- Email: [email protected]
Effects of nonylphenol on the gonadal differentiation of the hermaphroditic fish, Rivulus marmoratus
Aquatic Toxicology 57 (2002) 117– 125
www.elsevier.com/locate/aquatox
Effects of nonylphenol on the gonadal differentiation of the hermaphroditic fish, Ri6ulus marmoratus Jennifer N. Tanaka, John M. Grizzle * Southeastern Cooperati6e Fish Disease Project, Department of Fisheries and Allied Aquacultures, Auburn Uni6ersity, Auburn, AL 36849, USA Received 22 August 2000; received in revised form 8 March 2001; accepted 13 March 2001
Abstract Nonylphenol (NP) is an estrogenic degradation product of alkylphenol polyethoxylate surfactants. In this study, the effects of NP on gonadal differentiation and development in Ri6ulus marmoratus (Osteichthyes, Cyprinodontiformes), a self-fertilizing, hermaphroditic species, were examined. Starting at hatching, fish were exposed to 150 or 300 mg l − 1 NP (nominal concentrations) in a static system with daily renewal. The measured concentration of NP in the test water decreased rapidly; half-life was 8.0 h. After 60 d of exposure to NP, fish were kept in uncontaminated water for 20 d and were then preserved for histological examination. No fish exposed to 300 mg l − 1 NP (N= 8) and only two of nine fish exposed to 150 mg l − 1 NP developed testicular tissue, compared with nine of 13 water-control fish and five of nine solvent-control fish. Oogenesis was also significantly inhibited by NP. None of the fish exposed to 300 mg l − 1 and only two of nine fish exposed to 150 mg l − 1 NP had vitellogenic oocytes, compared with seven of 11 water-control fish (not including males) and six of nine solvent-control fish. Dysplasia of the gonadal lumen also occurred in fish exposed to 300 mg l − 1 NP. These changes, including testicular agenesis, have not been previously reported in fish exposed to NP. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Nonylphenol; Fish; Histopathology; Gonadal differentiation; Vitellogenesis
1. Introduction 4-Nonylphenol (NP) is a degradation product of a group of nonionic surfactants called nonylphenol ethoxylates (NPE), which are used globally in the production of plastics, pesticides, and cleaning products, and are present in sewage effluents around the world (Talmage, 1994). * Corresponding author. Tel.: + 1-334-844-3474; fax: +1334-844-9208. E-mail address: [email protected] (J.M. Grizzle).
Nonylphenol is more persistent in the environment than the parent NPE (Maguire, 1999), and is found in surface waters, aquatic sediments, and groundwater (Talmage, 1994; Bennie, 1999). Recent research has identified NP as the most important degradation product of NPE because of its enhanced resistance toward biodegradation, toxicity, ability to bioaccumulate in aquatic organisms, and estrogenicity (Ahel et al., 1994). Nonylphenol is estrogenic in various aquatic animals (see reviews by Talmage, 1994; Nimrod and Benson, 1996; Servos, 1999). Experiments
0166-445X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S0166-445X(01)00186-2
118
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
with fish have shown that NP causes elevated plasma vitellogenin and zona radiata protein concentrations in both males and females, inhibition of spermatogenesis, induction of ovotestes in males, and altered gonadosomatic indices (Jobling et al., 1996; Gray and Metcalfe, 1997; Arukwe et al., 1998, 2000; Kinnberg et al., 2000). Histological changes in the epidermis of NP exposed rainbow trout Oncorhynchus mykiss resembled changes in estradiol-exposed fish but in addition, mucosomes in goblet cells were enlarged and irregularly shaped (Burkhardt-Holm et al., 2000). Ri6ulus marmoratus is a euryhaline fish that inhabits coastal mangrove swamps from Brazil through central Florida (Taylor et al., 1995). Adult R. marmoratus are usually hermaphrodites, but males also occur. This species is the only known vertebrate normally reproducing by internal self-fertilization (Harrington, 1961). The clones that result from self-fertilization, as well as its hardiness, make R. marmoratus a useful species for aquatic chemical testing (Koenig and Chasar, 1984). Thus far, the only studies related to gonadal differentiation in fish exposed to NP were with Japanese medaka, Oryzias latipes (Gray and Metcalfe, 1997; Nimrod and Benson, 1998). The objective of this preliminary study was to determine the effects of chronic exposure to NP on the gonadal differentiation of R. marmoratus.
2. Materials and methods
2.1. Chemicals HPLC-grade methanol, acetone, and running buffer water were purchased from Fisher Scientific (Pittsburgh, PA). p-Nonylphenol (95.5% pnonylphenol; 4.34% o-nonylphenol; 0.05% 2,4-dinonylphenol) was a gift from Schenectady International, Inc. (Schenectady, NY). HPLCgrade 4-tert-butylphenol was purchased from Fluka Chemical Corporation (Milwaukee, WI). HPLC-grade dichloromethane was purchased from Sigma–Aldrich (St. Louis, MO). Brackish water was made with distilled water and synthetic sea salts (HW Marine Mix; Hawaiian Marine, Inc., Houston, TX).
2.2. Animals 2.2.1. Nonylphenol exposure experiments Newly hatched R. marmoratus were obtained from the DS clone (Kallman and Harrington, 1964) maintained at the Department of Fisheries and Allied Aquacultures at Auburn University and were descendants of specimens obtained from the Gulf Ecology Division of the Environmental Protection Agency Office of Research and Development, Gulf Breeze, FL. Fish were held singly in 5 cm diameter stackable glass bowls containing 10 ml of brackish water with 12–14 parts per thousand (ppt) salinity. Fish were fed to satiation with Artemia nauplii 6 d per week. Fish were held at 21–25°C and exposed to a 12 h light:12 h dark photoperiod. 2.2.2. Nonylphenol persistence experiment Adult R. marmoratus (mean total length= 44 mm; mean wet weight=0.97 g) were maintained singly in 10 cm diameter stackable glass bowls with approximately 150 ml of brackish water (12– 14 ppt salinity). Fish were fed to satiation two to three times per week with Artemia nauplii. 2.3. Study design Fish were randomly selected to be exposed to one of the two concentrations of NP, methanol (solvent control), or the water used to prepare other treatments (water control). Exposures began within 9 h of hatching. A stock solution of 500 mg l − 1 NP was prepared in methanol; aliquots of this stock solution were added to brackish water (12–14 ppt) to produce nominal concentrations of 150 or 300 mg l − 1 NP. Solvent-control fish were exposed to nominal concentrations of 600 ml l − 1 methanol, corresponding to the highest concentration of methanol used in the NP treatments. The number of fish exposed to each treatment were: water control, 13; solvent control, 9; 150 mg l − 1 NP, 9; and 300 mg l − 1 NP, 8. All solutions were renewed 6 d per week (following feeding). For water changes, fish were individually aspirated into a glass pipet, the old solution was discarded, and the bowl was scrubbed with a nylon bristle brush. The bowl was
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
rinsed with distilled water followed by a rinse with the appropriate test solution, and the fish was then replaced in fresh solution with less than 1% transfer of the old solution. Fish were fed only on days when solutions were changed. All treatment and control exposures were from hatching until 60 d post-hatching, which exceeds the time period when gonadal differentiation of some fish species can be influenced by exogenous compounds (Nakamura et al., 1998). Fish were maintained in a fume hood during exposures. After exposure, fish were kept in 12– 14 ppt brackish water outside the fume hood for 20 additional days to reduce the chances that alterations observed were acute or temporary effects of NP. Feeding and water changes were also conducted 6 d per week during this postexposure period.
