| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Toxicology |
Department of Zoology, Institute of Biological Sciences, University of Aarhus, DK-8000 Aarhus C, Denmark
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
behavior, fertilization, male sexual function, sperm, steroid hormones
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Despite these numerous effects of EDCs on fish from molecular to tissue and organ levels, very little evidence supports any direct effects of these chemicals on individual fertility, which might indicate serious population decline in the future. In the laboratory, fish reproduction has been reduced by sex hormones or their pharmaceutical analogues [14–16], octyl phenol [14–17], the aromatase-inhibitor fadrozole [18], and the pesticide methoxychlor, which is known to have estrogenic properties [14]. However, these laboratory studies only found evidence of reduced reproductive success at the highest concentrations tested, and two of these studies noted that effects on reproduction required exposure to more than eightfold the concentration causing effects at lower levels of biological organization, such as vitellogenesis or gonadal histology [15, 16]. Field data indicating effects of endocrine disruption on fish fertility are even more sparse, with only a single study showing a positive correlation between the incidence of ovotestes and reduced male fertility [1].
Although the majority of research into the effects of EDCs on fish has concerned chemicals suspected of emulating natural estrogens, it is now abundantly clear that reproductive physiology can be disrupted through a variety of mechanisms. One group that has received relatively little attention are chemicals with an anti-androgen action. Vinclozolin (3-(3,5-dichlorophenyl)-5-methyl-5-vinyloxazolidine-2,4-dione) is a dicarboximide fungicide that, in the field, is degraded to two metabolites, M1 (3-[[[3,5-dichlorophenyl]-carbamoyl]oxy]-2-methyl-3-butenoic acid) and M2 (3'5'-dichloro-2-hydroxy-2-methylbut-3-enanilide). Whereas vinclozolin itself is not persistent, these two metabolites have half-lives of more than 180 days and are likely to be highly mobile in the water phase [19]. Vinclozolin is rapidly metabolized in mammals into the same two metabolites, which have high affinity for the androgen receptor and can block gene expression, causing anti-androgenic effects [20, 21]. Both M1 and M2 have been shown to have a high affinity for androgen receptors in some fish [22]. Previous studies in our laboratory have shown that vinclozolin inhibits the development and maintenance of a variety of male traits in the guppy (Poecilia reticulata), including the development of male coloration patterns, sperm counts, and performance of sexual behavior [6, 7].
The aim of the present study was, first, to determine whether the effects of vinclozolin on the male guppy's secondary sexual characteristics translate into deleterious effects on reproduction and, second, to attempt to distinguish paternal and maternal influences on any possible reduction in reproduction.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The experimental animals were wild-type guppies (P. reticulata) originating from Columbia in 1997 and since maintained at our laboratory in 500-L, stainless-steel stock aquaria. These aquaria were kept at 25 ± 2°C and received water from an aerated header tank containing a mixture of deionized water, produced by a reverse-osmosis system, and chlorine-free tap water at a ratio of 5:1. Furthermore, 450 mg NaCl/L header tank water were added, resulting in a pH of 7.3 and a constant conductivity of 600 µS/cm. This water was used in all aquaria described below. The stock population was fed daily with newly hatched artemia and commercial fish feed (TetraMinRubin, TetraMin, and TetraPhyla; Tetra Werke, Melle, Germany).
Three experiments were performed in the present study to distinguish reproductive effects of vinclozolin (Riedel-de-Haen AG, Seelze, Germany) mediated through effects on the male sexual phenotype from effects on females before insemination and effects exerted during pregnancy. In the first two of these experiments, newly hatched individuals were removed daily from the stock aquaria refuges and placed in an 80-L, seamless glass aquarium (Struers KEBO lab A/S, Copenhagen, Denmark) containing a gravel-filled jar through which water was recycled. These juveniles were observed daily for the first appearance of sex characteristics, at which time they were separated by sex into two identical but smaller aquaria (25 L), allowing the collection of virgin females for experimentation, which was necessary because the guppy can retain sperm for a number of clutches from a single mating [23].
