Environ. Sci. Technol. 2004, 38, 6328-6332
Disruption of Rapid, Nongenomic Steroid Actions by Environmental Chemicals: Interference with Progestin Stimulation of Sperm Motility in Atlantic Croaker PETER THOMAS* AND KELLY DOUGHTY The University of Texas at Austin Marine Science Institute, 750 Channelview Drive, Port Aransas, Texas 78373
Several nongenomic steroid actions, like genomic ones, can be disrupted by estrogenic xenobiotics (xenoestrogens), but the extent and sensitivity of this alternative mechanism of steroid action to chemical interference remain unclear. The effects of environmentally realistic concentrations of a broad range of organic contaminants on the nongenomic action of a progestin (17,20β,21-trihydroxy-4-pregnen-3-one or 20β-S) to upregulate Atlantic croaker sperm motility were examined in an in vitro bioassay. Pretreatment of sperm for 10 min in vitro with estrogenic compounds (estradiol17β, o,p′-DDT derivatives, zearalenone, bisphenol A, 2′,3′,4′,5′PCB-4-OH, kepone, chlordane, methoxyclor) and nonestrogenic organic compounds (p,p′-DDT derivatives, atrazine, Aroclor 1254, naphthalene, benzene) at concentrations ranging from 0.01 to 10 µM did not decrease the percent of motile sperm, but all of them partially or completely blocked the response to 20β-S. Most of the compounds impaired this endocrine mechanism at a concentration of 0.1 µM (∼3040ppb), whereas o,p′-DDT and atrazine were effective at lower concentrations. The antagonistic actions of o,p′-DDT were partially reversed with 10-fold higher concentrations of 20β-S, which is consistent with a hormone receptormediated mechanism of DDT action. The finding that low concentrations of a wide range of organic environmental contaminants can interfere with a rapid, nongenomic steroid action suggests that this mechanism of endocrine disturbance is of toxicological importance.
Introduction Extensive evidence has been obtained over the past 20 yr for widespread developmental and reproductive problems in fish and wildlife exposed to environmental contaminants that interfere with endocrine function, especially those controlled by estrogens (1). Feminization of male birds, alligators, and fish and the production of estrogen-regulated proteins by male fish such as the yolk precursor (vitellogenin) have been reported after environmental exposure to xenoestrogens such as ortho,para derivatives of DDT, and sewage containing nonylphenols and ethynyl estradiol (2, 3). The majority of the xenoestrogens (including o,p′-DDT and its isomers, PCBs and their hydroxylated metabolites, phthalates, kepone, methoxychlor metabolites, nonylphenol, and bisphenol A) exert their estrogenic effects primarily by * Corresponding author phone: (361)749-6768; fax: (361)749-6777; e-mail:
[email protected]. 6328
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004
binding to and activating nuclear estrogen receptors, resulting in alterations in the transcription rates of estrogen-regulated genes (4, 5). Environmental chemicals have also been shown to interfere with androgen and progesterone functions by binding to their nuclear receptors, in most cases antagonizing their genomic actions (6, 7). Although interference with nuclear receptor-mediated steroid actions is the most common mechanism of endocrine disruption by xenobiotic chemicals identified to date, environmental contaminants could also potentially interfere with rapid, nongenomic steroid actions initiated at the cell surface by binding to steroid membrane receptors (8). Nongenomic steroid actions have been described in a broad range of cell types and animal models (9-12). For example, specific membrane receptors for estrogens have been identified in mammalian hypothalamic, pituitary, hepatic, and uterine tissues and fish testes and ovaries (13, 14); for glucocorticoids in liver, brain, and lymphocytes (9, 10); for androgens on lymphocytes (15) and in fish ovaries (16); and for progestogens in fish and amphibian oocytes (17, 18), in bovine ovaries, and in human and fish sperm membranes (19, 20). An important difference between the two steroid mechanisms is that genomic ones are typically slow, occurring over a time scale of hours, whereas the membrane receptorinitiated actions are rapid. Acute alterations in calcium concentrations occur within 30 s of steroid treatment at the cell surface and in G-proteins, other ions, and second messengers such as cyclic nucleotides within a few minutes (19, 21, 22). The significance of this alternative mechanism of steroid