119
1 and 2). The prominent features of the mid cortical alveolus stage was a ring of cortical alveoli (3-15 mm diameter) at the periphery of the oocyte (Figs. 1 and 3) and the presence of the zona radiata. Oocytes in the late cortical alveolus stage had cortical alveoli (10– 20 mm in diameter) throughout most of the cytoplasm (Fig. 2). In fish examined during this study, only early indications of vitellogenesis were present because fish were only 80 d old. The distinguishing feature of vitellogenesis was eosinophilic, homogenous yolk bodies in the cytoplasm (Fig. 3). Cortical alveoli were still present during vitellogenesis and filled most of the cytoplasm.
2.4. Histopathology Fish were starved for 1 d before fixation in Bouin’s solution for 24– 48 h. After decalcification in RDO (Apex Engineering Products Corporation; Plainfield, IL), whole fish were embedded in paraffin, serially sectioned (7 mm) transversely, and stained with hematoxylin and eosin. Sections of gonads not available for histological examination seldom spanned more than 35 mm and never more than 84 mm. All sections of gonads were examined by light microscopy. Oogenesis in each fish was categorized by the most developed oocyte present. Oocyte stages were modified from Soto et al. (1992) and Selman et al. (1993). Stages used were: (1) perinucleolus; (2) early cortical alveolus; (3) mid cortical alveolus; (4) late cortical alveolus; and (5) vitellogenesis. Oocytes in the perinucleolus stage had strongly basophilic cytoplasm (Figs. 1– 3). Cortical alveolus oocytes, characterized by the presence of unstained cortical alveoli but no eosinophilic yolk, were separated into early, mid, and late stages. Oocytes in the early cortical alveolus stage had small cortical alveoli (1– 8 mm diameter) located about midway between the nuclear membrane and the oocyte margin (Figs.
Fig. 1. Transverse histological section of a gonad in a watercontrol Ri6ulus marmoratus 80 d after hatching. Hermaphrodite with testicular tissue (T), gonadal lumen (L), and oocytes. In this section, oocytes are in the perinucleolus stage (P), early cortical alveolus stage (E), and mid cortical alveolus stage (M). Bar = 70 mm.
120
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
NP, or if they caused NP to leach out of any materials within the HPLC system. Adult fish were held singly in brackish water for 12–48 h to determine whether the fish excreted compounds that obscured the presence of, or could be mistaken for, NP during HPLC analysis. Water samples were extracted and analyzed by HPLC.
2.5.2. Nonylphenol concentrations in treatments Because only 10 ml of solution was used for exposure of each R. marmoratus to NP, a separate study was conducted to determine the persistence
Fig. 2. Transverse histological section of a gonad in a hermaphroditic water-control Ri6ulus marmoratus 80 d after hatching. Testis was not present in the portion of the gonad illustrated. Oocytes in the late cortical alveolus (C), early cortical alveolus (E), and perinucleolus (arrow) stages, as well as the gonadal lumen (L). Bar = 70 mm.
2.5. Quantification of nonylphenol concentrations 2.5.1. Preliminary water and chemical analyses For analysis of NP, samples were spiked with butylphenol, extracted in dichloromethane, dried, and then reconstituted in methanol. A Spherisorb S5 ODS2 bonded phase column (4.6 mm × 25 × cm) (Phase Separations, Inc., Franklin, MA) and Waters 474 scanning fluorescence detector were used to detect NP. Concentrations of NP were calculated based on peak areas of NP standards. Distilled water, brackish water, and solventcontrol water were extracted and analyzed by HPLC to determine if the water sources contained NP, or if the extraction procedure caused NP to leach out of the equipment used during extraction. Methanol, dichloromethane, and butylphenol dissolved in methanol were analyzed directly by HPLC to determine if the solvents contained
Fig. 3. Transverse histological section of a gonad in a hermaphroditic water-control Ri6ulus marmoratus 80 d after hatching. Testis was not present in the portion of the gonad illustrated. Oocyte early in the vitellogenesis stage, which is characterized by eosinophilic yolk bodies (arrowheads) in the cytoplasm and an irregular nuclear outline. Cortical alveoli fill most of the cytoplasm. Mid cortical alveolus (M) and perinucleolus stage oocytes are also present. The lumen (arrow) is distorted because of the large size of the vitellogenic oocyte. Bar= 70 mm.
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
of NP in sample solutions. One adult R. marmoratus in 300 ml of water provided a ratio between fish weight and water volume similar to the ratio during NP exposure of 60 d old fish. Time-0 solutions were made and immediately processed for analysis by HPLC; concentrations of these samples were used to calculate recovery efficiency. Adult R. marmoratus were exposed to a nominal concentration of 300 mg l − 1 NP for 6, 12, or 24 h in 300 ml of solution, and the solutions were analyzed by HPLC.
2.6. Statistical analyses Fisher’s exact test was used to determine whether NP affected the presence or absence of testicular tissue or the appearance of the gonadal lumen. Each treatment (including the solvent control) was compared to the water control. A numerical score was assigned to each oocyte stage, and then Kendall’s coefficient of rank correlation (~) was used to assess the effect of NP exposure on the development of oocytes. Analysis of variance was used to determine if fish weights and lengths were different among treatments or were different for hermaphrodites and females. Differences were considered significant at PB 0.05.
3. Results
3.1. Degradation of nonylphenol during fish exposure Nonylphenol was not detected in the brackish water or in any of the solvents used in this study. Furthermore, a detectable amount of NP did not leach out of any of the materials used during the extraction and HPLC procedures. Compounds that could interfere with measurement of NP did not accumulate in water containing R. marmoratus for 12 –48 h. The recovery efficiency for NP was 84.4% (N= 6, S.E.M. =1.3). During exposures of adult fish to a nominal concentration of 300 mg l − 1 NP (ratio between fish weight and solution volume as for exposure of test fish), measured concentrations decreased from an initial concentration of 2559
121
Table 1 Gender of 80 d old Ri6ulus marmoratus after a 60 d exposure to nonylphenol that started at hatching. The number of fish of each gender is listed Treatment
Water control Solvent control NP (150 mg l−1) NP (300 mg l−1)
Gender Hermaphrodite
Male
Female
7 5 2 0
2 0 0 0
4 4 7 8
3.9 mg l − 1 (mean9 S.E.M., N=6) to 1269 6.2 mg l − 1 after 6 h, 86911.3 mg l − 1 after 12 h, and 339 6.4 mg l − 1 after 24 h. The predicted time for loss of half the NP from solution was 8.0 h.
3.2. Nonylphenol exposure No fish died during the 60 d exposure to NP or during the 20 d post-exposure period. Fish lengths and weights were not significantly different among treatments. Mean total length of fish was 15.9 mm (S.E.M.= 0.3) and mean weight was 38.6 mg (S.E.M.= 1.7).