Vinclozolin Exposure
In all three experiments, guppies were exposed to vinclozolin through their food using acetone as the carrier solute in stock solutions. Food was prepared containing 0.1, 1.0, or 10.0 µg vinclozolin/mg dry food. Food for the controls was prepared in the same way using acetone alone. During exposure, the guppies received an average of 18 mg dry fodder per gram fish (start wet wt) per day, resulting in average nominal doses of 1.8, 18, and 180 mg vinclozolin/kg. Average start weights were determined by randomly removing 10 fish from each group of virgin males and females and gravid females.
Virgin Female Exposure Experiment
Seventy-five virgin females (age, 10 wk) were randomly chosen from the virgin female aquarium, and 15 were placed into each of five seamless glass aquaria (34.5 x 29 x 28 cm) containing water to a depth of 15 cm and a gravel filter jar as described above. Two of these aquaria were controls; fish in the remaining three aquaria received 0.1, 1.0, and 10.0 µg vinclozolin/mg dry food, respectively, for 30 days, during which time feces and uneaten food were removed daily by suction. At this time, the fish were distributed in groups of three fish from the same treatment into 23 seamless glass aquaria (28.5 x 20 x 22 cm) containing water to a depth of 7 cm, a gravel filter jar, and a stainless-steel net refuge (diameter, 8 cm) so that all aquaria contained three virgin females. Eight of these aquaria served as controls, whereas all other treatments where replicated five times. Following a day of acclimation, an adult male from the stock aquaria was added for 7 days for insemination, after which the adult male was removed. These aquaria were monitored three times daily for 45 days for newly hatched offspring, which were removed to avoid adult cannibalism.
Pregnant Female Exposure Experiment
Fully grown, adult females were randomly removed from the stock aquaria and placed in groups of three into 23 seamless glass aquaria as described above. Eight of these aquaria served as controls; all other treatments (receiving 0.1, 1.0, and 10.0 µg vinclozolin/mg dry food, respectively) were replicated five times. Exposure was continued over 45 days, during which time the aquaria were monitored three times daily for newly hatched offspring, which were removed to avoid adult cannibalism. Feces and uneaten food were removed daily by suction.
Male Exposure Experiment
Seventy-five virgin males (age, 10 wk) were randomly distributed in five seamless glass aquaria (34.5 x 29 x 28 cm) containing water to a depth of 15 cm and a gravel filter jar. Two of these aquaria were controls; the remaining three received 0.1, 1.0, and 10.0 µg vinclozolin/mg dry food, respectively, for 30 days, during which time feces and uneaten food were removed daily by suction. Subsequently, the males were kept for a further day in these aquaria, during which time they received uncontaminated fodder to reduce the risk of contaminating the females used in the reproduction experiment with vinclozolin from their feces. Twenty-three aquaria (28.5 x 20 x 22 cm) were set up containing water to a depth of 7 cm, a gravel filter jar, a stainless-steel net refuge (diameter, 8 cm), and three virgin females (age, 14 wk). Eight of these aquaria were designated to receive control males; the other 15 were designated for the three doses of vinclozolin treatment. Twenty-three males were randomly removed from the appropriate treatment and placed singly into the virgin female aquaria, to allow insemination, for 7 days. The average first-clutch size per aquarium (three females) was measured by inspecting the aquaria three times daily during the following 45 days. Newly hatched offspring were removed to avoid adult cannibalism.
Following their 7 days with females, the intensity of male courtship behavior of these 23 males was measured individually against an adult, pregnant female. Pregnant females were chosen because they are nonreceptive to copulatory attempts by males [24], hence maximizing the consistency of the female response. Two females were used alternately for all males. Following this recording of the male's sexual behavior, a sperm count was performed on each of the males.
