Effects of 17alpha-ethynylestradiol on early-life development, sex differentiation and vitellogenin induction in mummichog (Fundulus heteroclitus).

Abstract

Fertilized mummichog eggs retrieved from 17alpha-ethynylestradiol (EE(2)) exposed adult fish were raised in concentrations of EE(2) ranging from 0 to 100 ng/L (100 ng/L EE(2) estimated to have actual average exposure concentrations of 30% of nominal; 0.1-10 ng/L were below detect throughout 24-h exposure period) for 61 weeks post-hatch. Eggs exposed at 100 ng/L hatched sooner, the larvae were longer, and survival of juvenile fish from hatch to study termination was greater than all other treatments, though fewer hatched at this treatment. Sex ratios were skewed (>80% female phenotype) at 100 ng/L EE(2), and some gonadal male fish displayed female secondary sex characteristics. Condition factor, gonadosomatic index (GSI), and liver somatic index (LSI) were found to decrease in both sexes between 52 and 61 weeks post-hatch. Female fish had increased hepatic vitellogenin (VTG) at 52 weeks post-hatch. When exposed to 1, 10 and 100 ng/L EE(2), female fish had a higher proportion of vitellogenic follicles in the ovarian tissue. Males exposed at 100 ng/L may have had disruption at some endpoints (GSI, VTG) that is masked due to reduced sample size compared to other treatments. Fish exposed to concentrations of EE(2) at or below 10 ng/L showed inconsistent effects on development and reproductive potential. This study indicates the potential for population-level effects at the high range of environmental EE(2) at concentrations equivalent to those at which consistent effects in fecundity in the adult mummichog reproductive test have been measured. This work demonstrates that chronic EE(2) exposure causes developmental effects at concentrations similar to those which cause effects in the shorter-term adult mummichog reproductive test. Effects are at higher concentrations than have been noted for freshwater model species. Whether this is because of species sensitivity or due to differences between freshwater and saltwater availability of EE(2) or its uptake requires further study.

Full text

Page 1

Accepted Manuscript
Effects of 17α-ethynylestradiol on early-life development, sex differentiation
and vitellogenin induction in mummichog (Fundulus heteroclitus)
Rebecca E.M. Peters, Simon C. Courtenay, Mark L. Hewitt, Deborah L.
MacLatchy
PII: S0141-1136(09)00128-7
DOI: 10.1016/j.marenvres.2009.10.002
Reference:MERE 3377
To appear in:
Marine Environmental Research
Received Date:27 December 2007
Revised Date:6 September 2009
Accepted Date: 7 October 2009
Please cite this article as: Peters, R.E.M., Courtenay, S.C., Hewitt, M.L., MacLatchy, D.L., Effects of 17α-
ethynylestradiol on early-life development, sex differentiation and vitellogenin induction in mummichog (Fundulus
heteroclitus), Marine Environmental Research (2009), doi: 10.1016/j.marenvres.2009.10.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
peer-00565110, version 1 - 11 Feb 2011
Author manuscript, published in "Marine Environmental Research 69, 3 (2010) 178"
DOI : 10.1016/j.marenvres.2009.10.002

Page 2

ACCEPTED MANUSCRIPT

1
Effects of 17α-ethynylestradiol on early-life development, sex differentiation and 1
vitellogenin induction in mummichog (Fundulus heteroclitus) 2
3
Rebecca E. M. Peters1, Simon C. Courtenay2, Mark L. Hewitt3 and Deborah L. 4
MacLatchy1* 5
6
1Department of Biology and Canadian Rivers Institute 7
University of New Brunswick 8
Saint John, NB 9
Canada 10
E2L 4L5 11
12
2Fisheries and Oceans Canada at the Canadian Rivers Institute 13
Department of Biology, University of New Brunswick 14
Fredericton, NB 15
Canada 16
E3B 6E1 17
18
3National Water Research Institute 19
Environment Canada 20
Burlington, ON 21
Canada 22
L7R 4A6
23
peer-00565110, version 1 - 11 Feb 2011

