v04 . .rftlwx :- , s. v > . . h
n. I:

I ‘IIO . l. «7-K: . fink-x . Ill" .l
"$33 L» .4 dwfiv. .
an . .9... 1 3.

H .. V:
\btfiadhu . .
It . . n» w $1.0 o
N~3 3%.: \t V
-o....a;h.x..v....uv «ham.
F . .

W0. hfinfiflfifi§i . .. .. . . .. . .. \3WWwehmxnmew..ruu..ucw

. t .3. - . . . 33... , . . . . , - ..$rfl..n.i.. %
.1...4.w.a..m$ .m»«....b¢. 4.. emit. V . . _ .. . .. J»
9. run. .......!3V...ulhnl.1.§. gut... . S...“ .I.‘ : wt"... .. , . .. . . . . .U .v! .1 I
q .. taoh. (‘3:an .. . .. ‘ “H4 4‘ . ,.. . . . .. .
I. . . . . . .

a... ' . on V

.
kw .. . ....: .s.vd.r.u../..N.. “and: . 2.3‘ WM 93:
kWh. Miran". FIV .1. suflfinunmm .. .. . V ..
.h ”fink. .. $4... 1.». . - a . 4..
.. . I ».. 2). l u. . V . .. v . ..h.....o ., ‘.
figmmxfiki. . .. . . . . .. . . .qnam..-..»mmwn.ws.. ..V..
V .. . s 435...”? .X.” .. .w S}, : l......

» I .
4V"). - t
.5 . .
v

2 b5»- .
. mannh; u... . . . n......lrx
. «Vim.....1\...w..w.héwum......w.r I .
. u 0‘ is. u”.1u~.4av no“ .
.‘w..u«hui.buon.
a» .353.-.
. §U.III ‘ till I .
. _.....s..uno....v3r4.... 8&5? .. , , ..
.. 1‘»- .‘ov‘ 0|: I». .
, , . . . g .Dumvn. . . . .L~M+P..fiun.u.v.r$
1.3;} .V . .. .. . .Hnwmaé .
on . .. up... .\ Mo..- .
$.53... an: .. v -V 12...)...”
3 V» adA‘fiay annual.-
...L ... .II._.. or. {£3.29} a.»
nu... .55.. . . $3.339.
. x... - J»...
Jul? .v
I: '9

m

'2
LI

3
n?

.uun v.....u.u.t... .v...«..sh...hvm.
(Illlv I coltnvthhdzalet
u. . A r. .flfian.~(....l.l 1 .2. . . .
.\l| : 17. 1.411 lo. 1‘..- _
..l ‘.Hf. J $t|ocH .4 00‘ Ian
l...n‘J\A.Ir. [.Ivflfllwpi'
1.... .o 3.! .v‘ '3!!! 1...
103x... L)Il[~w i... u. .
... . I ...cl!...-:N»...ob.}.~
551 .d......!. .ol . 211 9.9....

:u. 5.
t. ..o .-
luffunlulumwn.

u".[o 1'0!‘

l :.s.}u.'.b to co.u.roo
. . .
. .

. g « J. ‘V‘D‘Vn...
.y:‘tnv.nnlly |IDFQyn..
u Han}-.. golfing .

.1t Ii. 00..

bLVhIOI-III '70.
J. Al...‘! .3

V . VI
n 0 o.A| ‘ 0‘11]!
“dammit-

r 15.... {.91 . c...“ I. ..... Y . q! 1c!(~...flt....l...uo‘...hn....h‘ .
N. J 6.13»? p...?.|:..ll. .:..t.. . . .....t..... w: .
FavT I) .. . In 1: . «ole: ..m..u..lnbn.3 t It'll... In... t
vor 1 4 .

Ute

 

. . . y... I . -t
I 1.4...Pd...... zruucai.
3.. mask Hanna.

9. .P. 0......

tiltiwq.
... E

1 ... . . .vtrt....h.. It. ..

_ Itnlu ‘rtlllenII

53H“.
L. H

22;!

“-r’

.- . N‘a..>
.uvrivnlp. .1110qu I‘rU L I V n y .
.nl.l .‘ .. .Io. 1|.“
. t a .

 

THESIS

t’f‘

lilllllfllUIIIIIIIIHHHIllllllllllHUItllHUllUlHlHlll

3 1293 0184190

This is to certify that the

thesis entitled

EFFECTS OF SUSPECT ENVIRONMENTAL ENDOCRINE
DISRUPTERS ON THE REPRODUCTIVE PHYSIOLOGY
OF FATHEAD MINNOWS, Pimephales promelas

 

presented by

Krista M. Nichols

has been accepted towards fulfillment
of the requirements for

M.S. degree in Fisheries & Wildlife

 

 

calm-s

Major Mom:

 

0-7639 MS U is an Aflirnma've Action/Equal Opportunity Institution

'v “w .' .‘I‘-vw v..—~' . *— w—V‘w

__ fi_.__ _ __ ._ _..__—...

'1

fi—fi- —.‘r if;

._._ H vww—

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m clowns-p44

 

Effects of Suspect Environmental Endocrine Disrupters on the Reproductive Physiology
of Fathead Minnows, Pimephales promelas

by

Krista M. Nichols

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Fisheries and Wildlife

1 997

ABSTRACT

EFFECTS OF SUSPECT ENVIRONMENTAL ENDOCRINE DISRUPTERS ON THE
REPRODUCTIVE PHYSIOLOGY OF FATHEAD MINNOWS, Pimephales promelas

BY

Krista M. Nichols

Male and female adult fathead minnows, Pimephales promelas, were exposed to
nonylphenol ethoxylate (NPEO) in the laboratory and to wastewater treatment plant
effluent (WWT P) in situ. NPEO exposure to environmentally relevant concentrations
elicited no concentration-dependent response for fecundity, plasma vitellogenin
concentrations, plasma estradiol or testosterone concentrations, or estradiol to
testosterone ratios. Laboratory NPEO exposure indicates that the relative risk of effects
on the reproductive physiology at concentrations less than 10 pg NPEOIL is low. Caged
fathead minnows exposed for three weeks below WWTP did not exhibit elevated male
plasma concentrations of vitellogenin. Thus, no estrogenic activity in the effluents is
indicated. Differences observed between 2 reference sites were likely due to site
physical characteristics and stress induced by cage confinement at the riverine site.
Differences among a riverine reference site and exposure sites were observed in female
vitellogenin, male and female plasma estradiol and testosterone, and estrogen to

androgen ratios.

Copyright by
Krista Marie Nichols
1997

ACKNOWLEDGEMENTS

The research contained herein would not have been possible without the
guidance and support of many people. I would like to thank Dr. John Giesy for serving
as the chair of my guidance committee and for the years of support and opportunity
given to me for professional development. Thanks to Dr. Tom Coon and Dr. Richard Hill
for serving as members to my graduate committee and for the time and insight offered
to the research at hand.

My research would not have been possible without the collaboration and
instruction of many researchers and acknowledgments for specific components of my
research are made in the following chapters. The MSU Aquatic Toxicology Laboratory
has provided me with a breadth of experience and knowledge from interactions among
peers with a variety of interests; thanks to Erin Snyder, Vince Kramer, Stephanie Miles-
Richardson, Sue Pierens, Shane Snyder, Lori Feyk, Dave Verbrugge, Dan Villeneuve,
Alex Villalobos, Alan Blankenship, and Thomas Sanderson for their guidance, support,
and cooperation. A special thanks to the undergraduates who often kept the lab running
smoothly and who assisted me frequently in the lab and field: Jason Griffith, Dan
Tuntevski, Lancie Dole, Rebecca Zmyslo, and Rose Cory. Dr. Glen Van Der Kraak at
the University of Guelph, Department of Zoology, has given his expertise in fish
physiology and endocrinology on countless occasions, and I am deeply grateful for his

insight and collaboration.

Throughout the past three years, my research has been generously supported
by fellowships from the Department of Fisheries and Wildlife and SC. Johnson Wax and
Co. Additional support has been provided by the Chlorine Chemistry Council of the
Chemical Manufacturer’s Association, the USEPA Office of Water, the National Institute
of Environmental Health Sciences, and the Institute for Environmental Toxicology.

Finally, I would like to thank my family for their patience, support, and
understanding throughout my graduate education and across the many miles. Much of
my success is owed to the support and patience of many close friends who have offered
me solace from daily frustrations, advice and consultation on my research, and

companionship.

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................... ix
LIST OF FIGURES ........................................................................................................ xi
INTRODUCTION ............................................................................................................. 1
Background lnfonnation ............................................................................................... 1
Endocrine disruption in feral fish populations ............................................................... 3
Biomarkers of estrogen exposure in fishes .................................................................. 5
Research objectives .................................................................................................... 7
REFERENCES ............................................................................................................ 7

CHAPTER 1. EFFECTS OF NONYLPHENOL ETHOXY LATE (NPEO) EXPOSURE ON
REPRODUCTIVE OUTPUT AND BIOINDICATORS OF ESTROGEN EXPOSURE IN

 

FATHEAD MINNOWS, PIMEPHALES PROMELAS - . _ _ .......... 11
INTRODUCTION ....................................................................................................... 1 1
MATERIALS AND METHODS ................................................................................... 14

Fish ........................................................................................................................ 14
Induction, separation, and purification of VTG ........................................................ 15
VTG polyclonal antibody production ....................................................................... 16
VTG ELISA optimization ......................................................................................... 18
Nonylphenol ethoxylate exposure .......................................................................... 19
NPEO water sampling ............................................................................................ 21
Sample collection ................................................................................................... 21
VTG competitive ELISA ......................................................................................... 22
Plasma steroid hormone measurements ................................................................ 24
Statistical methods .......................................................................................... - ....... 26
RESULTS .................................................................................................................. 27
Vite/logenin antigen characteristics ........................................................................ 27
VTG antibody specificity ......................................................................................... 28
VTG assay validation ............................................................................................. 28
Survival and fish health .......................................................................................... 30
NPEO water quality and chemistry ......................................................................... 30
Fecundity ............................................................................................................... 31
Plasma VTG ........................................................................................................... 32
Plasma sex steroid hormones ................................................................................ 33
Relationships between biomarkers of exposure and fish reproductive health ......... 36
DISCUSSION ............................................................................................................ 37
ACKNOWLEDGEMENTS .......................................................................................... 43

vi

REFERENCES .......................................................................................................... 44

CHAPTER 2. EFFECTS OF MUNICIPAL WASTEWATER EXPOSURE IN SITU ON
THE REPRODUCTIVE PHYSIOLOGY OF THE FATHEAD MINNOW, PIMEPHALES

 

 

PROMELAS-u ....... -- - - ..-.....63
INTRODUCTION ....................................................................................................... 63
MATERIALS AND METHODS ................................................................................... 66

Fish ........................................................................................................................ 66
Study site selection ................................................................................................ 66
Exposure ................................................................................................................ 67
Sample collection ................................................................................................... 68
Vitellogenin enzyme-linked immunosorbent assay (ELISA) .................................... 69
Plasma sex steroid hormone analysis .................................................................... 69
Statistical analyses ................................................................................................. 70
RESULTS .................................................................................................................. 71
Study site water quality characteristics ................................................................... 71
Fish survival and condition ..................................................................................... 72
Plasma VTG concentrations ................................................................................... 73
Plasma sex steroid hormone levels ........................................................................ 74
Relationships among biomarkers ........................................................................... 77
DISCUSSION ............................................................................................................ 78
ACKNOWLEDGMENTS ............................................................................................ 87
REFERENCES .......................................................................................................... 87

SUMMARY AND RECOMMENDATIONS ................................................................... 105
Recommendations for future evaluation .................................................................. 106

APPENDIX.

LABORATORY PROTOCOL. Induction, Purification, Identification, and Production

of Antibodies for Goldfish (Carassius auratus) Vitellogenin ------- -- 108
SCOPE .................................................................................................................... 108
SUMMARY OF PROTOCOL ................................................................................... 109
SIGNIFICANCE AND USE ...................................................................................... 109
PROTOCOL: Induction of vitellogenin in goldfish ................................................... 110
PROTOCOL: Separation of vitellogenin by HPLC ................................................... 111
PROTOCOL: SDS-PAGE for detection of vitellogenin ............................................ 113
PROTOCOL: Western immunoblotting for detection of vitellogenin ........................ 115
PROTOCOL: \frtellogenin antibody production ....................................................... 118
PROTOCOL: Slot-blot for detection of specific and nonspecific immune response in

immunized rabbits .............................................................................................. 120
REFERENCES ........................................................................................................ 122

vii

STANDARD OPERATING PROCEDURE. MEASUREMENT OF PLASMA
VITELLOGENIN IN FATHEAD MINNOWS AND GOLDFISH BY COMPETITIVE

 

ENZYME-LINKED IMMUNOSORBENT ASSAY (ELISA) ------ - - - ...... 125
SCOPE .................................................................................................................... 125
REFERENCES ........................................................................................................ 127
SUMMARY OF METHOD ........................................................................................ 128
SIGNIFICANCE AND USE ...................................................................................... 128
EQUIPMENT AND APPARATUS ............................................................................ 129
MATERIALS ............................................................................................................ 129
ELISA BUFFERS AND REAGENTS ........................................................................ 130
HAZARDS AND PRECAUTIONS ............................................................................ 131
PROCEDURE .......................................................................................................... 131

viii

LIST OF TABLES

Table 1. Oligomer distribution of nonylphenol ethoxylate (Surfonic N-95)
standard. Values are expressed as percent area by normal phase high-
perforrnance liquid chromatography ........................................................................ 52

Table 2. Plasma VTG for male and female fathead minnows exposed to 0, 0.3,
1, 3, and 10 pg NPEOIL for six weeks. Means (iSEM), ranges, and percent,
of observations greater than the detection limit (MDL) are presented for all
treatments. (T RTMT=ug NPEO/L). ........................................................................ 54

Table 3. Plasma concentrations of E2 and T and ratios of E2 to T (EZIT) in male
and female fathead minnows exposed to NPEO (mean 1 SEM). ............................ 57

Table 4. Correlation coefficients (r) for females (4A) and males (48) between
biomarkers of estrogenicity and indicators of general health and fecundity of
fathead minnows exposed to 0-10 pg NPEO/L. E2 and T concentrations
were log-transformed and coefficients between these, hematocrit, and E2fl'
were parametric Pearson product-moment correlations. All other
correlations were based upon ranks and are Speannan rank correlation
coefficients. (E2=estradiol, T=testosterone, VTG=vitellogenin,
'=Speannan’s Rho coefficients) ............................................................................. 61

Table 5. Characteristics of WWTP effluents at 7 study sites in mid-Michigan,
USA, for evaluation of effluent effects on reproductive physiology of fathead
minnows. ................................................................................................................ 93

Table 6. Ambient river and gross effluent water quality data from WWTP field
sites June-August 1996. Effluent water data represent numbers from
Discharge Monitoring Reports required by NPDES permits. All data are
averages unless otherwise indicated. ..................................................................... 94

Table 7. Survival and condition factors (K) for male and female fathead
minnows exposed in situ to WWTP effluent in mid-Michigan, June-August
1996. MeanstSEM are presented. ........................................................................ 95

Table 8. Plasma VTG concentrations and percent greater than the method
detection limit (MDL) as measured in the VTG ELISA for fathead minnow
exposed in situ to WWTP effluent, June-August 1996. Exposure was for 3
weeks. Values in parentheses are SEM. (N.D. = not detectable). .......................... 96

Table 9. Concentrations of E2 and T, and ratios of E2fT in male and female
fathead minnows exposed in situ to WWTP effluent. Means are reported :I:
SEM ........................................................................................................................ 99

Table 10. Correlation coefficients (r) between traditional biomarkers of
environmental endocrine disruption and between biomarkers and condition
factor (K) for fathead minnows exposed in situ to WWTP effluent, June-July
1996. Values for hormones and EZIT ratios were Iogw-transformed. Both
Pearson product moment correlation and Spearman’s Rho determinations
were made dependent upon the distribution of the data and are denoted in
the tables. ............................................................................................................. 103

Table A1. Schedule for immunization and blood collection for vitellogenin
antibody production in New Zealand white rabbits by ULAR. ................................ 124

LIST OF FIGURES

Figure 1. Anion exchange high performance liquid chromatogram for the
separation of VTG from E2-induced male and female goldfish. Arrow
indicates the VTG protein eluted from the column with a 0 to 0.50 M Tris-CI
linear gradient. Detection of the proteins was measured by absorbence at
wavelengths of 230, 254, and 280 nm. ................................................................... 49

Figure 2. SDS-PAGE Western immunoblot with rabbit anti-goldfish VTG
polyclonal antibodies. Protein bands represent E2 induced goldfish plasma
(Eng), male goldfish (mgf), female goldfish (fgf), and purified goldfish VTG
fractions (#2-26 and #26). Proteins were separated by continuous SDS-
PAGE (4-1 5% Tris-tricine). ..................................................................................... 50

Figure 3. VTG standard curve and dilution curves of goldfish and fathead
minnow male and female plasma for determination of curve parallelism.
Note that x-axis is the log transformed concentration of VTG (ng/mL) for
standard goldfish VTG or dilution factor of plasma samples. (+ = standard
GF VTG; O = E2 induced goldfish; * = female goldfish; U = female fathead
minnow (recrudesced); O = female fathead minnow (not recrudesced); A =
male goldfish; X = male fathead minnow) .............................................................. 51

Figure 4. Fecundity of fathead minnows exposed to NPEO. Data are means :t
standard errors for replicate blocks of NPEO treatments. Values in
parentheses above histograms represent the number of spawning events
contributing eggs to total fecundity. No differences in egg production were
observed among treatments. One replication of 3 pg NPEO/L was lost
completely to fungal infection and no females remained in one replication of
10 pg NPEO/L. ....................................................................................................... 53

Figure 5. Plasma VTG concentrations (pglmL) in fathead minnows exposed to
nonylphenol ethoxylate. A. Mean plasma VTG concentrations in females for
each NPEO exposure. B. Mean plasma VTG for males. Means :I: SEM are
illustrated. For all treatments, females had significantly higher
concentrations of plasma VTG than males. No differences in VTG were
observed for males or females among treatments. ................................................. 55

Figure 6. Plasma E2 (pg E2lmL) concentrations in male and female fathead
minnows following exposure to NPEO. Data are plotted as untransformed
means 1: SEM. (Statistical differences are for Iogw-transformed data:
*=significant differences between males and females within the same

xi

exposure group; O=significant difference of 1 pg NPEOIL males from all
other male treatments except for 0.3 pg NPEO/L males.) ....................................... 58

Figure 7. Plasma T (pg T/mL) concentrations in male and female fathead

minnows following exposure to NPEO. Data are plotted as untransformed

means :t SEM. (Statistical differences are based upon Iogm-transformed

data: *=significant differences between males and females within the same
exposure group (3 pg NPEOIL); No significant differences were observed

among males or females exposed to different concentrations of NPEO.) ............... 59

Figure 8. E2 to T (E2lT) ratios for male and female minnows following exposure

to nonylphenol ethoxylate. Data are plotted as means 1: SEM. No
differences between EZIT ratios were observed within or among treatments
and sexes. ....................................................................................................... ‘ ....... 60

Figure 9. Locations of wastewater treatment plant (VVWT P) sites in mid-Michigan, USA

where effects of WWTP effluent on reproductive physiology of fathead minnows
were surveyed, June-August 1996. Capital letters in bold are the WP sites
(BV=Bellevue, DL=Delhi Township/Holt, ER=Eaton Rapids, LA=Lansing,
LP=Limnology Pond, OW=Owosso, PT=PortIand, RS= Reference Site,
WM=erliamston). Boundary lines contrasted with different shading represent
watersheds. ............................................................................................................ 91

Figure 10. Plasma VTG concentrations for caged male and female fathead minnows

exposed in situ to VWVTP effluent. A. Female mean plasma concentrations
(iSEM) for VTG at each site. Differences among sites are represented by Tukey
letter grouping above the VTG concentration histograms. B. Male plasma VTG.
Histograms are mean VTG concentrations. Differences in the incidence of VTG
detection from both LP and RS rather than differences in absolute concentrations
are denoted by the symbols (*). Note the differences in y-axes scale (pg VTGImL)
between A and B. ................................................................................................... 97

Figure 11. Female and male plasma E2 (pg EZImL) concentrations following in situ

exposure to WWTP effluent. Data are plotted as nontransformed means :I: SEM.
(Statistical differences are for Iogm-transfonned data: *=significant difference
between males and females within the same site; A=difference from LP among
females; A=difference from LP among males; I=difference from RS among
females; [I=difference from RS among males) ..................................................... 100

Figure 12. Female and male plasma T concentrations (pg T/mL) following in situ

exposure to WWTP effluent. Data are plotted as nontransformed means :1: SEM.
(*=difference in logm-transfonned values between sexes within a given site;
A=difference from LP among females; A=difference from LP among males (p<0.10
for WM); I=difference from RS among females; [I=difference from RS among
males) ................................................................................................................... 1 01

Figure 13. Female and male plasma E2 to T ratios (E2lT) following in situ WWTP

effluent exposure. , Data are plotted as means i SEM of untransformed

xii

concentrations. *=difference in Iogw-trahsformed values between sexes within a
given site; A=difference from LP and RS among males only; p<0.10 for BV) ........ 102

Figure A1. Chromatogram for the separation of vitellogenin from estradiol
induced goldfish plasma. Note that the latest peak is the assumed to be
vitellogenin ............................................................................................................ 1 13

Figure A2. Example of immunoblotted proteins with anti-vitellogenin polyclonal
antibody detected with ECL reagents. Lanes represent different samples. .......... 118

Figure A3. Specific immune response for vitellogenin in New Zealand white
rabbits. Note columns represent a different antisera dilution or collection and
rows represent serially dilute pure vitellogenin or antigen. .................................... 122

Figure A4. Plate layout data sheet for determination of standard and sample
placement in the VTG ELISA. ............................................................................... 136

Figure A5. Example Microsoft Excel worksheet for a typical vitellogenin ELISA
plate with duplicate standard curves and duplicate samples ................................. 138

xiii

INTRODUCTION
Background Information

Observations of anomalies in the reproductive systems and physiology of
organisms have led to widespread concern for chemicals that may mimic or disrupt the
delicate balance of endogenous hormones (Colbom and Clement, 1992). Incidence of
decreased sperm count in human males and increased cancers of the humanfemale
and male reproductive tracts (Colbom and Clement, 1992), abnormalities in the
reproductive development, morphology, and endocrinology of alligators in Lake Apopka,
Florida (Guillette et al., 1994), unusual sexual development and mating behaviors in
avian species (Fry and Toone, 1981; Fox, 1992), and unusual occurrence of intersex
male fish in the United Kingdom (UK) (Purdom et al., 1994) have, only to name a few,
provided overwhelming evidence and urgency for research into questions regarding the
physiology of reproduction and how it may be altered by chemical insult.

The most widespread concern involving reproductive abnormalities in adult
organisms has been the issue of environmental endocrine disrupters, particularly
environmental ‘estrogen’ mimics. An endocrine disrupter has been defined 'as “an
exogenous agent that interferes with the production, release, transport, metabolism,
binding, action or elimination of natural hormones in the body responsible for the
maintenance of homeostasis and the regulation of developmental processes” (Kavlock

et al., 1996). Estrogens, primarily 17B—estradiol (E2) and to a lesser extent estrone (E1)

and estriol (E3), are produced endogenously in various tissues of all vertebrate animals.
Estrogen is produced in the greatest quantities by ovarian tissues during the female
reproductive cycle. Comparatively small quantities of E2 are also found in males but are
purported to primarily play a role in development of sexual behavior in the brain by
conversion of testosterone to E2 by aromatase (Selcer and Leavitt, 1991).

Endogenous E2 action has been well studied, especially in mammalian systems,
for its molecular mechanisms of action and effects during gametogenesis and
steroidogenesis. E2 elicits action by direct receptor-mediated control of gene
transcription. E2 can passively enter the cell and travels to the nucleus where it binds
the estrogen receptor (ER). Once formed, E2-ER complexes form homodimers that
bind to E2-responsive genes at a specific palindromic sequence known as the estrogen-
response element (ERE) upstream from the gene coding sequence. Binding of E2-ER
dimers to the ERE causes a conformational change in the gene chromatin structure and
recruitment and activation of specific transcriptional regulators. Subsequently, the E2-
responsive gene is activated and transcribed. An 'estrogen agonist’ has been defined
by endocrinologists and toxicologists as a substance that can bind to the ER and elicit
responses identical to endogenous E2 with differences only in potency (Selcer and
Leavitt, 1991).

Traditional measures of E2 agonist activity in mammalian systems have included
vaginal comification and uterine weight in female rats and mice in vivo (Cunha et al.,
1992). With the advancement of the field of environmental toxicology and physiology, in
vitro systems have been developed and used as models to study endogenous and
exogenous estrogen agonist activity. These have included cell proliferation assays in
human breast cancer cells (MCF-7 or E-Screen) (Soto et al., 1992), in vitro production of

vitellogenin (VT G) in teleost fish hepatocytes (Jobling and Sumpter, 1993), and

transfected cell reporter constructs where the ER, ERE, and nuclear machinery have
been inserted into yeast (Routledge and Sumpter, 1996), breast cancer, and hepatoma
cell lines including fish cells where a biochemical signal, transcribed by ER molecular
mechanisms, may be easily measured in a dose-responsive manner upon exposure to

E2 (Zacharewski, 1997).

Endocrine disruption in feral fish populations

Sublethal effects of contaminants on fishes have only recently been attributed to
direct endocrine modulating substances. Mosquitofish (Gambusia afi7nis) in a Florida
lake have been found to exhibit intersex characteristics, with female fish displaying male
secondary sexual morphological features (Bortone and Davis, 1994). Fish exposed to
effluent and effluent components from bleach kraft paper mills have exhibited stunted
growth, retarded gonadal recrudescence and maturation, altered secondary sex
characteristics, and decreased plasma concentrations of sex steroid hormones
(Munkittrick et al., 1991; McMaster et al., 1992; Mellanen et al., 1996). Recently,
investigations of environmental endocrine disruption for fishes have stemmed from
anomalies observed in the UK. Anglers noticed that adult feral male roach (Rutilus
rutilus) caught in one particular stream in the UK had abnormally great incidence of
intersex or evidence of both testicular and ovarian tissue in the gonads (Purdom et al.,
1994). Subsequent studies with caged rainbow trout (Oncorhyncus mykiss) and to a
lesser extent common carp (Cyprinus carpio) have confirmed the evidence of exposure
to estrogenic substances. The causative agents have been tentatively identified as
nonionic surfactant components and their degradation products (alkylphenol ethoxylates

and alkylphenols) as well as ethinyl estradiol from human birth control medication in

municipal and industrial wastewater effluents (Purdom et al., 1994; Harries et al., 1996,
1997)

Investigations of endocrine disruption below municipal wastewater treatment
plants (VVWT P) in the UK provide evidence for the effects of these effluents on fish
reproductive physiology. Increased plasma concentrations of the female-specific egg
yolk protein, vitellogenin (VT G), decreased testicular weight, and anomalies in testicular
histology have been observed in male rainbow trout following exposure to WWTP
effluents and suspected related compounds (Purdom et al., 1994; Harries et al., 1996,
1997; Jobling et al., 1996). In the United States, there is conflicting evidence that feral
male and female carp in surface waters and below WWTP have been impacted by
endocrine disruption. In Lake Mead located near Las Vegas, Nevada, male carp had
increased plasma VTG concentrations and male and females had altered sex steroid
hormone ratios (Bevans et al., 1996). These findings were similar to those observed in
Minnesota below an urban WWTP where male carp exhibited greater concentrations of
VTG and altered steroid hormone concentrations and ratios compared to fish from a
national scenic river in the same area (Folmar et al., 1996). A nationwide survey by the
Biological Resources Division of the United States Geological Service failed to find the
same induction of VTG in male carp although ratios of sex steroid hormones were
correlated with dissolved organic pesticide concentrations in surface waters (Goodbred

etaL,1997)

Biomarkers of estrogen exposure in fishes

Although in vitro measures of estrogenicity have helped toxicologists understand
the mechanisms and potencies of E2 and other estrogen agonists, in vivo potencies and
physiological actions of estrogen agonists are equally important in understanding the
anomalies seen in natural populations. For fishes, several biomarkers of exposure to
environmental estrogens have been used in laboratory and field exposures, including in
vitro and in vivo production of VTG and to a lesser extent, sex steroid hormone
concentrations and ratios.

Vitellogenesis is a model of E2 action in the liver of oviparous vertebrates and
measurements of VTG have been widely developed and validated, primarily by
radioimmunoassay (RIA) and enzyme-linked immunosorbent assays (ELISA) (Specker
and Sullivan, 1994; Heppell et al., 1995; Sumpter and Jobling, 1995). Estrogen binds
nuclear ER in the liver, dimen‘zes, and subsequently activates VTG genes that produce
VTG mRNA. Transcription of the VTG gene serves to regulate transcription of the
protein and stabilization of the VTG mRNA (Ren et al., 1996). In the cytoplasm, VTG
mRNA undergoes extensive modification in the Golgi apparatus where it is
phosphorylated on serine residues. The protein is immediately transported by the blood
to the developing oocytes where it is cleaved and incorporated for nutritional reserves
for the embryo (Mommsen and Walsh, 1988; Lazier and MacKay, 1993). Both male and
female fish have the VTG genes in the liver, but because males do not produce great
concentrations of E2 compared to females, under normal circumstances only females

produce significant concentrations of VTG. However, upon stimulation by. E2 or

environmental estrogens, male fish can produce VTG in vivo (Chen, 1983). Plasma
VTG can thus be us as a biomarker of in vivo estrogen exposure.