3.2.1. Testicular tissue Males, females, and hermaphrodites were identified histologically when R. marmoratus were sampled 80 d after hatching (Table 1). Percentages of water-control (69%) and solvent-control (56%) fish that had testicular tissue (i.e. hermaphrodites or males) were not significantly different. Testicular tissue in hermaphroditic fish was located in the ventral portion of the gonad and adjacent to the gonadal lumen (Fig. 1). In 92% of control hermaphrodites, testicular tissue was found in only one gonad. Testicular tissue was present in only the posterior two-thirds of the gonad and was found in one to three discontinuous areas. In 75% of all hermaphrodites, testicular tissue was present in less than 10% of the gonad length. Spermatids or spermatozoa were seen in five of the seven water-control hermaphrodites, and in two of the five solvent-control hermaphrodites.
122
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
Two water-control fish were males; these fish had no ovarian tissue when examined 80 d after hatching (Fig. 4). Both males had all stages of spermatogenesis, including spermatozoa. Males were not present in other treatments. None of the fish exposed to 300 mg l − 1 NP and only two fish exposed to 150 mg l − 1 NP developed testicular tissue (Fig. 5). This was significantly different from controls (P =0.0027) for the higher level of NP but was only marginally significant (P=0.054) for the lower concentration of NP. The location of testicular tissue in NP exposed fish was the same as in control fish that were hermaphrodites. Spermatids and spermatozoa were present in one of the two hermaphrodites in the 150 mg l − 1 NP treatment.
Fig. 5. Transverse histological sections of gonads in Ri6ulus marmoratus 20 d after a 60 d exposure to a nominal concentration of 300 mg l − 1 nonylphenol. The most advanced oocytes were in the perinucleolus stage and no testicular tissue was present. The gonadal lumen (arrow) was greatly reduced and did not have the normal tri-lobed appearance. Bar = 30 mm.
Fig. 4. Transverse histological section of a testis in a watercontrol Ri6ulus marmoratus 80 d after hatching. Oocytes were not present in this fish. Bar = 50 mm.
3.2.2. Gonadal lumen All water-control (except for the two males), solvent-control, and 150 mg l − 1 NP treated fish had lumens that were tri-lobed in transverse section (Fig. 1). The tri-lobe configuration was sometimes distorted by large oocytes (Fig. 3), but was apparent in other parts of the gonad where less developed oocytes were present. Exposure to 300 mg l − 1 NP significantly (PB 0.0001) affected development of the gonadal lumen. Only one fish in this treatment had a normal gonadal lumen; all other fish in this treatment had gonadal lumens that were misshapen and reduced in size compared with the lumens of fish in other groups. The gonadal lumens of fish exposed to 300 mg l − 1 NP did not have a tri-lobed configuration or were quite small (Fig. 5).
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
3.2.3. Effects of nonylphenol on oogenesis Most control fish had vitellogenic oocytes (64% of water control and 67% of solvent control), and all control fish had oocytes that had developed to at least the early cortical alveolus stage (Table 2). Exposure to 150 mg l − 1 NP significantly inhibited oogenesis (Kendall’s ~ = − 0.56, P = 0.0062); only 22% of fish in this treatment had vitellogenic oocytes. Exposure to 300 mg l − 1 NP had a stronger negative effect on oocyte development (Kendall’s ~= − 0.93, P B0.00001); no fish in this treatment had oocytes beyond the early cortical alveolus stage, and 50% had no oocytes that had progressed beyond the perinucleolus stage.
4. Discussion Nonylphenol significantly inhibited the differentiation of testicular tissue in R. marmoratus, and this effect was dose dependent. Comparison of the size of fish in the different treatment groups indicated that the testicular agenesis in fish exposed to NP was not an effect of slower overall development. Inhibition of testicular differentiation by NP has not been previously reported in fish or other vertebrates. Exposure to endocrinemodulating chemicals at critical points during early development can permanently alter sex determination in some fish species (Donaldson, 1996), but longer-term studies are required to determine whether the testicular agenesis seen in R. marmoratus in the present study is permanent. Table 2 Oogenesis in 80 d old Ri6ulus marmoratus exposed to nonylphenol. The number of fish is given for each stage, based on the most advanced oocyte present. Oogenesis was significantly inhibited in both of the NP treatments Treatment
Water control Solvent control NP (150 mg l−1) NP (300 mg l−1)
Oocyte stagea P
ECA
MCA
LCA
V
0 0 0 4
1 1 4 4
1 2 3 0
2 0 0 0
7 6 2 0
a P, perinucleolus; ECA, early cortical alveolus; MCA, mid cortical alveolus; LCA, late cortical alveolus; V, vitellogenesis.
123
The most commonly reported action of NP has been stimulation of vitellogenesis, measured by the concentration of plasma vitellogenin. However, in the present study, the ovarian aspect of vitellogenesis and earlier phases of oogenesis were inhibited, which has not been previously reported in fish exposed to NP. Previous studies demonstrated that high dosages of natural and synthetic hormones can inhibit oogenesis. For example, Piferrer and Donaldson (1992) exposed newly hatched chinook salmon in a single immersion to 400 mg l − 1 (nominal concentration) 17b-estradiol or 17a-ethynylestradiol before sexual differentiation, and found that 20.8–40% (17b-estradiol treated) and 11.1– 34.7% (17a-ethynylestradiol treated) of the females produced had reduced cross-sectional gonadal area and underdeveloped oocytes compared to control fish. Most R. marmoratus exposed to the nominal concentration of 300 mg l − 1 NP had a dysplastic gonadal lumen, which was not noted in previous studies of NP. A lumen normally forms in a teleost ovary about the time of ovarian differentiation, and eventually serves as a passageway for ovulated eggs. In hermaphroditic R. marmoratus, fertilization takes place in this lumen (Harrington, 1963). It is possible that the mechanism(s) responsible for inhibition of oogenesis were also responsible for the retarded development of the gonadal lumen. Sexual differentiation in gonochoric species of fish can be altered by hormonally active chemicals (Donaldson, 1996), but because the normal sexual identity of each individual is difficult or impossible to determine, evaluation of changes in gonadal differentiation must consider the sex ratio within a population. In contrast, the absence of testis in any adult R. marmoratus is abnormal because all adults are either hermaphrodites or males. This important advantage of R. marmoratus was not utilized in our study; fish were examined before testicular differentiation had occurred in all of the control fish, and the presence of testis in only one gonad indicated that testicular development was not complete in the control hermaphrodites. Waiting until R. marmoratus are 120 d old, when sexual differentiation is complete (Cole and Noakes, 1997), before evaluation of results would
124
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
improve the usefulness of R. marmoratus in future studies of testicular agenesis. In a previous study of gonadal differentiation in fish exposed to NP (nominal concentrations, 10–100 mg l − 1), ovotestis developed in Japanese medaka, and the sex ratio in the highest concentration of NP was skewed toward females (Gray and Metcalfe, 1997). Ovotestes consisted of an anterior testicular portion and a posterior ovarian portion. Oocytes present in the ovotestes appeared to be at the same stage of development as oocytes in females from the same treatments, but oocytes of NP exposed fish were not compared to those of control fish. Nimrod and Benson (1998) exposed Japanese medaka to measured concentrations of NP from 0.54 to 1.93 mg l − 1 for the first month after hatching. The medium level (0.77 mg l − 1 NP) caused a significant decrease in percentage of females (42.6%) compared to controls (53.7%), but hermaphrodites were not found. Loss of NP from aquaria without fish (Gray and Metcalfe, 1997) was slower (half-life 36.5 h) than in our experiment (half-life 8.0 h). The rapid loss of NP in our study probably included uptake by the fish, although this was not measured. Surface volatilization and adsorption to particulate matter and glass probably contributed to the loss. The difficulty in maintaining concentrations of NP is evident from the measured concentrations of NP that were only about 20% of nominal concentrations in a flow-through system with continuous administration of NP (Nimrod and Benson, 1998). Although the nominal concentrations of NP in our study were higher than typically measured in sewage and industrial effluents (Talmage, 1994), actual average concentrations were considerably lower than the nominal concentrations, and the fish did not have a dietary exposure, which would be likely in a contaminated ecosystem. Therefore, persistent concentrations that are lower than the nominal concentrations used in our study could affect sexual differentiation and development of fish. Additional research is required to determine the environmental levels of NP that would have estrogenic effects. As previously reported (see review by ArcandHoy and Benson, 1998) and seen in the present
study, exposure to an estrogenic chemical at an early life stage can lead to alterations in key developmental processes, including sexual differentiation. The effects documented in the present study could have deleterious effects on a population exposed to concentrations of NP sufficiently high to cause these changes. Future investigations on the effect of NP on reproduction in R. marmoratus should include longer postexposure periods to determine if the observed effects persist.