Page 3

ACCEPTED MANUSCRIPT

2

24
*Author to whom correspondence should be addressed. 25
Email: dmaclatchy@wlu.ca 26
27
Present address: 28
Department of Biology and Canadian Rivers Institute 29
Wilfrid Laurier University 30
Waterloo, ON 31
Canada 32
N2L 3C5 33
Phone: 519 884-0710 ext 2221 34
Fax: 519 746-2472 35
peer-00565110, version 1 - 11 Feb 2011

Page 4

ACCEPTED MANUSCRIPT

3
Abstract
36
Fertilized mummichog eggs retrieved from 17α-ethynylestradiol (EE2) exposed 37
adult fish were raised in concentrations of EE2 ranging from 0 to 100ng/L (100ng/L EE2
38
estimated to have actual average exposure concentrations of 30% of nominal; 0.1 to 39
10ng/L were below detect throughout 24-h exposure period) for 61 weeks post-hatch. 40
Eggs exposed at 100ng/L hatched sooner, the larvae were longer, and survival of juvenile 41
fish from hatch to study termination was greater than all other treatments, though fewer 42
hatched at this treatment. Sex ratios were skewed (>80% female phenotype) at 100ng/L 43
EE2, and some gonadal male fish displayed female secondary sex characteristics. 44
Condition factor, gonadosomatic index (GSI), and liver somatic index (LSI) were found 45
to decrease in both sexes between 52 and 61 weeks post-hatch. Female fish had 46
increased hepatic vitellogenin (VTG) at 52 weeks post-hatch. When exposed to 1, 10 and 47
100ng/L EE2, female fish had a higher proportion of vitellogenic follicles in the ovarian 48
tissue. Males exposed at 100ng/L may have had disruption at some endpoints (GSI, 49
VTG) that is masked due to reduced sample size compared to other treatments. Fish 50
exposed to concentrations of EE2 at or below 10ng/L showed inconsistent effects on 51
development and reproductive potential. This study indicates the potential for 52
population-level effects at the high range of environmental EE2 at concentrations 53
equivalent to those at which consistent effects in fecundity in the adult mummichog 54
reproductive test have been measured. This work demonstrates that chronic EE2 55
exposure causes developmental effects at concentrations similar to those which cause 56
effects in the shorter-term adult mummichog reproductive test. Effects are at higher 57
concentrations than have been noted for freshwater model species. Whether this is 58
peer-00565110, version 1 - 11 Feb 2011

Page 5

ACCEPTED MANUSCRIPT

4
because of species sensitivity or due to differences between freshwater and saltwater 59
availability of EE2 or its uptake requires further study. 60
peer-00565110, version 1 - 11 Feb 2011

Page 6

ACCEPTED MANUSCRIPT

5
Key Words 61
Ethynylestradiol, Fundulus heteroclitus, vitellogenin, sex differentiation, development, 62
lifecycle, bioassay, endocrine disruption 63
peer-00565110, version 1 - 11 Feb 2011

Page 7

ACCEPTED MANUSCRIPT

6
Introduction
64
Endocrine disrupting substances (EDSs) are exogenous substances or mixtures 65
that alter function of the endocrine system and consequently cause adverse effects in 66
individuals or their progeny (OECD, 1999). Effects of pulp mill effluents in Canada 67
(Hewitt et al. 2008) and sewage in Britain (Jobling and Tyler 2003) provide some of the 68
best examples of endocrine disruption in wild fish. Recently, in a whole-lake experiment, 69
chronic exposure of fathead minnow to low concentrations (5-6ng/L) of 17α-70
ethynylestradiol (EE2) caused feminization and intersex in males, induced vitellogenesis, 71
altered oogenesis in females, and caused a near collapse of the population (Kidd et al. 72
2007). EDSs may interfere with any part of endocrine control, including production, 73
release, transport, metabolism, receptor binding, action or elimination. The sum total of 74
these effects can be measured in fish laboratory bioassays designed to assess 75
reproduction, development and growth (Patyna et al., 1999; Zillioux et al., 2001; Parrott 76
and Wood, 2002; Seki et al., 2004). Chronic exposure to xenobiotics can induce changes 77
to organisms not noted by shorter-term exposures (Parrott and Blunt, 2005), including 78
factors that may be manifested at the population level, such as lower recruitment (Ankley 79
et al., 2001). 80
Lifecycle bioassays or multigenerational bioassays have been developed in 81
freshwater fish species, including the fathead minnow, Pimephales promelas (Ankley et 82
al., 2001; Länge et al., 2001), Japanese medaka, Oryzias laripes (Seki et al., 2003) and 83
zebrafish, Danio rerio (Olsson et al., 1999). Partial lifecycle and short-term bioassays 84
have been developed for estuarine and marine species; examples include sheepshead 85
minnow, Cyprinodon variegates (Folmar et al., 2000; Zillioux et al., 2001; Hemmer et 86
peer-00565110, version 1 - 11 Feb 2011