Although sex steroid hormones have been well studied for their roles in gonadal
steroidogenesis and gametogenesis, sexual development, and sexual behavior, their
validation and use as a biomarker of estrogen exposure has only recently been
attempted. Generally, 11-ketotestosterone (11-KT) and to a lesser extent, testosterone
(T), play critical roles in reproductive development and pre-spawning activities in teleost
males (Boume, 1991). In all vertebrates, E2 is the primary female sex steroid hormone
responsible for gonadal development and recrudescence during the spawning period
(Selcer and Leavitt, 1991). Previous investigators have indicated that the ratio of
estrogens to androgens, in most cases E2lT or E2111-KT, are even more important than
absolute concentrations of individual hormones to maintain endocrine balance for
gonadal recrudescence and normal spawning behavior (Folmar et al., 1996). For
common carp, sex differences in estrogen to androgen ratios have been observed;
ratios observed in females are typically greater than one, while ratios in males are
generally less than one. This ratio, however, has been observed to differ for fish in
various surface waters and upon exposure to WWTP effluents in the US (Folmar et al.,
1996; Bevans et al., 1996). Furthermore, the same investigators have found decreased
serum androgen concentrations in fish exposed to WWI'P effluents (Folmar et al.,
1996). The EZIT ratio and other hormone measures have neither been found to
correlate with plasma VTG concentrations (Folmar et al., 1996) nor validated for their

indication of overall fish health and reproductive fitness.

Research objectives

Although there is evidence of endocrine disruption in fish populations below
VWVTP, the extent of occurrence in the US below typical, mid-size, Midwestern plants
has not been evaluated. Furthermore, validation and calibration of traditional
biomarkers of exposure to endpoints indicative of overall fitness and reproductive output
for fishes remain to be conducted. The objectives of the research herein were aimed at
utilizing the fathead minnow, Pimephales pmmelas, to:

1) Calibrate the effects of suspect environmental estrogen agonists to reproductively
relevant endpoints such as egg production or fecundity in a controlled laboratory
exposure; and

2) Evaluate the in vivo effects of municipal VWVTP effluents on the reproductive
physiology of fish to determine the relative risk of municipal WWTP effluent for fish in

representative mid-Michigan streams.

The subsequent chapters correspond to the 2 objectives above and were written as
stand-alone manuscripts for submission to peer-reviewed journals. Specific objectives
and background information for each of these studies are explained there in further

detail.

REFERENCES

Bevans, H.E., S.L. Goodbred, J.F. Miesner, S.A. Watkins, T.S. Gross, N.D. Denslow, and T.
Schoeb (1996) Synthetic organic compounds and carp endocrinology and histology in Las
Vegas Wash and Las Vegas and Callville Bays of Lake Mead, Nevada, 1992 and 1995.
USGS Water-Resources Investigations Report 96-4266, 12 pp.

Bortone, SA. and WP. Davis (1994) Fish intersexuality as indicators of environmental stress.
Bioscience 44, 165-172.

Boume, A. (1991) Androgens. In: Vertebrate Endocrinology: Fundamentals and Biomedical
Implications, Vol. 4, Part B, edited by KT. Pang and MP. Schreibman, Academic Press, Inc.,
San Diego, pp. 115-148.

Chen, T.T. (1983) Identification and characterization of estrogen-responsive gene products in the
liver of rainbow trout. Can. J. Biochem. Cell Biol. 61, 802-810.

Colbom, T. and C. Clement (1992) Chemically-Induced Alteration in Sexual and Functional
Development: The Wildlife/Human Connection. Princeton Scientific Publishing Co., Inc.,
Princeton, NJ, 403 pp.

Cunha, G.R., E.L. Boutin, T. Turner, and AA. Donjacour (1992) Role of mesenchyme in the
development of the urogenital tract. In: Chemically-Induced Alteration in Sexual and
Functional Development: The Wildlife/Human Connection edited by T. Colbom and C.
Clement, Princeton Scientific Publishing Co., Inc., Princeton, NJ, pp. 85-105.

Folmar, L.C., N.D. Denslow, V. Rao, M. Chow, D.A. Crain, J. Enblom, J. Marcino, and L.J.J.
Guillette (1996) Vitellogenin induction and reduced serum testosterone concentrations in feral
male carp (Cyprinus carpio) captured near a major metropolitan sewage treatment plant.
Environ. Health Perspect. 104, 1096-1101.

Fox, GA (1992) Epidemiological and pathobiological evidence of contaminant-induced alterations
in sexual development in free-living wildlife. In: Chemically-Induced Alteration in Sexual and
Functional Development: The Wildlife/Human Connection edited by T. Colbom and C.
Clement, Princeton Scientific Publishing Co., Inc., Princeton, NJ, pp. 147-158.

Fry, 0M. and CK. Toone (1981) DDT-induced feminization of gull embryos. Science 213, 922-
924.

Goodbred, S.L., R.J. Gilliom, T.S. Gross, N.P. Denslow, W.B. Bryant, and TR. Schoeb (1997):
Reconnaissance of 17-beta estradiol, 11-ketotestosterone, vitellogenin, and gonad
histopathology in common carp of United States streams: Potential for contaminant-induced
endocrine disruption. US. Geological Survey Report, 47 pp.

Guillette, L.J., T.S. Gross, G.R. Masson, J.M. Matter, H.F. Percival, and AR. Woodward (1994)
Developmental abnormalities of the gonad and abnormal sex hormone concentrations in
juvenile alligators from contaminated and control lakes in Florida. Environ. Health. Perspect.
102, 680—688.

Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, E.J. Routledge, R. Rycroft, J.P.
Sumpter, and T. Tylor (1996) A survey of estrogenic activity in United Kingdom inland waters.
Environ. Toxicol. Chem. 15, 1993-2002.

Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, J.P. Sumpter, T. Tylor, and N.
Zaman (1997) Estrogenic activity in five United Kingdom rivers detected by measurement of
Vitellogenesis in caged male trout. Environ. Toxicol. Chem. 16, 534-542.

Heppell, S.A., N.D. Denslow, L.C. Folmar, and C.V. Sullivan (1995) Universal assay of vitellogenin
as a biomarker for environmental estrogens. Environ. Health Perspect. 103(Suppl 7), 9-15.

Jobling, 8., JP. Sumpter (1993) Detergent components in sewage effluent are weakly
oestrogenic to fish: an in vitro study using rainbow trout (Oncorhynchus mykiss) hepatocytes.
Aq. Toxicol. 27, 361-372.

Jobling, S., D. Sheahan, J.A. Osborne, P. Matthiessen, and JP. Sumpter (1996) Inhibition of
testicular growth in rainbow trout (Oncorhyncus mykiss) exposed to estrogenic alkylphenolic
chemicals. Environ. Toxicol. Chem. 15, 194-202.

Kavlock, R.J., G.P. Daston, C. DeRosa, P. Fenner-Cn'sp, LE. Gray, S. Kaattari, G. Lucier, M.
Luster, M.J. Mac, C. Maczka, R. Miller, J. Moore, R. Rolland, G. Scott, D.M. Sheehan, T.
Sinks, and HA. Tilson (1996) Research needs for the risk assessment of health and
environmental effects of endocrine disruptors: a report of the US. EPA-sponsored workshop.
Environ. Health Perspect. 104(Suppl. 4), 715-740.

Lazier, CB. and ME. MacKay (1993) Vitellogenin gene expression in teleost fish. In: Biochemistry
and Molecular Biology of Fishes, edited by Hochachka and Mommsen. Elsevier Science
Publishers, New York, pp. 391-405.

McMaster, M.E., C.B. Portt, K.R. Munkittrick, and DG. Dixon (1992) Milt characteristics,
reproductive performance, and larval survival and development of white sucker exposed to
bleached kraft mill effluent. Ecotox. Environ. Sat. 23, 103-117.

Mellanen, P., T. Petanen, J. Lehtimaki, S. Makela, G. Bylund, B. Holmbom, E. Mannila, A. Oikari,
and R. Santti (1996) Wood-derived estrogens: studies in vitro with breast cancer cell lines
and in vivo in trout. Toxicol. Appl. Pharmacol. 136, 381-388.

Mommsen, T.P. and P.J. Walsh (1988) Vitellogenesis and oocyte assembly. In: Fish Physiology,
Vol. XIA edited by W.S. Hoar and V.J. Randall. Academic Press, San Diego, pp. 347-406.

Munkittrick, K.R., C.B. Portt, G.J. Van Der Kraak, l.R. Smith, and DA. Rokosh (1991) Impact of
bleached kraft mill effluent on population characteristics, liver MFO activity, and serum steroid
levels of a Lake Superior white sucker (Catostomus commersonr) population. Can. J. Fish.
Aquat. Sci. 48, 1371-1389.

Purdom, C.E., P.A. Hardiman, V.J. Bye, N.C. Eno, C.R. Tyler, and JP. Sumpter (1994)
Estrogenic effects of effluents from sewage treatment works. Chem. Ecol. 8, 275-285.

Ren, L., S.K. Lewis, and J.J. Lech (1996) Effects of estrogen and nonylphenol on the postf
transcriptional regulation of vitellogenin gene expression. Chem-Biol. Interactions 100, 67-
76.

Routledge, E.J. and JP. Sumpter (1996) Estrogenic activity of surfactants and some of their
degradation products assessed using a recombinant yeast screen. Environ. Toxicol. Chem.
1 5, 241 -248.

Selcer, K.W. and WW. Leavitt (1991) Estrogens and progestins. In: Vertebrate Endocrinolgy:
Fundamentals and Biomedical Implications, Vol. 4, Part B, edited by KT. Pang and MP.
Schreibman, Academic Press, Inc., San Diego, pp. 67-114.

Soto, A.M., T.-M. Lin, H. Justicia, R.M. Silvia, and C. Sonnenschein (1992) An “in culture”
bioassay to assess the estrogenicity of xenobiotics (E-Screen). In: Chemically—Induced
Alteration in Sexual and Functional Development: The Wildlife/Human Connection edited by
T. Colbom and C. Clement, Princeton Scientific Publishing Co., Inc., Princeton, NJ, pp. 295-
309.

Specker, J.L. and C.V. Sullivan (1994) Vitellogenesis in fishes: status and perspectives. In:
Perspectives in Comparative Endocrinology, edited by K.G. Davey, R.G. Peter, and SS.
Tobe. National Research Council of Canada, Ottawa, pp. 304-315.

10

Sumpter, J. P. and S. Jobling (1995) VItellogenesis as a biomarker for estrogenic contamination of
the aquatic environment. Environ. Health Perspect.103(Suppl.7),173-178.

Zacharewski, T (1997) In vitro bioassays for assessing estrogenic substances. Environ. Sci.
Technol. 31, 613-623.

CHAPTER 1

Effects of Nonylphenol Ethoxylate (NPEO) Exposure on Reproductive Output and
Bioindicators of Estrogen Exposure in Fathead Minnows, Pimephales promelas

(To be submitted in modified form to Aquatic Toxicology)

INTRODUCTION

Alkylphenol ethoxylate (APEO) nonionic surfactants are manufactured for their
ubiquitous use as emulsifiers in industrial and household cleaning agents, agricultural
chemicals, and plastic polymerization processes (Nimrod and Benson, 1996). APEO
are produced at a rate of approximately 350,000 tons annually in the United States
(US), Western Europe, and Japan (Kvestak and Ahel, 1995) and the intensive use of
APEO results in release to wastewater (Nimrod and Benson, 1996). In the wastewater
treatment (VVWT P) process, APEO undergo biodegradation processes that remove
successive carboxyl terminals from the alkyl chain which results in less biodegradable
small chain ethoxylates, primarily mono- and di-ethoxylates (AP1EO, AP2EO),
carboxylates (APEC), and alkylphenols (AP) (Giger et al., 1984). APEO are removed
92.5 to 99.8% by the WWTP process in the US (Naylor et al., 1992); however, concern
stemming from evidence of endocrine disruption on fish below VWVTP (Purdom et al.,
1994; Bevans et al., 1996; Folmar et al., 1996; Harries et al., 1996, 1997) has led to

recent studies on the effects of APEO and their biodegradation products on fish.

11

12

Evidence of endocrine disruption by APEO and AP have been attributed
primarily to the para-substituted phenols and mono- and di-ethoxylates by direct action
at the estrogen receptor (ER) (White et al., 1994; Routledge and Sumpter, 1996).
Alkylphenols, primarily octylphenol (OP) and nonylphenol (NP) are the final phenolic
breakdown products of their respective APEO and are the most potent of the APEO and
APs based upon in vitro bioassays with OP greater than NP in potency. Because of the
para- substituted phenol, and thus purported similarity to endogenous estrogens, APs,
mostly OP and NP, have been studied most widely. Few studies have been conducted
on the parent APEO and the relatively more water-soluble constituents. Previous
studies with AP and some APEOs have reported increased in vitro production of the
female-specific egg yolk protein precursor, vitellogenin (VT G) (Jobling and Sumpter,
1993), in vitro proliferation of human breast cancer cells (MCF-7) (White et al., 1994),
and increased transcriptional activity of a yeast cell line under control of the human ER
(Routledge and Sumpter, 1996). Following exposure to OP, NP, NP1EC, or NPZEO,
rainbow trout (Oncorhyncus mykiss) have exhibited increased VTG production and
decreased testicular weight (Jobling et al., 1996). Similarly, NP has caused
abnormalities in the gonadal morphology and histopathology of medaka (Oryzias Iatipes)
(Gray and Metcalfe, 1997) and common carp (Cyprinus carpio) (Gimeno et al.,1996).
The effects of alkylphenols in vitro and in vivo, particularly for NP, have been evaluated
intensively, but the effects of APEO remain to be described in detail.

Disruption of the endocrine balance may severely compromise the reproductive
fitness and survival of an organism. Previously, biomarkers including plasma VTG and
sex steroid hormone levels have been used as measures of environmental estrogen
exposure (Purdom et al., 1994; Sumpter and Jobling, 1995; Bevans et al., 1996; Folmar

et al., 1996; Harries et al., 1997). Similarly, studies of feral English sole (Pleuronectes

13

vetulus) have been conducted to determine the potential impacts of contaminants on
ovarian development, fecundity, and egg quality with important implications for the
survival and natural propagation of the species in the Puget Sound, Washington
(Johnson et al., 1988, 1997). These authors also evaluated VTG (as plasma alkaline
labile phosphorus; ALP), E2, T, and other biochemical endpoints in exposed versus
nonexposed populations. The implications of anomalies in these biomarkers have not
been realized nor have they been calibrated to the general health and reproductive
fitness of the fish. Investigations remain to attempt calibration of these biomarkers to
endpoints relevant to the reproductive fitness and condition of fishes. Recently,
calibration of 17B—estradiol (E2) exposure to biomarkers including plasma ALP, plasma
E2, and fecundity of fathead minnows (Pimephales promelas) has been reported
(Kramer, 1996). The potency of E2 far exceeds that of xenoestrogens such as NP and
NPEO and the need for calibration of traditional biomarkers of exposure with model
xenoestrogens found in the environment remains.

NPEO was tested in this experiment to evaluate the effects of environmentally-
relevant concentrations on the reproductive physiology of adult fathead minnows,
Pimephales promelas. The aims of the research were to determine: 1) the frequency of
detection and quantity of VTG in male and female exposed fish, respectively, 2) plasma
concentrations of reproductive sex steroid hormones, and 3) egg production and iviability
when adult fathead minnows were exposed to a mixture of NPEO oligomers (NP, NP1-
17EO). The experiment required the development of a reliable method for VTG
measurement in fathead minnows and a competitive enzyme-linked immunosorbent
assay (ELISA) is presented. A discussion of the relationships between biomarkers of
exposure and fecundity and a relative assessment of risk for NPEO in the environment

is given.

14

MATERIALS AND METHODS

Fish

Fathead minnows age 8 to 18 months were cultured and reared in the Aquatic
Toxicology Laboratory at Michigan State University (MSU) from mixed stocks obtained
from the Limnology Ponds at the MSU Inland Lake Teaching and Research facility, the
United States Environmental Protection Agency, Duluth, Minnesota, and the Dow
Chemical Company, Environmental Laboratories, Midland, Michigan. Fathead minnows
were maintained at temperatures ranging from 15°C to 21°C during normal culture
conditions. All fish were fed a mixture of TetraMin flakes and dense culture food (1:1;
v/v) and Artemia and were maintained in a 16:8 h lightzdark cycle at the University
Research Containment Facility at MSU.

Goldfish (Carassius auratus) were used as a source of VTG for antibody
production. Goldfish ranging in weight from 21-53 g were received from Grassyforks
Fisheries Co. (Martinsville, IN). The fish were acclimatized in the laboratory for several
months at 15°C before beginning of VTG induction for separation and purification of the

protein.

15

Induction, separation, and purification of VTG

Induction

Twenty male and female goldfish were administered 2 mg/kg-week 17B-estradiol
(E2) dissolved in ethanol and suspended in corn oil for induction and separation of VTG
as described by Silversand and Haux (1989). Fish were injected intra-peritoneolly (i.p.)
2 times each week with 0.1 mL E2 following anesthetization with MS-222 (tricaine
methane sulfonate, Argent Chemicals, Redmond, WA). At the end of the 2 week
exposure to E2, goldfish were injected with 0.1 mL of aprotinin (10% v/v in corn oil) 0.5 h
prior to euthanization to prevent proteolytic degradation of the protein during sample
collection and storage. All fish were euthanized with a lethal concentration of MS—222
and were exsanguinated by drawing blood from the caudal vein with a heparinized
needle and syringe. Blood from all the goldfish was pooled, allowed to clot for 1 h on
ice, and then centrifuged at 3000 Xg for 10 min at 4°C. Plasma was drawn off and

stored at ~80°C until VTG was separated.

Purification

VTG was isolated from the plasma by high-performance liquid chromatography
(HPLC) (Silversand and Haux, 1989). One milliliter of the pooled EZ-induced goldfish
plasma was filtered and diluted to 10.0 mL with 20 mM Tris-HCI. The dilute plasma (500
pL) was loaded onto a DEAE anion exchange column (15 mm X 20 cm; PJICobert
Assoc, St. Louis, MO) equilibrated with 20 mM Tris-HCI. Proteins bound to the column
following an initial wash with 20 mM Tris-HCI were eluted with a linear gradient of 20

mM Tris-HCI to 0.5 M NaCl in 50 min at a flow rate of 1.0 mUmin. Eluted proteins were

16

monitored by ultraviolet (UV) fluorescence with a photodiode array fluorescence

detector; absorption was measured at 230, 254 and 280 nm.

Separation and identifica_tio_n

The identity of protein eluted from the HPLC and suspected to be VTG was
confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE; 4-
15% Tris-Tricine continuous Laemmeli method; Gallagher, 1995). Due to its large size
and highly positive charge, VTG is one of the last proteins eluted from the column.
HPLC fractions, plasma from un-induced females or E2-induced positive controls, and
plasma from males were run simultaneously on each gel. Gels were transferred to a
polyvinylidene difluoride (PVDF) or nitrocellulose membrane for Western immunoblotting
(Gallagher et al., 1993). Membranes were stained to detect all proteins with Coomassie
Brilliant Blue or Ponceau S to determine the purity of samples and HPLC fractions. The
largest protein was positively identified by subsequent immunoblotting with a previously
developed goldfish VTG polyclonal antibody. Total protein in each fraction was
quantified with the fluorescamine protein assay using bovine serum albumin (BSA) for

the standard curve (Lorenzen and Kennedy, 1993).

VTG polyclonal antibody production

Two female New Zealand white rabbits were injected with the purified goldfish
VTG to produce polyclonal antibodies. Immunizations were conducted by personnel of
the University Laboratory Animal Resources at Michigan State University. Prior to initial
injection, the rabbits were anaesthetized with acepromazine (0.7-1.5 mg/kg) and 20 mL

blood was collected. Purified antigen (15 pg VTG/mL in filtered 0.1 M PBS) was

17

emulsified in an equal volume of Hunter's TitreMaxTM adjuvant (Sigma Chemicals, St.
Louis, M0) for an injection volume of 1.0 lerabbit. A total of not more than 10
subcutaneous injections were made for each rabbit.

One of the rabbits (rabbit #1) was given a booster injection with the same
concentration of antigen 28 d following initial injection. Forty-two days after the initial
immunization, blood samples were collected from both rabbits and serum was checked
for non-specific (immunoglobulin G; IgG) and specific immune response (VT G
antibodies). Subsequent boosting of the rabbits with 15 pg VTGlmL was made
periodically to ensure maximum specific antibody production (see schedule, Appendix A,
Table A1). Rabbits were anesthetized and exsanguinated approximately 6 months after
the initial immunization. All blood collected from the rabbits was stored at 4°C overnight
to allow clotting. The following day, blood was centrifuged in 4°C for 10 min at 5000 X9.
Antiserum was collected, aliquoted, and stored at -80°C until further characterization
and use in a competitive ELISA to quantify VTG.

Nonspecific and specific VTG immune responses were determined by slot-blot
Western immunoblotting (BioRad, Hercules, CA). Briefly, a nitrocellulose membrane
equilibrated and rinsed in distilled, deionized water for 15 min was assembled in the slot-
blot apparatus and each slot was rinsed with a volume of 250 pL Tris-buffered saline
(TBS: 2 mM Tris, 137 mM NaCl, pH=7.6) two times.

For determination of nonspecific lgG antibodies, antisera collected from the
rabbits was serially diluted (0, 10", 10°, 10’, 10°, 10", 101°, and 10") in TBS by rows and
100 pL added to each slot. After pulling the dilute antisera through the membrane by
vacuum and subsequent rinsing 2 times each with 250 pL TBS, the nitrocellulose was

removed from the slot-blot apparatus. The membrane was blocked with a 5% powdered

18

milk solution in TBS for one hour and then incubated with 1:5000 donkey anti-rabbit lgG
linked to horseradish peroxidase (HRP) for 1-2 h.

For specific or VTG antibody determinations, goldfish VTG purified as above was
serially diluted by rows (0, 10 pg, 1 pg, 0.1 pg, 10 ng, 1 ng, and 0.1 ng) and 100 pL
added to each slot after rinsing of the membrane as above. The pure antigen was
pulled through the membrane which was then removed and blocked with 5% milk
solution. Following blocking, the membrane was cut into strips according to columns to
give the entire range of VTG concentrations. Each strip of the membrane was
incubated in a different concentration of antiserum diluted in TBS. HRP conjugated
secondary antibody was added for incubation to detect the primary, or specific antibody
bound to the membrane. Luminescence was developed by enzymatic reaction with the
HRP-conjugated antibody using the ECL Western blotting detection system (Amersham
Corp., Arlington Heights, IL) and subsequent exposure to radiographic film.

Specificity and affinity of the VTG antisera for male and female plasma proteins
was determined by SDS-PAGE and Western immunoblotting. Plasma from male and
female goldfish, HPLC-purified goldfish VTG fractions, and male and female fathead
minnow plasma were used to determine molecular weights, specificity of the antisera for

VTG, and non-specific cross-reactivity with male proteins.

VTG ELISA optimization

A competitive VTG ELISA was adapted and optimized from previously developed
methods (Yao and Van Der Kraak, pers. comm.; Maisse et al., 1991; Mourot and Le
Bail, 1995). Prior to VTG quantification, the ELISA was optimized to minimize

nonspecific binding and to maximize the sensitivity and discrimination of the assay.

19

Parallelism of dilution curves for female fathead minnow and goldfish plasma was tested
against standard curves produced with purified goldfish VTG.

VTG antisera was diluted and used in the ELISA to determine the optimal rabbit
antisera dilution for incubation with fathead minnow or goldfish samples and standards.
The dilution that gave the widest range of absorbance and that allowed 50% maximum
absorbance (Bo) for the mid-range of the standard curve was chosen for use in the

ELISA.

Nonylphenol ethoxylate exposure

Fathead minnows were exposed to an industrial mixture of NPEO (Surfonic N-
95, CAS 9016-45-9, Huntsman Corporation, Port Neches, TX) in a proportional flow-
through diluter (Ace Glassware, Vineland, NJ). Concentrations of 0, 0.3, 1, 3, and 10 pg
NPEOIL were chosen for the exposure to coincide with environmental concentrations of
ethoxylates reported in effluent waters from municipal wastewater facilities (VVVVT P) and
surface waters of the US (Naylor et al., 1992; Nimrod and Benson, 1996; Weeks et al.,
1996). All concentrations were mixed by the proportional flow-through diluter system
from one aqueous stock solution of NPEO. The diluter delivered approximately 666 mL
of the test solutions to each of three replicates (blocks or water baths) for each
concentration at a rate of 2-4 cycles/h. The experimental design was a randomized
complete block design with each of three replicate concentrations represented once in
each of three water baths (blocks).

Three adult female and 3 male fathead minnows were placed into each 19 L

aquarium at the beginning of the experiment and were acclimatized for approximately 2

20

weeks. Temperature in the chambers was maintained at 24-27°C. During the
experiment, fish were fed daily 1:1 (vlv) flake fish foodzdense culture pelleted food and
brine shrimp (Artemia) at a rate of 0.1% body weight and were maintained under a 16:8
h Iight:dark cycle. Dissolved oxygen (DO) and temperature were measured in each of
the chambers weekly. Calcium carbonate hardness was measured 3 times throughout
the duration of the exposure.

Following the acclimation period, NPEO was added to the diluter for an exposure
lasting 42 d (6 weeks). Each chamber was observed daily for mortality and fish health.
Three terraocotta tiles were placed in each aquarium as spawning substrata beginning
on day 3 of the exposure period. "files were removed and examined daily for the
presence of eggs. Spawning tiles were kept in the tanks for periods of 5 d at a time and
then removed from the tanks for 2 d throughout the experiment. If eggs were present,
numbers of eggs laid on the first day and numbers of viable eggs on subsequent days
were counted and the spawning event recorded. Viable eggs included those that were
fertilized and were not otherwise infested with fungus, eaten, or popped. The number of
eggs laid was intended to be a functional measure of fecundity and is hereinafter
referred to as fecundity or egg production. Egg production is expressed as eggs
produced normalized to the number of females in each exposure chamber at the time
eggs were laid. The tiles were left in the chambers until just prior to hatching since
paternal care and protection from predation and fungal infection is important for maximal
egg survival (pers. obs.). Tiles with eggs were left in the chambers for 4 d until the eyed
stage of development. On the fourth day, eggs were placed in separate aquaria
designated for each exposure concentration for hatching. The fry were maintained in

our laboratory for future second-generation evaluations.

21

NPEO water sampling

During the 6 week exposure, 1 L water samples were taken to determine actual
NPEO oligomer distributions of the ethoxylates and phenols in the exposure chambers.
NPEO was extracted from the water samples with Empore® styrene divinyl benzene
discs (3M Corporation, St. Paul, MN) by vacuum filtration. The discs were stored at

-20°C until extraction of the test compound.

Sample collection

On the last day of exposure to NPEO, fish were euthanized with a lethal dose of
MS—222. To ensure comparability for hormone analysis, blood and tissue samples were
collected between 07 00 h and 10 00 h. Sex determinations were made visually by
gross morphology and were later confirmed by histology. Standard length (cm) and
weight (9) measurements were recorded along with observations of obvious health and
morphological abnormalities. The caudal peduncle was severed with a razor blade and
blood was immediately collected into heparinized hematocrit tubes. Hematocrit tubes
were stored on ice for 1-2 h to allow clotting before centrifugation in 4°C at 3000 X9 for 5
min. Packed cell volume or blood hematocrit values were recorded for each fish and the
plasma was pulled off and stored in a microcentrifuge tube at -80°C until further
analyses. Fish were preserved in Bouin's fixative for gross morphology and gonadal
histopathology examination by S. Miles-Richardson, Aquatic Toxicology Laboratory,
MSU. Blood plasma VTG was measured by a competitive ELISA. Any plasma

remaining after VTG measurement was used for E2 and testosterone hormone analyses

22

also by ELISA. Plasma VTG and hormone determinations were limited by the small

quantities of blood and thus plasma (1-100 pL) sampled from the fish.