action has become more widely appreciated in the past few years with the publication of many rapid, nongenomic steroid actions in vertebrate tissues. In contrast, there are only a few reports on the effects of xenobiotic chemicals on these nongenomic steroid actions, so the toxicological importance of this novel mechanism of endocrine disruption remains unclear. The observation that two xenoestrogens, kepone and o,p′DDD, interfered with the nongenomic action of a progestin, 20β-S, on croaker oocytes in vitro to induce oocyte maturation provided initial evidence for this mechanism of endocrine disruption (23). Subsequently, it was shown that kepone and o,p′-DDD disrupt this nongenomic mechanism in the oocytes of a closely related species, spotted seatrout, by binding to the progestin membrane receptor that mediates this action of 20β-S (24). Kepone and an ortho,para DDT isomer, o,p′DDE, are also able displace bound [H3 ]-20β-S from the sperm progestin membrane receptor in Atlantic croaker (25). Moreover, it was demonstrated that high concentrations of kepone block 20β-S upregulation of sperm motility in croaker (25). Preliminary evidence that xenoestrogens can interfere with nongenomic steroid actions has also been obtained in other vertebrate models (13, 26-29). However, the actions of only a limited number of xenoestrogens were examined in these studies, and the effects of nonestrogenic compounds were not evaluated. The purpose of the present study was to examine the potential of a broader range of estrogenic compounds to interfere with a nongenomic steroid action in a vertebrate model. The progestin upregulation of sperm motility in croaker was selected for these studies because this nongenomic progestin mechanism is equivalent to that described in other vertebrate models, including humans, and therefore the results are likely to be broadly applicable (22, 19). Several nonestrogenic compounds were also evaluated to determine whether the mechanism of endocrine disruption is specific to estrogenic compounds. Low environmentally relevant 10.1021/es0403662 CCC: $27.50
2004 American Chemical Society Published on Web 09/01/2004
concentrations of the compounds were tested to determine whether this mechanism is of potential importance in environmental toxicology.
Materials and Methods Chemicals. 17,20β,21-Trihydroxy-4-pregnen-3-one (20β-S) was purchased from Steraloids, Inc. (Wilton, NH). o,p′-DDT, o,p′-DDE, p,p′-DDT, atrazine, and Aroclor 1254 were obtained from Chem Services (Westchester,PA). Aroclor 1254 and 2′,3′,4′,5′-tetrachloro-4-biphenylol (2′,3′,4′,5′-PCB-4-OH), were purchased from Ultra Scientific (Kingston, RI). Chemicals and salts used for making the buffers, bovine serum albumen for coating the slides, estradiol-17β, and zearalenone were purchased from Sigma Chemical Company (St. Louis, MO) and from Fisher Scientific (Pittsburgh, PA). Benzene and naphthalene were purchased from Aldrich Chemical Co. (Milwaukee, WI). Kepone and methoxychlor were obtained from the NIEHS repository. Animals. Adult 1-year-old male Atlantic croaker (approximately 70 g body weight) were collected by commercial trawl in the vicinity of the University of Texas Marine Science Institute, Port Aransas, TX, at the beginning of the reproductive season in September and transported to holding facilities at the institute. Fish were maintained in 500-gal re-circulating seawater tanks and fed a commercial pellet diet daily. The fish were exposed to a photoperiod/temperature regime that mimicked the seasonal changes for 2 months, which resulted in the completion of spermatogenesis and spermiogenesis, and the fish had freely flowing milt. Thereafter, the environmental conditions were kept constant in fall conditions for an additional 3 months to maintain spermiation. Treatment of Sperm with 20β-S and Xenobiotics. Milt samples were obtained at the beginning and end of the reproductive season outside the peak of spawning activity. Milt was collected from the cloaca of spermiating croaker with a syringe after applying gentle pressure to the abdomen to express the milt, care being taken to avoid contamination of the milt with urine or seawater, which would cause premature activation of the sperm. Milt from 2-3 individuals was pooled to reduce interassay variation, and 5 µL was transferred to a glass vial and kept at 4° C for up to 30 min prior to experimentation. Pooled milt was diluted 20-fold in physiological saline (in 95 µL of pre-dilutant buffer) as described previously (30) to ensure the subsequent activation of sperm was uniform. Aliquots of milt were incubated in pre-dilutant buffer (0.5 mL) containing 20β-S (20 or 200 nM) in the presence or absence of the test chemicals over a range of concentrations (0.01-10 µM) for 10-20 min prior to further dilution (20-fold) and activation of the sperm with activating buffer and assessment of sperm motility. 