Page 8

ACCEPTED MANUSCRIPT

7
al., 2008) and mummichog, Fundulus heteroclitus (MacLatchy et al., 2003; Boudreau et 87
al., 2004; Peters et al. 2007; Bosker et al., 2009). Effects of EDSs on species living in 88
salt water may be quite different than those in fresh water because of differences in 89
biological availability of the contaminants based on physical water chemistry and fish 90
physiology (e.g., in contaminant uptake). Therefore, studies focused on the effects of 91
EDSs on estuarine species are warranted, especially given the extent of coastal activity by 92
humans and the importance of estuaries as spawning and nursery grounds (Oberdörster 93
and Cheek, 2001). 94
Mummichog are the numerically dominant fish species in salt marshes along the 95
east coast of Canada and the United States (Armstrong and Child, 1965), and have 96
demonstrated sensitivity to EDSs in laboratory (MacLatchy et al., 2003; Peters et al., 97
2007), artificial stream (Dubé et al., 2002), and field (Leblanc et al., 1997) assessments. 98
The mummichog is a good candidate for lifecycle bioassay development due to its size, 99
ease of husbandry and ability to manipulate its reproductive cycles. Mummichog are 100
relatively sedentary, exhibiting small home ranges in the wild (Skinner et al. 2005) and 101
they are potentially exposed to environmental EDSs throughout their life cycle. 102
EE2 has been chosen as a model EDS for developing bioassays due to its 103
environmental relevance as well as its confirmed effects on the reproductive endocrine 104
system via estrogen receptor-mediated pathways (OECD, 1999; Ankley et al., 2001; 105
Metcalfe et al., 2001; MacLatchy et al., 2003). EE2 is one component of sewage effluent 106
associated with increased incidence of intersex in male fish exposed downstream of 107
sewage treatment plants (STPs) (Jobling and Tyler, 2003). It is a synthetic 108
pharmaceutical (birth control pill and hormone replacement therapy) that is not broken 109
peer-00565110, version 1 - 11 Feb 2011

Page 9

ACCEPTED MANUSCRIPT

8
down in sewage treatment; concentrations of EE2 present in Canadian STPs are usually 110
between 1-10ng/L EE2, although levels have been documented as high as 42ng/L 111
(Desbrow et al., 1998; Ternes et al., 1999). In earlier short-term (7- or 15-day) exposure 112
studies using EE2, adult mummichog displayed endocrine impacts at low, 113
environmentally-relevant concentrations, as well as similar responses at higher 114
pharmaceutical concentrations (MacLatchy et al., 2003). In longer-term (21- or 28-day) 115
EE2 exposures, reproductive cycling was shifted in females, sex steroid production and 116
circulating levels were altered and at environmentally-relevant (approximately 20% of 117
nominal 100ng/L exposures), fecundity and fertility were reduced (Peters et al., 2007). 118
The objective of this study was to determine the impact of chronic EE2 exposure 119
on offspring development. Embryos derived from mummichog parents exposed during 120
pre-spawning and spawning phases to EE2 were continuously exposed to EE2 for 15 121
months, through their development to pre-spawning juveniles/adults. Embryonic/larval 122
endpoints (time to hatch, hatch success, length at hatch), larval/juvenile endpoints 123
(growth, survival, vertebral abnormalities) and yearling endpoints (liver vitellogenin, 124
gonad and liversomatic indices, condition factor, sex ratios) were evaluated for 125
anomalies. This study, in conjunction with our previous studies (MacLatchy et al., 2003, 126
Boudreau et al., 2004; MacLatchy et al., 2005; Boudreau et al., 2005; Sharpe et al., 2004; 127
Peters et al., 2007) furthers our ability to understand the effects of EDSs on various life 128
stages of mummichog. This study demonstrates that developmental stages of 129
mummichog are sensitive to EE2 at exposure levels similar to those that interfere with 130
reproduction in adult mummichog (Peters et al. 2007), and that the concentrations at 131
peer-00565110, version 1 - 11 Feb 2011