Acknowledgements We thank Mary Delaney for her assistance with water sample preparation procedures, and Bruce Manning for his assistance with HPLC procedures and analyses. Helen Emory Young assisted with histological techniques and fish husbandry. Theodore Henry assisted with maintenance of experimental fish. Terra Knapp, Kristi Huels, Andrew Noyes, Anthony Jackson, and Lisa Scully maintained broodstock. Statistical advice was provided by Jason Osborne. This study was supported by the Southeastern Cooperative Fish Disease Project.
References Ahel, M., Giger, W., Koch, M., 1994. Behaviour of alkylphenol polyethoxylate surfactants in the aquatic environment — I. Occurrence and transformation in sewage treatment. Water Res. 28, 1131 – 1142. Arcand-Hoy, L.D., Benson, W.H., 1998. Fish reproduction: an ecologically relevant indicator of endocrine disruption. Environ. Toxicol. Chem. 17, 49 – 57. Arukwe, A., Celius, T., Walther, B.T., Goksøyr, A., 1998. Plasma levels of vitellogenin and eggshell zona radiata proteins in 4-nonylphenol and o, p%-DDT treated juvenile Atlantic salmon (Salmo salar). Mar. Environ. Res. 46, 133 – 136. Arukwe, A., Celius, T., Walther, B.T., Goksoyr, A., 2000. Effects of xenoestrogen treatment on zona radiata protein and vitellogenin expression in Atlantic salmon (Salmo salar). Aquat. Toxicol. 49, 159 –170. Bennie, D.T., 1999. Review of the environmental occurrence of alkylphenols and alkylphenol ethoxylates. Water Qual. Res. J. Can. 34, 79 – 122. Burkhardt-Holm, P., Wahli, T., Meier, W., 2000. Nonylphenol affects the granulation pattern of epidermal mucous cells in
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125 rainbow trout, Oncorhynchus mykiss. Ecotoxicol. Environ. Safety 46, 34 – 40. Cole, K.S., Noakes, D.L.G., 1997. Gonadal development and sexual allocation in mangrove killifish, Ri6ulus marmoratus (Pisces: Atherinomorpha). Copeia 1997, 596 –600. Donaldson, E.M., 1996. Manipulation of reproduction in farmed fish. Anim. Reprod. Sci. 42, 381 –392. Gray, M.A., Metcalfe, C.D., 1997. Induction of testis-ova in Japanese medaka (Oryzias latipes) exposed to p-nonylphenol. Environ. Toxicol. Chem. 16, 1082 –1086. Harrington, R.W. Jr., 1961. Oviparous hermaphroditic fish with internal self-fertilization. Science 134, 1749 –1750. Harrington, R.W. Jr., 1963. Twenty-four-hour rhythms of internal self-fertilization and of oviposition by hermaphrodites of Ri6ulus marmoratus. Physiol. Zool. 36, 25– 341. Jobling, S., Sheahan, D., Osborne, J.A., Matthiessen, P., Sumpter, J.P., 1996. Inhibition of testicular growth in rainbow trout (Oncorhynchus mykiss) exposed to estrogenic alkylphenolic chemicals. Environ. Toxicol. Chem. 15, 194 – 202. Kallman, K.D., Harrington, R.W. Jr., 1964. Evidence for the existence of homozygous clones in the self-fertilizing hermaphroditic teleost Ri6ulus marmoratus (Poey). Biol. Bull. 126, 101 – 114. Kinnberg, K., Korsgaard, B., Bjerregaard, P., Jespersen, A, ., 2000. Effects of nonylphenol and 17b-estradiol on vitellogenin synthesis and testis morphology in male platyfish Xiphophorus maculatus. J. Exp. Biol. 203, 171 –181. Koenig, C.C., Chasar, M.P., 1984. Usefulness of the hermaphroditic marine fish, Ri6ulus marmoratus, in carcinogenicity testing. Natl. Cancer Inst. Monogr. 65, 15 –33.
125
Maguire, R.J., 1999. Review of the persistence of nonylphenol and nonylphenol ethoxylates in aquatic environments. Water Qual. Res. J. Can. 34, 37 – 78. Nakamura, M., Kobayashi, T., Chang, X.T., Nagahama, Y., 1998. Gonadal sex differentiation in teleost fish. J. Exp. Zool. 281, 362 – 372. Nimrod, A.C., Benson, W.H., 1996. Environmental estrogenic effects of alkylphenol ethoxylates. Crit. Rev. Toxicol. 26, 335 – 364. Nimrod, A.C., Benson, W.H., 1998. Reproduction and development of Japanese medaka following an early life stage exposure to xenoestrogens. Aquat. Toxicol. 44, 141 – 156. Piferrer, F., Donaldson, E.M., 1992. The comparative effectiveness of the natural and a synthetic estrogen for the direct feminization of chinook salmon (Oncorhynchus tshawytscha). Aquaculture 106, 183 – 193. Selman, K., Wallace, R.A., Sarka, A., Qi, X., 1993. Stages of oocyte development in the zebrafish, Brachydanio rerio. J. Morphol. 218, 203 – 224. Servos, M.R., 1999. Review of the aquatic toxicity, estrogenic responses and bioaccumulation of alkylphenols and alkylphenol polyethoxylates. Water Qual. Res. J. Can. 31, 123 – 177. Soto, C.G., Leatherland, J.F., Noakes, D.L.G., 1992. Gonadal histology in the self-fertilizing hermaphroditic fish Ri6ulus marmoratus (Pisces, Cyprinodontidae). Can. J. Zool. 70, 2338 – 2347. Talmage, S.S., 1994. Environmental and Human Safety of Major Surfactants: Alcohol Ethoxylates and Alkylphenol Ethoxylates. Lewis Publishers, Boca Raton, FL. Taylor, D.S., Davis, W.P., Turner, B.J., 1995. Ri6ulus marmoratus: ecology of distributional patterns in Florida and the central Indian River Lagoon. Bull. Mar. Sci. 57, 202 – 207.