VTG competitive ELISA

The VTG ELISA was adapted from previously developed methods (Yao and Van
Der Kraak, pers. comm.; Maisse et al., 1991; Mourot and Le Bail, 1995). The assay is a
competitive ELISA in which sample or standard in the well compete with pure antigen
coated on the well for the primary antibody (rabbit anti-VTG). More antibody in the well
and thus more color during enzymatic color development indicates less analyte (VT G) In
the well. Briefly, the protocol for the VTG ELISA follows: (a) Plate Coating. . Round
bottom 96-well ELISA plates (Corning, Cambridge, MA) were coated with 25 nglwell
purified goldfish VTG in sodium bicarbonate buffer (50 mM, pH 9.6) with 5 mg
gentamycin/mL. Plates were coated for at least 3 h or overnight at 37°C. Plates were
removed and wells washed 4 times with 200 pL wash buffer (T BS-T: 10 mMTris-HCI,
0.15 M NaCl, pH 7.5; Tween-20 0.1% and 5mg gentamycinlL). (b) Plate saturation.
Blocking of any unbound sites on the surface of the wells was accomplished with a
nonspecific protein, 2% goat serum (Sigma Biosciences, St. Louis, MO) in TBS-T (TBS-
T-SG). After addition of 200 prell the plates were incubated for 30 min at 37 °C. (0)
Primary antibody incubation. Upon saturation of the wells, the blocking buffer was
removed and 50 pL of samples or standards diluted in TBS-T-SG were added to each
well. Standards and samples were run in duplicate on each plate. Rabbit VTG antisera
was diluted 1:50.000 in TBS-T-SG and 100 pL was added to each well and plates were
incubated overnight at room temperature. The following morning, plates were washed

to remove unbound antibody and sample. (d) Secondary antibody incubation. The

23

secondary antibody, goat anti-rabbit lgG linked to HRP was diluted 1:2000 in TBS-T-SG
and 150 pL added to each well. Plates were covered and incubated for 2 h at 37°C
followed by washing. (9) Color development. Incubation of the HRP-conjugated
antibody (linked to the primary antibody bound to the plate) with the substrate of the
enzymatic reaction was accomplished by addition of 1,2-phenylene diamine (or o-
phenylene diamine, OPD, 0.5 glL, Sigma) in 50 mM ammonium acetate adjusted to pH
5.0 with citric acid (50 mM) and 0.5 leL 30% hydrogen peroxide. Each well received
150 pL of the OPD solution and plates were incubated at room temperature for 30 min
in complete darkness. (0 Color reaction stop and OD determination. The reaction was
stopped with the addition of 50 pL of 5 M sulfuric acid. Absorbances were measured
with a 96-well plate reading spectrophotometer, the Cayman Autoreader (OEM Version,
Cayman Chemical, Ann Arbor, MI) executed by Cayman EIA software (version 2.0).
After 10 min the optical density (OD) was measured at 492 nm for the measured
wavelength and 650 nm for reference. (g) Calculation of results. The standard curve
was linearized with a log-logit transformation of ng VTG/well versus absorbance. The

log-logit regression equation was:
logit BIB0 = m*log(VTG(nglwell)) + b (1)

where:

m = slope of the least squares regression line;

b = y-intercept of the least squares regression line;

B = sample absorbance (OD) corrected for nonspecific binding (NSB);

B0 = total binding in the absence of primary antibody corrected for N88;

24

IOQMB/Bo) = '09I(B/Bo)/(1-(B/Bo))l-

The least squares regression line (1) was used to calculate pg VTG/mL for samples with
correction for sample volumes and dilution factors.

Standards ranging from 0.135 to 75.3 ng VTG/well and dilute samples were
assayed in at least duplicate and most of the time triplicate on each 96-well plate.
Samples or standards that exceeded 15% variability among absorbance values were
discarded and/or re-analyzed dependent upon availability of sample. Samples less
than 20% B0 on the standard curve were re-assayed at greater dilutions. Inter-assay
coefficients of variation were determined from an average of VTG measured standard
concentrations on each plate (standard 4.34 ng VTG/well). All samples for VTG
measurement were assayed on the same day with intra- and inter-assay coefficients of

variation of 7.6% and 10.9% (n=3), respectively.

Plasma steroid hormone measurements

E2 and testosterone (T) were assayed in the plasma of exposed fish by ELISA
with Enzyme ImmunoAssay (EIA) kits (Cayman Chemicals, Ann Arbor, MI). Remaining
plasma after VTG determination but not more than 50 pL was transferred quantitatively,
diluted to 1.0 mL with nanopure water, and then extracted twice with 5 mL of diethyl
ether. Extracts were blown to complete dryness under nitrogen and reconstituted in
300-500 pL EIA buffer (0.1 M PBS) provided in the kit depending on the volume of
plasma used for extraction.

Protocols for the E2 and T ELISAs followed the instruction manuals from the

manufacturer of the EIA kits (Cayman Chemicals, 1992a,b). Briefly, 50 pL of standards

25

or dilute sample extracts were added to each well of a 96-well flat bottom polystyrene
plate pre-coated with mouse monoclonal anti-rabbit antibody and blocking proteins.
Following addition of samples or standards, E2 or T linked to an acetylcholinesterase
(ACE) tracer was added to the wells followed by addition of rabbit specific E2 or T
antiserum (50 pL each). The plates were incubated for 1 h allowing competition of free
hormone in the standards or samples and the ACE-linked hormone for the monoclonal
antibody bound to the plate. Following incubation of the samples, the plates were
washed five times with PBS-Tween (0.1 M PBS, 0.1% Tween-20). Color development
of the wells was accomplished by addition of Ellman's reagent containing the substrate
for ACE, acetylthiocholine, and 5,5'-dithio-bis-(2-nitrobenzoic acid). The enzymatic
cleavage of acetylcholine to thiocholine and the nonenzymatic reaction of thiocholine
with 5,5'-dithio—bis—(2-nitrobenzoic acid) produces a yellow color (5-thio-2-nitrobenzoic
acid) absorbed at 414 nm. Absorbance of each of the wells was measured at 414 nm
by a plate reading spectrophotometer as in the VTG ELISA.

Estradiol (7.8 to 1000 pglmL) and testosterone (3.9 to 500 pglmL) standard
curves were assayed in duplicate on each plate. The specificity of the E2 antibody was:
17(-E2, 100%; estrone, 7.5%; estriol, 0.3%; T, 0.1%; 5a-dihydrotestosterone (Cayman
Chemical, 1992a). For the T monoclonal antibody, specificity was: testosterone, 100%;
5a-dihydrotestosterone, 21%; 58-dihydrotestosterone, 10%; androstenedione, 3.6%;
11B-hydroxytestosterone, 1.2%; 5a-androstane-38,17B—diol, 0.4%; 5a-androstane-
3a,17B—diol, 0.2%; SIS-androstane-3,17-dione, 0.08%; E2, 0.02% (Cayman Chemical,
1992b). An internal standard of pooled male and female goldfish plasma was extracted,
diluted, and assayed in at least triplicate on each plate for determination of variability

between assays. Log-logit transformation of standard hormone concentrations (pglmL)

26

on % maximum binding (%BIB0) calculated from absorbance units as for VTG were
plotted to calculate a linear regression model used for determination of steroid hormone
concentrations in plasma (Equation 1). Samples were analyzed in at least duplicate and
samples or standards exceeding 20% coefficient of variation (CV) were re-assayed.

Samples below 20% binding on the standard curve were diluted 50% and re-analyzed.

Statistical methods

Fecundity, VTG concentrations, and concentrations of E2 and T were tested for
assumptions of normality and homogeneous variance. Rejection of these assumptions
for fecundity and VTG data resulted in the use of nonparametric statistical comparisons
among blocks and treatments. Kruskal-Wallis one-way ANOVA was used on ranks of
these data. A chi-square test was used to test for differences among treatments in the
incidence of VTG above the method detection limit (MDL) for males. ELISA detection
limits were quantified using a paired t-test to determine the lowest concentration of the
standard curve statistically different from maximum binding (Bo) or zero VTG. Log“,-
transformed hormone values and untransformed hormone ratios were analyzed for
differences among treatments, blocks, and sexes with a general linear model (PROC
GLM), SAS statistical software (Cary, NC); subsequent pairwise comparisons were
made with a Tukey's HSD test. Unless otherwise indicated that differences in blocks
contributed to the variability of the treatment means, all data reported were pooled
across blocks for each treatment. Nonparametric (Spearman's Rho) and parametric
(Pearson product-moment) correlation coefficients (r) were calculated with Plot-It
software (Scientific Programming Enterprises, Haslett, MI) to identify trends between

traditional biomarkers of exposure (VT G, T, E2, E2lT) and more general indicators of

27

fish condition and reproductive health (hematocrit, fecundity). All significance values
were set at p=0.05 unless otherwise indicated. Means 1: standard errors of the means

(SEM) are reported and plotted in figures.

RESULTS

Vitellogenin antigen characteristics

VTG was eluted on the HPLC after approximately 26 min with a linear gradient
and was the last protein eluted from the column (Figure 1). Subsequent identification of
VTG by SDS-PAGE and Western immunoblotting indicated that the purified proteins
correspond to the large molecular weight protein also present in the E2—induced crude
plasma. Female fathead minnow VTG had an approximate molecular weight of 146,000
and 74,500 Da. The molecular weight for the largest proteins ranged from
approximately 132,000 to 156,000 Da for goldfish. One or two smaller molecular weight
proteins (~69,000 to 118,500 Da) in the crude E2-induced goldfish plasma and the
purified fractions were identified as degradation products of VTG cross-reacting with a
previously characterized goldfish VTG antibody (Figure 2). In HPLC fractions,
nonspecific proteins that did not cross-react with the specific goldfish VTG antibody
were not evident. All proteins detected by nonspecific protein staining in purified HPLC
fractions also cross-reacted with polyclonal VTG antibody. Fractions #2-26 and #26
exhibited relatively great quantities of protein and proportionally less degradation
products and were subsequently used to immunize the rabbits for polyclonal antisera

production.

28

VTG antibody specificity

The VTG antisera produced was specific for the VTG protein, cross-reacting with
the protein from several species including uninduced goldfish, fathead minnow, and
mirror or common carp (Cyprinus carpio) females and E2-induced goldfish. From SDS-
PAGE and Western immunoblotting, the antibody did not appear to cross react with
male proteins for any of these three species. Slot-blotting techniques indicated that
within 14 d after boosting the rabbits, the VTG antisera could detect as little as 1 ng of

purified VTG.

VTG assay validation

The VTG ELISA was optimized for use with fathead minnow and goldfish
plasma. A criss-cross method of dilutions for the primary antibody and the antigen
coating rate determined the optimal primary antibody dilutions to be approximately
1:50.000 with a VTG coating concentration of 25 ng VTG/well. These dilutions allowed
maximum discrimination among samples, the greatest linear standard working range,
and the least background interference. Previous methods for detection and quantifying
VTG have employed various measures for nonspecific binding (NSB) (Maisse et al.,
1991; Goodwin et al., 1992; Mananos et al., 1994). By coating the NSB wells with an
equivalent protein content (25 ng VTG/well) of male control plasma, no difference was
observed in the absorbances between this method or determining the NSB from
nonspecific cross-reactivity of the secondary antibody in the assay.

Since the volume of plasma collected from fathead minnows was limited,
dilutions of minnow plasma were optimized prior to assaying the samples from

laboratory male and female control plasma. Female fathead minnow plasma was

29

diluted 1:1000 to 1:5000 while male plasma was diluted 1:50. Previous authors have
indicated the interference and nonspecific cross-reactivity of male nonspecific proteins
with polyclonal antisera for VTG (Rodriguez et al., 1989; Goodwin et al., 1992). SDS-
PAGE did not indicate any cross-reactivity of male plasma proteins with VTG antisera.
Serial dilutions of male control plasma did exhibit displacement of VTG antisera binding
to the wells at the least dilution (1/50), but this displacement or cross-reactivity did not
overlap with the linear range of the standard curve (Figure 3).

Due to the small quantities of blood obtained from a fathead minnow, goldfish
VTG was used in the ELISA both to coat the wells and for the standard curve. In order
to assay VTG in fathead minnows, the antigenicity of the antisera for the VTG of goldfish
and minnows must be similar to provide reliable results. Dilution curves of fathead
minnow, goldfish, E2-induced goldfish, and purified goldfish VTG in the range of the
standard curve were assayed and compared to the standard curves. Statistical analysis
by F-test on the mean squares of the regression equations on log-logit transformed data
indicated that the samples were not parallel to the standard curve even though, visually,
the curves appear to be parallel (Figure 3). The rigidity of this parametric test was
further tested by a chi-square analysis of the standard curve using the regression line
slope of the standards and comparing these values to those obtained using the mean
slope of fathead minnow and/or goldfish plasma. The chi-square analysis indicated that
the regression lines were bordering nonsignificance from approximately 30% to greater
than 90% maximum binding (q=15.62073, df=6, chi-square=14.45). Dilutions of fathead
minnow plasma were made such that absorbances remained in this comparable range

of the standard curve.

30

Survival and fish health

Survival of adult fathead minnows in each concentration ranged from 67% for 3
pg NPEO/L, to 72% for 10 pg NPEOIL, and to 89% for 0, 0.3, and 1 pg NPEO/L, thus
not occurring in a concentration-dependent manner. All of the fish comprising one
replicate of the 3 pg NPEO/L exposure treatment died due to a fungal infection. Most
other mortalities were also due to fungal infections or were observed with emaciation
and hemorrhages near the base of the fins. None of the females in the third replicate of
10 pg NPEO/L survived to the end of the experiment.

Packed cell volume (hematocrit), an index of overall fish health (Goede and
Barton, 1990), was similar for all males and females among all treatments. Differences
that were observed within treatments or among sexes were attributed only to variability

among replicates.

NPEO water quality and chemistry

Mean water temperature in the exposure chambers was 24.7:0.11°C throughout
the experiment with no differences observed among NPEO treatments. Mean dissolved
oxygen was maintained between 6.9 and 7.3 mg/L in the treatment chambers. The
lowest mean DO (6.9 mg/L) was observed for the 1 pg NPEO/L treatment and was
lower than all other treatment chamber DOs except for 3 pg NPEO/L. Mean water
hardness was 90.0:456 mg CaCOalL for all of the treatment chambers.

An examination of the oligomer distribution of the mixture of NPEO in the water
indicated no NPEO degradation to smaller carbon chain ethoxylates or subsequently
nonylphenol. Oligomer distributions within the chambers were the same as undiluted

technical grade standards and the parent compound (Table 1). The technical mixture of

31

NPEO consists primarily of 7 to 11 carbon chain ethoxylates with only 0.58% by weight
of the compound consisting of combined nonylphenol (NP) and mono- and di-
ethoxylates (NP1EO, NP2EO), the primary degradation products during bacterial
degradation in wastewater. Contributing more than any other single constituent of the

technical mixture is NP9EO (10.73%).

Fecundity

Fecundity did not exhibit a statistically significant concentration-dependent
relationship to NPEO exposure (Figure 4). The concentration-dependent trend for both
total number of eggs laid and numbers of eggs normalized to surviving females
appeared to be an inverted-U pattern with increasing nominal concentrations of NPEO,
but the sample sizes and variances were too great to discriminate any statistical
differences among treatments. Minnows in the least concentration, 0.3 pg NPEO/L,
produced the greatest numbers of eggs and had the greatest number of spawning
events. The number of spawning events, or frequency which eggs were laid, is
recorded above the histograms of egg numbers and illustrates a parallel change with
fecundity (Figure 4); the greatest total numbers of eggs/female were produced by those
fish that had more spawning events. The total number of spawning events and
eggs/female for the control chambers were contributed by two of the three blocks and
exhibited no differences in fecundity when compared to exposure chambers. For 3 and
10 pg NPEOIL, only one chamber of fish was actively laying and fertilizing eggs.
Evaluation of female body weight and covariance with fecundity showed no statistical

evidence that body weight contributed significantly to differences in fecundity. Mean

32

body weight for females (n=39) was 1.65:0.078 g and for males (n=35) was 3.03i0.145
9.

Observations of average egg viability for each treatment were quite variable
within and among treatments and replicates and indicated no statistically significant
trend. Average egg viabilities 2-4 d post-spawning for chambers with fish producing
eggs were 44.7%, 34%, 64%, 73.5%, and 54% for 0, 0.3, 1, 3, and 10 pg NPEO/L
treatments, respectively. Most of the eggs that were counted as nonviable had
apparently been damaged or eaten during the period in which the tiles remained with the
adults. Eggs were left for paternal care and maintenance and it is unknown whether the
NPEO exposure contributed to changes in parental behavior. Furthermore, keeping the
eggs in the exposure tanks for 4 d will allow future evaluation of a short term exposure
during development of the embryos on later sex ratios and reproductive performance

and physiology once these fish reach sexual maturity.

Plasma VTG

Mean concentrations for plasma VTG in females ranged from 291.7 pg VTG/mL
in fish exposed to 10 pg NPEO/L to 895.1 pg VTGlmL for control fish (Table 2). VTG
was detectable in all female fathead minnow plasma tested for VTG by ELISA.
Concentrations of VTG in females appear be similar among the lesser concentrations of
NPEO, while exposure to 3 or 10 pg NPEOIL caused up to an approximate 3-fold
decrease in VTG concentration (Figure 5A). Female VTG concentrations among
treatments were not statistically different and were difficult to discern with the numbers

of samples collected from the exposure.

33

Male plasma VTG concentrations ranged from 0.614 pg VTG/mL in the greatest
NPEO treatment (10 pg NPEOIL) to 3.17 pg VTGlmL in 0.3 pg NPEO/L (Table 2).
However, VTG concentrations for most males (SO-87.5%) within each treatment were
non-detectable in the ELISA. There were no differences in conclusions when values of
0, ‘A MDL, or MDL were used for sample values less than the VTG assay detection limit
of 0.27 pg VTG/mL. Incidence of detecting VTG in the males was not contingent upon
exposure concentration (chi-square analysis, q=2.75, chi-square=11.14, df=4). The
greatest concentration of VTG measured for any male in the experiment was 21.1 pg
VTG/mL in 1 pg NPEO/L treatment. Furthermore, the two greatest values of VTG
obtained for males were for fish that were mis-sexed at the termination of the
experiment as females and were subsequently identified as males by histology.

Although there is some overlap in the greatest concentrations of VTG in males
with the least VTG concentrations in females, mean VTG concentrations in males were
significantly less than that for females in every treatment. This indicates that mean VTG
concentrations in males was not induced up to or even a significant proportion of
concentrations observed in females at the concentrations of NPEO tested in this
experiment. In fact, the average VTG concentrations in all male fathead minnows were
10- to 100-fold less than those of females. No significant differences in male incidence

of VTG detection among treatments were detected (Figure 5B).

Plasma sex steroid hormones

Plasma concentrations of E2 for females ranged from 4058:863 pg E2lmL in the
lowest NPEO treatment (0.3 pg NPEO/L) to 989412887 pg E2lmL for fish exposed to 1

pg NPEO/L. E2 concentrations measured in the plasma of males were 5413;159 pg

34

E2lmL for fish exposed to 3 pg NPEO/L to 3891:1208 pg E2lmL for the 1 pg NPEO/L
treatment (Table 3). The detection limit in the E2 ELISA was 15.6 pg EZImL. E2 was
not detectable in 8 fish, 7 of which were males (3 from 0 pg NPEOIL, 1 from 0.3 pg
NPEO/L, 1 from 3 pg NPEO/L, and 2 from pg NPEOIL) and 1 of which was a female
from 3 pg NPEOIL. Detection of E2 in these samples was limited by the volume of
plasma available for extraction and ELISA. Contingent upon different extracts used as
internal standards, the CV for inter-assay variability for E2 was 40% but when calculated
at the mid-range or 50% of the standard curve, this variability was only 19.6% (n=4).
lntra-assay variation for E2 was 3.23%. i

All samples contained detectable concentrations of T in the ELISA with a
detection limit of 3.9 pg TlmL. Mean concentrations of T in females ranged from
31 9511097 pg TlmL (0.3 pg NPEO/L) to 9671i5178 pg TlmL (10 pg NPEOIL) (I' able 3).
Concentrations of male T overlapped the concentrations found in females and ranged
from 12591562 pg TlmL (0.3 pg NPEOIL) to 9998i5065 pg TlmL (1 pg NPEOIL). Intra-
and inter-assay coefficients of variation were 5.4 and 10.7% for the T assay (n=5).

Overall, fish sex contributed most significantly to treatment differences in male
and female plasma concentrations of E2 (F=62.20, p=0.0001) and T (F=6.53, p=0.0134)
during the exposure. Treatment differences for E2 were more evident than those for T.

For all treatments, there was a difference between male and female plasma E2
values; female E2 concentrations were always greater than those for males (Figure 6).
Plasma concentrations of 52 were similar among treatments for the females (F=1.23,
p=0.3311). Male plasma E2 concentrations were significantly greater for the 1 pg
NPEOIL exposed fish compared to all other treatments at p<0.05 except for 0.3 pg

NPEOIL, but this difference was significant at p<0.10 (p=0.0635). The greater E2

35

concentrations for these fish was similar to the greater values observed in the VTG
response in the 0.3 and 1 pg NPEO/L males.

Among all treatments except 3 pg NPEO/L, male and female fish had similar
plasma concentrations of T (Figure 7). Trends in female T concentrations reflect the
similarity observed for E2 among treatments (F=0.42, p=0.7902). Generally, males
exposed to different concentrations of NPEO also had similar circulating plasma
concentrations of T. Male T concentrations indicate a significant block effect. A lesser
mean concentration of T for males was observed from block one of 10 pg NPEO/L
(905.71 pg TlmL, n=2) than the third block where T concentrations were 3278.78 pg
TlmL (n=2). These sample sizes were small however, and the differences for 10 pg
NPEO/L in male hormone values between blocks were not observed for E2 or for E2lT
ratios. One male from 10 pg NPEO/L was excluded prior to analysis as an outlier as it
was not reliably quantified in the T EIA (BlBo<20%).

For all treatments except for 10 pg NPEO/L, mean E2lT values were greater for
females than for males (Table 4). This difference, however, was not evident for 10 pg
NPEO/L for which males and females were more similar in the ratios of the two
hormones (Figure 8). Differences among E2fl' ratios were not distinguishable with
statistical tests due to small sample sizes and great variability. The trend for female
minnows, however, appeared to be similar to the same trend observed for male VTG
and fecundity in that the middle concentrations of exposure, particularly 1 pg NPEO/L,
had greater E2lT ratios. The males did not show the same trend in E2lT but appeared

more similar among treatments.

36

Relationships between biomarkers of exposure and fish reproductive health

Relationships between traditional biomarkers of environmental endocrine
exposure and more general indicators of fish health and reproductive fitness were tested
for correlations that may be useful for further evaluation or calibration of biomarkers to
indicators more relevant on a large scale of fish fitness (Table 4). Correlation
coefficients (r) were not calculated to assign causality between biomarkers or health
indicators, but to provide some insight on calibrations that may be made in future
studies.

Overall, values for female (Table 4A) biomarkers and reproductive and health
indicators were correlated more frequently than those for males (Table 48). E2
concentrations in females was positively correlated both with T and with E2fT as well as
with VTG. Similarly, plasma T concentrations for females were negatively correlated
with E2lT. The correlation of each hormone with the ratio indicates that not one
hormone but both contribute significantly to fluctuations that are seen in the absolute
ratio of E2lT. Plasma T concentrations in the females were also positively correlated
with VTG and fecundity, but the nature of this correlation will be further explored in
examination of the role of T in reproductive physiological processes (see Discussion).
VTG was also correlated with packed cell volume of the blood (hematocrit) but not with
the ratio of E2IT or eggs/female for the females. The ratio of E2lT was not correlated
with either VTG, the typical biomarker of estrogen exposure, or with eggs/female
(fecundity).

Traditionally, these biomarkers of endocrine disruption have been applied most
widely to males. For males, none of the biomarkers were correlated with VTG or

eggs/female (Table 43). Furthermore, the only correlation evident from these analyses

37

was that T was negatively correlated with E2lT. The fact that E2 is not correlated with
the ratio implies that changes in T were more responsible for variations observed in the

E2lT ratio than changes in E2 concentrations.

DISCUSSION

Evidence of endocrine disruption in feral and caged fishes has led to the
hypothesis that alkylphenols and their ethoxylates, together with natural and
pharmaceutical estrogens, may be responsible for the effects observed below WWTP
(Purdom et al., 1994; Harries et al., 1996, 1997; Nimrod and Benson, 1996; Lye et al.,
1997). This evidence has led to many studies regarding the effects of these compounds
on estrogen receptor binding, in vitro response, and in vivo biomarkers of exposure.
Aside from studies of acute toxicities and sublethal exposures to mono— and di-
ethoxylates, NP1-17EO has not been evaluated for in vivo effects. NP9EO, the primary
constituent of the commercial blend used in this study, is a very weak estrogen agonist
as measured by its ability to stimulate in vitro production of VTG compared to
endogenous E2 and alkylphenols (Jobling and Sumpter, 1993). Although nonylphenol is
the most widely studied and the most potent estrogen agonist of the alkylphenol
ethoxylates and degradation products, the fact that the ethoxylates and carboxylates are
more water soluble, are more bioavailable to water-dwelling organisms, and are found in
greater concentrations in effluents and surface waters compared to the phenols,
suggested that it is prudent to investigate the potential endocrine disruption of these
compounds on aquatic fauna.

In the United States, the greatest concentration of NPEO in the form of NP3-

17EO from a survey of 30 rivers with suspected great concentrations of the compounds

38

was 15 pg NP3-17EOIL (Naylor et al., 1992; Nimrod and Benson, 1996). Nominal
concentrations of the exposure herein ranged from 0-10 pg NPEO/L. Since most of
these compounds require anaerobic and aerobic microbial degradation or UV photolysis
(Ahel et al., 1994a, b, c; Kvestak and Ahel, 1995), it is unlikely that the compounds are
being degraded in controlled laboratory studies. The fact that NPEO was not being
degraded was confirmed in our study in that the oligomer distribution in exposure
chambers was the same as that for the parent test compound. With only 60-75% of the
30 rivers tested having detectable levels of any of these compounds, and ranges of
NP3-17EO not exceeding 15 pg/L (Naylor et al., 1992), we are confident that our
exposure of fathead minnows overlaps with the same low concentrations found in the
environment.

The overall objective of this study was to assess the impacts, if any, of NPEO on
the reproductive output of fathead minnows, and to calibrate any effects observed with
traditional biomarkers of environmental estrogen exposure and indicators of fish' health.
Among the concentrations tested (0-10 pg NPEO/L), no significant effects were
observed on fecundity, male plasma VTG, or male or female sex steroid concentrations
or ratios. Our in vivo monitor of estrogen exposure was not intended to elucidate
questions of molecular, biochemical, and physiological mechanisms whereby NPEO
may act, but the following discussion offers some possible explanations for the trends
observed.

Although no statistically significant effects on fecundity were observed following
NPEO exposure, the response pattern of fecundity appeared to be inverted-U shaped;
the least concentration of exposure, 0.3 pg NPEO/L, had both the greatest fecundity

and the greatest number of spawning events. This trend is similar to that observed in

39

studies of metal and toxicant exposures to organisms resulting in increased growth and
fecundity of aquatic organisms. The least exposure concentrations of fish to cadmium,
DDT, or PCBs, only to name a few, have offered a stimulatory effect on fecundity,
plasma sex steroid hormone concentrations, and growth (Macek, 1968; Pickering and
Gast, 1972; Stebbing, 1981a; Weis and Weis, 1986; Kime, 1995). The inverted-U
response has not been attributed to toxicant properties but has been postulated to be an
overcompensation of homeostatic regulatory processes (Stebbing, 1981a, b). However,
the exact mechanism of toxicant-induced increases in growth and fecundity has not
been fully explored. In this study, we cannot ascertain whether the observed inverted-U
response for fecundity is real or an artifact of small sample numbers and natural
variability.

VItellogenesis has been used widely in the past two decades as a functional
indicator of in vivo and in vitro estrogen exposure, and thus has been implemented as a
biomarker of environmental estrogen exposure (Specker and Sullivan, 1994; Sumpter
and Jobling, 1995). Vitellogenesis is under control of E2 produced by follicular tissue
upon stimulation by pituitary gonadotropins. E2 produced by the follicular cells is
transported by plasma hormone binding globulins to liver hepatocytes. In the liver, EZ
binds to nuclear hepatocyte estrogen receptors (ER) causing dimerization of the
hormone receptor complex, interaction with the estrogen responsive elements (ERE) on
E2-responsive genes, and subsequent recruitment and activation of transcription factors
(Mommsen and Walsh, 1988). Transcription of estrogen-responsive and VTG genes
serves to produce and stabilize VTG mRNA (Ren et al., 1996). VTG undergoes
extensive post-translational modification before being rapidly exported to systemic
circulation where it is transported to developing oocytes. In the ovaries, VTG is

incorporated into oocytes by receptor-mediated pinocytosis where it is cleaved to

40

important egg yolk proteins and nutritional reserves for embryonic development
(Mommsen and Walsh, 1988; Lazier and MacKay, 1993).

A reliable ELISA for detection of VTG in goldfish and fathead minnows has been
developed. Molecular weight values of fathead minnow and goldfish VTG were similar
to that observed for goldfish by previous investigators (DeVIaming et al., 1980). The
similarity in the VTGs between the two species has allowed use of goldfish plasma for
standards and well-coating, and has greatly facilitated measurement of plasma VTG in a
species with small volumes of plasma. Further refinement of the method and
determination of the differences between goldfish and fathead minnow VTGs would be
useful for future analyses. The objective for developing the VTG competitive ELISA in
our laboratory was for comparative quantitation of VTG in toxicant-exposed and non-
exposed male and female fathead minnows and goldfish. Previous studies have
reported induction of VTG in males to levels of significant proportion and equal to values
for females upon exposure to environmental estrogens (Purdom et al., 1994), an effect
that was not observed in our laboratory NPEO exposure.

Concentrations of VTG following exposure to NPEO were statistically the same
across all treatments for females, although there was an observed decrease in the
amount of VTG in 10 pg NPEOIL relative to controls. The fact that E2 and T
concentrations were not statistically different among treatments does not offer an
intuitive explanation for the lesser concentration VTG. Previous investigators have
hypothesized that estrogen agonists, if acting directly in endocrine signaling pathways,
could be binding directly to ER in the liver thus preventing full-scale endogenous E2
action or could elicit negative feedback on the hypothalamic-pituitary-gonadal axis

(Folmar et al., 1996).