Page 10

ACCEPTED MANUSCRIPT

9
which effects occur are higher than those noted in freshwater model species (Länge et al. 132
2001; Andersen et al., 2003). 133
134
Materials and Methods
135
Chemicals
136
The 17α-ethynylestradiol (EE2; 98% purity) was purchased from Sigma-Aldrich 137
Canada (Oakville, ON, Canada). EE2 was stored at -20°C in 100% ethanol (Les Alcools 138
de Commerce, Boucherville, QC, Canada) at stock concentrations of 3ng/mL, 30ng/mL, 139
300ng/mL and 3000ng/mL EE2 for adult exposures and 10ng/mL, 100ng/mL, 1000ng/mL 140
and 10000ng/mL EE2 for larval and juvenile exposures. Unless otherwise indicated, 141
chemicals and reagents were purchased from Sigma-Aldrich and laboratory supplies from 142
Fisher Scientific (Nepean, ON, Canada). 143
144
Experimental Conditions 145
Collection and breeding protocols for adult mummichog used in this study have 146
previously been described (Peters et al., 2007). Offspring were maintained at the same 147
exposure conditions as their parental groups at 0, 0.1, 1, 10, or 100ng/L EE2, in static 148
conditions with daily water changes and treatment renewal for the 61-week study period. 149
As developing embryos, larvae, fry and juveniles grew, photoperiod was adjusted to 150
simulate seasonal day length: 16:8 h light:dark at initiation (July); 14:10 h light:dark at 151
week 14 (October); 12:12; light:dark at week 20 (November); 14:10 light:dark at week 32 152
(February); 15:9 light:dark at week 40 (April); and 16:8 light:dark from week 46 (May) 153
to termination of the experiment. Temperature was held at room temperature, which 154
peer-00565110, version 1 - 11 Feb 2011

Page 11

ACCEPTED MANUSCRIPT

10
decreased slightly from 18-21˚C in the summer (weeks 1-20) to 16-18˚C for the winter 155
(weeks 21-40), returning to 18-21˚C for the remainder of the study. 156
Larvae were fed live, newly hatched Artemia sp. nauplii (Bohai Bay Salt Ponds 157
Artemia Cysts, Aquatic Ecosystems, Apopka, FL, USA) enriched with Roti-rich™ 158
(Aquatic Ecosystems) twice daily (1 mL of concentrated Artemia per L of water) and Fry 159
Food (Rolf C. Hagen, Montreal, QC, Canada) to satiation once daily for 14 weeks. 160
Beginning at 8 weeks, freeze-dried Red Grubs (Rolf C. Hagen) were used to supplement 161
the Fry Food diet as the juveniles were weaned from the Artemia. Flaked Staple Food 162
(Rolf C. Hagen) was introduced as the primary feed at 22 weeks, fed 2-3 times daily, 163
supplemented by Red Grubs or Cichlid Food (Rolf C. Hagen) once daily. 164
165
Exposures
166
Naturally-spawned, fertilized mummichog eggs were collected from adults 167
exposed to nominal exposure concentrations of 0, 0.1, 1, 10, or 100ng/L EE2 for 21 and 168
28-days as previously reported (Peters et al., 2007). The fertilized eggs were transferred 169
to glass Petri dishes at an initial density of 30 eggs per dish and held in 50 mL of 20‰ 170
salinity EE2-treated water. Each dish was examined daily at 4X magnification, and dead 171
embryos and hatched larvae were removed. The water was then removed from each dish 172
and replaced with 50 mL of water treated with the appropriate amount of EE2. Time to 173
hatch, survival to hatch and length at hatch were monitored over the hatching period. 174
Upon hatch, larvae were maintained in 50mL beakers containing the appropriate 175
EE2 concentrations and held at a maximum density of 10 larvae per beaker. Once the 176
yolk sac was absorbed (1-3 days) and swimming began, the larvae were randomly 177
peer-00565110, version 1 - 11 Feb 2011