www.elsevier.com/locate/aquatox
Effects of nonylphenol on the gonadal differentiation of the hermaphroditic fish, Ri6ulus marmoratus Jennifer N. Tanaka, John M. Grizzle * Southeastern Cooperati6e Fish Disease Project, Department of Fisheries and Allied Aquacultures, Auburn Uni6ersity, Auburn, AL 36849, USA Received 22 August 2000; received in revised form 8 March 2001; accepted 13 March 2001
Abstract Nonylphenol (NP) is an estrogenic degradation product of alkylphenol polyethoxylate surfactants. In this study, the effects of NP on gonadal differentiation and development in Ri6ulus marmoratus (Osteichthyes, Cyprinodontiformes), a self-fertilizing, hermaphroditic species, were examined. Starting at hatching, fish were exposed to 150 or 300 mg l − 1 NP (nominal concentrations) in a static system with daily renewal. The measured concentration of NP in the test water decreased rapidly; half-life was 8.0 h. After 60 d of exposure to NP, fish were kept in uncontaminated water for 20 d and were then preserved for histological examination. No fish exposed to 300 mg l − 1 NP (N= 8) and only two of nine fish exposed to 150 mg l − 1 NP developed testicular tissue, compared with nine of 13 water-control fish and five of nine solvent-control fish. Oogenesis was also significantly inhibited by NP. None of the fish exposed to 300 mg l − 1 and only two of nine fish exposed to 150 mg l − 1 NP had vitellogenic oocytes, compared with seven of 11 water-control fish (not including males) and six of nine solvent-control fish. Dysplasia of the gonadal lumen also occurred in fish exposed to 300 mg l − 1 NP. These changes, including testicular agenesis, have not been previously reported in fish exposed to NP. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Nonylphenol; Fish; Histopathology; Gonadal differentiation; Vitellogenesis
1. Introduction 4-Nonylphenol (NP) is a degradation product of a group of nonionic surfactants called nonylphenol ethoxylates (NPE), which are used globally in the production of plastics, pesticides, and cleaning products, and are present in sewage effluents around the world (Talmage, 1994). * Corresponding author. Tel.: + 1-334-844-3474; fax: +1334-844-9208. E-mail address: [email protected] (J.M. Grizzle).
Nonylphenol is more persistent in the environment than the parent NPE (Maguire, 1999), and is found in surface waters, aquatic sediments, and groundwater (Talmage, 1994; Bennie, 1999). Recent research has identified NP as the most important degradation product of NPE because of its enhanced resistance toward biodegradation, toxicity, ability to bioaccumulate in aquatic organisms, and estrogenicity (Ahel et al., 1994). Nonylphenol is estrogenic in various aquatic animals (see reviews by Talmage, 1994; Nimrod and Benson, 1996; Servos, 1999). Experiments
0166-445X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S0166-445X(01)00186-2
118
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
with fish have shown that NP causes elevated plasma vitellogenin and zona radiata protein concentrations in both males and females, inhibition of spermatogenesis, induction of ovotestes in males, and altered gonadosomatic indices (Jobling et al., 1996; Gray and Metcalfe, 1997; Arukwe et al., 1998, 2000; Kinnberg et al., 2000). Histological changes in the epidermis of NP exposed rainbow trout Oncorhynchus mykiss resembled changes in estradiol-exposed fish but in addition, mucosomes in goblet cells were enlarged and irregularly shaped (Burkhardt-Holm et al., 2000). Ri6ulus marmoratus is a euryhaline fish that inhabits coastal mangrove swamps from Brazil through central Florida (Taylor et al., 1995). Adult R. marmoratus are usually hermaphrodites, but males also occur. This species is the only known vertebrate normally reproducing by internal self-fertilization (Harrington, 1961). The clones that result from self-fertilization, as well as its hardiness, make R. marmoratus a useful species for aquatic chemical testing (Koenig and Chasar, 1984). Thus far, the only studies related to gonadal differentiation in fish exposed to NP were with Japanese medaka, Oryzias latipes (Gray and Metcalfe, 1997; Nimrod and Benson, 1998). The objective of this preliminary study was to determine the effects of chronic exposure to NP on the gonadal differentiation of R. marmoratus.
2. Materials and methods
2.1. Chemicals HPLC-grade methanol, acetone, and running buffer water were purchased from Fisher Scientific (Pittsburgh, PA). p-Nonylphenol (95.5% pnonylphenol; 4.34% o-nonylphenol; 0.05% 2,4-dinonylphenol) was a gift from Schenectady International, Inc. (Schenectady, NY). HPLCgrade 4-tert-butylphenol was purchased from Fluka Chemical Corporation (Milwaukee, WI). HPLC-grade dichloromethane was purchased from Sigma–Aldrich (St. Louis, MO). Brackish water was made with distilled water and synthetic sea salts (HW Marine Mix; Hawaiian Marine, Inc., Houston, TX).
2.2. Animals 2.2.1. Nonylphenol exposure experiments Newly hatched R. marmoratus were obtained from the DS clone (Kallman and Harrington, 1964) maintained at the Department of Fisheries and Allied Aquacultures at Auburn University and were descendants of specimens obtained from the Gulf Ecology Division of the Environmental Protection Agency Office of Research and Development, Gulf Breeze, FL. Fish were held singly in 5 cm diameter stackable glass bowls containing 10 ml of brackish water with 12–14 parts per thousand (ppt) salinity. Fish were fed to satiation with Artemia nauplii 6 d per week. Fish were held at 21–25°C and exposed to a 12 h light:12 h dark photoperiod. 2.2.2. Nonylphenol persistence experiment Adult R. marmoratus (mean total length= 44 mm; mean wet weight=0.97 g) were maintained singly in 10 cm diameter stackable glass bowls with approximately 150 ml of brackish water (12– 14 ppt salinity). Fish were fed to satiation two to three times per week with Artemia nauplii. 2.3. Study design Fish were randomly selected to be exposed to one of the two concentrations of NP, methanol (solvent control), or the water used to prepare other treatments (water control). Exposures began within 9 h of hatching. A stock solution of 500 mg l − 1 NP was prepared in methanol; aliquots of this stock solution were added to brackish water (12–14 ppt) to produce nominal concentrations of 150 or 300 mg l − 1 NP. Solvent-control fish were exposed to nominal concentrations of 600 ml l − 1 methanol, corresponding to the highest concentration of methanol used in the NP treatments. The number of fish exposed to each treatment were: water control, 13; solvent control, 9; 150 mg l − 1 NP, 9; and 300 mg l − 1 NP, 8. All solutions were renewed 6 d per week (following feeding). For water changes, fish were individually aspirated into a glass pipet, the old solution was discarded, and the bowl was scrubbed with a nylon bristle brush. The bowl was
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
rinsed with distilled water followed by a rinse with the appropriate test solution, and the fish was then replaced in fresh solution with less than 1% transfer of the old solution. Fish were fed only on days when solutions were changed. All treatment and control exposures were from hatching until 60 d post-hatching, which exceeds the time period when gonadal differentiation of some fish species can be influenced by exogenous compounds (Nakamura et al., 1998). Fish were maintained in a fume hood during exposures. After exposure, fish were kept in 12– 14 ppt brackish water outside the fume hood for 20 additional days to reduce the chances that alterations observed were acute or temporary effects of NP. Feeding and water changes were also conducted 6 d per week during this postexposure period.