41

In situ and laboratory fish exposures to suspect environmental endocrine
disrupters have indicated decreases in plasma sex steroids, namely T in male fish and
E2 in female fish (Kime, 1995; Folmar et al., 1996). Typically, the concentrations of E2
in females are much greater than concentrations in males, but low levels of E2 are
found in males. Plasma hormones measured by ELISA and radioimmunoassay (RIA)
are frequently extracted to prevent interference from plasma lipids and other nonspecific
agents. Extraction of plasma steroid hormone also serves to free the hormone from sex
steroid binding globulins that control the concentrations available to tissues while in
circulation (Lim et al., 1991). In goldfish, it has been estimated that only 5% of sex
steroids remain unbound from plasma proteins and are able to cross cell boundaries for
physiological processes (Pasmanik and Callard, 1986).

Concentrations of E2 observed for males were 2 to 3-fold less than female
concentrations for all treatments of NPEO. The significantly greater concentration of E2
in males exposed to 1 pg NPEO/L and the similarity to the mean concentration for 0.3
pg NPEO/L corresponds also to the greatest VTG concentrations observed for males,
albeit several orders of magnitude less than females. Previous investigators have
observed small but measurable concentrations of VTG in the plasma of male fish, with
concentrations up to 79.8 ng VTG/mL for fathead minnows (Tyler et al., 1996) and as
much as 1 mg VTGImL in Siberian sturgeon (Acipenser baen) (Goodwin et al., 1992).
The detection of VTG in male control plasma has been postulated to be the result of
method artifacts and lack of detection of nonspecific cross-reactivity by traditional
immunoprecipitation and gel electrophoresis methods for polyclonal antisera (Goodwin
et al., 1992; Tyler et al., 1996). However, with the presence of E2 and evidence of de
novo synthesis and presence of VTG mRNA in the livers of male and immature fish

(Ren et al., 1996), it seems likely that male fish could contain low levels of plasma VTG.

42

Whether this is true remains to be investigated in light of low levels of dietary exposure
to phytoestrogens in commercially formulated diets that can also mimic the effects of
endogenous E2 (Pelissero et al., 1991).

Previously, effects of environmental estrogens have been assessed most
frequently in males stemming from concern that environmental concentrations would
greatly exceed small concentrations of endogenous estrogens and therefore have
impacts on male reproductive physiology (Fry and Toone, 1981; Guillette et al., 1994;
Jobling et al., 1996). Evaluation of relationships between traditional biomarkers of
estrogen exposure (VT G, T, E2, E2lT), and indicators of general and reproductive
health of fishes such as hematocrit and fecundity indicated stronger correlations for
females than for males. This is not surprising in view of the fact that most of these
bioindicators are more direct measures of female reproductive status. For females, both
E2 and T contributed significantly to differences observed for E2/T ratios. Similarly, E2
and T each correlated significantly with concentrations of plasma VTG, but the E2lT
ratio did not. E2 is known to be the primary hormone involved in exogenous
Vitellogenesis and oocyte assembly and is produced by aromatase enzymatic
conversion of T to E2 in follicular cells (Mommsen and Walsh, 1988). The directactions
of T in females is not completely understood but appears to peak in plasma of cyprinids
near oocyte maturation and ovulation when T is no longer needed for aromatization to
estrogen in the ovaries. Plasma T may be an important signaling factor for
neuroendocrine release of maturation hormone (gonadotropin II) and subsequent
ovulation (Rinchard et al., 1997). The observation that T was related to fecundity during
NPEO exposure may be a result of this role of T in ovulation. Finally, the relationship
between VTG and hematocrit is not surprising as much of the blood plasma is

dominated by VTG during Vitellogenesis.

43

For males, significant relationships between biomarkers and bioindicators were
not observed under the nonstimulatory effects of the low NPEO exposure. Only T was
correlated with E2rI' ratios which indicates the relative proportion and dominance of T in
the blood compared to E2 in males. None of the biomarkers or bioindicators correlated
with the reproductively relevant endpoint of fecundity, indicating that even under no
effect and control exposures, these measures are not indicative of reproductive output
of male and female pairs.

In summary, no effects were observed at nominal concentrations up to 10 pg
NPEO/L on plasma vitellogenin or fecundity of exposed fathead minnows. Since no
effects were observed, calibration of biomarkers of estrogenicity to reproductive output
and indicators of general fish health for NPEO, a xenoestrogen found in the
environment, was not relevant. The relative risk of NPEO in the surface waters of the
US on adult fathead minnows and similar species is low given environmental
concentrations and a no-observable effect concentration assumed greater than 10 pg
NPEOIL from this study. Further investigations remain to determine effects of these
exposures on more sensitive developmental stages and upon subsequent generations.
Research in the fields of endocrinology and physiology will greatly enhance our
understanding of mechanisms for the modulation of reproductive endocrinology by
environmental estrogenic substances and may better elucidate the implications of these

biomarkers for the reproductive output and fitness of fish.

ACKNOWLEDGEMENTS

Work contained herein would not have been possible without the collaborative

efforts and technical assistance from personnel in the Aquatic Toxicology Laboratory,

44

Michigan State University. A special thanks to Mr. Joe Leykam, Macromolecular
Structure Facility, MSU for guidance and use of laboratory facilities for the separation of
VTG. Previously characterized polyclonal goldfish VTG antisera was generously
donated by Dr. Glen Van Der Kraak, Department of Zoology, University of Guelph,
Ontario, who also shared invaluable VTG ELISA protocols and training. Dr. Norbert
Kaminski and Mr. Bob Crawford provided guidance for preparation of antigen and
characterization of immune response in VTG polyclonal antibody production. Carp
plasma was provided by Jean Smeets, University of Utrecht, Netherlands. This work
was funded in part by Chemical Manufacturer's Association, National Institute of Health
and Environmental Sciences (NIEHS-ES-04911), and the United States Environmental

Protection Agency Office of Water (CR-822983-01-0).

REFERENCES
Ahel, M., W. Giger, and C. Schaffner (1994a) Behaviour of alkylphenol polyethoxylate surfactants
in the aquatic environment - II. Occurrence and transformation in rivers. Wat Res. 28, 1143-
1 152.

Ahel, M., D. Hrsak, and W. Giger (1994b) Aerobic transformation of short-chain alkylphenol
polyethoxylates by mixed bacterial cultures. Arch. Env. Contam. Toxicol. 26, 540-548.

Ahel, M., F .E. Scully, J. Hoigne, and W. Giger (1994c) Photochemical degradation of nonylphenol
and nonylphenol polyethoxylates in natural waters. Chemosphere 28, 1361-1368.

Bevans, H.E., S.L. Goodbred, J.F. Miesner, S.A. Watkins, T.S. Gross, N.D. Denslow, and T.
Schoeb (1996) Synthetic organic compounds and carp endocrinology and histology in Las
Vegas Wash and Las Vegas and Callville Bays of Lake Mead, Nevada, 1992 and 1995.
USGS Water-Resources Investigations Report 96-4266, 12 pp.

Cayman Chemical. 1992a. Estradiol enzyme immunoassay kit Cayman Chemical, Ann Arbor,
MI, 28 pp.

Cayman Chemical. 1992b. Testosterone enzyme immunoassay kit. Cayman Chemical, Ann
Arbor, MI, 28 pp.

DeVlaming, V.L., H.S. WIley, G. Delahunty, and RA. Wallace (1980) Goldfish (Carassius auratus)
vitellogenin: Induction, isolation, properties and relationship to yolk proteins. Comp. Biochem.
Physiol. 67B, 613-623.

45

Folmar, L.C., N.D. Denslow, V. Rao, M. Chow, D.A. Crain, J. Enblom, J. Marcino, and L.J.J.
Guillette (1996) Vitellogenin induction and reduced serum testosterone concentrations in feral
male carp (Cyprinus carpio) captured near a major metropolitan sewage treatment plant.
Environ. Health Perspect. 104, 1096-1101.

Fry, OM. and CK. Toone (1981) DDT-induced feminization of gull embryos. Science 213, 922-
924.

Gallagher, SR. 1995. Electrophoretic separation of proteins. In: Current Protocols in Molecular
Biology, edited by FM. Ausubel, Brent, Kingston, Moore, Seidman, Smith, and Struhl, John
WIley & Sons, Inc., Harvard, Ma., pp. 10.2.1-10.2.35. ‘

Gallagher, S.R., S.E. Vlfinston, S.A. Fuller, and J.G.R. Hurrell. 1993. Immunoblotting and
immunodetection. In: Current Protocols in Molecular Biology, edited by FM. Ausubel, Brent,
Kingston, Moore, Seidman, Smith, and Struhl, John WIley 8 Sons, Inc., Harvard, Ma., pp.
10.8.1-10.8.17.

Giger, W., P.H. Brunner, and C. Schaffner (1984) 4-Nonylphenol in sewage sludge: accumulation
of toxic metabolites from nonionic surfactants. Science 225, 623-625.

Gimeno, S., A. Genitsen, T. Bowmer, and H. Komen (1996) Feminization of male carp. Nature
384, 221-222.

Goede, R.W. and BA. Barton (1990) Organismic indices and an autopsy-bases assessment as
indicators of health and condition of fish. American Fisheries Society Symposium 8, 93-108.

Goodwin, A.E., J.M. Grizzle, J.T. Bradley, and B.H. Estridge (1992) Monoclonal antibody-based
immunoassay of vitellogenin in the blood of male channel catfish (lctalurus punctatus). Comp.
Biochem. Physiol. 101B, 441-446.

Gray, MA. and CD. Metcalfe (1997) Induction of testis-ova in Japanese medaka (Oryzias Iatipes)
exposed to p-nonylphenol. Environ. Toxicol. Chem. 16, 1082-1086.

Guillette, L.J., T.S. Gross, G.R. Masson, J.M. Matter, H.F. Percival, and AR. Woodward (1994)
Developmental abnormalities of the gonad and abnormal sex hormone concentrations in
juvenile alligators from contaminated and control lakes in Florida. Environ. Health. Persp.
102, 680-688.

Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, E.J. Routledge, R. Rycroft, J.P.
Sumpter, and T. Tylor (1996) A survey of estrogenic activity in United Kingdom inland waters.
Environ. Toxicol. Chem. 15, 1993-2002.

Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, J.P. Sumpter, T. Tylor, and N.
Zaman (1997) Estrogenic activity in five United Kingdom rivers detected by measurement of
Vitellogenesis in caged male trout. Environ. Toxicol. Chem. 16, 534-542.

Jobling, S. and JP. Sumpter (1993) Detergent components in sewage effluent are weakly
oestrogenic to fish: an in vitro study using rainbow trout (Oncorhynchus mykiss) hepatocytes.
Aq. Toxicol. 27, 361-372.

46

Jobling, S., D. Sheahan, J.A. Osborne, P. Matthiessen, and JP. Sumpter (1996) Inhibition of
testicular growth in rainbow trout (Oncorhyncus mykiss) exposed to estrogenic alkylphenolic
chemicals. Environ. Toxicol. Chem. 15, 194-202.

Johnson, L.L., E. Casillas, T.K. Collier, B.B. McCain, and U. Varanasi (1988) Contaminant effects
on ovarian development in English sole (Parophrys vetulus) from Puget Sound, Washington.
Can. J. Fish. Aquat. Sci. 45, 2133-2146.

Johnson, L.L., S.Y. Sol, D.P. Lomax, G.M. Nelson, C.A. Sloan, and E. Casillas (1997) Fecundity
and egg weight in English sole, Pleuronectes vetulus, from Puget Sound, Washington:
influence of nutritional status and chemical contaminants. Fish. Bull. 95, 231-249.

Kime, DE. (1995) The effects of pollution on reproduction in fish. Rev. Fish Biol. Fish. 5, 52-96.

Kramer, V.J. (1996) Xeno—estrogens in the aquatic environment: development and application of
in vitro and in vivo bioassays. Ph.D. Dissertation, Michigan State University, East Lansing,
MI, 129 pp.

Kvestak, R. and M. Ahel (1995) Biotransforrnation of nonylphenol polyethoxylate surfactants by
estuarine mixed bacterial cultures. Arch. Environ. Contam. Toxicol. 29, 551-556.

Lazier, CB. and ME. MacKay (1993) Vitellogenin gene expression in teleost fish. In: Biochemistry
and Molecular Biology of Fishes, edited by P.W. Hochachka and T.P. Mommsen. Elsevier
Science Publishers, New York, pp. 391-405.

Lim, E.H., J.L. Ding, and T.J. Lam (1991) Estradiol-induced vitellogenin gene expression in a
teleost fish, Oreochromis aureus. Gen. Comp. Endocrinol. 82, 206-214.

Lorenzen, A. and SW. Kennedy (1993) A fluorescence-based protein assay for use with a
microplate reader. Anal. Biochem. 214, 346-348.

Lye, C.M., C.L.J. Frid, M.E. Gill, and D. Mcconnick (1997) Abnormalities in the reproductive health
of flounder Platichthys flesus exposed to effluent from a sewage treatment works. Mar. Poll.
Bull. 34, 34-41.

Macek, K.J. (1968) Reproduction in brook trout (Salvelinus fontinalis) fed sublethal concentrations
of DDT. J. Fish. Res. Board Can. 25, 1787-1796.

Maisse, G., B. Mourot, B. Breton, A Fostier, O. Marcuzzi, P.Y. Le Bail, J.L. Bagliniere, and A.
Richard (1991) Sexual maturity in sea trout, Salmo trutta L., running up the River Calonne
(Normandy, France) at the 'finnock' stage. J. Fish Biol. 39, 705-715.

Mananos, E., J. Nunez, S. Zanuy, M. Carrillo, and F. Le Menn (1994) Sea bass (Dicentrarchus
labrax L.) vitellogenin. II - Validation of an enzyme-linked immunosorbent assay (ELISA).
Comp. Biochem. Physiol. 107B, 217-223.

Mommsen, T.P. and P.J. Walsh (1988) Vitellogenesis and oocyte assembly. In: Fish Physiology,
Vol. XIA, edited by W.S. Hoar and V.J. Randall. Academic Press, San Diego, pp. 347-406.

Mourot, B. and P.-Y. Le Bail (1995) Enzyme-linked immunosorbent assay (ELISA) for rainbow
trout (Oncorhyncus mykiss) vitellogenin. J. Immunoassay 16, 365-377.

Naylor, C.G., J.P. Mieure, W.J. Adams, J.A. Weeks, F.J. Castaldi, L.D. Ogle, and RR. Ramano
(1992) Alkyl ethoxylates in the environment. JAOCS 69, 695-703.

47

Nimrod, AC. and W.H. Benson (1996) Environmental estrogenic effects of alkylphenol
ethoxylates. Crit. Rev. Toxicol. 26, 335-364.

Pasmanik, M. and G. Callard (1986) Characteristics of a testosterone-estradiol binding globulin
(TEGB) in goldfish serum. Biol. Reprod. 35, 838-845.

Pelissero, C., F. Le Menn, and S. Kaushick (1991) Estrogenic effect of dietary soya bean meal on
Vitellogenesis in cultured Siberian Sturgeon Acipenser baeri. Gen. Comp. Endocrinol. 83, 447-
457.

Pickering, Q.H. and M.H. Gast (1972) Acute and chronic toxicity of cadmium to the fathead
minnow (Pimephales promelas). J. Fish. Res. Board Can. 29, 1099-1106.

Purdom, C.E., P.A. Hardiman, V.J. Bye, N.C. Eno, C.R. Tyler, and J.P. Sumpter (1994)
Estrogenic effects of effluents from sewage treatment works. Chem. Ecol. 8, 275-285.

Ren, L., S.K. Lewis, and J.J. Lech (1996) Effects of estrogen and nonylphenol on the post-
transcriptional regulation of vitellogenin gene expression. Chem-Biol. Interactions 100, 67-76.

Rinchard, J., P. Kestemont, and R. Heine (1997) Comparative study of reproductive biology in
single and multiple-spawner cyprinid fish. ll. Sex steroid and plasma protein phosphorus
concentrations. J. Fish Biol. 50, 169-180.

Rodriquez, J.N., O. Kah, M. Geffard, and F. Le Menn (1989) Enzyme-linked immunosorbent
assay (ELISA) for sole (Solea vulgaris) vitellogenin. Comp. Biochem. Physiol. 928, 741 -746.

Routledge, E.J. and J.P. Sumpter (1996) Estrogenic activity of surfactants and some of their
degradation products assessed using a recombinant yeast screen. Environ. Toxicol. Chem.
15, 241-248.

Silversand, C. and C. Haux (1989) Isolation of turbot (Scophthalmus maximus) vitellogenin by
high-performance anion-exchange chromatography. J. Chromatogr. 478, 387-397.

Specker, J.L. and C.V. Sullivan (1994) Vitellogenesis in fishes: status and perspectives. In:
Perspectives in Comparative Endocrinology. edited by K.G. Davey, R.G. Peter, and SS.
Tobe. National Research Council of Canada, Ottawa, pp. 304-315.

Stebbing, A.R.D. (1981a) Horrnesis - stimulation of colony growth in Campanularia fiexuosa
(Hydrozoa) by copper, cadmium, and other toxicants. Aq. Toxicol. 1, 227-238.

Stebbing, A.R.D. (1981b) Stress, health and homeostasis. Mar. Poll. Bull. 12, 326-329.

Sumpter, J.P. and S. Jobling (1995) VItellogenesis as a biomarker for estrogenic contamination of
the aquatic environment. Environ. Health Perspect. 103(Suppl. 7), 173-178.

Tyler, OR, B. van der Eerden, S. Jobling, G. Panter, and J.P. Sumpter (1996) Measurement of
vitellogenin, a biomarker for exposure to oestrogenic chemicals, in a wide variety of cyprinid
fish. J. Comp. Physiol. B 166, 418-426.

Weeks, J.A., W.J. Adams, P.D. Guiney, J.F. Hall, and CG. Naylor (1996) Risk assessment of
nonylphenol and its ethoxylates in US. river water and sediment. SC Johnson 8 Son, Inc.
Report, 17.

48
Weis, P. and J.S. Weis (1986) Cadmium acclimation and horrnesis in Fundulus heteroclitus during
fin regeneration. Environ. Res. 39, 356-363.

White, R., S. Jobling, S.A. Hoare, J.P. Sumpter, and MG. Parker (1994) Environmentally
persistent alkylphenolic compounds are estrogenic. Endocrinol. 135, 175-182.

49

i
230n M

 

 

 

 

 

 

 

 

 

254nm a

8

3

280nm 8
o' " '5' '7 '1'3'7'15772'07' ' '257'33' ' 3540 45'” '53

Time (min)

Figure 1. Anion exchange high performance liquid chromatogram for the separation of
VTG from E2-induced male and female goldfish. Arrow indicates the VTG protein
eluted from the column with a 0 to 0.50 M Tris-Cl linear gradient. Detection of the

proteins was measured by absorbance at wavelengths of 230, 254, and 280 nm.

50

e2gf mfiwm ffhm #26 #2-26 mgf e29f
3:125" “"7"? ’1‘"

’ y."

T‘ I

    

Figure 2. SDS-PAGE Western immunoblot with rabbit anti-goldfish VTG polyclonal
antibodies. Protein bands represent E2 induced goldfish plasma (E29f), male
goldfish (mgf), female goldfish (fgf), and purified goldfish VTG fractions (#2-26 and

#26). Proteins were separated by continuous SDS-PAGE (4-15% Tris-tricine).

51

1.4 n-

1.2 ..

0.8 --

% bound
0
O)

0.4 4-

0.2 ,.

 

 

log dilution or log vtg (nglmL)

Figure 3. VTG standard curve and dilution curves of goldfish and fathead minnow male
and female plasma for determination of curve parallelism. Note that x-axis is. the log
transformed concentration of VTG (nglmL) for standard goldfish VTG or dilution
factor of plasma samples. (+ = standard GF VTG; O = E2 induced goldfish; * =
female goldfish; CI = female fathead minnow (recrudesced); O = female fathead

minnow (not recrudesced); A = male goldfish; x = male fathead minnow)

52

Table 1. Oligomer distribution of nonylphenol ethoxylate (Surfonic N-95) standard.
Values are expressed as percent area by normal phase high-performance liquid

chromatography.

 

Component Standard

 

(% area)

NP 0.14
NP1EO 0.12
NPZEO 0.32
NP3EO 0.95
NP4EO 1.72
NP5EO 3.69
NP6EO 7.27
NP7EO 9.90
NPBEO 10.53
NP9EO 10.73
NP10EO 10.45
NP11EO 9.38
NP12EO 8.47
NP13EO 7.64
NP14EO 5.97
NP15EO 5.42
NP16EO 4.61
NP17EO 2.71

TOTAL 100.02

 

53

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

250 .. (10)
200 .. (4)
'I'
o 156 (4)
- _, 5
g 150 a... (3) I— I ) ..
“- " 113
R
g 100 .. "I— 93
: ‘—— b . _.
8 T
2 50 ..
IL
0 T i 3 IL i
0 03 1 3 10
NPEO pg/L

Figure 4. Fecundity of fathead minnows exposed to NPEO. Data are means 3:
standard errors for replicate blocks of NPEO treatments. Values in parentheses
above histograms represent the number of spawning events contributing eggs to
total fecundity. No differences in egg production were observed among treatments.
One replication of 3 pg NPEO/L was lost completely to fungal infection and no

females remained in one replication of 10 pg NPEOIL.

Table 2. Plasma VTG for male and female fathead minnows exposed to 0, 0.3, 1, 3,

and 10 pg NPEO/L for six weeks.

Means (1SEM), ranges, and percent of

observations greater than the detection limit (MDL) are presented for all treatments.

(T RTMT=pg NPEO/L).

 

 

 

TRTMT FEMALES MALES
n Range Mean %>MDL‘ n Range Mean %>MDL"
(nglmL) (nglmL) (nglmL) (nglmL)
0 9 3.40-2.266.1 895.055 100 6 n.d.-7.59 1.7225 33.3
(1271.67) (11.22)
0.3 8 27.8-25482 869.78 100 7 n.d.-9.36 3.1697 50
(£296.07) (11.576)
1 8 58.1-2,051.7 885.307 100 8 n.d.-21.1 2.7526 12.5
($262.59) (12.618)
3 8 4.28-1,535.0 651.186 100 4 n.d.-4.35 1.191 25
(i229-05) (11.056)
10 5 531-6993 291.745 100 5 n.d.-3.15 0.6149 43
(1133.62) (10.424)

 

'Percent of total greater than the method detection limit (%>MDL) are calculated as percent of
total number of fish measured for VTG for each sex.

55

Figure 5. Plasma VTG concentrations (pglmL) in fathead minnows exposed to
nonylphenol ethoxylate. A. Mean plasma VTG concentrations in females for each
NPEO exposure. B. Mean plasma VTG for males. Means 1 SEM are illustrated.
For all treatments, females had significantly higher concentrations of plasma VTG
than males. No differences in VTG were observed for males or females among

treatments.

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A lmu

.1 I.

. 1

u l.
i- .
._
p — b
. F.
F-bP-p
qdqqdq
mmmmmmo
2 8642
11

3253 05 95%

NPEO pg/L

 

 

 

3.59: 0.5 mEmma

NPEO pglL

57

Table 3. Plasma concentrations of E2 and T and ratios of E2 to T (E2lT) in male and

female fathead minnows exposed to NPEO (mean 1 SEM).

 

 

Trtmt. FEMALES MALES
n E2 T E2/T n E2 T E2/T
(pglmL) (pglmL) (Pg/ml-I (99’le

 

0 8 5485.42 7331.21 1.422 7 1094.11 5374.67 0.838
(1914.17) (13291.80) (10.297) (1539.02) (12628.04) (10.574)

0.3 6 4057.89 3195.39 1.798 8 1183.58 1381.65 1.063

(1863.02) (11097.07) (10.506) (1373.15) (1264.22) (10.584)
1 6 9894.10 7102.58 3.002 7 3891.22 9997.89 1.586
(12886.89) (13739.45) (10.971) (11207.98) (15065.04) (10.778)

3 8 5956.58 4608.42 1.447 4 541.40 1258.81 1.202
(11824.55) (11514.20) (10.286) (1159.30) (1562.34) (10.854)
10 5 6477.28 9670.70 1.228 8 630.28 2342.37 1.456

(14191.52) (15177.60) (10.301) (1139.90) (1603.97) (10.854)

 

58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ufemales

14000 q- Ima|es

12000 ..
A * WP
75' 10000 .. -—
3::
g 8000 .. J .. *
0) * ——
g 6000 ..
g ‘ *
O.

4000 -- I
0 1 3 10
NPEO treatment (pg/L)

Figure 6. Plasma E2 (pg E2lmL) concentrations in male and female fathead minnows
following exposure to NPEO. Data are plotted as untransformed means 1 SEM.
(Statistical differences are for Iogw-transfon'ned data: *=significant differences
between males and females within the same exposure group; O=significant
difference of 1 pg NPEOIL males from all other male treatments except for 0.3 pg

NPEOIL males.)

59

16000 -

I

-- ,_ afemales

14000 - .males

I

12000 -

 

10000 -

8000 4

 

6000 --

 

 

 

4000 -r ‘-

plasma testosterone (pglmL)

 

 

 

2000 1

 

 

 

 

 

 

 

 

j-

0 0.3 1 3 10
NPEO 1130013001 (ngL)

 

Figure 7. Plasma T (pg TlmL) concentrations in male and female fathead minnows
following exposure to NPEO. Data are plotted as untransformed means 1 SEM.
(Statistical differences are based upon logo-transformed data: *=significant
differences between males and females within the same exposure group (3 pg

NPEOIL); No significant differences were observed among males or females

exposed to different concentrations of NPEO.)

60

3.5 «-

Dfemales

2.5 ~-

 

EZIT

 

1.5 .L

 
 

0.5 ~-

 

 

 

 

 

 

     

1

NPEO treatment (pglL)

Figure 8. E2 to T (E2lT) ratios for male and female minnows following exposure to
nonylphenol ethoxylate. Data are plotted as means 1 SEM. No differences

between E2lT ratios were observed within or among treatments and sexes.

61

Table 4. Correlation coefficients (r) for females (4A) and males (48) between
biomarkers of estrogenicity and indicators of general health and fecundity of
fathead minnows exposed to 0-10 pg NPEO/L. E2 and T concentrations were log-
transformed and coefficients between these, hematocrit, and E2lT were parametric
Pearson product-moment correlations. All other correlations were based upon
ranks and are Spearman rank correlation coefficients. (E2=estradiol,

T=testosterone, VTG=vitellogenin, '=Spearman’s Rho coefficients).

Effie?
til.) 1'
re 509-
met:

I up?

. 4: A.
IIGUW‘

62

A. Correlation coefficients for females exposed to NPEO.

 

 

Comparison E2 T E2lT VTGa
E2 1 .000
T 0.41 9“ 1 .000
E2fl' ratio 0.486” -0.450“ 1.000
VTG" 0.329” 0.381 ** 0.039 1.000
Eggs/Female 0.333 0.561“ -0.383 0.089
Hematocrit -0.093 -0.135 0.053 0.322'

 

*Significance is 0.05<p<0.10 for correlations.
”Significance is p<0.05.

B. Correlation coefficients for males exposed to NPEO.

 

 

E2 T EZIT VTG‘
E2 1 .000
T 0.186 1.000
E2lT ratio 0.315 -0.643** 1.000
VTG" -0.139 -0.079 0.171 1.000
Eggs/Female 0.239 0.105 0.112 -0.087
Hematocrit -0.259 -0.094 0.004 0.060

 

“Significance is p<0.05.

CHAPTER 2

Effects of Municipal Wastewater Exposure in situ on the Reproductive Physiology
of the Fathead Minnow, Pimephales promelas

(To be submitted in modified form to Environmental Toxicology and Chemistry)

INTRODUCTION

Evidence of intersex in feral male roach (Rutilus nrtilus) and observations of
estrogenic activity below wastewater effluents in the United Kingdom (UK) [1-3] have, led
to in situ evaluations of municipal and industrial wastewater treatment plant (VVWT P)
effluents on the reproductive physiology of fishes. Alkylphenols and their parent
alkylphenol ethoxylate compounds used in industrial cleaning and wool processing
plants as well as natural endogenous and pharmaceutical estrogens, estradiol-17B (E2),
estrone, and ethinyl estradiol, have all been implicated in the anomalies observed
primarily in feral or caged male fishes below WWTP in the UK [1,2]. Investigators have
found up to 500 to 500,000 and even one-million-fold induction of the female-specific
' egg-yolk protein, vitellogenin (VT G), with the greatest inductions in caged male rainbow
trout (Oncorhyncus mykiss) [1-3] and lesser inductions in carp (Cyprinus carpio) [1], and
in feral male flounder (Platyichthys flesus) [4] maintained in and downstream from
VWVTP effluents. The observed effects are consistent with exposure to estrogen
agonists, however, the definitive cause and mechanisms of the observed endocrine
disruption below WWTP remains to be determined. The overwhelming evidence of

63

64

endocrine disruption in male fishes in the UK has led to a flurry of similar studies in the
United States (US).