Page 12

ACCEPTED MANUSCRIPT

11
allocated to one of four aerated, static 37-L aquaria per treatment. Daily water renewals 178
were done by completely replacing the water and adding new treatment solution in each 179
tank after removing fish with a dip net and placing them in a temporary holding 180
aquarium; fish were returned to their original treatment tanks following the water change. 181
Upon initiation of the growout phase, each aquarium contained 5 L of EE2-treated 20‰ 182
saline water (dissolved oxygen>80% saturation). At five weeks, the volume of water in 183
each aquarium was increased to 10 L to accommodate growing larvae. Beginning at 10 184
weeks, volumes of water were adjusted separately for each tank to minimize the effect of 185
density differences among the tanks due to differential survival. Water volume was set to 186
1 L per 1 g total wet weight of the fish in the aquarium and was adjusted every 3-4 weeks 187
for the remainder of the experiment. Each tank was replaced 100% daily throughout the 188
exposure period. Due to a dosing error in week 14, replicate tanks for 0ng/L and 100ng/L 189
EE2 were reduced to three tanks for subsequent endpoints. 190
Growth was determined weekly for the initial seven-week period, and then 191
approximately every three weeks for the remainder of the experiment by collecting total 192
length measurements (to the nearest mm) from 25 fish removed, measured and returned 193
to each aquarium. Survival was determined from the number of fish in each tank on each 194
measurement day, and calculated as the proportion of the original number of fish in the 195
tank. Vertebral abnormalities were assessed at weeks 15, 48 and 61 and ranged from 196
mild (one or two slight bends in the spine) to severe (one or more bends in the spine that 197
dramatically altered body shape and/or affected swimming ability) and included scoliosis 198
(lateral curvature) and lordosis (dorsoventral curvature) (Boudreau et al., 2004). 199
peer-00565110, version 1 - 11 Feb 2011

Page 13

ACCEPTED MANUSCRIPT

12
At week 48 (May 2004), all fish were assessed for weight (to 0.001g), standard 200
and total length (mm), vertebral abnormalities and external sex. Sex was determined 201
based upon colouration and secondary sex characteristics including yellow bellies, dark 202
dorsal fin spot, silver vertical stripes, white or yellow spots in males, and brown or green 203
backs and pale bellies in females. At week 52 (22 June 2004), less than a week following 204
the full moon, 20 fish were randomly sampled from each aquarium except one 1ng/L 205
aquarium and one 100ng/L aquarium where 0 and 15 fish were selected respectively, due 206
to low numbers of fish in these tanks. These fish were anaesthetized (buffered 0.05% 207
tricaine methane sulfonate; Syndel Laboratories, Vancouver, BC, Canada), assessed for 208
weight (to 0.001g), standard and total length (mm), vertebral abnormalities and sex based 209
upon secondary characteristics and gonadal assessment at 4X magnification. 210
Vitellogenin (VTG) in the liver was assessed at week 52 using an ELISA 211
modified from MacLatchy et al. (2003). Liver tissue was dissected out, weighed, frozen 212
on dry ice, and then stored at -80˚C. For each mg of liver tissue, 1 µL of aprotinin 213
(1KIU/µL) was added to the microfuge tube. Tissue was thawed and homogenized with a 214
Kontes Pellet Pestle hand-held homogenizer. A small amount of the slurry (5 µL) was 215
diluted 100X with Tris buffered saline containing Tween and bovine serum albumin and 216
assayed as per MacLatchy et al. (2003). Prior to analysis, the method was tested by 217
spiking liver slurry samples with a known concentration of VTG. Recovery of the spiked 218
amount of VTG [recovered VTG = 0.549 + 1.071 * (spiked VTG); R2=0.774; p<0.001] 219
was 88%. Interassay variability was 7% and intra-assay variability was 6%. 220
Gonadal tissue collected at week 52 was fixed in 10% buffered formalin and 221
stored in 70% ethanol. Tissues were embedded in paraplast and sectioned at 7 µm using 222
peer-00565110, version 1 - 11 Feb 2011