119
1 and 2). The prominent features of the mid cortical alveolus stage was a ring of cortical alveoli (3-15 mm diameter) at the periphery of the oocyte (Figs. 1 and 3) and the presence of the zona radiata. Oocytes in the late cortical alveolus stage had cortical alveoli (10– 20 mm in diameter) throughout most of the cytoplasm (Fig. 2). In fish examined during this study, only early indications of vitellogenesis were present because fish were only 80 d old. The distinguishing feature of vitellogenesis was eosinophilic, homogenous yolk bodies in the cytoplasm (Fig. 3). Cortical alveoli were still present during vitellogenesis and filled most of the cytoplasm.
2.4. Histopathology Fish were starved for 1 d before fixation in Bouin’s solution for 24– 48 h. After decalcification in RDO (Apex Engineering Products Corporation; Plainfield, IL), whole fish were embedded in paraffin, serially sectioned (7 mm) transversely, and stained with hematoxylin and eosin. Sections of gonads not available for histological examination seldom spanned more than 35 mm and never more than 84 mm. All sections of gonads were examined by light microscopy. Oogenesis in each fish was categorized by the most developed oocyte present. Oocyte stages were modified from Soto et al. (1992) and Selman et al. (1993). Stages used were: (1) perinucleolus; (2) early cortical alveolus; (3) mid cortical alveolus; (4) late cortical alveolus; and (5) vitellogenesis. Oocytes in the perinucleolus stage had strongly basophilic cytoplasm (Figs. 1– 3). Cortical alveolus oocytes, characterized by the presence of unstained cortical alveoli but no eosinophilic yolk, were separated into early, mid, and late stages. Oocytes in the early cortical alveolus stage had small cortical alveoli (1– 8 mm diameter) located about midway between the nuclear membrane and the oocyte margin (Figs.
Fig. 1. Transverse histological section of a gonad in a watercontrol Ri6ulus marmoratus 80 d after hatching. Hermaphrodite with testicular tissue (T), gonadal lumen (L), and oocytes. In this section, oocytes are in the perinucleolus stage (P), early cortical alveolus stage (E), and mid cortical alveolus stage (M). Bar = 70 mm.
120
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
NP, or if they caused NP to leach out of any materials within the HPLC system. Adult fish were held singly in brackish water for 12–48 h to determine whether the fish excreted compounds that obscured the presence of, or could be mistaken for, NP during HPLC analysis. Water samples were extracted and analyzed by HPLC.
2.5.2. Nonylphenol concentrations in treatments Because only 10 ml of solution was used for exposure of each R. marmoratus to NP, a separate study was conducted to determine the persistence
Fig. 2. Transverse histological section of a gonad in a hermaphroditic water-control Ri6ulus marmoratus 80 d after hatching. Testis was not present in the portion of the gonad illustrated. Oocytes in the late cortical alveolus (C), early cortical alveolus (E), and perinucleolus (arrow) stages, as well as the gonadal lumen (L). Bar = 70 mm.
2.5. Quantification of nonylphenol concentrations 2.5.1. Preliminary water and chemical analyses For analysis of NP, samples were spiked with butylphenol, extracted in dichloromethane, dried, and then reconstituted in methanol. A Spherisorb S5 ODS2 bonded phase column (4.6 mm × 25 × cm) (Phase Separations, Inc., Franklin, MA) and Waters 474 scanning fluorescence detector were used to detect NP. Concentrations of NP were calculated based on peak areas of NP standards. Distilled water, brackish water, and solventcontrol water were extracted and analyzed by HPLC to determine if the water sources contained NP, or if the extraction procedure caused NP to leach out of the equipment used during extraction. Methanol, dichloromethane, and butylphenol dissolved in methanol were analyzed directly by HPLC to determine if the solvents contained
Fig. 3. Transverse histological section of a gonad in a hermaphroditic water-control Ri6ulus marmoratus 80 d after hatching. Testis was not present in the portion of the gonad illustrated. Oocyte early in the vitellogenesis stage, which is characterized by eosinophilic yolk bodies (arrowheads) in the cytoplasm and an irregular nuclear outline. Cortical alveoli fill most of the cytoplasm. Mid cortical alveolus (M) and perinucleolus stage oocytes are also present. The lumen (arrow) is distorted because of the large size of the vitellogenic oocyte. Bar= 70 mm.
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
of NP in sample solutions. One adult R. marmoratus in 300 ml of water provided a ratio between fish weight and water volume similar to the ratio during NP exposure of 60 d old fish. Time-0 solutions were made and immediately processed for analysis by HPLC; concentrations of these samples were used to calculate recovery efficiency. Adult R. marmoratus were exposed to a nominal concentration of 300 mg l − 1 NP for 6, 12, or 24 h in 300 ml of solution, and the solutions were analyzed by HPLC.
2.6. Statistical analyses Fisher’s exact test was used to determine whether NP affected the presence or absence of testicular tissue or the appearance of the gonadal lumen. Each treatment (including the solvent control) was compared to the water control. A numerical score was assigned to each oocyte stage, and then Kendall’s coefficient of rank correlation (~) was used to assess the effect of NP exposure on the development of oocytes. Analysis of variance was used to determine if fish weights and lengths were different among treatments or were different for hermaphrodites and females. Differences were considered significant at PB 0.05.
3. Results
3.1. Degradation of nonylphenol during fish exposure Nonylphenol was not detected in the brackish water or in any of the solvents used in this study. Furthermore, a detectable amount of NP did not leach out of any of the materials used during the extraction and HPLC procedures. Compounds that could interfere with measurement of NP did not accumulate in water containing R. marmoratus for 12 –48 h. The recovery efficiency for NP was 84.4% (N= 6, S.E.M. =1.3). During exposures of adult fish to a nominal concentration of 300 mg l − 1 NP (ratio between fish weight and solution volume as for exposure of test fish), measured concentrations decreased from an initial concentration of 2559
121
Table 1 Gender of 80 d old Ri6ulus marmoratus after a 60 d exposure to nonylphenol that started at hatching. The number of fish of each gender is listed Treatment
Water control Solvent control NP (150 mg l−1) NP (300 mg l−1)
Gender Hermaphrodite
Male
Female
7 5 2 0
2 0 0 0
4 4 7 8
3.9 mg l − 1 (mean9 S.E.M., N=6) to 1269 6.2 mg l − 1 after 6 h, 86911.3 mg l − 1 after 12 h, and 339 6.4 mg l − 1 after 24 h. The predicted time for loss of half the NP from solution was 8.0 h.
3.2. Nonylphenol exposure No fish died during the 60 d exposure to NP or during the 20 d post-exposure period. Fish lengths and weights were not significantly different among treatments. Mean total length of fish was 15.9 mm (S.E.M.= 0.3) and mean weight was 38.6 mg (S.E.M.= 1.7).