Sperm Counting Procedure
The male was lightly anesthetized in an aqueous solution of 200 µg/ml of Ethyl-4-aminobenzoate (Sigma-Aldrich, Vallensboek Strand, Denmark) for approximately 1 min and gently placed on a clean glass plate. The gonopodium was carefully teased to a 70° angle (relative to the abdominal surface), and 10 µl of 175 mM KCl were added in the angle between the gonopodium and the body of the fish. The guppy was stripped of sperm according to the method described by Toft and Baatrup [24], and sperm were counted according the general guidelines for counting human sperm [25].
Quantification of Male Courtship Behavior
Normal adult male guppies are extremely active in their courtship of females, performing a highly stereotyped and characteristic swimming pattern toward adult females often more than once a minute during the daytime. This swimming pattern, which is termed sigmoid behavior because of the male's adoption of the characteristic tensed body shape resembling the Greek letter sigma during courtship, has been intensely studied and described in detail [26]. Because of its stereotyped nature, it has been possible to develop computerized image-analysis software (DISPLAY ver 3.01; Institute of Biological Sciences, University of Aarhus, Denmark) capable of recognizing this behavior from video sequences [6].
A fully grown, nonreceptive female was placed in a seamless glass aquarium (28.5 x 20 x 22 cm), which was sand-blown to eliminate reflections, containing 3 cm of water from the stock aquarium. This water was chosen because of the suspected presence of pheromones released during birth, which are stimulatory to male courtship behavior [27, 28]. Nonreceptive females were chosen to maximize the similarity of the female response to different males during measurement. Adult female guppies are receptive to copulation for 3 days after giving birth to a clutch of young [29], and nonreceptive females can be chosen from stock by finding obviously gravid individuals. The male was placed in the aquarium for a 15-min acclimation period before recording the courtship behavior for 10 min.
The courtship behavior of the 23 males whose reproductive capability was measured as described above was quantified in this manner. Because of our experience with high variance in this endpoint, caused in part by occasional males in all groups who failed to perform courtship behavior during the recording period, the courtship behavior of a further 36 identically treated males (15 control males and 7 males in each exposure group) was quantified.
Statistical Analysis
Differences between treatment groups for all parameters were tested using one-way ANOVA. When treatment differences were detected, differences between the controls and vinclozolin treatments were tested using the Dunnett post-hoc test with SPSS software (SPSS for Windows, release 10.0; SPSS, Inc., Chicago, IL). In the case of courtship behavior, differences between the two groups of males (experiment with 23 males and, subsequently, with 36 males) were tested using the Mann-Whitney test. Because no significant difference was detected between these groups (P = 0.398), data were pooled for further analysis. Correlations between reproductive output and sexual behavior or sperm count were investigated using the Pearson correlation test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
No mortality was observed in any of the treatments during experimentation. Exposure of young adult males to the two highest vinclozolin doses for 30 days had a significant (F3,19 = 4.22, P = 0.019) negative effect on the size of the first clutch produced by the unexposed virgin females they inseminated. Furthermore, the effect of vinclozolin followed a standard dose-response pattern (Fig. 1a). Thus, uncontaminated males sired, on average, 3.3 offspring per female, and males exposed to the lowest dose sired 2.4 offspring per female (not significantly different from controls). However, the group exposed to the intermediate and high doses sired 1.4 and 1.1 offspring per female, respectively. In contrast, exposure of young adult, virgin females to vinclozolin, which were subsequently paired with unexposed males, resulted in only a slight reduction in first-clutch size, and in no case were the clutches of these exposed females significantly smaller (F3,19 = 0.294, P = 0.829) than those of controls (Fig. 1b). Similarly, no significant treatment effect (F3,19 = 1.407, P = 0.272) was observed on the clutches produced during gravid exposure (Fig. 1c).
|
Sperm Counts
The 30-day exposure of young adult males to vinclozolin caused a highly significant (F3,51 = 17.1, P < 0.001) reduction in the number of sperm cells in the provoked ejaculate, and the magnitude of reduction in sperm count followed a dose-response curve (Fig. 2). Thus, control males produced an ejaculate containing an average of 3.02 million sperm cells, which in males receiving the highest vinclozolin dose was reduced by almost two-thirds (to 1.06 million sperm cells). In addition to this reduction in the number of sperm cells in the ejaculate of vinclozolin-exposed males, a high proportion of the spermatozoa of exposed males were immobile, a result that, unfortunately, was not quantified.