Page 14

ACCEPTED MANUSCRIPT

13
a rotary microtome. Tissue sections were stained with Mallory stain and examined with a 223
light microscope at 10X magnification for female tissue and 40X magnification for male 224
tissue. For each fish that had sufficient tissue, six fields of view (at different depths into 225
the gonad) were assessed for intersex and to determine developmental stage at week 52. 226
Using the terminology for developmental stage of Blazer (2002), the number of 227
previtellogenic and vitellogenic follicles were counted in females and the developmental 228
stage(s) of the testes were determined for each field of view. 229
At week 61 (August 2004), four days prior to the full moon, all remaining fish 230
were sampled and assessed for weight (0.001g), standard and total length (mm), vertebral 231
abnormalities and sex based upon secondary characteristics and gonadal assessment at 232
4X magnification. At 52 weeks, the fish were not reproductive and by 61 weeks were 233
beginning to regress; therefore, a breeding trial could not be carried out and the 234
experiment was terminated. 235
236
EE2 Analysis
237
Analysis of the concentration of EE2 in the water was unavailable for this study 238
due to insufficient water volumes collected for analysis. Therefore, a subsequent two- 239
week exposure of adult mummichog (with equivalent volume of water and mass of fish 240
as per juvenile exposures) was conducted in duplicate aquaria at nominal concentrations 241
of 0 and 100ng/L EE2. Volumes of water extracted and analyzed were 1L. EE2 levels 242
were extracted and analyzed as previously described (Peters et al., 2007) with additional 243
method standardization to ensure EE2 recovery in salt water. The water concentrations 244
for the 100ng/L EE2 treatment aquaria were highest at the point of delivery (74.7ng/L) 245
peer-00565110, version 1 - 11 Feb 2011

Page 15

ACCEPTED MANUSCRIPT

14
and decreased with time to 66.8ng/L at 1 h and 30.7ng/L at 6 h; at 12 and 24 h EE2 levels 246
were undetectable (detection limit at 10ng/L). Because of the inability to determine 247
exposure concentrations at exposures <10ng/L nominal, all exposure concentrations are 248
given as nominal values.
249
250
Statistics
251
Statistical analyses were conducted using SigmaStat 3.0 (SPSS, Chicago, IL, 252
USA) or Systat 9.0 (Systat Software Inc., Richmond, CA, USA). Significant differences 253
(p<0.05) among the data were assessed separately for each sex when it was possible to 254
determine gonadal sex. Prior to parametric analysis, assumptions of normality and 255
variance homogeneity were tested on morphological data using normal probability plots 256
and Levene’s test. Where the data failed to meet the assumptions, they were log10 257
transformed and the assumptions retested. If the data still failed to meet the assumptions, 258
an equivalent non-parametric test was used. Measures of time to hatch, hatch success 259
rate, and length at hatch were assessed using single factor ANOVA, using Petri dishes as 260
the replication units. Mortality, vertebral abnormalities and liver VTG were assessed 261
using single factor ANOVA, with aquarium as the unit of replication. Vitellogenin data 262
were examined for outliers using Dixon’s test prior to analysis. Analyses of weight and 263
length were performed using a nested analysis of variance (ANOVA) followed by 264
Tukey’s or Dunn’s tests, with aquarium as the factor nested within treatment group. 265
Kruskal-Wallis was used when the data failed to meet parametric assumptions. 266
Differences in gonadal and liver weights relative to body weights (gonadosomatic index; 267
GSI; [gonad wt/body wt]·100; and liversomatic index; LSI; [liver wt/body wt]·100), body 268
peer-00565110, version 1 - 11 Feb 2011