In the US, feral male and female carp have been evaluated for exposure to
environmental estrogens by measurement of plasma VTG concentrations, estrogen and
androgen plasma concentrations and ratios, and gonadal histopathological lesions [5-7].
Results have indicated inductions in serum VTG concentrations concomitant with
significantly decreased concentrations of serum androgen concentrations for male carp
[5,6]. However, a widespread survey of carp in US streams across the nation failed to
detect significant inductions in male plasma VTG to those concentrations found in
females. From this survey the authors discovered that the most obvious effects and
differences among sites tested for environmental endocrine disruption were modulations
in ratios of E2 to testosterone (1') or 11-ketotestosterone (11-KT) for females and males
(respectively) and subsequent correlations of these ratios with body burdens of
organochlorine pesticides and dissolved water concentrations of pesticides [7].
However, estrogen to androgen ratios have been only poorly associated with VTG, the
most widely used biomarker of exposure of fish to estrogen agonists [1,7].

Vitellogenin is a high molecular weight glycolipophosphoprotein that is
synthesized in the liver of oviparous vertebrates by E2-reponsive genes for
incorporation into developing oocytes. Both male and female fish have the capacity to
produce VTG, but great quantities of the protein are only produced by females due to
greater endogenous concentrations of plasma E2 [8,9]. Modulation of VTG gene
expression by E2 has been relatively well-characterized in teleost fish. The binding of
E2 to the estrogen receptor (ER) in the nucleus of hepatocytes, subsequent dimerization
and binding to estrogen responsive elements (ERE) on the VTG gene(s), transcriptional

activation, and post-translational modification of the protein have been well-described

65

and subsequently have been a model of estrogen action in vitro and in vivo [10]. Since
male hepatocytes are responsive to E2 and can produce VTG upon stimulation by E2 or
E2-Iike substances [11], VTG has been used as a biomarker of endogenous and
exogenous estrogen agonist exposure in fishes [12,13]. Similarly, the sex steroid
hormones, T, 11-KT, and E2 all play roles in gametogenesis and steroidogenesis during
gonadal recrudescence and maturation and spawning in fish. The roles of the
hormones have been intensively evaluated, individually and collectively, for their
generalized roles on the reproductive physiology of fishes, and thus have been utilized
for evaluation of reproductive function [5].

The use of these biomarkers of exposure to environmental endocrine modulating
substances, particularly xenoestrogens, has been widespread with the evidence of
endocrine disruption below WWTP. The results of these studies have indicated that
endocrine disruption by municipal wastewater in the US is not nearly as definitive and
widespread as in the UK. Prior to regulation of suspect environmental endocrine
modulating chemicals, it is imperative to observe the effects of these compounds
singularly and in combination, under laboratory conditions and in situ to determine the
relative risk of the effluents for the overall reproductive physiology, fitness, and survival
of fishes. The objective of the investigation reported here was to observe the effects of
mid-Michigan WWTP effluents, in situ, on the reproductive physiology of caged male
and female fathead minnows, Pimephales promelas, with traditional ecotoxicological
biomarkers of endocrine disrupter exposure and indices of fish health. This investigation
was intended to provide further information on the incidence and degree of endocrine
disruption below VWVTP for a region typical of Midwestern US streams and WWTP

treatment technologies and types of consumer influents.

66

MATERIALS AND METHODS

Fish

Adult fathead minnows were obtained from natural populations in Pond 1 at the
Inland Lakes Teaching and Research property on the campus of Michigan State
University, East Lansing, Michigan (southeastem-most pond; hereinafter referred to as
the Limnology Pond; LP). Fathead minnows were chosen for use because of their
representation of fishes indigenous to mid-Michigan streams, small size, and suitability
for comparison to laboratory exposures with single chemicals purported to be
estrogenic. Minnows were collected from baited aluminum minnow traps set overnight
immediately before transporting the fish to WWTP field sites. Fish selected for field
exposure were randomly chosen mature adults with obvious secondary sexual

characteristics.

Study site selection

Municipal WWTP in mid-Michigan were chosen based upon type and extent of
treatment, volume treated per day, types of wastewater influents received by the
facilities (i.e. industry, commercial, residential), accessibility, and susceptibility for cage
vandalism. Seven WWTP located in 3 different Michigan watersheds were chosen for
study (Figure 9). Average effluent discharges ranged from 174 to 88,950 m3lday (Table
5). These facilities included the Bellevue (BV), Delhi Township (DL), Eaton Rapids
(ER), Lansing (LA), Owosso (OW), Portland (PT), and Williamston (WM) municipal

WWTPs. All plants were using at least secondary treatment technologies as required by

67

the Clean Water Act and some sites (DL, LA, and OW) employed tertiary treatment such
as sand or charcoal filters to remove excess biological oxygen demand (BOD). Two
WWTP (BV and OW) were utilizing tricking filter secondary treatment technologies while
the remainder of the sites employed activated sludge treatment. One reference site
(RS) on the Looking Glass River in Eagle, Michigan was chosen based upon similarity of
stream characteristics to receiving waters at WWTP sites and distance downstream
from probable point-sources of municipal or industrial wastewater. LP was also used as

a reference site for comparison of pre-treatment and pre-confinement conditions.

Exposure

Fathead minnows were transported to field sites in June and July 1996 and were
placed at the sites in 0.60 m X 0.60 m X 0.45 m cages constructed from marine
plywood, galvanized threaded rod and polyethylene mesh and weathered prior to use.
One cage was placed at each exposure site. Approximately 35 male and 35-female
fathead minnows were placed in each cage. Cages were placed in the receiving river
directly in the flow of the WWTP effluent stream to maximize exposure. This preliminary
study of the effects of WWTP effluents evaluated only the effects on fish placed directly
in the effluent. The study design was such that, if significant effects in the VTG and
hormone responses of males and females were observed, more extensive studies
would be conducted including necessary upstream controls and downstream grades of
exposure.

Cages were left in place for 21 d. During the exposure, sites were checked

weekly for fish mortality and health and cage vandalism. Fish were not fed and obtained

68

food only from that which was present at the site. Water quality parameters including
dissolved oxygen (DO), temperature, and water hardness were measured weekly at
each site. Effluent water quality data was collected from Discharge Monitoring Reports
(DMRs) required by the National Pollutant Discharge Elimination System (NPDES)
permits obtained from the Department of Environmental Quality (BBQ) in the State of

Michigan.

Sample Collection

At the end of the 3 week exposure, fish were removed from cages and
immediately euthanized with a lethal concentration of Finquel® MS-222 (tricaine
methane sulfonate, Argent Chemical Laboratories, Redmond, WA). All sampling was
conducted in the early morning (0700h to 0900h) at the sites to ensure comparability of
VTG and hormone concentrations for fish at each site. Standard length (L), weight (W),
and sex were recorded for each fish. The general health and condition of the fish,
including any gross morphological abnormalities were noted. Condition factors (K) were
calculated using standard lengths and weights (K=WIL3 *10,000) [14].

Following severance of the caudal fin at the caudal peduncle with a sharp blade,
blood was immediately collected into heparinized hematocrit tubes and sealed.
Hematocrit tubes were placed in 15 mL centrifuge tubes and stored on ice for clotting
and transport to the laboratory. Fish were placed in Bouin’s fixative for later gross
morphological and histological examination. Upon return to the laboratory, blood was
centrifuged at 3000 X9 for 10 min in 4°C to separate plasma. Plasma was removed

from the hematocrit tubes and stored at -80°C until VTG and hormone analyses.

69

Vitellogenin enzyme-linked immunosorbent assay (ELISA)

Goldfish (Carassius auratus) VTG was induced in males and females and was
isolated by anion exchange high performance liquid chromatography (HPLC) according
to methods adapted from Silversand and Haux [15] and previously described [16].
Polyclonal antibodies were raised in New Zealand white rabbits against the [purified
goldfish VTG [16].

Concentrations of plasma VTG were determined by a competitive enzyme-linked
immunosorbent assay (ELISA) developed for goldfish and fathead minnows and was
previously described [G. Van Der Kraak, pers. comm.; 16-18]. VTG was quantified by a
linear regression equation calculated from the log-logit transformation of ng VTG/well
goldfish standard versus absorbance calculated as a percent maximum binding (%BlBo)
and subsequent correction for sample dilution [16,19]. Purified goldfish VTG standards
(0.14-69.4 ng VTG/well) were assayed at least in duplicate on each plate. All samples
were assayed also in duplicate and any sample or standard exceeding 20% variation
between readings was discarded and/or re-assayed. Intra- and inter-assay coefficients

of variation were 10.3% (n=5) and 25.6% (n=11), respectively.

Plasma sex steroid hormone analysis

E2 and T were assayed in the plasma of fathead minnows by ELISA (Cayman
Chemical, Ann Arbor, MI). Remaining plasma after VTG determination and not more
than 50 pL was transferred volumetrically, diluted to 1.0 mL with nanopure water, and

then extracted twice with 5 mL of diethyl ether. Extracts were blown to complete

70

dryness under nitrogen and reconstituted in 300-500 pL EIA buffer (0.1 M PBS, 0.1%
Tween-20) provided in the kit depending on the volume of plasma used for extraction.
Protocols for E2 and T ELISAs are previously described [16] and follow instructions
provided in the EIA kits.

E2 (7.8 to 1000 pg E2lmL) and T (3.9 to 500 pg TlmL) standard curves were
assayed in duplicate on each plate. An internal standard of pooled male and female
goldfish plasma extracted as above was diluted and assayed in at least triplicate on
each plate for determination of variability between assays. Log-logit transformation of
standard hormone concentrations (pglmL) on % maximum binding (%B/Bo) calculated
from absorbance units as that for VTG were plotted to calculate a linear regression
model used for determination of steroid hormone concentrations in the plasma.
Samples were run in at least duplicate and any sample or standard exceeding 20%
coefficient of variation (CV) was rerun. Samples below 20% binding on the standard
curve were diluted 50% and re-assayed. Intra- and inter-assay CV's were calculated
from pooled goldfish internal standards. lntra- and inter-assay CV‘s were 5.0 and 17.3%

(n=9) for T and 4.3 and 29.6% (n=11) for E2.

Statistical analyses

Overall comparisons of VTG and hormone concentrations among WWTP sites
were made with SAS statistical software (Cary, North Carolina, USA). Tests for
homogeneity of variance and normality (PROC UNIVAR-IATE) resulted in the use of a
nonparametric one-way ANOVA (Kruskal-Wallis) which was run for ranks on VTG and
female E211” for fish at each site. Painrvise comparisons were made between each of

the sites using a modified nonparametric t statistic or Tukey’s HSD. Plasma T and E2

71

values were logo-transformed and one-way ANOVA (PROC GLM) was used to assess
differences among sites and sexes. Painrvise comparisons were subsequently made
with Tukey-Kramer tests using least squares means of the log-transformed values. The
method detection limits (MDL) for the VTG and hormone ELISA's were determined by
paired t-tests to determine the lowest standard different in absorbance units compared
to maximum binding (8,) values. For values less than the VTG or hormone ELISA
MDLs, a value of ‘/z MDL was substituted for determination of treatment means and
subsequent statistical analyses. Incidence of male plasma VTG above the MDL among
sites was determined with chi-square analysis. Pairwise comparisons for incidence
above the detection limit between reference sties (LP and RS) and exposure sites were
made with a Bonferroni chi-square contingency analysis with continuity correction.
Correlation coefficients were calculated between pairs of biomarkers (VT G, E2, T, E2lT.
and K). Pearson product-moment correlations were made for all data with bivariate
normal distributions. All non-normal data, including VTG, and female E2lT were ranked
and the Spearman's Rho correlation test was applied. For all statistical tests, probability
for type I error was set to 0.05. All data were reported as means 1 the standard error of

the means (SEM) unless otherwise indicated.

RESULTS

Study site water quality characteristics

In general, effluent and stream water quality parameters and temperatures were
similar among sites throughout the study period (June-August 1996) (T able 6). Effluent
water quality values appear different, but in-stream water quality values illustrate that

ambient exposure waters abated extreme values in DO and pH, particularly for PT and

72

ER with very low values of DO._ OW had relatively high total ammonia (0.4 mglL) and
high pH compared to other WWTP. Chlorination, often used for disinfection of the final
effluent, was greatest at PT at a concentration of 0.9 mglL. Mean water temperature
ranged from 20°C at WM to 236°C at LP; water temperature was consistently greater
in LP compared to the streams into which the fish were placed. Variations in mean
temperatures, however, could not explain the differences or variability in biomarker
values observed among sites. From the WWTP Discharge Monitoring Reports
summarized in Table 6, it appears that all WWTP were efficiently removing BOD, a
general measure of wastewater treatment efficiency; for all VWVTP selected and
reporting percent removal, BOD was on average 90.3% to 95.3% removed by the
treatment process. Eighty-five percent removal is the minimum accepted by NPDES

permit requirements.

Fish survival and condition

Survival at each of the sites ranged from 20% (BV and OW) to 68% (RS) (Table
7). Fish that did not survive were not completely accounted for since most of the fish
were decomposed or were forced through the mesh of the cages by high stream flows.
Moreover, survival during the study was not completely indicative of mortality of the
organisms due to exposure to the effluent; there was evidence of escapement, mortality
due to predation by other aquatic organisms, and some vandalism. Vandalism was a
problem at several of the sites due to the location and conspicuous nature of the cages.
The fish placed below the Lansing WWTP were lost completely due to vandalism. Only
one female survived at the OW WWTP preventing any statistical analyses to distinguish

differences between OW females and other sites.

73

Condition factors (K) among exposure sites, overall, were not significantly
different than the reference sites, LP or R8 (T able 7). Variability in K among sites was
attributed mostly to differences among sites rather than between sexes. For DL
(p<0.05) and BV (p<0.10) only, K was lower for females than for males. Differences in
exposure site K compared to LP and RS are indicated in Table 7 by Tukey grouping.
RS had a significantly lesser K than LP, and was similar to values of K observed at all
WWTP sites except for OW and WM. Conversely, K at all sites different from RS were

similar to that observed at LP.

Plasma VTG concentrations

Exposure of fathead minnows to WWTP effluent did not induce VTG in males
relative to reference sites and concentrations of VTG in the plasma of males remained
much less than those in females. Vitellogenin concentrations (pg VTG/mL) in females
from RS and LP ranged from 324.8 to 2463 pg VTG/mL and 777.7 to 3406.3 pg
VTGlmL, respectively (Table 8). Females at BV, ER, and PT had significantly lesser
plasma VTG concentrations than those at the two reference sites (Figure 10A).
Females from sites DL and WM also had lower plasma VTG levels than the reference
sites; this difference was statistically significant when compared with LP but not with
RS.

Male plasma VTG concentrations ranged from less than the MDL of 0.27 pg
VTG/mL to maximum concentrations that were an order of magnitude less than those in
females (Table 8). Seventy-three percent of all male VTG measurements from all sites
combined were not detectable by the VTG ELISA. Proportions of male fish in which

VTG was detectable in the ELISA ranged from 0% at BV and DL to 62% at LP. The

74

proportion of male fathead minnows with detectable concentrations of VTG varied
among sites (Q=309.45, p<<0.001). Pairwise comparisons of values less than the MDL
indicated a significant difference between the two reference sites (LP and RS), followed
by a significant greater number of non-detects for BV, DL, and for PT compared to LP
but not to RS. The power of this pairwise chi-square analysis was not compromised by
the fact that the two reference sites were not similar in the percentage of non-detects.
The differences in incidence of detection rather than differences in VTG concentrations
are denoted in Figure 10B. Values of detectable plasma VTG in males did not overlap
with VTG concentrations found in females within any site. Male VTG concentrations at
the reference sites (LP and RS) ranged from less than the MDL for both sites to 4.79
and 0.94 pg VTGlmL, respectively. The maximum concentration of male VTG among

all sites was 29.61 pg VTG/mL for an individual at the OW WWI'P.

Plasma sex steroid hormone levels

Plasma hormone concentrations were detectable for almost all fish evaluated by
ELISA (Table 9). Only 4 fish had non-detectable concentrations of E2 and all of these
fish were females from RS and WM for which detection was limited by small volumes of
plasma. For determinations of total plasma T, only one male (WM) was non-detectable
in the assay, and re-assaying was not possible due to insufficient extract. Overall
comparisons among sites and between sexes indicated that both sex and site
contributed to variations observed in concentrations of E2 and T, and the E21T ratio.

Mean E2 concentrations for females ranged from 3615 pg E2/mL at BV to

11,120 pg E2lmL at LP. Males generally overlapped the females in total E2 within each

75

site and means ranged from 1139 pg E2lmL at WM to 5014 pg EZImL at LP (Table 9).
E2 concentrations in the plasma of male and female fish from the reference sites (LP
and RS) indicated a difference between the sexes; for these two sites, males had lesser
mean concentrations of E2 than females (Figure 11). This same trend was observed for
WM, ER, and PT.

Plasma concentrations of E2 for females were significantly greater in fish from
LP relative to those from RS and BV, but not from any of the other sites. This was not
evident from that illustrated in Figure 11, but when considering the sample sizes and
distributions of these hormone concentrations, LP female E2 data points were widely
distributed and crossed the entire range and median of all sites except for RS and BV.
Plasma EZ concentrations for males from LP and RS were different indicating the
natural variability observed for fish in the environment even between unexposed
populations. Plasma E2 in male fish from all other sites were similar to those from LP
with the exception of WM which did not differ from E2 concentrations observed in fish
from RS.

Plasma T concentrations for females ranged from 949.4 pg TlmL (BV) to 6775
pg TlmL (LP) and often overlapped and sometimes exceeded values observed in males
within each site. Mean concentrations of T in males ranged from 774.5 pg TlmL (WM)
to 6859 pg TlmL (LP) (Table 9). Plasma T concentrations were similar between males
and females within sites for all but RS and WM, sites for which concentrations of T in
females were generally greater (Figure 12). These values indicate that both males and
females from LP both had greater T concentrations than the same sex at any other site,
except for PT. The large standard error for PT females may be explained by one female

that had greater concentrations of both T (16,461 pg TlmL) and E2 (10,489 pg EZImL)

76

compared to all other females but that could not be excluded as an outlier for statistical
analysis.

Generally, plasma T concentrations were less at RS and the exposure sites
when compared to LP (Figure 12). Plasma concentrations of T for females was
significantly less for all sites including RS, except for OW and PT. RS females had
significantly less plasma T than females from LP. All exposure sites were similar to RS
except for BV, for which plasma T was significantly less than RS. For males, plasma T
was greater in LP and PT than all other sites. Plasma T concentrations for RS was
similar to most exposure sites except for PT and LP, mentioned above, and WM which
had significantly lower plasma T levels.

Ratios of estrogen to androgen, here measured as E2lT. were significantly
different between the sexes for the reference sites LP and RS (Figure 13). For LP, the
difference in E2lT was dependent only upon differences observed in E2 between males
and females as T concentrations were similar between the sexes. However, for R8 both
E2 and T were different between males and females and the differences in the ratios
cannot be apportioned to one hormone or the other. WM and PT were the only sites
with differences in EZIT between males and females and similar to the trend observed in
the reference sites; generally the E2lT ratio was greater for females than for‘ males.
Although E2 and T concentrations in the fathead minnows separately reflected
differences among females from reference sites and exposure sites, the ratio of E2fl'
indicated that the relative proportion of the hormones remained the same for females
among all sites. Males at all sites except for PT, however, had significantly greater E2fl'
than the reference sites; this difference together with individual E2 and T concentrations

in PT males indicated greater concentrations of T relative to E2 similar to that found at

77

RS and LP and greater concentrations of T when compared to all other exposure sites.

DL had a greater mean E2lT than males at all other sites.

Relationships among biomarkers

In an attempt to calibrate traditional biomarkers of endocrine disrupter exposure
such as VTG and hormone concentrations to effects on health and physiology of fishes,
correlations (r) between E2, T, E2lT. VTG, and K were calculated (T ables 10A and
10B). The correlations were calculated to determine any relationships or trends that
may occur in situ but do not imply causality by simple correlation alone.

Female biomarker values were inter-correlated more frequently than those for
males (T able 10A). For females, E2 but not T was correlated with E2lT indicating that
ratio was dictated mostly by variations in E2 rather than T. Both E2 and T were
positively correlated with concentrations of VTG observed in females. K correlated only
with total plasma T concentrations.

The incidence of biomarker correlations for males exposed to WWTP were fewer
than that for females (Table 10B). Correlation coefficients for males indicated that both
E2 and T influenced and correlated with the E2fl' ratios. The field exposure and
reference sites could not distinguish any correlations between male VTG
concentrationfincidence and any other biomarker or measure of reproductive and

physical health.

78

DISCUSSION

Induction of VTG in male fish has been observed in rivers below VWVTP in the
UK and from large urban water systems in the US. In some cases concentrations of
VTG in male fish overlap concentrations measured in females [2,3,5,6]. However, the
occurrence of induced VTG in male fish below WWTP in the US has not been fully
elucidated for average flow and treatment technologies utilized by municipal WWTP. In
this study of caged male and female fathead minnows exposed to WWTP effluent, no
significant induction of VTG concentrations in males was observed. Differences were
observed in concentrations of hormones and female VTG, and explanation for
differences observed among sites will be further explored.

All of the fathead minnows used in this study were taken from the same stock at
the MSU Limnology Ponds and were presumed at the same stage of sexual maturity.
There were obvious differences among the 6 exposure sites (BV, DL, ER, OW, PT, WM)
and 2 reference sites (LP, RS) in observed biotic and abiotic factors influencing the fish
concomitant with differences in endogenous VTG and hormone production, but an
account of all of these factors was not possible and the causes of the in vivo differences
among the sites may be difficult to discern. The objectives of the study were to
determine the incidence, if any, of endocrine disruption as indicated by measures of
Vitellogenesis and reproductive steroid hormone concentrations in the blood plasma of
fish. The study was not conducted to elucidate physiological, biochemical, and
molecular mediators and mechanisms of the responses seen in the WWTP. However,
the following discussion will present some possible explanations for the differences that

were observed.

79

Fathead minnows from LP had the greatest VTG concentrations in females, the
greatest incidence of VTG detection in males, and the greatest hormone concentrations
compared to most other sites. The apparent difference of LP from the river reference
site (RS) and all other sites may be elucidated by site biotic and abiotic characteristics
alone. LP is a highly eutrophic, shallow pond inhabited primarily by fathead minnows
and some redear sunfish (Lepomis microlophus) in addition to abundant micro- and
macroinvertebrates. The fish obtained from the LP for reference were trapped in the
ponds only 24 h preceding sampling, and thus were not exposed to the confinement
stress and limited food availability that may have been present for caged fish. The
biomarkers measured for fathead minnows at LP are indicative of a population that was
relatively undisturbed by sampling and exposure devices, or by fluctuations in stream
flows caused by changes in effluent discharge or rain events during the exposure
period.

When compared to those from LP, fathead minnows caged at RS were under
confinement stress as well as competition for a less abundant food resource in the
confines of the cage and that provided by the stream. The effects of generalized and
confinement stress on reproductive physiology, primarily fecundity, plasma E2, T, and
11-KT, and on plasma VTG concentrations has been studied intensively and provides
some insight for the decreases in biomarker concentrations relative to that seen at LP
[20]. In previous studies, stress and the induction of cortisol (F), the primary steroid
produced by neuroendocrine stimulation of the interrenal glands upon stimulation by a
stressor, exhibit a marked inverse relationship with concentrations of plasma E2, T, and
VTG and to a lesser extent, 11-KT [21-23]. When compared to LP, all of the sites had
lesser concentrations of E2, T, and VTG. Investigators have determined that the effects

of F for several species of fish including the goldfish are not elicited by direct action on

80

ovarian steroidogenesis; rather, the effects that are observed may more likely be
mediated at levels other than the gonads in the hypothalamic-pituitary-gonad axis [24].
Induction of stress, as measured by plasma F concentrations, was not possible in the
study discussed herein given the limited volumes of plasma from the fathead minnows.
Nevertheless, it appears that stress is the most obvious implication for the decreased
concentrations of steroid hormones and female VTG observed for all sites compared to
LP, but the differences observed for survival, E2, T, and VTG between RS and some
exposure sites remain to be explored.

The differences between RS and that observed at other sites is difficult to
discern, as biotic and abiotic factors including food availability and concentrations of
suspect environmental estrogens and other chemical stressors were not directly
measured during the exposure. From the effluent and stream water quality parameters
measured, there were few obvious associations between these factors and the
incidence of mortality or fluctuations in concentrations of VTG, E2, or T or ratios of E2lT.
Mortality was greatest for caged minnows below the BV and OW WWTP, the only 2
trickling filter treatment plants, but causality of this type of treatment with increased
mortality is only circumstantial. At high pH, total ammonia tends towards the more toxic
unionized ammonia (NH3). The concentration in the effluent of OW was high and
approaching values for acute toxicity in more sensitive species. Even at low
concentrations, NH, can compromise the physiology and immune function of fish [25].
The low survival at OW could be due to this high concentration of ammonia during the
exposure. Additionally, although not reflected in effluent water quality values from
DMRs, the plant operator at BV reported a pump failure in the plant during the exposure
period that could have resulted in decreased effluent water quality (T. Kesler, pers.

comm).

81

Based upon the W6 response alone, males were not affected by endocrine
modulating substances during the 21 d exposure; male plasma VTG concentrations
were not found in the same range as females at the same site or compared with
reference sites. In fact, LP, the reference site from which all of the minnows originated
had the greatest incidence of male VTG detection. Quantities of VTG in the plasma of
males of various species of fishes in the cyprinid or minnow family including the fathead
minnow, carp, and goldfish have been previously detected [26]. For most of these
species, the concentrations of male VTG were less than 20 ng VTG/mL, but the greatest
concentrations were observed in fathead minnows (up to 79.8135 ng VTG/mL) [26].
This detection of VTG in male fathead minnow plasma is similar to observations from
our field and laboratory exposures. Differences in the absolute concentrations
measured in this study and by previous investigators are not of great concern in light of
the use of VTG detection as a biomarker of exposure and more frequent comparisons of
male VTG concentrations to levels found in females. Furthermore, differences in the
sexual state, maturity, and ELISA or RIA assay sensitivity and specificity may explain
these differences.

The mechanisms causing lesser plasma VTG for females at exposure sites
compared to LP and RS is not known. The differences in female VTG concentrations
are most likely due to differences in stress response during the exposure and
subsequent decrease in the VTG produced by hepatocytes. Previous authors
investigating the effects of stress and increased plasma concentrations of F, have
indicated that a reduction of female plasma VTG may be explained by decreases in the
number of nuclear ER binding sites in hepatic tissues and the relative increase in

plasma sex steroid binding proteins together with unchanged E2 levels [27]. The fact

82

that concentrations of E2 and T for females were not significantly different between RS
and most exposure sites indicates that the stress induced decrease in ER and thus a
decline in the transcription of VTG may be a viable explanation for the decreased VTG
concentrations at exposure sites when compared to RS. Partial estrogen agonists, or
environmental estrogens are usually only a fraction of the potency of endogenous E2
when evaluated for in vitro hepatocyte VTG production [28,29]. Previous authors have
suggested the possibility that environmental estrogens could decrease VTG response in
females, if estrogen agonist binding was in direct competition with endogenous E2 [5].
These mechanisms warrant further investigation as to the role of stress response and
weak estrogen agonists in direct modulation of female reproductive endocrine function.
The fathead minnow is an asynchronous spawner, with gametogenic and
spawning activity in situ from May through August in LP (pers. obs). Cyprinids typically
have several populations of oocytes at different stages of development at any given time
for a relatively continuous recruitment of oocytes for maturation and spawning during
periods of reproductive activity. The sperrnatogenic and oogenic activity in the testis
and ovary during the spawning period of fathead minnows and most cyprinids are
relatively constant when compared to synchronous spawners such as the rainbow trout
(Oncorhyncus mykiss) [30]. For minnows, sperrniation and oocyte maturation and the
associated transient rises in maturation hormones and associated factors can occur
within a matter of one day [31,32]. The primary female hormones involved in
exogenous Vitellogenesis and oocyte maturation are E2 and T, respectively. T is
produced by a cascade of enzymatic conversions of cholesterol in the thecal and
granulosa cells of the follicles; E2 is produced by aromatase conversion of T. When T
is no longer needed for conversion to E2 for VTG production, a rise in plasma T is

thought to signal the release of gonadotropin II for stimulation of oocyte maturation and

 

83

ovulation [31,33]. For males, the hormonal regulation of spermatogenesis and
spermiation are less well understood than that for ovarian endocrinology. There is
evidence that in goldfish the hormonal milieu required for maturation and release of
male gametes may be different from salmonids [34]. 11-KT has previously been defined
as the most androgenic teleost male hormone and thus the most important in the
reproductive physiology of male fishes [35]. However, 11-KT and T have been reported
to be produced in similar quantities as evidenced by in vitro and in vivo concentrations;
both 11-KT and T have been implicated to play roles in spermatogenesis and
spermiation in cyprinids such as the goldfish and the carp [32], but the relative
contribution and roles of each of these androgens is poorly understood and has not
been described for the fathead minnow. I chose to measure T based upon the ease of
measuring the hormone in both sexes with a commercially available EIA kit.