|
Because the same males were used in the reproduction test and in the subsequent tests of male sexual behavior and sperm counts, it was possible to test for correlations between these aspects of male fitness and the reproductive outcome. A significant correlation (Pearson correlation coefficient R = 0.45, adjusted R2 = 0.162, P = 0.033) was found between the sperm count and the number of juveniles subsequently produced by the inseminated females (Fig. 3).
|
Sigmoid Display
Two nonreceptive females from the stock aquarium were used alternately to stimulate male sexual behavior in the test arena. No significant difference was found in the male response to the two females, and the data for the two females were therefore pooled within treatment groups.
The level of sigmoid display by the males to these nonreceptive females was also negatively affected, in a dose-response manner, by vinclozolin (Fig. 4). Hence, control males performed an average of 0.96 displays per minute toward the female, whereas the males that received the highest dose performed only 0.35 displays per minute. However, this parameter was less sensitive to vinclozolin than either female fecundity or the male's sperm count, and it was only significantly different from the controls at the highest vinclozolin dose. The number of sigmoid displays was correlated neither to the sperm count (R = 0.275, P = 0.22) nor to the clutch size (R = -0.033, P = 0.885).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The reduction in reproductive output with increasing vinclozolin dose was paralleled by reduced sperm counts in males. It is often assumed that the energetic cost of sperm production in fish is negligible compared with the magnitude of the female investment and that males therefore have a massive overproduction of sperm. In this context, there may be little need for concern about fish populations with reduced sperm counts. However, Warner [33] argued that male sperm production might well place limitations on male fitness in promiscuous fish species. The guppy male is particularly sexually active, with a large number of attempted daily copulations [24]. Therefore, it can be argued that the reduced sperm production of up to 63% measured in the present study made a direct contribution to the reduction in the size of the first clutch of juveniles. In addition, the sperm count measured after the male had inseminated the virgin female was significantly correlated to size of the first clutch, indicating that the reduced sperm production in males contributed to the reduced fertility. This correlation should be treated with caution, however, because the measured sperm count does not necessarily represent the sperm reserves that are available to the male immediately before the copulation that resulted in fertilization of the female. It was noted, although unfortunately not quantified, that a considerable number of the sperm cells in the vinclozolin-treated males were entirely immobile, which may well have had considerable significance in the reduced male fertility caused by this anti-androgen.
It has previously been shown in our laboratory that exposure of adult male guppies from the same strain, but of unknown age (range, 3–18 mo), to the same doses of vinclozolin in food had no significant effect on sperm count [6]. In contrast, exposure of juveniles from hatching to 26 wk of age [7] caused a reduction in sperm count similar to that found in the present study. In addition, the males from the adult exposure study [6] were dissected and the testis examined histologically [34]. No histological aberrations were detected, even at the highest dose of 100 µg vinclozolin/g food. The fish in the present study were exposed from 10 to 14 wk, indicating that the window of sensitivity for the effects of anti-androgens, such as vinclozolin, is still open during this period. As noted by Evans et al. [35], little is known about the ontogeny of the various sexual traits in the guppy. Those authors studied the incidence of sexual behavior, sperm counts, and development of the male coloration in guppies of 7–10 wk of age, which they termed juveniles; at 13 wk; and again at 26 wk. They found juvenile males at 7–10 wk of age had adult coloration and had started performing sexual display, but 17 of the 18 juvenile males had sperm counts of zero. In addition, the study revealed that all three traits continued to develop throughout the 26 wk and that the sperm count reached an average 2.4 million per ejaculate at that time, which is of an order of magnitude similar to the 3.0 million sperm cells per ejaculate in the controls of the present study. Thus, testicular development in the guppy is still incomplete during the juvenile period (7 and 10 wk) and early adult age, and during this time, it is still sensitive to the action of anti-androgen chemicals.
