percolating filters and fine bubble-activated sludge processes. WwTWs B had a population equivalent of 312 700 with 24% of the influent derived from trade sources. At WwTWs B after primary treatment, the influent was treated using a combination of bubble-diffused air-activated sludge treatment and a biological phosphorus removal activated sludge plant. Both WwTWs contained a variety of environmental estrogens, including alkylphenols and natural and synthetic steroid estrogens, and bisphenol A (WwTWs A, 24, 47, 48; WwTWs B, our own unpublished data). On the basis of their estrogenic contents, it was envisaged that the effluent from WwTWs A would invoke a strong vitellogenic response for the exposure period and the effluent from WwTWs B a weaker induction. Previous studies have shown seasonal differences in the potency of effluent discharges at the study WwTWs (16, 24, 48), and the studies undertaken here took place in quick succession during late November and early December. Study Protocol. The objectives of this study were to determine the responses of two fish species to estrogenic WwTWs effluents and to assess how the vitellogenic responses related to uptake of estrogen from the water (estrogen in bile). Vitellogenin production in individual fish was determined over an exposure period of between 8 (WwTWs B) and 10 days (WwTWs A) from blood samples collected from each fish at time 0 and day 8/10. The level of estrogenic activity was measured in hydrolyzed bile at the end of the exposures. The responses to estrogens may potentially vary depending on the gender and sexual status of the exposed fish. To allow the responses of individuals to be tracked, personal identity tags (PIT) were implanted in the fish. Effluents were similarly analyzed for estrogen content on day 0 and day 8/10. Exposure to both effluents was conducted in flow-through systems. At WwTWs A, three 1 m3 tanks were supplied with either full strength treated effluent, river water (0% effluent; river water control), or dechlorinated tap water (absolute control). At WwTWs B, two 1 m3 tanks were supplied with either full strength treated effluent or dechlorinated tap water (absolute control). Total flow of effluent, river water, or dechlorinated tap water to each tank was 9-10 L/min. The tanks were aerated to ensure sufficient oxygen supply to support the biomass of fish, and the temperature was monitored daily in all tanks. Each tank was divided into two with a mesh partition to keep the rainbow trout and roach separated in the same exposure tanks. The roach and rainbow trout used at both WwTWs were supplied as mixed sex. Under anaesthesia, between 20 and 26 fish of each species were blood sampled, weights and lengths were measured, and the fish were tagged with PITs placed in musculature anterior to the dorsal fin. Fish were fed commercial trout pellets (BOCM Pauls Limited, Renfrew, U.K.) once daily throughout the trial until 48 h prior to the final sampling, when food was withheld to maximize the
amount of collectable bile. The food source had been tested in the laboratory in the YES (recombinant yeast estrogen receptor assay, see below) and was free of any estrogenic contamination. On day 10 (WwTWs A) and day 8 (WwTWs B), all fish were anaesthetized, blood sampled, and body weights and lengths were recorded. The condition factor (K; a measure of the body form) was calculated for individual fish as the ratio of bodyweight (g) × 100/(total length mm) (3). Fish were killed humanely by a sharp blow to the head and by macerating the brain tissue, according to Home Office regulations, and the liver and gonads were dissected and weighed to determine the hepatosomatic index (HSI; liver weight as a percentage of body weight) and gonadosomatic weight (GSI; gonad weight as a percentage of body weight), respectively. Bile was collected using sterile 1-mL syringes fitted with a 23-gauge needle and dispensed into sterilizedacid-washed glass vials and deep-frozen for subsequent hydrolysis for estrogen activity assessment. Sample Extraction and PreparationsEffluent. Effluent, river or tap water samples (1 L), were collected on day 0 and on the last day of the exposures (day 8 or 10), and were acidified with glacial acetic acid (25 mL) prior to extraction on OASIS (500 mg) SPE cartridges (Waters Ltd, Herts, U.K.). The columns were conditioned with methanol (10 mL) and water (5 mL), and the sample was eluted with methanol (8 mL), ethyl acetate (6 mL), and hexane (6 mL). The combined extracts were evaporated to dryness and then redissolved in 1 mL ethanol for YES assay analysis. Sample Extraction and PreparationsBile. Bile was deconjugated with β-glucuronidase (1000 units/mL), sulfatase (2 units/mL), and β-glucosidase (20 units/mL). The enzymes (200 uL of each) were added to a mixture of bile (100 uL), 0.1 M phosphate buffer at pH 6.0 (1500 uL), and water (800 uL). The solution was incubated for 16 h at 37 °C. Glacial acetic acid (300 uL) and water (2 mL) were added to the samples and assayed for estrogenic activity directly in the YES assay. Measurement of Estrogenic Activity in Effluents and Bile. The intrinsic estrogenic activity of environmental samples (effluent and hydrolyzed bile) was extracted as described above and measured in a recombinant yeast estrogen receptor assay (49). The “estrogenic” activity of the samples was determined by directly comparing the concentrationresponse curves of the individual effluent samples with the concentration-response curves obtained for reference standard chemicals. The amount of estrogenic activity is presented as ng/estradiol-17β equivalents (E2eq). The reference standard chemicals for the estrogen assay was 17β-estradiol (E2; >98% pure, Sigma Chemical Company Limited, Dorset, U.K.). To determine estrogen receptor agonist activity, E2 was serially diluted in ethanol and 10 µL aliquots were transferred to a 96-well flat bottom microtiter plate (Linbro/ Titertek, ICN FLOW, Bucks, U.K.). Assay Procedure. The yeast assays were performed as described by Routledge and Sumpter (49). Briefly, test chemicals were serially diluted in ethanol and 10-µL aliquots were transferred to 96-well flat bottom microtiter plates. The ethanol was allowed to evaporate to dryness, after which aliquots (200 µL) of assay medium (containing the recombinant yeast and the chromogenic substrate, CPRG) were dispensed into each sample well. The plates were sealed, shaken for 2 min, and then incubated at 32 °C. After an incubation period of 3 days, color development in the medium was measured at an absorbance of 540 nm and turbidity of the yeast was measured at 620 nm (using a Spectramax Plus, microtiter plate reader). A toxicity, identification, and evaluation approach was used to identify estrogenic components in effluent and bile (reported in Gibson et al., ref 45).
TABLE 1. Estrogenic Activity of Tap Water, River Water, and Effluent Determined in the Yeast Estrogen Screen (YES)a estrogenic activity (ng E2eq/L) sample
site
tap water river water
WwTW A WwTW A
effluent
WwTW A
tap water effluent
WwTW B WwTW B
a
at start of exposure
at time of fish sacrifice