3.2.1. Testicular tissue Males, females, and hermaphrodites were identified histologically when R. marmoratus were sampled 80 d after hatching (Table 1). Percentages of water-control (69%) and solvent-control (56%) fish that had testicular tissue (i.e. hermaphrodites or males) were not significantly different. Testicular tissue in hermaphroditic fish was located in the ventral portion of the gonad and adjacent to the gonadal lumen (Fig. 1). In 92% of control hermaphrodites, testicular tissue was found in only one gonad. Testicular tissue was present in only the posterior two-thirds of the gonad and was found in one to three discontinuous areas. In 75% of all hermaphrodites, testicular tissue was present in less than 10% of the gonad length. Spermatids or spermatozoa were seen in five of the seven water-control hermaphrodites, and in two of the five solvent-control hermaphrodites.
122
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
Two water-control fish were males; these fish had no ovarian tissue when examined 80 d after hatching (Fig. 4). Both males had all stages of spermatogenesis, including spermatozoa. Males were not present in other treatments. None of the fish exposed to 300 mg l − 1 NP and only two fish exposed to 150 mg l − 1 NP developed testicular tissue (Fig. 5). This was significantly different from controls (P =0.0027) for the higher level of NP but was only marginally significant (P=0.054) for the lower concentration of NP. The location of testicular tissue in NP exposed fish was the same as in control fish that were hermaphrodites. Spermatids and spermatozoa were present in one of the two hermaphrodites in the 150 mg l − 1 NP treatment.
Fig. 5. Transverse histological sections of gonads in Ri6ulus marmoratus 20 d after a 60 d exposure to a nominal concentration of 300 mg l − 1 nonylphenol. The most advanced oocytes were in the perinucleolus stage and no testicular tissue was present. The gonadal lumen (arrow) was greatly reduced and did not have the normal tri-lobed appearance. Bar = 30 mm.
Fig. 4. Transverse histological section of a testis in a watercontrol Ri6ulus marmoratus 80 d after hatching. Oocytes were not present in this fish. Bar = 50 mm.
3.2.2. Gonadal lumen All water-control (except for the two males), solvent-control, and 150 mg l − 1 NP treated fish had lumens that were tri-lobed in transverse section (Fig. 1). The tri-lobe configuration was sometimes distorted by large oocytes (Fig. 3), but was apparent in other parts of the gonad where less developed oocytes were present. Exposure to 300 mg l − 1 NP significantly (PB 0.0001) affected development of the gonadal lumen. Only one fish in this treatment had a normal gonadal lumen; all other fish in this treatment had gonadal lumens that were misshapen and reduced in size compared with the lumens of fish in other groups. The gonadal lumens of fish exposed to 300 mg l − 1 NP did not have a tri-lobed configuration or were quite small (Fig. 5).
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
3.2.3. Effects of nonylphenol on oogenesis Most control fish had vitellogenic oocytes (64% of water control and 67% of solvent control), and all control fish had oocytes that had developed to at least the early cortical alveolus stage (Table 2). Exposure to 150 mg l − 1 NP significantly inhibited oogenesis (Kendall’s ~ = − 0.56, P = 0.0062); only 22% of fish in this treatment had vitellogenic oocytes. Exposure to 300 mg l − 1 NP had a stronger negative effect on oocyte development (Kendall’s ~= − 0.93, P B0.00001); no fish in this treatment had oocytes beyond the early cortical alveolus stage, and 50% had no oocytes that had progressed beyond the perinucleolus stage.
4. Discussion Nonylphenol significantly inhibited the differentiation of testicular tissue in R. marmoratus, and this effect was dose dependent. Comparison of the size of fish in the different treatment groups indicated that the testicular agenesis in fish exposed to NP was not an effect of slower overall development. Inhibition of testicular differentiation by NP has not been previously reported in fish or other vertebrates. Exposure to endocrinemodulating chemicals at critical points during early development can permanently alter sex determination in some fish species (Donaldson, 1996), but longer-term studies are required to determine whether the testicular agenesis seen in R. marmoratus in the present study is permanent. Table 2 Oogenesis in 80 d old Ri6ulus marmoratus exposed to nonylphenol. The number of fish is given for each stage, based on the most advanced oocyte present. Oogenesis was significantly inhibited in both of the NP treatments Treatment
Water control Solvent control NP (150 mg l−1) NP (300 mg l−1)
Oocyte stagea P
ECA
MCA
LCA
V
0 0 0 4
1 1 4 4
1 2 3 0
2 0 0 0
7 6 2 0
a P, perinucleolus; ECA, early cortical alveolus; MCA, mid cortical alveolus; LCA, late cortical alveolus; V, vitellogenesis.
123
The most commonly reported action of NP has been stimulation of vitellogenesis, measured by the concentration of plasma vitellogenin. However, in the present study, the ovarian aspect of vitellogenesis and earlier phases of oogenesis were inhibited, which has not been previously reported in fish exposed to NP. Previous studies demonstrated that high dosages of natural and synthetic hormones can inhibit oogenesis. For example, Piferrer and Donaldson (1992) exposed newly hatched chinook salmon in a single immersion to 400 mg l − 1 (nominal concentration) 17b-estradiol or 17a-ethynylestradiol before sexual differentiation, and found that 20.8–40% (17b-estradiol treated) and 11.1– 34.7% (17a-ethynylestradiol treated) of the females produced had reduced cross-sectional gonadal area and underdeveloped oocytes compared to control fish. Most R. marmoratus exposed to the nominal concentration of 300 mg l − 1 NP had a dysplastic gonadal lumen, which was not noted in previous studies of NP. A lumen normally forms in a teleost ovary about the time of ovarian differentiation, and eventually serves as a passageway for ovulated eggs. In hermaphroditic R. marmoratus, fertilization takes place in this lumen (Harrington, 1963). It is possible that the mechanism(s) responsible for inhibition of oogenesis were also responsible for the retarded development of the gonadal lumen. Sexual differentiation in gonochoric species of fish can be altered by hormonally active chemicals (Donaldson, 1996), but because the normal sexual identity of each individual is difficult or impossible to determine, evaluation of changes in gonadal differentiation must consider the sex ratio within a population. In contrast, the absence of testis in any adult R. marmoratus is abnormal because all adults are either hermaphrodites or males. This important advantage of R. marmoratus was not utilized in our study; fish were examined before testicular differentiation had occurred in all of the control fish, and the presence of testis in only one gonad indicated that testicular development was not complete in the control hermaphrodites. Waiting until R. marmoratus are 120 d old, when sexual differentiation is complete (Cole and Noakes, 1997), before evaluation of results would
124
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125
improve the usefulness of R. marmoratus in future studies of testicular agenesis. In a previous study of gonadal differentiation in fish exposed to NP (nominal concentrations, 10–100 mg l − 1), ovotestis developed in Japanese medaka, and the sex ratio in the highest concentration of NP was skewed toward females (Gray and Metcalfe, 1997). Ovotestes consisted of an anterior testicular portion and a posterior ovarian portion. Oocytes present in the ovotestes appeared to be at the same stage of development as oocytes in females from the same treatments, but oocytes of NP exposed fish were not compared to those of control fish. Nimrod and Benson (1998) exposed Japanese medaka to measured concentrations of NP from 0.54 to 1.93 mg l − 1 for the first month after hatching. The medium level (0.77 mg l − 1 NP) caused a significant decrease in percentage of females (42.6%) compared to controls (53.7%), but hermaphrodites were not found. Loss of NP from aquaria without fish (Gray and Metcalfe, 1997) was slower (half-life 36.5 h) than in our experiment (half-life 8.0 h). The rapid loss of NP in our study probably included uptake by the fish, although this was not measured. Surface volatilization and adsorption to particulate matter and glass probably contributed to the loss. The difficulty in maintaining concentrations of NP is evident from the measured concentrations of NP that were only about 20% of nominal concentrations in a flow-through system with continuous administration of NP (Nimrod and Benson, 1998). Although the nominal concentrations of NP in our study were higher than typically measured in sewage and industrial effluents (Talmage, 1994), actual average concentrations were considerably lower than the nominal concentrations, and the fish did not have a dietary exposure, which would be likely in a contaminated ecosystem. Therefore, persistent concentrations that are lower than the nominal concentrations used in our study could affect sexual differentiation and development of fish. Additional research is required to determine the environmental levels of NP that would have estrogenic effects. As previously reported (see review by ArcandHoy and Benson, 1998) and seen in the present
study, exposure to an estrogenic chemical at an early life stage can lead to alterations in key developmental processes, including sexual differentiation. The effects documented in the present study could have deleterious effects on a population exposed to concentrations of NP sufficiently high to cause these changes. Future investigations on the effect of NP on reproduction in R. marmoratus should include longer postexposure periods to determine if the observed effects persist.