For their ease of measurement and characterized roles in gametogenesis and
spawning, plasma sex steroid hormones and the ratios of estrogens to androgens (EIA)
in the form of E2lT, E2l11-KT, or E2ltotal androgens have recently been utilized as
biomarkers of exposure to suspected endocrine-modulating substances [5-7]. These
investigators hypothesize that the relative ratio, rather than the absolute values of
individual steroid hormones, preclude the reproductive health and fitness of fish and
exhibit an apparent differences between males and females. In general, feral-female
carp evaluated in situ and in reference areas had ratios greater than one, while feral
males had ratios that were generally less than one [6,7]. This difference in E2IT
between males and females was observed In the two reference sites (LP and RS) and
was significant for 2 of the 6 exposure sites (PT and WM). Among all sites, there were
no statistical differences detected in E2lT ratios for the females, but male E2lT ratios

were significantly greater in WWTP exposure sites DL, ER, OW, and WM and to a

84

lesser extent for BV (p<0.10) than references. In fact, DL males had an extremely great
E2lT, a result that cannot be fully explained with the data at hand except that DL males
had a relatively greater mean value of E2 compared to other sites and a lesser
concentration of T. The differences observed in E2lT ratios between the reference sites
and exposure sites is unexplained. It is not known whether the relative proportions of
the hormones are indicative of greater stress reponses, real exposures to sublethal
chemicals that may be directly modulating the synthesis, catabolism, and action of
endogenous hormones, or results of natural fluctuations and differences in the gonadal
cyclicity on the day and times that the fish were sampled.

In the past, VTG has been the primary biomarker of exposure to environmental
estrogens in male fishes. However, from this study, VTG and hormone biomarkers
appear only to inter-correlate with each other for females, a result that is to be expected
since Vitellogenesis is normally a female-specific response to endogenous E2
production for oocyte assembly [8]. The incidence of VTG detection in male fathead
minnows did not correlate with any of the reproductive steroid hormones known to
stimulate VTG production or play a role in male gametogenesis. Under unstimulated
conditions observed in male fathead minnow and carp VTG [7] and under stimulated
VTG production in situ with feral male carp by others [5,6], there is an apparent lack of
calibration of this widely used biomarker with endocrine function and general fish
reproductive behavior and function. The anomalies in VTG and hormone production for
exposed male and female fishes, and the implications of these reproductive anomalies
for the health, survival, and propagation of the species remains to be explored. The use
of VTG as a biomarker of exposure appears to have the greatest utility in discerning
only exposed versus non-exposed in the most drastic situations, as that observed in the

UK and below relatively large urban municipal VWVTP in the US. The utility of these

85

biomarkers of exposure in predicting the reproductive and general health of fish remains
to be determined.

The relative risk of exposure to endocrine modulating substances in these
representative mid-Westem US municipal WWTP and in random surface waters of the
US [7] appears to be small relative to the reproductive abnormalities observed in the
UK. In the UK freshwater resources are relatively limited and during drought conditions
WWTP effluents can comprise up to 36% of the river flow regimes [2]. The US has
abundant water resources. WWTP discharge permits required by the Clean Water Act
are dictated first by technology (secondary treatment is required for most publicly owned
VWVTP) and then by the water quality and quantity of the effluent receiving stream. The
greater the contribution of effluents to the flows in US streams, the more stringent the
effluent water quality standards. Furthermore, with the advent of the Clean Water Act
and the allocated funding provided to municipalities in the 1970’s, the most economically
feasible and best treatment technologies were employed to remove pollutants from
wastewater [36]. In Michigan and all of North America, trickling filter treatment plants
still operate, particularly in small communities, but the preferred technology is activated
sludge [37]. Many environmental contaminants are lipophilic and are often found
absorbed or distributed in greatest concentrations in sediments in the aquatic
community. For these properties, the activated sludge process has been observed to
be a more efficient sink for very lipophilic contaminants such alkylphenols [38]. In the
UK, the prevalence of trickling filter technologies, the relatively great volumesof flow
from most WWTP, and the lesser dilution of the effluents by receiving streams are
possible explanations for the differences observed in endocrine disruption compared to

the US.

86

In summary, overt endocrine disruption in mid-Michigan VWVTP effluents was not
observed as evidenced by very low concentrations or low incidence of VTG detection in
male fathead minnows. The differences in the two reference sites (LP and RS) indicate
the natural variability observed among fish in different physical environments (stillwater
pond to riverine, respectively) and upon exposure to confinement stress. The relative
decrease in VTG, E2, and T in females and E2 and T in males at exposure sites and RS
compared to LP were most likely due to generalized stress imposed by biotic and abiotic
factors characteristic of the ambient streams and confinement. The specific cause of
the anomalies in these biomarkers and greater E2lT ratios for males compared to RS,
exposed to the same confinement stress, remains to be determined. The reproductive
endocrinology of fathead minnows has not been described in detail in the literature. The
literature indicates that the sensitivity and biology of salmonids and cyprinids differ
greatly and the conclusions of evaluations of endocrine disruption are dependent upon
the sensitivity of the species used and the representation that the species offers for
indigenous stream species. The fathead minnow is a more representative species of
warm water stream communities in lower Michigan. In vivo exposure to WWTP effluent
did not indicate overt endocrine disruption, but histopathological examination of the
gonads remains to be determined. Although it is assumed that the fathead minnows are
very similar to closely related goldfish and carp, the differences in species and life stage
sensitivities among cyprinids and between more sensitive species such as salmonids,
as well as further characterization of the roles of E2, T and related steroidogenic and
nonsteroidogenic reproductive factors will greatly contribute to understanding these site

differences.

 

87

ACKNOWLEDGMENTS

The research presented would not have been possible without the collaborative
efforts within the Aquatic Toxicology Laboratory at MSU. Sylvia Heaton, Michigan
Department of Environmental Quality was helpful in providing information regarding
Michigan WWTP. Thanks to Glen Van Der Kraak and 2.x. Yao at the University of
Guelph for technical assistance and training for development and application of the VTG
ELISA. Municipal WWTP owners and operators were extremely helpful in their
cooperation and interest in the study at hand. The research conducted herein was
funded by the Chlorine Chemistry Council of the Chemical Manufacturer’s Association,
the National Institute for Environmental Health Sciences (NIEHS-ES-0491 1), and the US
Environmental Protection Agency Office of Water (CR-822983-01—0). Partial funding for

KMN was generously provided by the Department of Fisheries and VWldlife, MSU, and

by SC Johnson Wax & Co.
REFERENCES
1. Purdom, C.E., P.A. Hardiman, V.J. Bye, N.C. Eno, C.R. Tyler, and J.P. Sumpter.
1994. Estrogenic effects of effluents from sewage treatment works. Chem. Ecol. 8:275-
285.

2. Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, E.J. Routledge, R.
Rycroft, J.P. Sumpter, and T. Tylor. 1996. A survey of estrogenic activity in United
Kingdom inland waters. Environ. Toxicol. Chem. 15:1993-2002.

3. Harries, J.E., D.A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, J.P. Sumpter, T.
Tyler, and N. Zaman. 1997. Estrogenic activity in five United Kingdom rivers detected by
measurement of Vitellogenesis in caged male trout. Environ. Toxicol. Chem. 16:534-542.

4. Lye, C.M., C.L.J. Frid, M.E. Gill, and D. Mccorrnick. 1997. Abnormalities in the
reproductive health of flounder Platichthys flesus exposed to effluent from a sewage
treatment works. Mar. Poll. Bull. 34:34-41.

5. Folmar, L.C., N.D. Denslow, V. Rao, M. Chow, D.A. Craln, J. Enblom, J. Marcino, and
L.J.J. Guillette. 1996. Vitellogenin induction and reduced serum testosterone

10.

11.

12.

13.

14.

15.

16.

17.

88

concentrations in feral male carp (Cyprinus carpio) captured near a major metropolitan
sewage treatment plant. Environ. Health Perspect. 104:1096-1101.

Bevans, H.E., S.L. Goodbred, J.F. Miesner, S.A. Watkins, T.S. Gross, N.D. Denslow,
and T. Schoeb. 1996. Synthetic organic compounds and carp endocrinology and
histology in Las Vegas Wash and Las Vegas and Callville Bays of Lake Mead, Nevada,
1992 and 1995. USGS Water-Resources Investigations Report 96-4266z12 pp.

Goodbred, S.L., R.J. Gilliam, T.S. Gross, N.P. Denslow, W.B. Bryant, and TR.
Schoeb. Reconnaissance of 17-beta estradiol, 11-ketotestosterone, vitellogenin, and
gonad histopathology in common carp of United States streams: Potential for
contaminant-induced endocrine disruption. US. Geological Survey Report, 47 pp.

Mommsen, T.P. and P.J. Walsh. 1988. Vitellogenesis and oocyte assembly. In W.S.
Hoar and V.J. Randall, eds., Fish Physiology, Vol. XIA. Academic Press, San Diego, pp.
347-406.

Lazier, 0.8. and ME. MacKay. 1993. VIteIlogenin gene expression in teleost fish. In
P.W. Hochachka and T.P. Mommsen, eds., Biochemistry and Molecular Biology of
Fishes, ed., Vol. 2. Elsevier Science Publishers, New York, pp. 391-405.

Specker, J.L. and C.V. Sullivan. 1994. VItelIogenesis in fishes: status and perspectives.
In K.G. Davey, R.G. Peter, and SS. Tobe, eds., Perspectives in Comparative
Endocrinology, ed., Vol. . National Research Council of Canada, Ottawa, pp. 304-315.

Chen, T.T. 1983. Identification and characterization of estrogen-responsive gene
products in the liver of rainbow trout. Can. J. Biochem. Cell Biol. 61:802-810.

Heppell, S.A., N.D. Denslow, L.C. Folmar, and C.V. Sullivan. 1995. Universal assay of
vitellogenin as a biomarker for environmental estrogens. Environ. Health Perspect.
103(Suppl 7):9-15.

Sumpter, J.P. and S. Jobling. 1995. Vitellogenesis as a biomarker for estrogenic
contamination of the aquatic environment. Environ. Health Perspect. 103(Suppl. 7):173-
178.

Anderson, R.A. and RM. Neumann. 1996. Length, weight, and associated structural
indices. In B.R. Murphy and ON WIIIis, eds., Fisheries Techniques, 2nd ed., American
Fisheries Society, Bethesda, MD, pp. 447-482.

Silversand, C. and C. Haux. 1989. Isolation of turbot (Scophthalmus maximus)
vitellogenin by high-performance anion-exchange chromatography. J. Chromatogr.
478:387-397.

Nichols, K.M. 1997. The Effects of Suspect Environmental Endocrine Disrupters on the
Reproductive Physiology of Fathead Minnows, Pimephales promelas. M.S. Thesis.
Michigan State University, East Lansing, MI.

Maisse, 6., B. Mourot, B. Breton, A. Fostier, O. Marcuzzl, P.Y. Le Ball, J.L.
Bagliniere, and A. Richard. 1991. Sexual maturity in sea trout, Salmo trutta L., running
up the River Calonne (Normandy, France) at the 'finnock' stage. J. Fish Biol. 39:705-715.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

89

Mourot, B. and P.-Y. Le Ball. 1995. Enzyme-linked immunosorbent assay (ELISA) for
rainbow trout (Oncorhyncus mykiss) vitellogenin. J. Immunoassay 16:365-377.

Mananos, E., J. Nunez, S. Zanuy, M. Carrillo, and F. Le Menn. 1994. Sea bass
(Dicentrarchus labrax L.) vitellogenin. II - Validation of an enzyme-linked immunosorbent
assay (ELISA). Comp. Biochem. Physiol. 1078:217-223.

Donaldson, E.M. 1990. Reproductive indices as measures of the effects of
environmental stressors in fish. American Fisheries Society Symposium 8:109-122.

Clearwater, S.J. and N.W. Pankhurst. 1997. The response to capture and confinement
stress of plasma cortisol, plasma sex steroids and vitellogenic oocytes in the marine
Teleost, red gumard. J. Fish Biol. 50:429-441.

Carragher, J.F., J.P. Sumpter, T.G. Pottinger, and AD. Pickering. 1989. The
deleterious effects of cortisol implantation on reproductive function in two species of trout,
Salmo trutta L. and Salmo gairdneri Richardson. Gen. Comp. Endocrinol. 76:310-321.

Pickering, A.D., T.G. Pottinger, J. Carragher, and J.P. Sumpter. 1987. The effects of
acute and chronic stress on the levels of reproductive hormones in the plasma of mature
brown trout, Salmo trutta L. Gen. Comp. Endocrinol. 68:249-259.

Pankhurst, N.W., G. Van Der Kraak, and RE. Peter. 1995. Evidence that the inhibitory
effects of stress on reproduction in Teleost fish are not mediated by the action of cortisol
on ovarian steroidogenesis. Gen. Comp. Endocrinol. 99:249-257.

Russo, R.C. 1985. Ammonia, nitrite, and nitrate. In G.M. Rand and SR. Petrocelli, eds.,
Fundamentals of Aquatic Toxicology, Taylor & Francis, Bristol, PA, pp. 455-471.

Tyler, C.R., B. van der Eerden, S. Jobling, G. Panter, and J.P. Sumpter. 1996.
Measurement of vitellogenin, a biomarker for exposure to oestrogenic chemicals, in a
wide variety of cyprinid fish. J. Comp. Physiol. B 166:418-426.

Pottinger, T.G. and AD. Pickering. 1990. The effect of cortisol administration on hepatic
and plasma estradiol-binding capacity in immature female rainbow trout (Oncorhyncus
mykiss). Gen. Comp. Endocrinol. 80:264-273.

Jobling, S. and J.P. Sumpter. 1993. Detergent components in sewage effluent are
weakly oestrogenic to fish: an in vitro study using rainbow trout (Oncorhynchus mykiss)
hepatocytes. Aq. Toxicol. 27:361-372.

White, R., S. Jobling, S.A. Hoare, J.P. Sumpter, and M.G. Parker. 1994.
Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinol. 135:175-
182.

Kobayashl, M., K. Aida, and I. Hanyu. 1986. Gonadotropin surge during spawning in
male goldfish. Gen. Comp. Endocrinol. 62:70-79.

Kime, DE. 1995. The effects of pollution on reproduction in fish. Rev. Fish Biol. Fish.
5:52-96.

Billard, R. A. Fostier, C. Well, and B. Breton. 1982. Endocrine control of
spermatogenesis in teleost fish. Can. J. Fish. Aquat. Sci. 39:65-79.

33.

35.

36.

37.

38.

90

Rlnchard, J., P. Kestemont, and R. Heine. 1997. Comparative study of reproductive
biology in single and multiple-spawner cyprinid fish. ll. Sex steroid and plasma protein
phosphorus concentrations. J. Fish Biol. 50:169-180.

Wade, M.G. and G. Van Der Kraak. 1991. The control of androgen production in the
goldfish: effects of activators of different intracellular signaling pathways. Gen. Comp.
Endocrinol. 83:337-344.

Boume, A. 1991. Androgens. In K.T. Pang and MP. Schreibman, eds., Vertebrate
Endocrinology: Fundamentals and Biomedical Implications, Vol. 48. Academic Press,
Inc., New York, pp. 115-147.

Water Environment Federation. 1993. The Clean Water Act of 1987. Water
Environment Federation, Alexandria, VA.

Ganczarczyk, J.J. 1983. Activated Sludge Process: Theory and Practice. Marcel
Dekker, Inc., New York.

Giger, W., P.H. Brunner, and C. Schaffner. 1984. 4-Nonylphenol in sewage sludge:
Accumulation of toxic metabolites from nonionic surfactants. Science 225:623-625.

91

Figure 9. Locations of wastewater treatment plant (WWTP) sites in mid-
Michigan, USA where effects of WWTP effluent on reproductive physiology
of fathead minnows were surveyed, June-August 1996. Capital letters in
bold are the WWTP sites (BV=BeIIevue, DL=Delhi Township/Holt, ER=Eaton
Rapids, LA=Lansing, LP=Limnology Pond, OW=Owosso, PT=PortIand, RS=
Reference Site, WM=WIIIiamston). Boundary lines contrasted with different

shading represent watersheds.

92

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

93

Table 5. Characteristics of WWI'P effluents at 7 study sites in mid-Michigan,

USA, for evaluation of effluent effects on reproductive physiology of fathead

 

 

minnows.
WWTP River Daily flow (m’) Treatment Type
mean max.

Bellevue (BV) Battle Creek 174 458 Trickling filter
Delhi Township (DL) Grand 6,850 11,200 Activated sludge"
Eaton Rapids (ER) Grand 3,110 8,020 Activated sludge
Lansing (LA) Grand 88,950 195,460 Activated sludge‘

Owosso (OW) Shiawassee 15,330 49,210 Trickling filter‘
Portland (PT) Grand 1,805 1.930 Activated sludge

Williamston (Willi) Red Cedar 1,787 8,330 Activated sludge

 

’lndicates treatment beyond traditional secondary technology to further remove biological oxygen
demand and phosphorus

Table 6. Ambient river and gross effluent water quality data from WWTP field

sites June-August 1996.

Effluent water data represent numbers from

Discharge Monitoring Reports required by NPDES permits. All data are

averages unless othenIvise indicated.

 

 

 

AMBIENT EXPOSURE WATER' EFFLUENT WATERh

Site Depth Temp 00 pH Hardness DO‘ pH‘ Amm.‘ CIz BODf

(m) (°C) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) mg/mL % rem.
BV 0.7 20.4 7.20 7.43 280.44 6.83 7.8 0.016 0* 8.3 92.7
DL 0.5 21.0 8.25 7.80 212.3 8.29 8.54 0.117 0* 1.17
ER 0.7 20.7 8.53 7.18 251.27 5.43 7.8 0.003 5 95.3
LP 23.6 7.35 7.88
OW 0.4 20.6 8.17 7.57 235.78 8.23 8.38 0.4 0* 6.8
PT 0.9 21.6 7.20 7.30 4.5 7.67 0.9 15.33 90.3
RS 0.8 20.7 6.40 6.80
WM 0.8 20.0 8.63 6.82 336.4 6.33 7.85 0.165 0* 6.8

 

*Chlorine was not detectable below the MDL of 0.0001 mglL.

“Ambient water quality parameters measured randomly on days of site check-ups (n=1-4).

bEffluent water quality parameters are means of monme averages derived from daily grab
samples of gross treated effluent.

°Average of minimum monthly values June through August.

°Average of maximum monthly values.

°Amm.=total ammonia concentrations in gross effluent.
'BOD=biological oxygen demand and is a measure of 5-day carbonaceous BOD at 20°C.

95

Table 7. Survival and condition factors (K) for male and female fathead
minnows exposed in situ to WWI'P effluent in mid-Michigan, June-August

1996. Means1SEM are presented.

 

 

 

FEMALES MALES K Tukey Total

SITE % survival K‘ % survival K° groupb % survival

LP° - 019610.005 - 021910.005 A -

RS 82.5 020110.008 52.5 019110.005 B 68

BV 20 0.15410.010 20 018010.007 B 20

DI. 10 015010.023 37.5 021810.009 B 24

ER 47.5 0.19910.005 40 020710.006 B' 44
CW 2.9 0.198 37 017810.008 A 20

PT 22.5 0.19510.007 71 018510.006 B 46
WM 17 0. 1 9210.007 45.7 019110.006 A 31

 

'K=condition factor calculated as K=WIL3 *10,000 where W=weight in g and L=standard length in
mm.

”Tukey grouping for K represents difference from controls, LP and RS, only.

cSurvival for LP is not applicable as fish were trapped only 24 h prior to sampling.

*For DL, there was a statistical difference between male and female condition factors.

96

Table 8. Plasma VTG concentrations and percent greater than the method
detection limit (MDL) as measured in the VTG ELISA for fathead minnow
exposed in situ to WWTP effluent, June-August 1996. Exposure was for 3

weeks. Values in parentheses are SEM. (ND. = not detectable.)

 

 

FEMALES MALES
SITE n Range Mean %>MDL" 11 Range Mean %>MDL‘
(pglmL) (pglmL) (nglmL) (nglmL)

 

LP 25 656.1-3,406.3 1460.3 100 23 N.D.-4.79 0.783 62

(1137.46) (10.224)

RS 33 324.8-2,463.0 1120.3 100 21 N.D.-O.94 0.19 14
@8951) (10.039)

BV 8 4235-8576 336.44 100 7 ND. ND. 0
(194.52)

DL 4 34.37-1,130.02 811.97 100 13 ND. ND. 0
(1260.27)

ER 17 75.95 - 1,747.4 640.07 100 16 ND. - 9.31 1.089 , 62

(1133.7) (10.667)
ow 1 — 1926.83 100 13 ND. - 25.99 5.859 50
- (12.4)

PT 7 173.17 - 766.69 553.61 100 27 ND. - 16.27 0.883 15

(171.52) (10.62)
WM 5 224-18957 608.28 100 15 N.D.-29.61 3.24 13
(1341.4) (11.97)

 

'Percent >MDL are calculated as percent of total number of fish measured for VTG for each sex.

97

Figure 10. Plasma VTG concentrations for caged male and female fathead
minnows exposed in situ to WWTP effluent. A. Female mean plasma
concentrations (:tSEM) for VTG at each site. Differences among sites are
represented by Tukey letter grouping above the VTG concentration
histograms. B. Male plasma VTG. Histograms are mean VTG
concentrations. Differences in the incidence of VTG detection from both LP
and RS rather than differences in absolute concentrations are denoted by
the symbols (*). Note the differences in y-axes scale (pg VTG/mL) between

A and B.

98

 

 

 

m1

 

 

 

 

 

 

 

 

 

 

 

 

 

m:
11.0
MIT.
mrfl.
«Ln. .
arfl

3.59: 5:82.35

RS BV DL ER OW PT

LP

9.-

.
d
8

7...

p p h P
# J u u
5 4 3 2

6...

3.59.. 5:82.85

 

RS BV DL ER OW PT

LP

99

Table 9. Concentrations of E2 and T, and ratios of EZIT in male and female

fathead minnows exposed in situ to WWTP effluent. Means are reported 1

 

 

 

SEM.
FEMALES MALES
SITE 11 E2 T E2/T n 52 T E2/T
(Pg/mL) (pglmL) (Pg/ML) (Pg/ML)
BV 7 3615.4 949.4 3.711 7 3641.32 1832.96 2.299
(11277.1) (1235.6) (11.102) (11021.21) (1436.50) (10.854)
DL 4 4327.7 1330.4 3.977 10 5282.44 1336.12 6.358
(1584.0) (1289.8) (11.253) (1728.84) (1351.61) (11.688)
ER 16 5281.3 2585.4 2.667 14 2707.24 1676.97 2.336
(11107.1) (1404.4) (10.527) (1389.23) (1370.92) (10.427)
LP 22 11120.0 6774.9 2.017 15 5013.83 6859.01 0.920
(11405.4) (1799.2) (10.252) (11055.21) (11043.25) (10.197)
ow 1 8521.8 5398.6 1.578 12 3175.61 2456.79 2.328
(11035.41) (1270.17) (10.306)
PT 7 7075.7 5295.3 2.575 25 4431.55 3488.30 1.213
(11294.5) (12127.7) (10.786) (1839.58) (1341.64) (10.158)
RS 31 4859.2 3045.4 2.201 21 1312.92 1726.79 0.814
(11170.0) (1445.3) (10.372) (1245.86) (1302.06) (10.080)
WM 4 4648.3 1790.7 3.276 14 1138.82 774.50 2.209
(11708.1) (1168.1) (10.542) (1133.50) (1150.39) (10.524)

 

100

 

 

 

   

 

 

 

14000 1-
I
12000 -- afemales
A * Imales
3 10000 ,_
E _
m 8000 --
I *
S .
E 6000 - D
*
n 11-
4000 -
2000 1 A
o 1 —.-1

 

 

 

 

 

 

 

 

 

DL ER OW PT WM

Figure 11. Female and male plasma E2 (pg EZImL) concentrations following in
situ exposure to VWVTP effluent. Data are plotted as nontransformed
means :1: SEM. (Statistical differences are for logo-transformed data:
*=significant difference between males and females within the same site;
A=difference from LP among females; A=difference from LP among males;

I=difference from RS among females; U=difference from RS among males) -

101

8000 .1 I

7000 ..
nfemales

6000 .. 'ma'”

5000 .-

4000 u:-

 

Plasma T (pglmL)

3000 4.

 

2000 J)-

1000 «-

 

 

 

 

 

 

 

 

Figure 12. Female and male plasma T concentrations (pg TlmL) following in sftu
exposure to VWVTP effluent. Data are plotted as nontransformed means 1
SEM. (*=difference in logw-transformed values between sexes within a
given Site; A=difference from LP among females; A=difference from LP
among males (p<0.10 for WM); I=difference from RS among females;

Cl=difference from RS among males)

102

g .-

3 ..

7 nfemales
" Imales

5 ._

E2"
01

 

 

 

 

 

 

 

 

 

 

 

 

4 ..
A
3 " *
* 11:

2 «-

1 ..

0 1

LP RS BV PT

Figure 13. Female and male plasma. E2 to T ratios (EZIT) following in situ
WWTP effluent exposure. Data are plotted as means 1 SEM of
untransformed concentrations. (*=difference in logw-transfon'ned values
between sexes within a given site; A=difference from LP and RS among

males only; p<0.10 for BV)

103

Table 10. Correlation coefficients (r) between traditional biomarkers of
environmental endocrine disruption and between biomarkers and condition
factor (K) for fathead minnows exposed in situ to WWTP effluent, June-July
1996. Values for hormones and E2/T ratios were logw-transformed. Both
Pearson product moment correlation and Spearman’s Rho determinations
were made dependent upon the distribution of the data and are denoted in

the tables.

104

A. Correlation coefficients (r) between biomarkers and condition factors in female

fathead minnows exposed to WWTP effluents.

 

 

Biomarkel' E2 T EZIT' VTGa
E2 1.000
T 0523* 1.000
E2/T‘ 0.481' -0.439 1.000
VTG‘I 0.187” 0383* -0.148 1.000
K 0.169 0399* -0.217 0.039

 

'Values of an and VTG for females were ranked and correlation coefficients between these
biomarkers and others are determined by Spearman’s Rho.
“p<0.05 for correlation coefficients

*0.05<p<0.10 for correlation coefficients

B. Correlation coefficients (r) between biomarkers of exposure and condition factors in

for male fathead minnows exposed to WWTP effluents.

 

 

Blomarkers E2 T EZIT VTGb
E2 1 .000
T 0572" 1.000
E2lT 0.468‘ -0.457* 1 .000
VTG" 0.058 0.019 0.118 1.000
K 0.052 -0.010 0.080 -0.023

 

bCorrelations for VTG with other biomarkers are based upon ranks and are Spearman's Rho
coefficients ‘

*0.05<p<0.10 for correlations

SUMMARY AND RECOMMENDATIONS

Development of a VTG ELISA for detection of VTG in the plasma of goldfish and
fathead minnows was successful. The goldfish VTG antiserum ELISA was able to
detect ng quantities of VTG with defined accuracy and precision sufficient for monitoring
for exposure to environmental estrogens. The ability to use goldfish VTG for assays
involving measures of fathead minnow plasma VTG greatly enhances the utility of the
ELISA for both species.

Exposure of fathead minnows to the suspect environmental estrogen, NPEO, did
not produce a significant concentration-dependent relationship for fecundity, female or
male plasma VTG concentrations, female or male plasma E2 and T, or steroid hormone
ratios, E2lT. It appears that the NOEL for NP1-17EO is above 10 pg NPEOIL.
Calibration of these biomarkers to NPEO exposure and reproductively important
endpoints was not relevant. The relative risk of NPEO in surface waters of the US
appears to be small in view of the nondetectable to small concentrations of these
compounds in a nationwide survey. The natural variability in egg production and steroid
hormone concentrations greatly exceeded the variability observed among NPEO
treatments and work remains to determine the mechanisms controlling egg production
and how other xenobiotics may impair this function. Furthermore, the implication and
calibration of male VTG production with environmental estrogen exposure remains to be
elucidated for effects on the overall health and reproductive performance of male fish

and male and female pairs.

105

106

Upon in situ exposure to WWTP effluent, fathead minnows caged below 6 mid-
Michigan WWTP were not observed to have the same overt effects as those observed
in feral and caged fish in the UK and below large urban WWTP outfalls in the US. Male
plasma VTG concentrations were not induced to that found in females. Differences
among Sites in reproductive steroid hormone concentrations and mechanisms
responsible for these differences could not be completely described within the confines
of this research. Decreases in plasma steroid hormone concentrations could be
attributed to a variety of effects including generalized stress response, but exact
mechanisms whereby WWTP effluent or confinement stress affect fishes remain to be
determined in controlled laboratory in vitro and in vivo experiments. From this study and
observations from studies conducted in the US, evidence of endocrine disruption below
municipal WWTP in the US does not appear to be as widespread and frequent as in the
UK. In the US, endocrine disruption in fish populations below WWTP appears to occur
in isolated incidences of great exposures to endocrine modulating substances in WWTP
effluents and/or receiving streams. The differences are most likely due to the greater
dilution offered by typical receiving streams in the US and the efficiency of municipal

WWTP treatments.

Recommendations for future evaluation

. The VTG ELISA, although reliable for quantifying VTG especially for goldfish, could
use refinement to identify or eliminate any nonspecific background that may be
detected in male plasma samples and further validation for use with the relatively
small volumes of plasma and small quantities of VTG produced by the fathead

minnow.

107

A similar laboratory exposure of fathead minnows to the mono- and di-ethoxylated
and carboxylated degradation products of APEO would be useful for a more
complete risk assessment of APEO in surface waters, as these are constituents with
the greatest concentrations of municipal WWTP and surface waters where APEO
and AP have been measured.

Calibration of biomarkers of environmental estrogen exposure in teleost fish,
especially VTG, should be made to determine the feasibility and predictability of the
biomarker for Short and long-term effects on reproductive behavior, function, and
output.

Species sensitivities and mechanisms for differential effects observed in
estrogenicity of compounds in vitro and in vivo should be evaluated to ensure

comparability or to describe variability among species and geographical locations.