The tendency of males to perform sigmoid display toward a nonreceptive female was also affected by vinclozolin exposure, although to a lesser extent than the other two parameters. Only the highest exposure dose caused a significant reduction in the number of displays performed per minute. Previous studies of the effect of vinclozolin on guppies in our laboratory have shown that the intensity of guppy behavior is more readily affected if exposure proceeds from birth [7]. Thus, exposure of juveniles to a dose of only 0.1 µg vinclozolin/mg food was enough to eliminate the sigmoid display later in adulthood. In contrast, when exposed during adult life [6], a significant reduction was seen in animals receiving 1 µg/mg but not 100 µg/mg.
In studies during which several males are observed in the company of a single female, the female will copulate with the male performing the most and the longest sigmoid displays [36]. In the present study, only a single male was present, both during fertilization of the virgin females, which are known to be extremely eager to copulate [23], and during the measurement of sigmoid display with the nonreceptive females. Competition between males was therefore not an element in the present experiment. In this light, it is hardly surprising that we found no correlation between the intensity of an individual male's sexual behavior and his success in fertilizing a receptive virgin female. Furthermore, it has previously been shown that the male guppy advertises his fertility through his coloration and his sexual behavior such that the intensity of his sexual behavior is correlated to the number of sperm cells per ejaculate [37]. In the present study, we could find no evidence of this type of correlation, and the difference may well be found in the lack of male competition in the experimental protocol used.
Preliminary evidence from our laboratory supports the hypothesis that testosterone is the primary androgen in the guppy; hence, from this point of view, the guppy bears closer similarity to higher vertebrates, such as rats, which use testosterone as their primary androgen, than to many commercial fish species, which use 11-keto androstenedione [32]. The doses used in the present study were in the same range as or lower than those used during studies in which pregnant rats were fed between vinclozolin at between 100 and 200 mg/kg, resulting in a variety of malformations in the sexual characteristics of male offspring [38, 39]. In the present study, the fish received an estimated average dosage of between 1.8 and 180 mg vinclozolin/kg, and significant effects were seen at more than 18 mg vinclozolin/kg. However, an analysis of the hazard posed by vinclozolin to field populations of fish is not possible at present because of the lack of knowledge of the environmental behavior of vinclozolin itself and, especially, its two active metabolites, M1 and M2, which are likely to bear direct responsibility for the anti-androgenic effects. As noted by U.S. Environmental Protection Agency [19], the information for these two metabolites is extremely sparse, but in terrestrial field studies, vinclozolin itself exhibited half-lives of between 34 and 94 days whereas its metabolites had half-lives of between 179 and 1000 days, resulting in an availability in surface waters of several weeks to months after application. Possible deleterious effects of vinclozolin on fish populations therefore cannot be ruled out at present.
|
FOOTNOTES |
|---|
2 Correspondence: Department of Zoology, Institute of Biological Sciences, University of Aarhus, Building 135, Ole Worms Allé, Dk-8000 Aarhus C, Denmark. FAX: ++45 86125175; mark.bayley{at}biology.au.dk
Received: 9 April 2003.
First decision: 24 April 2003.
Accepted: 12 August 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Hamers, J. H. Kamstra, E. Sonneveld, A. J. Murk, M. H. A. Kester, P. L. Andersson, J. Legler, and A. Brouwer In Vitro Profiling of the Endocrine-Disrupting Potency of Brominated Flame Retardants Toxicol. Sci., July 1, 2006; 92(1): 157 - 173. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R. Propper The Study of Endocrine-Disrupting Compounds: Past Approaches and New Directions Integr. Comp. Biol., January 1, 2005; 45(1): 194 - 200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kristensen, E. Baatrup, and M. Bayley 17{alpha}-Ethinylestradiol Reduces the Competitive Reproductive Fitness of the Male Guppy (Poecilia reticulata) Biol Reprod, January 1, 2005; 72(1): 150 - 156. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