Acknowledgements We thank Mary Delaney for her assistance with water sample preparation procedures, and Bruce Manning for his assistance with HPLC procedures and analyses. Helen Emory Young assisted with histological techniques and fish husbandry. Theodore Henry assisted with maintenance of experimental fish. Terra Knapp, Kristi Huels, Andrew Noyes, Anthony Jackson, and Lisa Scully maintained broodstock. Statistical advice was provided by Jason Osborne. This study was supported by the Southeastern Cooperative Fish Disease Project.
References Ahel, M., Giger, W., Koch, M., 1994. Behaviour of alkylphenol polyethoxylate surfactants in the aquatic environment — I. Occurrence and transformation in sewage treatment. Water Res. 28, 1131 – 1142. Arcand-Hoy, L.D., Benson, W.H., 1998. Fish reproduction: an ecologically relevant indicator of endocrine disruption. Environ. Toxicol. Chem. 17, 49 – 57. Arukwe, A., Celius, T., Walther, B.T., Goksøyr, A., 1998. Plasma levels of vitellogenin and eggshell zona radiata proteins in 4-nonylphenol and o, p%-DDT treated juvenile Atlantic salmon (Salmo salar). Mar. Environ. Res. 46, 133 – 136. Arukwe, A., Celius, T., Walther, B.T., Goksoyr, A., 2000. Effects of xenoestrogen treatment on zona radiata protein and vitellogenin expression in Atlantic salmon (Salmo salar). Aquat. Toxicol. 49, 159 –170. Bennie, D.T., 1999. Review of the environmental occurrence of alkylphenols and alkylphenol ethoxylates. Water Qual. Res. J. Can. 34, 79 – 122. Burkhardt-Holm, P., Wahli, T., Meier, W., 2000. Nonylphenol affects the granulation pattern of epidermal mucous cells in
J.N. Tanaka, J.M. Grizzle / Aquatic Toxicology 57 (2002) 117–125 rainbow trout, Oncorhynchus mykiss. Ecotoxicol. Environ. Safety 46, 34 – 40. Cole, K.S., Noakes, D.L.G., 1997. Gonadal development and sexual allocation in mangrove killifish, Ri6ulus marmoratus (Pisces: Atherinomorpha). Copeia 1997, 596 –600. Donaldson, E.M., 1996. Manipulation of reproduction in farmed fish. Anim. Reprod. Sci. 42, 381 –392. Gray, M.A., Metcalfe, C.D., 1997. Induction of testis-ova in Japanese medaka (Oryzias latipes) exposed to p-nonylphenol. Environ. Toxicol. Chem. 16, 1082 –1086. Harrington, R.W. Jr., 1961. Oviparous hermaphroditic fish with internal self-fertilization. Science 134, 1749 –1750. Harrington, R.W. Jr., 1963. Twenty-four-hour rhythms of internal self-fertilization and of oviposition by hermaphrodites of Ri6ulus marmoratus. Physiol. Zool. 36, 25– 341. Jobling, S., Sheahan, D., Osborne, J.A., Matthiessen, P., Sumpter, J.P., 1996. Inhibition of testicular growth in rainbow trout (Oncorhynchus mykiss) exposed to estrogenic alkylphenolic chemicals. Environ. Toxicol. Chem. 15, 194 – 202. Kallman, K.D., Harrington, R.W. Jr., 1964. Evidence for the existence of homozygous clones in the self-fertilizing hermaphroditic teleost Ri6ulus marmoratus (Poey). Biol. Bull. 126, 101 – 114. Kinnberg, K., Korsgaard, B., Bjerregaard, P., Jespersen, A, ., 2000. Effects of nonylphenol and 17b-estradiol on vitellogenin synthesis and testis morphology in male platyfish Xiphophorus maculatus. J. Exp. Biol. 203, 171 –181. Koenig, C.C., Chasar, M.P., 1984. Usefulness of the hermaphroditic marine fish, Ri6ulus marmoratus, in carcinogenicity testing. Natl. Cancer Inst. Monogr. 65, 15 –33.
125
Maguire, R.J., 1999. Review of the persistence of nonylphenol and nonylphenol ethoxylates in aquatic environments. Water Qual. Res. J. Can. 34, 37 – 78. Nakamura, M., Kobayashi, T., Chang, X.T., Nagahama, Y., 1998. Gonadal sex differentiation in teleost fish. J. Exp. Zool. 281, 362 – 372. Nimrod, A.C., Benson, W.H., 1996. Environmental estrogenic effects of alkylphenol ethoxylates. Crit. Rev. Toxicol. 26, 335 – 364. Nimrod, A.C., Benson, W.H., 1998. Reproduction and development of Japanese medaka following an early life stage exposure to xenoestrogens. Aquat. Toxicol. 44, 141 – 156. Piferrer, F., Donaldson, E.M., 1992. The comparative effectiveness of the natural and a synthetic estrogen for the direct feminization of chinook salmon (Oncorhynchus tshawytscha). Aquaculture 106, 183 – 193. Selman, K., Wallace, R.A., Sarka, A., Qi, X., 1993. Stages of oocyte development in the zebrafish, Brachydanio rerio. J. Morphol. 218, 203 – 224. Servos, M.R., 1999. Review of the aquatic toxicity, estrogenic responses and bioaccumulation of alkylphenols and alkylphenol polyethoxylates. Water Qual. Res. J. Can. 31, 123 – 177. Soto, C.G., Leatherland, J.F., Noakes, D.L.G., 1992. Gonadal histology in the self-fertilizing hermaphroditic fish Ri6ulus marmoratus (Pisces, Cyprinodontidae). Can. J. Zool. 70, 2338 – 2347. Talmage, S.S., 1994. Environmental and Human Safety of Major Surfactants: Alcohol Ethoxylates and Alkylphenol Ethoxylates. Lewis Publishers, Boca Raton, FL. Taylor, D.S., Davis, W.P., Turner, B.J., 1995. Ri6ulus marmoratus: ecology of distributional patterns in Florida and the central Indian River Lagoon. Bull. Mar. Sci. 57, 202 – 207.