APPENDIX

LABORATORY PROTOCOL
Induction, Purification, Identification, and Production of Antibodies for Goldfish

(Carassius auratus) Vitellogenin

Krista M. Nichols
Michigan State University
Aquatic Toxicology Laboratory

Version 1.0
29 April 1997

SCOPE

Vitellogenin is a large glycolipophospoprotein synthesized in the liver under control of
estrogen responsive vitellogenin genes. From the hepatocytes, the protein is cam'ed by the blood
to maturing oocytes in female oviparous vertebrates where it is cleaved to smaller proteins that
are the primary nutritional reserves for developing embryos (Mommsen and Walsh, 1988).

Both male and female fish possess the genes in the liver and have the capacity to
produce vitellogenin. However, because males do not have significant levels of endogenous
estrogen, the protein is not expressed in the males. Recently, in the advent of environmental
toxicology and endocrine disruption in the environment, induction of the vitellogenin protein in
males has become an in vivo model of exposure to environmental estrogenic compounds. Caged
and feral male fish have been evaluated in streams below suspected point sources of
environmental estrogens resulting in evidence, particularly in the United Kingdom, that the
reproductive physiology of fishes below these outfalls are indeed impacted by anthropogenic
effluents (Bevans et al., 1996; Folmar et al., 1996; Harries et al., 1996, 1997).

In efforts to provide a screening tool for evaluation of vitellogenin induction in
environmental and laboratory samples, the vitellogenin ELISA has been developed for rapid and
inexpensive screening of samples (Heppell et al., 1995; Sumpter and Jobling 1995; Tyler et al.,
1996). The use of polyclonal antibodies for piscine vitellogenin ELISA, however, is limited by the
lack of a commercially availabile antibody. The ELISA calls for Specific and sensitive antibodies
for the vitellogenin protein. Polyclonal antibodies have been proven to be relatively easy and
inexpensive to produce for piscine vitellogenins. The following protocol was used for the
induction, separation, identification, and antibody production for goldfish vitellogenin for use in the

108

109

vitellogenin ELISA described in the “Standard Operating Procedure for Measurement of Plasma
\fitellogenin in Fathead Minnows and Goldfish by Competitive Enzyme Linked lmmunosorbent
Assay (ELISA)," Appendix A.

SUMMARY OF PROTOCOL

Male and female goldfish were injected with 17-B-estradiol and the blood plasma was
collected for separation of the induced protein, vitellogenin, by high performance liquid
chromotagraphy (HPLC). Upon separation of the protein, the fractions collected during the
suspected elution time for vitellogenin were evaluated by sodium dodecyl sulfate-polyacrylamide
gel electrophoresis (SDS-PAGE) and subsequent Western immunoblotting with rabbit anti-goldfish
vitellogenin characterized and donated by Dr. Glen Van Der Kraak and Zuxu Yao at the University
of Guelph. Upon positive identification and purity evaluation (no cross-reactivity of the antibody
with nonspecific male proteins) of the separated proteins, the protein was emulsified with Hunters
TitreMaxTM adjuvant and injected into two New Zealand white rabbits. Rabbits were subsequently
boosted on several occasions and 15 mL plasma collected every two weeks until exanguation.

SIGNIFICANCE AND USE

The following protocol was followed to ultimately provide polyclonal antiserum for goldfish
vitellogenin. Production of vitellogenin antibodies has been demonstrated throughout the
literature and has been used for a variety of purposes including determination of the sexual
maturity of female fish and screening for environmental estrogenic effects on male fish (Maisse et
al., 1991; Heppell et al., 1995; Sumpter et al., 1995; Tyler et al., 1996). The lack of commercial
availability of piscine anti-vitellogenin has promulgated numerous experiments that must begin
first with induction, purification, and preparation of Species-specific polyclonal antibodies. The
protocol contained herein is a conglomeration and an adaptation of a variety of protocols and cited
literature. This protocol may be followed in the future for preparation of antibodies for goldfish
vitellogenin. The methods presented here are organized categorically rather than strictly
sequential as one complete protocol as these methods were used several times and out of
sequence for some purposes during raising and characterizing the antibodies.

110

PROTOCOL: Induction of Vitellogenin in Goldfish
*Notes and daily logs for fish condition, injection, and blood collection may be found on pp. 18-19

of volume 1, K. Nichols Laboratory Notebook.

The following describes the protocol following Silversand and Haux (1989) for induction
and purification of plasma vitellogenin. Fish were to be injected at a rate of 20 mg/kg/week
according to Silversand and Haux methods, but due to calculation error, the fish were only
injected with 2 mg/kg/week (see calculations below). This amount was quite adequate for
induction of vitellogenin.

Materials

Aprotinin (7.6 TlU/mL, Sigma A-6012)

Bucket

Corn oil

Estradiol-17B (stored under N2, Sigma E-8875)

Goldfish

Hypodermic needles (25G5/8, Becton Dickinson)
Syringes (1 cc Tuberculin, Becton Dickinson)

Tricaine methane sulfonate (MS-222, Argent Chemicals)

Methods

1. Goldfish previously received from Grassyforks (Martinsville, Indiana) were weighed
(range 21-53 9; mean 37 g) and then placed in one 7 feet X 2 feet X 2 feet fiberglass tank.
The fish were acclimatized in the University Research Containment Facility Room 1608
laboratory at a water temperature of approximately 15°C and a light cycle of 16 hours
light: 8 hours dark.

2. A 10mM solution of 17B-estradiol (E2) in ethanol prepared previously by Vince Kramer
and stored at -20°C was used to prepare an estradiol in corn oil solution for injection of
the goldfish. The prepared solution contained 747 pL of the 10 mM stock E2 solution in
5.5 mL corn oil each week and stored at 4°C. Calculations for corn oil solution have been
back calculated below:

10 mM of E2 = 2.2739 g EZIL (MW=272.39)

(2.7239 mg/mL)* 0.747 mL = 2.0347 mg E2

2.0347mg E2 in 5.5 mL results in 0.37 mg/mL
Only 100 pL of this solution (0.37mg/mL) was injected into each fish once a week
resulting in 0.037mg/injection and a rate of.
(0.037mglinjection)*(2 injections/fish)*(fiSh/0.037 kg) = 2mglkgMeek

3. Immediately prior to injection, fish (two at a time) were anaesthetized in MS-222
(120mgl1.5 L culture water). Fish were deemed anaesthetized when the gill operculum

111

stopped moving and the fish lost equilibrium. lntraperitoneal (i.p.) injections of 0.1 mL
were made with sterile 1.0 mL syringes and a 25 gauge needles for each fish.

4. Immediately after injection, fish were placed in a clean bucket of culture water, allowed to
recover, and then placed into the containment tank. Recovery water was replaced after
each set of 2 fish had recovered.

5. The goldfish were injected four times in the course of two weeks with the above E2
solution.
6. Two weeks after the first hormone injection, fish were anaesthetized with MS-222 and

given an injection of aprotinin (0.1 mL of 10% (vlv) aprotinin in corn oil) 0.5 hour before
being euthanized and all blood was collected.

7. Upon administration of a lethal dose of MS-222 for euthanization, the fish were bled using
a sterile syringe and needle from the caudal vein. Blood was immediately transferred to a
1.5 mL eppendorf vial and stored on ice for clotting until centrifugation.

8. Blood plasma was spun at 3000 X g for 10 minutes at 4°C. Plasma was collected with a
100 11L pipette and transferred to a 15 mL orange capped centrifuge tube (on ice) to pool
all plasma.

9. Pooled plasma from the goldfish was stored at -80°C until separation.

PROTOCOL: Separation of vitellogenin by HPLC

'Adapted from methods by Silversand and Haux (1989) and assisted by Mr. Joe Leykam,
Macromolecular Structure Facility, Department of Biochemistry, Michigan State University.

*Notes and chromatograms for the protocol may be found on pp. 22-26, 38-40, volume 1, K.
Nichols Laboratory Notebook.

Materials

0.2 11M filter

Aprotinin (7.6 TlU/mL, Sigma A—6012)

Milli-Q water (dH20)

Gilman syringe filter, 0.2 11M

Glass Hamilton syringe, 500 pL

HPLC (Waters 600) with photodiode array fluorescence detector
Diaminoethylamine HPLC column (DEAE anion exchange column, 15 mm X 20 cm)
Methanol

NaCl (JT Baker 3624-05)

Syringe, 5.0 mL plastic

Tris (Mallincrodt 7732)

Vacuum pump

Hydrochloric acid, concentrated

112

 

Buffers
Prior to separating the protein on the HPLC. the following buffers were prepared:
A. 20 mM Tris-Cl

Tris 121 .14g

Aprotinin (1%) 1.0 mL

dH20 fill to 1000 mL Adjust pH to 8.0 with concentrated HCI
B. 0.50 M NaCl in 29 mM Tris

NaCI 58.44 g

Tris 121.14 9

Aprotinin (1%) 1.0 mL

dHZO fill to 1000 mL Adjust to pH 8.0 with concentrated HCI
Methods

1.

The DEAE-Sephadex column was washed and equilibrated with 20 mM Tris-Cl (buffer A).
A blank was run on the HPLC with a linear gradient from 100% buffer A to 100% 0.5 M
NaCI (buffer B) in 50 minutes with 1.0 mL per minute flow rate with baseline monitoring on
the fluorescence detector. The fluorescence detector was programmed to measure 230
nm, 254 nm, and 280 nm. Sparge with N2 was set at 20%.

During the column washing, 1.5 mL screw-cap eppendorf tubes were labelled 1 through
60 and placed in order in the HPLC automated fraction collector.

A portion (1.0 mL) of the pooled estradiol induced goldfish plasma stored at -80°C was
filtered with a 0.2 pm Gilman syringe filter and diluted in 9.0 mL of buffer A (for a total of
10.0 mL) and then kept on ice.

Five hundred microliters (500 pL) of dilute plasma was pulled into the clean Hamilton
syringe and injected into the HPLC injection loop. After 30 s another injection of 500 pL
was made. Immediately following injection, unbound substances were eluted with buffer
A.

Once the baseline had returned to zero, the HPLC and fraction collector were started for a
linear gradient of 0 to 0.5 M NaCl in 50 minutes to elute bound proteins at a flow rate of
1.0 mL per minute. The fraction collector began collection of fractions simultaneously with
1.0 mUmin/tube that were chronologically ordered. The fluorescence detector was set as
above.

Once the HPLC had completed the 50 mL linear gradient, buffer B was run through the
column for an additional 10 minutes. The column was rinsed with 20 mM buffer A for 20
minutes and then rinsed with dHZO to ensure all salt and buffer are removed from the
equipment and column.

113

7. The complebd chromatogram for the elution of the proteins is seen in Figure A1. The last
and largest fraction coincides with that seen in Silversand and Haux (1989) and was
assumed to be vitellogenin as it was the largest retained plasma protein. All samples
were capped and stored at -80°C until electrophoresis and positive identification of the

eluted protein.

Flgure A1. Chromatogram for the separation of vitellogenin from estradiol induced goldfish

plasma. Note that 018 latest peak is the assumed to be vitellogenin.

 

 

 

 

 

 

 

 

 

PROTOCOL: SOS-PAGE for Detecflon of Vltollogenln

‘Adapted from methods from Cunent Protocols in Molecular Biology. Unit 10.2 (Electrophoretic

Separation of Proteins)

'Notes are found in volume 1, pp. 30-32, 60-61 and volume 2, pp. 15-17, 19 K Nichols' laboratory

notebook

The following protocols, adapted from Laemmeli’s methods for continuous sodium
dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) in Cun'ent Protocols, were used
for separation and positive detection of vitellogein prior to nonspecific protein staining and specific
antibody immunoblotting. This method was used with Western immunoblotting to determine the
purity of the HPLC fractions as well as to determine and compare cross-reactivity of the antibody
produced with nonspecific male plasma proteins and plasma proteins from other species.

Materials

Constant-current power supply (BioRad Model ZOO/2.0)
Dry-block heating unit (Therrnolyne Dri-Bath)

Electrophoresis apparatus: Mini-Protean ll (BioRad 165-2940)
Gel loading pipette tips

Pipettes

114

Razor blade
Tris-glycine gels, 4-15% (BioRad Ready gels, 161-0902)

Chemicals

Bromophenol blue (Sigma B-1026)

Dithiothreitol (DTT, Sigma D-9799)

Glycerol (SigmaUltra, G—6279)

Milli-Cl water (dHZO)

Protein molecular weight standards (high molecular weight range, Multi-Mark Multi-Colored
Standard, Novel Experimental Technology (recommended), LCS725)

Sodium dodecyl sulfate (SDS, electrophoretic gratde, Boehringer Manheim 1667262)

Sodium phosphate, dibasic, hepathydrate (NazHPOfl H20, Sigma S-9390)

Sodium phosphate, monobasic, monohydrate (NaH2P04-H20, Aldrich 22352-2)

 

 

Bufl'ers and Reagents
1. 4X Phosphate/SDS electrophoresis buffer (0.4 M sod_ium phosphate/0.4% SDS. DH 7.2)
NaH2P04-HZO 7.8 g
Na,HPO.-7 H20 38.6 9
SDS 29
dHZO Fill to 500 mL Adjust pH to 7.2 with NaOH or HCI

*Mix fresh each time as SDS will precipitate when stored.

2. 1X Phosphate/SDS electrophoresis buffer (0.1 M sodium phosphate/0.1% SDSJH 7.2)
4X phosphate/SDS 500 mL

dH20 1500 mL
*Can be stored at room temperature for a week

3. 2X Phosphate/SDS sample buffer (20 mM sodium phosphate, 2% SDS)
4X phosphate/SDS 0.5 mL

 

SDS 029
bromophenol blue 0.1 mg
DTT 0.31 9
glycerol 2.0 mL
dHZO Fill to 10 mL
*Make fresh each time
Methods

Methods for SDS-PAGE follow that which is found in Alternate Protocol 3 for Continuous
SDS-PAGE described in Current Protocols in Molecular Biology, Analysis of Proteins, p.
10.2.14. Only deviations and adaptations of this method will be described in detail below:

1. Gel buffers were prepared and gels removed from 4°C and prepared for electrophoresis
as described by manufacturing instructions on the box. The electrophoresis apparatus
was assembled with the gel and the gel combs removed. Phosphate/SDS
electrophoresis buffer was poured into the apparatus and pipetted into the wells to allow
the gels to equilibrate during sample preparation.

115

2. Samples were prepared in phosphate/SDS sample buffer. Molecular weight markers
were prepared according to manufacturer's instructions on the packaging depending upon
the MW marker used. Sample volume was determined based on a gel loading rate of 1 to
10 pg total protein (p. 10.2.8). Generally for plasma samples, 5 pL of plasma in 20 pL
dHZO with 25 pL of phosphate/SDS sample buffer was a good mixture. For induced or
pure fractions, smaller amounts of protein were needed.

3. After preparation and heating to deactivate proteolytic enzymes, samples were put on ice
until loaded into the gel with flat-tip gel loading pipette tips. Twenty microliters (20 pL) of
the mixture was added to each well.

4. Gels were mn at approximately 40 V for several hours to prevent “smiling’ at room
temperature until the blue dye reached the bottom of the gel.

5. Gels were removed from the gel electrophoresis apparatus and were immediately
transferred for Western blotting analysis.

PROTOCOL: Western lmmunoblottlng for Detectlon of Vitellogenin

*Adapted from methods from Current Protocols in Molecular Biology, Unit 10.8 (Immunoblotting
and lmmunodetection) .

*Notes are found in volume 1, pp. 30-32, 60-61 and volume 2, pp. 15-1 7, 19, 21 K. Nichols’
laboratory manual

The following methods were used for protein transfer and Western immunoblotting for
nonspecific and specific detection of plasma proteins and determination of the specificity of
polyclonal antibody for vitellogenin.

Materials

Bio-ice cooling unit (BioRad 170-3934)

Constant-current power supply (BioRad Model ZOO/2.0, 165-4761)

Fiber pads (BioRad 170-3933)

Filter paper (BioRad 170-3932)

Hyperfilm (Amersham RPN 1674)

Ice

Milli-Q water (dHZO)

Mini Trans-blot electrophoretic transfer cell apparatus: Mini-Protean Il (BioRad 170—3935)
Nitrocellulose membrane (BioRad 160-0145)

Pipettes

Plastic trays or lids

Tris-glycine gel run from SDS-PAGE, 4-15% (BioRad Ready gels, 161-0902)

Chemicals

3-Amino-9-ethylcarbazole (AEC, 20 mg tablets, Sigma A-6926)
Acetic acid
Coomassie blue G-250 (Brilliant Blue R, Sigma B-0149)

116

Donkey anti-rabbit lgG-horseradish peroxidase (HRP) conjugated antibody (Amersham NA934)
ECL detection reagents (Amersham RPN 2106)

Glycine (Sigma G-7126)

Methanol

Nonfat dry milk (Carnation)

Ponceau S (Sigma P-3504)

Rabbit anti-goldfish vitellogenin polyclonal antibody

Sodium chloride (NaCI, JT Baker 3624-05)

Tris (Mallincrodt 7732)

Tween-20 (SigmaUltra P-7949)

Buffers and Reagents
1. Gel transfer buffer
tris 6.06 g
glycine 28.63 g
methanol 400 mL
dHZO Fill to 2000 mL

*Recipe for use with nitrocellulose membranes. Decrease concentration to 300 mL MetOH when
using PVDF. CAPS transfer buffer can also be used - see p. 10.8.13 of Current Protocols

2. 10X Tris-buffered saline (20 mM Tris. 137 mM NaCl, TBS, pH 7.5)

 

Tris 20.02 9
NaCl 80.0 g
deo Fill to 1000 mL Adjust pH to 7.5 with NaOH or HCI

*May be stored at room temperature indefinitely

3. 1X Twe_§n ZONE-buffered saline (1X TTBS)

10X TBS 100 mL
Tween 20 (0.1%) 1.0 mL
dHZO Fill to 1000 mL

*Store at 4°C not more than three months

4. Blocking buffer

nonfat dry milk (5% wlv) 5.0 9

1X TTBS Fill to 100 mL
*Mix immediately before use

5. Ponceau 8 protein stain

Ponceau S 0.5 g
glacial acetic acid 1.0 mL
dHZO Fill to 100 mL

*Filter with 1 pm filter and store at 4°C. May be re-used.

6. Coomassie blue protein stain

acetic acid (10%) 100 mL
Coomassie brilliant blue G-250 0.06 g
dHZO Fill to 1000mL

*Store indefinitely at room temperature

117

Methods

Methods for protein transfer and immunoblotting follow that laid out in Current Protocols in

Molecular Biology, lmmunoblottlng and lmmunodetection, Unit 10.8. Only adaptations and

specific information not included in this protocol are specified below:

1.

Following equilibration in transfer buffer, gels were assembled in the gel transfer
apparatus with nitrocellulose membranes and transferred with cooling using the Bio-Ice
cooling unit at 75 V for approximately 1.5-2 hours. Ice was changed periodically to
prevent overheating and to allow efficient transfer of protein.

Upon completion of the transfer, gels were discarded and blots were detected for
nonspecific protein transfer with Ponceau S and destaining or Coomassie Brilliant Blue as
described in Cunont Protocols.

Blots were placed in blocking buffer for the start of Immunoproblng with Directly
Conjugated Secondary Antibody, p. 10.8.7.

Following incubations and rinses indicated in the protocol, primary antibody (rabbit anti-
goldfiSh vitellogenin) was diluted usually 1:5000 to 1:10.000. Secondary antibody
(donkey anti-rabbit IgG) was always diluted 1:5000.

Upon final rinsing after the secondary antibody incubation, bands were visualized using
one of two methods described below:

0 Chromogenic detection with AEC substrate for HRP: 1 AEC tablet was dissolved in
2.5 mL DMF and then added to 47.5 mL of 50 mM acetate buffer, pH 5.0 with stirring.
2.5 pL of 30% H202 was added just before use. This reagent was poured onto
nitrocellulose membrane and allowed to develop until bands were visualized. AEC is
poured off and nitrocellulose is rinsed with water and then dried.

0 Memo det_ection with ECL detection reagents (SQ example, Figure A3): Equal
volumes of detection reagents 1 and 2 (1.5 mL of each for one mini-blot) were mixed

together and were pipetted onto the membrane to cover the entire surface. The
membrane was incubated precisely one minute with the ECL reagents and then the
edge blotted onto kim—wipes to remove excess. Blots were wrapped in plastic wrap
with protein side facing the smooth Side of the plastic wrap. Membranes were
exposed to HyperfiIm-ECL and then developed. Exposure time may need to be
manipulated for best picture. For more complete description, see ECL Western
Blotting Protocols (1993).

 

118

Figure A2. Example of immunoblotted proteins with anti-vitellogenin polyclonal antibody detected
with ECL reagents. Lanes represent different samples.

pure male fem. male fem. male fem. male
vtg fhm fhm RBT RBT carp carp GF eng

-— 1 11

'v-ni

 

PROTOCOL: Vitellogenin Antibody Production

*Maintenance, injections, bleeding, and exanguation of the rabbits for antibody production were
done by University Laboratory Animal Resources at Michigan State University

'Notes are found in ”Vitellogenin Antibody Production ”, on pp. 35, 46—4 7, 53, vol. 1 K. Nichols' lab
notebook, and in the injection logs from ULAR

The following method was followed for preparation of antibodies by ULAR. A more
complete description of antibody production and titre checks are described in Current Protocols in
Molecular Biology, Preparation of Polyclonal Antlsera, Unit 11.12.

Materials

1 mL plastic syringes

22-G needles, sterile

3 mL Luer-Lok plastic syringes
3-way plastic locking hub stopcock
Centrifuge

Eppendorf vials (1.5 mL)
Gilman 0.2 pm syringe filters
Hunter's TitreMaxT” adjuvant
New Zealand white rabbits
Phosphate buffered saline
Pipette Tips

Pipettes

Purified antigen (vitellogenin)

119

Buffers

Phosphatp bufferfi saline (10 X)

NaCI (137 mM) 80 9

KCl (2.7 mM) 2.0 g

NazHPO4-7H20 (4.3 mM) 11.5 g

KHZPO4 (1.4 mM) 2.0 9 Adjust to pH 7.3 with 1 M NaOH or HCI

'May be stored at room temperature indefinitely

Methods

1.

10.

Pre-immune serum (20 mL)was taken from each rabbit 2 weeks prior to the initial
immunization to determine background immune response and comparisons after antigen
injection.

Pure antigen was diluted 15 pg/mL in 1.0 mL filtered PBS.

TrtreMax was mixed thoroughly by vortexing. 1.0 mL of adjuvant was pulled into one 3
mL luer-lok syringe and the needle discarded. One end of the 3-way stopcock was
securely attached to the syringe

The dilute antigen mixture (1.0 mL) was pulled into another 3 mL luer-lok syringe, the
needle discarded, and the second syringe attached to the 3-way stopcock.

The antigen and adjuvant were emulsified together ensuring that only the two stopcock
ends to the syringes were open by gently injecting one into the other back and forth. The
antigen and adjuvant were mixed until a stable emulsion was formed or when the
emulsion would not disperse when dropped into water.

Two New Zealand white rabbits (females) were each given a total of 1.0 mL of emulsified
antigen at a total of ten Sites subcutaneously.

Following initial injection and four weeks (28 days) to allow the immune response to
develop in the rabbits, one rabbit was boosted with the same concentration of pure
antigen in adjuvant.

Forty-two days (42 d) after the initial injection and subsequently every 2 weeks, rabbits
were anaesthetized and bled (15.0 mL) by ULAR staff. Rabbits were boosted according
to the schedule in the the vitellogenin laboratory notebook, K. Nichols (see Table A1 ).
Blood was allowed to clot overnight at 4°C. The following morning the serum was
separated from the clot by centrifugation at 4°C for 10 minutes at 5000 X9. Serum was
separated and 500 pL aliquots were stored in 1.5 mL eppendorf vials at -80°C.

Initial immune response before boosting was determined by slot-blot detection of antibody
titre (see protocol below). It was determined from this that boosting was needed to
produce the maximum immune response for vitellogenin in the rabbits.

120

11. The strength of the immune response throughout blood collections was evaluated by slot-
blot (see following protocol). 1

12. Antibody specificity for purified vitellogenin and multi-Species vitellogenin samples was
determined by SDS-PAGE and immunoblotting described above.

13. Optimal antibody dilutions for use in the vitellogenin ELISA described in the “Standard
Operating Procedure: Measurement of Plasma Vitellogenin in Fathead Minnows and
Goldfish by Competitive Enzyme—Linked Immunosorbent Assay (ELISA)" was determined
by standard curves incubated with different dilutions of primary antibody. The dilution that
provided 50% of maximum OD at the 50% binding point of the curve, was chosen as the
most optimum dilution. Antisera with the greatest slope gave the widest linear range for
discrimination and determination of plasma vitellogenin concentrations.

PROTOCOL: Slot-Blot for Detection of Specific and Nonspecific Immune Response In
Immunized Rabbits

“Instruction and laboratory equipment was initially generously provided by Dr. Norbert Kaminski
and Bob Crawford, Department of Pharmacology and Toxicology

*Notes for slot blot determinations of immune response may be found on pp. 56-59, 84-85, K.
Nichols Laboratory Notebook Vol. 1

The following protocol was used for determination of initial immune response before
boosting the rabbits and for evaluation of nonspecific immune response (or production of
immunoglobulin G). Once the rabbits were confirmed to have responded to the injections, Slot
blots were done to determine the specific response and the strength of the immune response and
antigenicity of the vitellogenin protein injected. Essentially, the protocol follows that for the
Western immunoblotting techniques after the slot-blot is made with method adaptations described
below.

Materials

Western immunoblotting materials and reagents (see above Protocol)
Hyperfilm (Amersham RPN 1674)

ECL Western blot detection reagents (Amersham RPN 2106)

Slot blot apparatus (BioRad 170-6542)

Vacuum pump

Vitellogenin antisera from immunized rabbits

Erlenmeyer flask with side arm

Vacuum tubing

Chemicals
Same as for Western immunoblotting.

Method

121

A nitrocellulose membrane was cut the same size as the slot blot apparatus. The
membrane was soaked in Milli-Q water for at least 15 minutes ensuring that no air
bubbles were trapped in the membrane.

For lgG determination, dilutions of the rabbit sera were made for the slot blot. For specific
antibody determination for vitellogenin, pure vitellogenin was diluted. Generally, Six
different antisera were checked (6 columns) over a logarithmic range of 8 concentrations
(8 rows) for IgG including pre—immune sera. Concentrations made for nonspecific lgG
response were 0, 105, 10", 10’ 10°, 10°, 10‘°, 10" antisera in TBS. A dilution of 102 was
first made. One hundred microliters (100 pL) will be added to each slot and dilutions must
be made to allow for this quantity. I

For vitellogenin response, all Six columns were made to contain the same concentration
range of pure goldfish vitellogenin (0, 100 pg, 10 pg, 1 pg, 0.1 pg, 0.01 pg, 1 ng, and 0.1
ng). Later, the blot was cut into strips according to columns and incubated with different
concentrations of primary antibody (rabbit anti-vitellogenin) to determine the specificity
and strength of the immune response for the protein.

The Slot-Blot apparatus was connected to tygon tubing and a 3-way stopcock exiting to
the side arm flask attached to a vacuum pump. The apparatus was opened up and
assembled with 3 Milli-Q wet pieces of blotting paper, then the nitrocellulose membrane
ensuring that no air bubbles remained between the nitrocellulose and the blotting paper.
The slot blot apparatus was then tightened on opposite corners.

The vacuum pump was turned on and the stopcock arranged so that the vacuum pump
was pulling air through the blot and then the apparatus was tightened more.

The vacuum pressure pulling through the blot was removed by turning the stopcock. Vlfith
a repeater pipette, 250 pL of TBS was added to each of the slots and then pulled‘through
with the vacuum. This step was repeated 2 times.

100 pL of sample was added to the designated Slot with a pipette ensuring that no air
bubbles were trapped in the slot. The vacuum was applied to pull the samples through
after all have been added.

Each slot was rinsed as in step 5. After the last wash, the vacuum was allowed to pull
each slot dry.

The blot was removed from the apparatus with forceps and Western immunoblotting was
begun by blocking the blot in a 5% milk solution as above.

For lgG determination, only one incubation with donkey anti-rabbit lgG antisera was
needed using a concentration of 1:5000.

122

10. For vitellogenin determination, two incubations were required. Each column of
vitellogenin dilution series were cut from the blot and incubated with different
concentrations or different collections of antisera (rabbit anti-vitellogenin) in a sealable
baggy or small container (ranging from 1:500 to 125000). Subsequently, HRP conjugated
secondary antibody (1:5000) was added to detect the primary antibody. Initially, this
method was used to compare pre- and post-immune sera for specific immune response.
Alter boosting, the method was used to determine the relative strength of the immune
response for each rabbit on a given blood collection day.

11. After prescribed incubations and rinses described above, blots were detected with ECL
detection reagents and exposed to film. The strength of the nonspecific immune
response (IgG) was determined by the presence of dark slots and compared between
boosted and non-boosted rabbits. Specific immune response strength was indicated by
the lowest VTG concentration that the antibody recognized. Subsequently, the antisera
was used to determine specificity for vitellogenin versus other plasma protein via SDS-
PAGE as described above.

Figure A3. Specific immune response for vitellogenin in New Zealand white rabbits. Note
columns represent a different antisera dilution or collection and rows represent serially dilute
pure vitellogenin or antigen.