20β-S and chemicals were dissolved in ethanol and added to the pre-dilutant buffer in 10-µL aliquots. An equal volume of ethanol was added to the controls so that all treatments received the same volume of ethanol (final concentration: 1%) that did not affect motility (data not shown). The activating buffer contained the same ratio of ions as the pre-dilutant but at double the concentration (680 mOs/kg) to mimic the hyper-osmotic stimulus croaker sperm are exposed to upon their release into seawater resulting in their activation. Activation: a sample (1 µL) of the diluted sperm sample was pipetted onto a microscope slide coated with bovine serum albumen (to prevent sperm from sticking to the slide), 20 µL of activating solution was added, a coverslip was placed on the sperm preparation, and the sperm motility was recorded immediately at 400× magnification using a compound microscope equipped with a video camera (black and white Cohu CCD) and VHS recorder for later analysis. Percent motile sperm was calculated from direct counts of the total number of alive sperm (i.e., showing any discernible movement) and of the number of actively swimming sperm (i.e., showing
normal, high motility; 40-80 active sperm counted/treatment replicate) 10 s after activation. Both the velocity and rate of change of direction of sperm changed in parallel with the increase in sperm motility as observed previously (30), but only the percent of motile sperm was determined from the videotape. The percent motility of pooled untreated sperm from different groups of donors varied after treatment with 20β-S. Therefore, to combine the results from replicate assays, all data for each sperm sample were normalized to percent of control untreated values (set at 100%).
Results The percent of untreated sperm (not treated with 20β-S) that were motile at the beginning and end of the reproductive season varied from 51 to 61% (mean 56 ( 0.5%, N ) 29). Treatment of croaker sperm with 20 and 200 nM 20β-S caused 36% and 32% increases, respectively, in the percent motile sperm in six separate experiments; both 20β-S treatments significantly different from controls but not from each other (p < 0.05). Incubation of croaker sperm for 10 min with o,p′DDT did not cause a decrease in percent motile sperm below control levels at concentrations of 0.1 µM (Figure 1) up to 10 µM (results not shown). However, co-incubation of 0.1 µM o,p′-DDT with 20 nM 20β-S tested consistently blocked the response to the hormone (o,p′-DDT + 20β-S treatment significantly different from 20β-S alone, p < 0.05, Figure 1A). One-tenth of this concentration of o,p′-DDT (0.01 µM) also significantly decreased the response to 20 nM 20β-S, but the block was only partial (Figure 1B, p < 0.05). Other estrogenic compounds also decreased the sperm motility response to 20 nM 20β-S (Table 1). The response to 20β-S in the presence of bisphenol A, 2′,3′,4′,5′-PCB-4-OH, zearalenone, and estradiol-17β was significantly impaired at concentrations of 0.1 µM (Table 1, Figure 1C,D, p < 0.05), whereas antagonism of 20β-S action was only observed with 10-fold higher concentrations (1 µM) of chlordane and methoxyclor and with 10 µM kepone (Table 1). Several nonestrogenic endocrine-disrupting chemicals also interfered with the sperm motility response to 20β-S (Figure 2). Interestingly, at a concentration of 1 µM the nonestrogenic p,p-DDT isomer was equally effective as the estrogenic o,p-DDT isomers in blocking the stimulatory action of 20β-S (Figure 2A). Atrazine completely blocked the response to 20 nM 20β-S at a concentration of 0.05 µM (Figure 2B, p < 0.05) but was ineffective at 0.01 µM (results not shown). Naphthalene and benzene were also effective antagonists of 20β-S action at a concentration of 0.1 µM, whereas significant inhibition with Aroclor 1254 was only at 10 µM (Figure 2C, Table 1). Co-incubation of 0.1 µM o,p′-DDT with a 10-fold higher concentration of 20β-S (200 nM) partially reversed the inhibitory effect of the pesticide (Figure 3A, p < 0.05); however, this treatment was ineffective in restoring motility in the presence of 0.1 µM 2′,3′,4′,5′-PCB-4-OH (Figure 3B).