#6 #6 #6 #1 #1 #1
pre 98d 126d 56d 112d 126d
10ug
1ug
100ng

10 ng

1 ng
0.1 ng

 

REFERENCES

Bevans HE, Goodbred SL, Miesner JF, Watkins SA, Gross TS, Denslow ND, Schoeb T (1996)
Synthetic organic compounds and carp endocrinology and histology in Las Vegas Wash and Las
Vegas and Callville Bays of Lake Mead, Nevada, 1992 and 1995. USGS Water-Resources
Investigations Report 96-4266:12 pp.

123

ECL Western Blotting Product Manual. 45 pp.

Folmar LC, Denslow ND, Rao V, Chow M, Crain DA, Enblom J, Marcino J, Guillette LJJ (1996)
Vitellogenin induction and reduced serum testosterone concentrations in feral male carp (Cyprinus
carpio) captured near a major metropolitan sewage treatment plant. Environmental Health
Perspectives 104:1096-1101

Gallagher, SR (1995) Electrophoretic separation of proteins. In Current Protocols for Molecular
Biology, FM Ausubel et al., eds. John Vlfiley & Sons, Inc., Massachussetts, pp. 1022-10235.

Harries JE, Sheahan DA, Jobling S, Matthiessen P, Neall P, Routledge EJ, Rycroft R, Sumpter
JP, Tylor T (1996) A survey of estrogenic activity in United Kingdon inland waters. Environmental
Toxicology and Chemistry 15:1993—2002

Harries JE, Sheahan DA, Jobling S, Matthiessen P, Neall P, Sumpter JP, Tylor T, Zaman N
(1997) Estrogenic activity in five United Kingdom rivers detected by measurement of
Vitellogenesis in caged male trout. Environmental Toxicology and Chemistry 16:534-542

Heppell SA, Denslow ND, Folmar LC, Sullivan CV (1995) Universal assay of vitellogenin as a
biomarker for environmental estrogens. Environmental Health Perspectives 103(Suppl 7):9-15

Hombeck, P., S.E. Winston, and SA. Fuller. 1995. Enzyme-linked immunosorbent assays
(ELISA). In Current Protocols for Molecular Biology, F.M. Ausubel et al., eds. John Wiley 8. Sons,
Inc., Massachussetts, pp. 11.2.1-11.2.22.

Mommsen TP, Walsh PJ (1988) Vitellogenesis and oocyte assembly. In: Hoar WS, Randall VJ
(eds) Fish Physiology, Vol. XIA. Academic Press, San Diego, pp. 347-406

Silversand C, Haux C (1989) Isolation of turbot (Scophthalmus maximus) vitellogenin by high-
perfon'nance anion-exchange chromotography. Journal of Chromatography 478:387-397

Sumpter JP, Jobling S (1995) Vitellogenesis as a biomarker for estrogenic contamination of the
aquatic environment. Environmental Health Perspectives 103(Suppl. 7):173-178

Tyler CR, van der Eerden B, Jobling S, Panter G, Sumpter JP (1996) Measurement of
vitellogenin, a biomarker for exposure to oestrogenic chemicals, in a wide variety of cyprinid fish. J
Comp Physiol B 166:418-426

124

Table A1. Schedule for immunization and blood collection for vitellogenin antibody production in

New Zealand white rabbits by ULAR.

 

Date

Actlvlty

 

9—May-96

#1, #6 injected 30 pglmL Vtg (#2-26 vtg fraction)

 

6-Jun-96

#1 Boosted with 30 pglmL (1.0 mL of #26 from 3-14)

 

20-Jun-96

Blood collected #1 , #6 - aliquoted (500 pL and stored at -80 deg C)

 

5-JuI-96

Blood collected #1, #6 - aliquoted (500 pL and stored at -80 deg C)

 

16-Jul-96

Slot blot for nonspecific immune response of #1, #6 at 56 days

 

1 9-JuI-96

#6 boosted with 30 pglmL (1.0 mL of #2-26), #1 blood collected

 

25-Jul-96

Slot blot specific immune response of rabbit #1 56 d to Vtg

 

29-Jul-96

SDS-PAGE for rabbit #1 56 d antisera with Vtg protein v. GVDK Ab

 

31-JuI-96

Blood collected #1, #6 - aliquoted and stored -80 deg C

 

1 5-Aug-96

Blood collected #1, #6 - aliquoted; Rabbit #1 boosed with 0.7 mL (15 pglmL)

 

29-Aug-96

Blood collected #1, #6 - aliquoted and stored —80 deg C

 

1 2-Sep-96

Blood collected #1, #6 - aliquoted by EMS and stored -80 deg C

 

26-Sep-96

Blood collected #1, #6 - aliquoted and stored -80 deg C; Slot blot #1, #6

 

1 0-Oct-96

Blood collected #1, #6 - aliquoted and stored -80 deg C

 

24-Oct-96

Blood collected #1, #6 - aliquoted and stored -80 deg C

 

7-Nov-96

Rabbit #1, #6 injected with 15 pglmL each

 

 

21 -Nov-96

 

Rabbits exanguated by ULAR and all blood collected

 

 

STANDARD OPERATING PROCEDURE
Measurement of Plasma Vitellogenin in Fathead Minnows and Goldfish by

Competitive Enzyme-Linked lmmunosorbent Assay (ELISA)

Adapted from method by Z.X. Yao and G. Van Der Kraak
University of Guelph

Krista M. Nichols
Michigan State University
Aquatic Toxicology Laboratory

Version 3.0
23 March 1997

SCOPE

Vitellogenin (VT G) is a high molecular weight glycolipophosphoprotein synthesized in the
liver of oviparous vertebrate females during exogenous Vitellogenesis (for review, see Ho, 1991;
Lazier and MacKay, 1993). The protein is under strict control of estrogen (E2); estrogen binds to
the estrogen receptor (ER) in the nucleus of hepatocytes and subsequently binds the estrogen
responsive element (ERE), a palindromic sequence upstream of the vitellogenin gene(s) (Landel
et al., 1993). Binding of the E2-ER complex to the ERE allows a DNA conformational change
whereby transcriptional factors associate with the initiation complex in the promoter region for
transcriptional activation (Tsai and O’Malley, 1994). Transcriptional inititiation of one or several
vitellogenin genes (not yet fully identified in fish) begins and stabilization of vitellogenin mRNA
may occur (Lazier and MacKay, 1993; Nielsen and Shapiro, 1990). The vitellogenin mRNA is
translated by the ribosomes on the rough endoplasmic reticulum and undergoes extensive
processing in the Golgi apparatus before transport in the blood to developing oocytes.
Investigators have indicated that VTG may circulate both as a dimer and a monomer (DeVIaming
et al., 1980; Lazier and MacKay, 1993). In the oocytes, VTG is cleaved to lipovitellin and
phosvitin where it is incorporated into the yolk as the primary proteinaceous reserve for
developing embryos.

The vitellogenin competitive Enzyme-Linked Immunosorbent Assay (ELISA) determines
the amount of circulating plasma vitellogenin in fishes. Traditionally, the ELISA was developed for

125

126

measurement of vitellogenin in fishes for determination of female reproductive status during
maturation and spawning (Kishida et al., 1992; Mananos et al., 1994; Mourot and Le Bail, 1995;
Chang et al., 1996). Both male and female fish have the hepatic vitellogenin genes. Because
males have very low or no levels of endogenous estrogen, the protein is not normally expressed
in males. However, upon administration of estradiol or environmental estrogens has been shown
to induce Vitellogenesis in some cases up to levels found in females (Chen, 1983; Purdom et al.,
1994). Vlfith evidence of endocrine disruption on fish and wildlife populations, induction of
Vitellogenesis in male fish has become a biomarker of exposure to environmental estrogen. Thus,
the assay has been developed and used as a screening tool for fishes exposed to exogenous
environmental estrogens, both in the laboratory and in the field, that may be impacting
reproductive health and success (Heppell et al., 1995; Sumpter and Jobling, 1995; Tyler et al.,
1996).

A universal vitellogenin assay would be advantageous for identifying the reproductive
status in females and determining exposure and effects of environmental endocrine disruptors on
fishes. Production of a universal, immortal antibody for vitellogenin across a variety of classes
and families of fishes would require much time and effort. Folmar et al. (1995) have investigated
the possibility of an universal vitellogenin antibody by determining a consensus sequence for fish
vitellogenins in the N terminal region translating for the protein and subsequently producing a
monoclonal antibody that will recognize vitellogenins across species and families of fishes and
other organisms. However, to save time and expense, polyclonal antibodies may also be used
provided the antibodies are specific and sensitive enough to recognize only VTG and at the low
levels of induction that may be found in males.

Although the vitellogenin genes and proteins are highly conserved among the oviparous
vertebrates, there are considerable differences in the sequence of the proteins isolated from
different species that provide the epitope or site of antigenicity for production of polyclonal
antibodies. Polyclonal antibodies used in VTG ELISA in the literature have indicated little cross-
reactivity to species not within the same family (Tyler et al., 1996). The following ELISA was
developed for goldfish (Carassius auratus) and fathead minnows (Pimephales promelas) — two
Cyprinid fish - with a polyclonal antibody for goldfish vitellogenin. The vitellogenin competitive
ELISA described in this standard operating procedure was developed with the following
objectives:

. To quantify vitellogenin in fathead minnow and goldfish plasma after in situ and laboratory
exposure to environmental endocrine disruptors

a To provide an ELISA optimized and capable of using purified goldfish vitellogenin and VTG
antisera in the competitive assay to accurately detect and quantify the protein in other‘cyprinid
species, namely the fathead minnow.

127

REFERENCES

Chang, C.-F., E.-L. Lau, B.-Y. Lin, and S.-R. Jeng (1996) Characterization of vitellogenin induced
by estradiol-17 beta in protandrous black porgy, Acanthopagrus schlegeli. Fish Physiol. Biochem.
15, 11-19.

Chen, T.T. (1983) Identification and characterization of estrogen-responsive gene products in the
liver of rainbow trout. Can. J. Biochem. Cell Biol. 61, 802-810.

DeVlaming, V.L., H.S. erey, G. Delahunty, and R.A. Wallace (1980) Goldfish (Carassius auratus)
vitellogenin: Induction, isolation, properties and relationship to yolk proteins. Comp. Biochem.
Physiol. 678, 613-623.

Folmar, L.C., N.D. Denslow, R.A. Wallace, G. LaFleur, T.S. Gross, S. Bonomelli, and C.V.
Sullivan (1995) A highly conserved N-tenninal sequence for teleost vitellogenin with potential
value to the biochemistry, molecular biology, and pathology of vitellogenin. J. Fish Biol. 46, 255-
263.

Heppell, SA, ND. Denslow, L.C. Folmar, and C.V. Sullivan (1995) Universal assay of vitellogenin
as a biomarker for environmental estrogens. Environmental Health Perspectives 103(Suppl 7), 9-
15.

Ho, Shuk-M. (1991) Wtellogenesis. In: Vertebrate Endocrinology: Fundamentals and Biomedial
Implications, edited by P.K.T. Pang and MP. Schreibman. Academic Press, Inc., New York, pp.
91-125.

Kishida, M., T.R. Anderson, and J.L. Specker (1992) Induction by B—estradiol of vitellogenin in
sptriped bass (Morone saxatilis): characterization and quantification in plasma and mucus. Gen.
Comp. Endocrinol. 88, 29-39.

Landel, C.C., P.J. Kushner, and G.L. Greene (1995) Estrogen receptor accessory proteins:
effects on receptor-DNA interactions. Environmental Health Perspectives 103(Suppl. 7), 23—28.

Lazier, CB. and ME. MacKay (1993)V1tellogenin gene expression in teleost fish. In: Biochemistry
and Molecular Biology of Fishes, edited by Hochachka and Mommsen. Elsevier Science
Publishers,, pp. 391-405.

Maisse, G., B. Mourot, B. Breton, A. Fostier, O. Marcuzzi, P.Y. Le Bail, J.L. Bagliniere, and A.
Richard (1991) Sexual maturity in sea trout, Salmo trutta L., running up the River Calonne
(Normandy, France) at the 'finnock' stage. J. Fish Biol. 39, 705-715.

Mananos, E, J. Nunez, S. Zanuy, M. Carrillo, and F. Le Menn (1994) Sea bass (Dicentrarchus
labrax L.) vitellogenin. Il - Validation of an enzyme-linked immunosorbent assay (ELISA). Comp.
Biochem.Physiol.1073,217-223.

Mourot, B. and P.-Y. Le Bail (1995) Enzyme-linked immunosorbent assay (ELISA) for rainbow
trout (Oncorhyncus mykiss) vitellogenin. Journal of Immunoassay 16, 365-377.

Nielsen, D.A. and D.J. Shapiro (1990) Insights into hormonal control of messenger RNA stability.
Mol. Endocrinol. 4, 953-957.

Purdom, C.E., P.A. Hardiman, V.J. Bye, N.C. Eno, C.R. Tyler, and J.P. Sumpter (1994)
Estrogenic effects of effluents from sewage treatment works. Chemistry and Ecology 8, 275-285.

128

Sumpter, J.P. and S. Jobling (1995) Vitellogenesis as a biomarker for estrogenic contamination of
the aquatic environment. Environmental Health Perspectives 103(Suppl. 7), 173-178.

Tsai, M.-J. and B.W. O'Malley (1994) Molecular mechanisms of action of steroid/thyroid receptor
superfamily members. Annu. Rev. Biochem. 63, 451-486.

Tyler, C.R., B. van der Eerden, S. Jobling, G. Panter, and J.P. Sumpter (1996) Measurement of
vitellogenin, a biomarker for exposure to oestrogenic chemicals, in a wide variety of cyprinid fish.
J. Comp. Physiol. B 166, 418-426.

SUMMARY OF METHOD

This ELISA, adapted from methods by Z. Yao , Maisse et al. (1991), Mourot and Le Bail
(1995), and Mananos et al. (1994) was designed to quantify plasma vitellogenin for two species of
Cyprinids, fathead minnows and goldfish for determination and calibration of environmental
endocrine exposure. The assay is a competitive ELISA whereby antigen (VT G) is coated onto the
well surfaces of 96-well plates. Dilute samples compete with the pure VTG bound to the surface
of the wells with the primary rabbit anti-VTG antibody in an overnight incubation. Vitellogenin in
the samples with inhibit binding of the primary antibody to the pure VTG coated on the plate.
Upon removal of the samples and primary antibody with washing, VTG antibody bound to the
plate is quantified by adding a second antibody, goat anti-rabbit lgG linked to horseradish
peroxidase (HRP) and subsequent color development by the HRP and o-phenylenediamine
(OPD) enzymatic reaction. Logit-log transformation of the standard curves and linear regression
provide the equation whereby vitellogenin in the samples can be quantified. The assay is rapid
and inexpensive with little interference from nonspecific male plasma proteins. This ELISA
provides efficient and rapid determination of VTG in large numbers of samples thus providing an
excellent screening tool for environmental endocrine effects.

SIGNIFICANCE AND USE

Traditionally VTG measurements have been made by nonspecific, time consuming, or
expensive methods such as plasma alkaline labile phosphorus measurements, high performance
liquid chromatography (HPLC), gel electrophoresis, radioimmunoassay, and measurement of VTG
mRNA levels in the liver of fish. The alkaline labile phosphorus determination depends on the
high degree of phosphorylation of VTG (0.79%, DeVlaming et al., 1980) but is nonspecific for the
target protein. Electrophoresis and HPLC require much time and expense for sample preparation,
separation, and quantifications. Although measurement of VTG mRNA in the liver may be useful
for detection of the strength and stability of the protein message or signal, these methods of
detection are also time consuming, involve expensive methods requiring cDNA clones of specific

129

mRNA, and are not efficient for routine screening. Radioimmunoassay has, traditionally, been
more sensitive in quantification of hormones and proteins in blood plasma. Recently, however,
investigators are trying to use less expensive and more safe methods that do not involve
radioactive labels.

Disadvantages of the VTG ELISA discussed herein include perhaps decreased senSitivity of
the method compared to radioimmunoassay. The dynamics of the assay are dependant upon the
stability, specificity, and sensivity of the polyclonal antibodies produced to the vitellogenin and
care must be taken to Optimize the assay before addition of new immunochemicals and antisera in
the assay.

Overall, the development of this ELISA for VTG quantification provides Specific, rapid, and
inexpensive means for detection and screening for the protein in environmental samples.

EQUIPMENT AND APPARATUS

96-well plate Spectra Reader with 492 nm and 650 nm filters (OEM Version; Cayman Chemical
Autoreader, Cayman Chemical, Ann Arbor, MI)

200 pL multi-channel pipette

Incubator

Rainin pipettes (2 pL, 10 pL, 20 pL, 100 pL, 200 pL, 1000 pL)

Eppendorf repeater pipette

Shaker

MATERIALS

Chemicals

1,2-phenylene diamine or o-phenylene diamine (OPD; Sigma P3804)
30% hydrogen peroxide (H202; Baker 2186-01)

Ammonium acetate (Baker 0596-01 or 432 BCH Stores)

Citric acid (Columbus Chemical Industries, Inc. 11993-010 or BCH Stores 2368)
Gentamycin (Boehringer Manheim 1059467)

Normal goat serum (Sigma G-9023)

Sodium bicarbonate (NaHCOa; Baker 3506-01)

Sodium chloride (NaCI; Baker 3624-05)

. Sodium hydroxide or hydrochloric acid, concentrated (NaOH or HCI)
10. Sulfuric acid (H2804)

11. Tris (Mallinckrodt 7732)

12. Tween-20 (Sigma P-7949)

”SPNP’SPF‘P’N?

Supplies

0.5 mL and 1.5 mL eppendorf vials

96 well round bottom ELISA plates (Corning 25802 or VWR Scientific 62407-903)
Eppendorf repeater pipetter combitip syringes, 2.5 mL capacity (Baxter P5063-26)
Paper towels

Pipette boat

Plastic wrap

P’SPPP’N.‘

130

7. Rainin pipette tips (1-10 pL, 1-200 pL, 200-1000 pL)
8. Ultrapure or deionized water (dH20)

ELISA BUFFERS AND REAGENTS

All of the buffers below should be prepared before beginning the VTG ELISA with the
exception of the AACA solution for color development. SBB and TBS-T may be stored at 4°C for
3 months but the pH must be checked and corrected before each use. TBS-T-SG may be stored
at 4°C only for a few days if made in excess. Ammonium acetate and citric acid solutions may be

stored separately at room temperature for extended lengths of time. AACA solution must be

mixed immediately before use.

1. Sodium bica_rbona_te buffer (50 mM SBB)

NaHCOa 4.20 g
gentamycin 5.0 mg
dH20 fill to 1000 mL Adjust pH to 9.6 with 1 M NaOH or HCI ~

*Can be stored at 4°C, check and adjust pH before use

2. Tris buffer (10 mMTBS-T)

Tris (10 mM) 1.211 9

NaCl (0.15 M) 8.766 g

Tween-20 (0.1%) 1.0 mL

gentamycin 5.0 mg

deo fill to 1000 mL Adjust pH to 7.5 with 1 M NaOH or HCI
*Can be stored at 4°C, check and adjust pH before use

3. Blocking and sample buffer (TBS-T-SG; 2% goat serum)

 

Normal goat serum 2.0 mL
TBS-T buffer 98.0 mL
*Mix just before use

4. Ammonium acetate-citric acid solution (AACA)
(a) Ammonium acetate (50 mM)
Ammonium acetate 0.385 g

 

deO fill to 100 mL pH 6.68
(b) Citric acid solution (50 mM)
Citric acid 0.525 g
dH20 fill to 50 mL pH 2.13 _
*Measure the amount of (a) needed for the number of plates and adjust (a) to pH 5.0 with (b)
5. OPD solution 1 pla_te 2 plates 3 plates
o-phenylene diamine 10 mg 15 mg 25 mg
30 % H202 10 pL 15 pL 25 pL
AACA solution 20 mL 30 mL 50 mL

““OPD is TOXIC, wear gloves

131

HAZARDS AND PRECAUTIONS

Strong acid and strong base solutions (citric, hydrochloric, sulfuric acids and NaOH) are
caustic; be sure to wear protective clothing, gloves, and eyewear. OPD is toxic, be sure to wear
gloves at all time when in use.

Vitellogenin in the purified fractions and in the samples are very sensitive to proteolysis
when exposed to room temperature and repeated freezing and thawing. Immunochemicals and
antisera are also sensitive to degradation particularly to freezing and thawing. All antisera, pure
protein, and samples should be stored at -80°C in working aliquots and Should not be exposed to
room temperature before the assay for long periods of time. Care should be taken to minimize
repeated freezing of samples and antisera.

PROCEDURE

NOTE THAT THIS PROCEDURE FOLLOWS FOR ONE (1) - 96 WELL PLATE.
CALCULATIONS SHOULD BE MADE ACCORDING TO THE NUMBER OF PLATED NEEDED
TO ASSAYAT ONE TIME.

A. Samples and standard curve layout

Before beginning the assay and dilution of samples and standards, use the 96-well plate
layout form attached to determine placement of standard curves and samples (Figure A4).
Standard curves should be at least duplicated on each plate. Samples should be at least
duplicated for each dilution. Generally, more than 34 plates run at one time can become
cumbersome and constraining on timing of incubations.

8. Plate Coating

Check the pH of SBB and adjust if needed to 9.6.

2. Aliquots of purified goldfish vitellogenin should be stored in at least -20°C (frost
free). Remove one aliquot (5lealiquot, 1.506 mgImL) of purified VTG from
freezer and thaw at room temperature.

3. Pipette 1.66 pL VTG solution into 15 mL SBB and mix well by aspiration with the
repeater pipette and 2.5 mL combih'p.

4. Pipette 150 pL of solution in 3 into each well of a 96 well plate to achieve a
coating rate of 25 nglwell.Cover the plates with plastic wrap and incubate for at
least 3 hours at 37°C or overnight.

132

C. Standard curve dilutions

During plate coating the following standard curve dilutions should be made:

1.

Pipette 1. 0 pL purified goldfish VTG from thawed aliquot above- (1. 506 mglmL)
into 999 pL of TBS-T-SG to get a 3012 ng/mL dilution. This rs tube #1. .

Do serial dilutions to get the remaining 9 standard solution concentrations. For
example, take 500 pL of the first dilution and add to 500 1.1L of TBS-T-SG buffer
for dilution #2 and so on. The following table gives the standard curve dilutions:

lube—ti IJ_Uw_e|| nil/Len Lia/ml;

1 50 75.3 1506

2 50 37.65 753

3 50 18.825 376.5

4 50 9.4125 188.25
5 50 4.70625 94.125
6 50 2.35 47.0625
7 50 1.177 23.5

8 50 0.5883 1 1 .77

9 50 0.2941 5.883
10 50 0.147 2.941

11 0 (+Ab) Total count

12 0 (-Ab) Background (non-specific binding)

D. Sample dilutions

Samples should be diluted during the three hour plate coating incubation:

1.

Goldfish or fathead minnow plasma (thawed from storage at -80°C) can be diluted
appropriately in 1.5 mL eppendorf vials with TBS-T-SG. Typically a 1:5000
dilution for fathead minnow females, a 1:8000 dilution for goldfish females, and
1:100 dilution for low expression (or no expression) of vitellogenin for males of
both species.

For a 1 :5000 dilution, first make a 1:1000 dilution by adding 1 pL of plasma to 999
pL TBS-T-SG. Take 200 11L of the 1:1000 dilution and add to 800 1.1L TBS-T-SB
for 1:5000.

For 1:8000, first make a 1:1000 dilution by adding 1 1.1L of plasma to 999 1.1L TBS-
T-SG. Take 125 pL of the 1:1000 dilution and add to 875 pL for a final 1:8000.
For 1 :100, dilute 5 11L of plasma in 495 pL of TBS-T-SG. ‘

133

Plate washing #1 - removal of excess plate coating.

1.

Discard the content of each well in sink and pat plate on paper towels.

2. Wash well 4 times with 200 1.1L TBS-T buffer with a multi-channel pipette, patting
on paper towels each time.
Plate saturation

To block or coat the remainder of the well surface not coated by purified GF VTG:

1.
2.
3.

Add 200pL TBS-T-SG to each well with the multi-channel or repeater pipette.
Incubate at 37°C for 30 minutes.

Discard the solution in the sink and pat the plates dry on paper towels. Do not
wash.

Addition of first antibody (Ab)

1.

Add 50pL VI'G standard solution or diluted samples to designated well aCcording
to the plate layout Be sure that for standard 11 (Total count), 50 pL of TBS-T-SG
is added and 150 1.1L is added to standard 12 (NSB) well.

Dilute primary antibody in TBS-T-SG and add 100 pL VTG diluted antiserum to
each well. For this assay, optimum antibody dilution of v9912rb1 antisera is
1245,000. DO NOT ADD FIRST ANHBODY TO STANDARD #12 WELLS FOR
MEASURING NSB. Note these dilutions have been made to achieve a 50% max
OD at the middle dilution for the standard curve. Antiserum and 2nd antiserum
dilutions should be determined with this in mind.

Shake briefly to stir the contents of each well together. Cover the plates with
plastic wrap and incubate at room temperature (20-25°C) overnight.

Plate washing #2

1.

Proceed as for plate washing #1 for removal of sample and unbound primary
antibody.

Addition of 2nd antibody

1.

Dilute the goat anti-rabbit gamma-globulin (horseradish peroxidase) diluted
1:2000 in TBS-T-SG buffer. Goat anti-rabbit-HRP should be stored in working
aliquots of 25 pL in the -80°C freezer.

Add 150 1.1L of dilute secondary antibody-HRP to each well with a repeater

pipette.

134

3. Cover plates in plastic wrap and incubate 37°C for 2 hours.

Plate washing #3

1. Proceed as for plate washing #1 to remove excess secondary antibody.

Color development

1. Mix the AACA solution (adjust to pH 5.0 with citric acid) with the appropriate
amount of CPD and then add the hydrogen peroxide just before you are ready to
use it. ‘

Add 150 1.1L OPD solution into each well with a repeater pipette.
Incubate plates at room temperature for 0.5 hour in the dark. Light will cause the
coloring reaction to occur regardless of the amount of VTG present.

Stop coloring
1. Add 50 pL of 5 M sulphuric acid to each well with a repeater pipette.
2. Shake the plates on an auto-shaker for 10 minutes.

Optical Density readings (OD) on a plate-reading spectrophotometer

1. While the plates are being shaken, turn the computer and plate-reading
spectrophotometer on. The software for the spectrophotometer is Cayman EIA
software Version 2.0. _

2. Go to “Read a Plate” in the main menu and ensure that all settings are correct:
<F4> Reader Specific Options: Choose dual wavelength and enter 492 nm for
the Measured wavelength and 650 nm for Reference; make sure Blanking is
turned OFF
<F1> Plate ID: Name the plate

3. Scan each plate ensuring that after each is read, a new plate name is entered for
the next plate before reading.
4. Print the data for a given file by <F8> Print file to printer when the appropriate

name is written in the Plate ID space.
5. After closing the ELISA software, use the File Manager in Windows to copy the
.Slt files from the c:/cayman drive for data analysis and backup.

Data Analysis

1. In Microsoft Excel, the data must be arranged so that it may be manipulated for

135

calculations (see Figure A5). A macro has been made to do this for routine
sample runs. For the standards and samples, coefficient of variation, mean, %
bound, and logit for the Optical Density (OD) readings are calculated with
formulas in the Spreadsheet. Calculations for % bound and logit OD are as

follows:
a. % bound = (sample mean OD-NSB OD)/(Total count OD-NSB OD)
b. logit OD = log [(%bound )l(1-%bound)]

In the spreadsheet, plot the standard curve with log VTG concentration (nglwell)
on x-axis and logit OD on the y-axis and perform a linear regression resulting with
an equation that can be used to determine sample concentrations of VTG. The
logit transformation places less emphasis on the tails of the standard curve
(compresses the error) and more emphasis in the range of the 50% maximum
binding. Due to this error in the tails of the standard curve, it is important that
samples of fish producing vitellogenin (induced or females) remain as close to
50% bound as possible (between 15 and 85% binding).

a. The regression equation on the log/logit transformation of the standard
curve is:
logit OD=m*log vtg + b,
m=slope of standard curve regression line 8 b=y-intercept

b. Sample vitellogenin concentrations may be made by rearranging
equation in 2a to: VTG = 10"((Iogit OD - b)lm). This VTG concentration
will be expressed as nglwell and the nglmL can be determined by dilution
rates of samples.

c. For the samples, the amount of VTG derived from the equation will be
expressed in nglwell and should be converted for the dilution factor to
pglmL. For example, for a sample that is diluted 1l500, VTG (pglmL) =
(nglwell)l50 pL‘500

136

Figure A4. Plate layout data sheet for determination of standard and sample placement in the
VTG ELISA.

 

 

137

PLATE LAYOUT Aquatic Toxicology Laboratory
Vitellogenin ELISA Michigan State University

Date:

 

Plate:

 

 

 

 

 

I'I'IUOW

 

‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

138

Figure A5. Example Microsoft Excel worksheet for a typical vitellogenin ELISA plate with
duplicate standard curves and duplicate samples

 

~v-.. .1.

 

139

as =33: no. at; 4 ENS..- . :2

 

rem
3&5? «>50 26:8». 9>

 

             

 

 

HICHIGRN ST

RTE UNIV. LIBRRRIES
III III IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
9 01 841 901 4

312 3