Discussion The results of the present study demonstrate that a wide range of organic contaminants can interfere with a nongenomic steroid action that may disrupt reproductive function. Previous studies had indicated the potential for chemicals to disrupt endocrine function by interfering with this steroid mechanism, but its possible toxicological importance remained uncertain, particularly in comparison to genomic mechanisms mediated by binding to nuclear steroid receptors. For example, information was only available from these earlier studies on the effects of several xenoestrogens (13, 23-29). The present results show that nonestrogenic as well as estrogenic organic compounds can interfere with a rapid, nongenomic progestin action to upregulate sperm motility VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6329
TABLE 1. Chemical Concentrations Causing Attenuation of 20β-S-Induced Upregulation of Croaker Sperm Motility xenobiotic bisphenol A chlordane estradiol kepone methoxyclor o,p′-DDD o,p′-DDE o,p′-DDT 2′,3′,4′,5′-PCB-4OH zearalenone
minimum effective concn (µM)a 0.1 1b 0.1 10 1b 1 0.1c 0.01 0.1 0.1
nonyphenol
1d
Aroclor 1254 atrazine benzene p,p′-DDE p,p′-DDT naphthalene
10 0.05 0.1 0.1 0.1 0.1
a Lowest concentration tested that significantly (p < 0.05) reduced upregulation of sperm motility by 20 nM 20β-S. Range of concentrations tested usually ranged from 0.1 to 10 µM with the exception of o,p′-DDD (not tested at 0.1 µM). Each concentration was tested three times in an assay, and three replicated assays were conducted. b Decrease not significant at 1 or 10 µM. c Not tested at 0.01 µM. d Only two experiments.
FIGURE 1. Effects of exposure to (A) 0.1 µM o,p′-DDT ( 20 nM 20β-S; (B) 0.01 µM o,p′-DDT ( 20 nM 20β-S; (C) 0.1 µM zearalenone ( 20 nM 20β-S; and (D) 0.1 µM estradiol-17β ( 20 nM 20β-S, on subsequent percent motility of Atlantic croaker sperm in vitro. Sperm were pretreated with the compounds, alone and in combination with 20 nM 20β-S for 10 min in predilutant buffer prior to the addition of activating solution. Percent motile sperm was assessed 10 s after activation and normalized to controls (100%) to account for differences in the percent motility of the sperm from different donors. Each bar represents the mean ( SEM of three replicate estimations from three separate experiments. Asterisks denote means for combined treatment of 20 nM 20β-S and each compound significantly different from means for treatment with 20 nM 20β-S alone (p < 0.05). 6330
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004
in Atlantic croaker. Moreover, the finding that this progestin action can be partially blocked by xenobiotic concentrations as low as 10 nM raises the possibility that nongenomic steroid actions may be as susceptible as genomic ones to chemical interference. The majority of the chemicals tested completely blocked this hormone action at 0.1 µM, equivalent to 30-40 ppb, which is a tissue burden frequently reported for many of them in vertebrates collected from contaminated aquatic ecosystems (31). However, the environmental relevance of these findings is unclear because information is currently lacking on the concentrations of these compounds in the milt of fish collected from contaminated environments. Limited data for the general human population and for domestic animals suggest semen concentrations of pesticides such as p,p′-DDE typically range from
