DRAFT
FINAL
REPORT
on
COMPARATIVE
EVALUATION
OF
VITELLOGENIN
METHODS
EPA
CONTRACT
NUMBER
68­
W­
01­
023
WA
3­
5,
Task
13
(
For
the
study
conducted
under
WA
2­
19)

May
2,
2003
Prepared
for
LES
TOUART,
PH.
D.
WORK
ASSIGNMENT
MANAGER
U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
DC
20460
BATTELLE
505
King
Avenue
Columbus,
Ohio
43201
Battelle
Draft
May
2003
i
TABLE
OF
CONTENTS
Page
ACKNOWLEDGMENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
iii
1.0
INTRODUCTION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
2.0
SAMPLE
PREPARATION
AND
HANDLING
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
3.0
SAMPLE
METHODS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
4.0
ANALYTICAL
METHODS
OF
THE
PARTICIPATING
LABORATORIES
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4
5.0
PARTICIPATING
LABORATORIES
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
6.0
DATA
ANALYSIS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
7.0
RESULTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
8.0
DISCUSSION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
9.0
SUMMARY
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
44
10.0
REFERENCES
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
46
APPENDIX
A.
PARTICIPATING
LABORATORY
PROTOCOLS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
A­
1
APPENDIX
B.
PURIFIED
VITELLOGENIN
PROTOCOL
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
B­
1
APPENDIX
C
DESCRIPTIVE
STATISTICS
OF
THE
WITHIN­
RUN
VTG
RESULTS.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
C­
1
APPENDIX
D
DESCRIPTIVE
STATISTICS
OF
THE
INTRA­
LABORATORY
RESULTS
.
.
.
.
.
.
.
.
.
.
.
.
.
D­
1
APPENDIX
E
DESCRIPTIVE
STATISTICS
OF
THE
INTRA­
ASSAY
RESULTS.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
E­
1
APPENDIX
F
GC­
MS
ANALYTICAL
PROTOCOL
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
F­
1
LIST
OF
TABLES
Table
1.
Summary
of
Reporting
Laboratories
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
Table
2.
Summary
of
Antibody
Methods
Used
by
Participating
Laboratories
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
9
Table
3.
Summary
of
Standards
Employed
by
Participating
Laboratories
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
Table
4.
Summary
of
the
Concentration,
Standards,
Antibody,
and
Assay
Codes
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
Table
5.
Descriptive
Statistics
of
the
CVs
of
the
Within­
run
Analytical
Results
where
Q1
and
Q3
are
the
1st
and
3rd
Quartiles,
Respectively
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
Table
6.
Descriptive
Statistics
of
the
CVs
of
the
Intra­
Assay
Analytical
Results
where
Q1
and
Q3
are
the
1st
and
3rd
Quartiles
Respectively
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
Table
7.
Descriptive
Statistics
for
the
Mean
VTG
Results
Averaged
Over
Laboratory,
Antibody,
Standard,
and
Assay
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17
Table
8.
Results
of
the
Linear
Regression
on
Within­
Laboratory,
Sample
Type,
Standard,
and
Antibody
Ranked
VTG
Concentrations
Observed
in
the
Series
Codes
0
Through
4
.
.
.
.
.
.
24
Battelle
Draft
May
2003
ii
Table
9.
Results
of
Tukey's
HSD
Multiple
Comparison
Test
On
the
Ranked
VTG
Concentrations
Using
Laboratories
as
Replicates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
25
Table
10.
Descriptive
Statistics
for
the
VTG
Concentrations
Obtained
from
the
Carp
sandwich
ELISA
Each
Sample
Type,
Standard,
Laboratory,
and
Concentration
Code
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
28
Table
11.
Upper
and
Lower
95%
Confidence
Limits
for
the
Mean
VTG
Concentrations
Obtained
from
the
Carp
Sandwich
ELISA
Using
Laboratories
as
Replicates
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
30
Table
12.
Summary
of
the
reported
mRNA
results
for
the
liver
samples
from
unexposed
and
exposed
male
and
female
fathead
minnows.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
32
Table
13.
The
top
sequence
matches
for
the
VTG
peak
from
exposed
female
plasma
sample
219PT18.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
38
Table
14.
Summary
of
the
results
of
GC­
MS
analysis
of
unexposed
and
exposed
male
and
female
fathead
minnow
plasma
samples.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
LIST
OF
FIGURES
Figure
1.
Distribution
of
the
Within­
run
Coefficient
of
Variation
(
CV)
of
the
VTG
Analytical
Result
on
Plasma
Samples
for
Each
Concentration
Code
(
0
­
5)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
Figure
2.
Distribution
of
the
Within­
run
Coefficient
of
Variation
(
CV)
of
the
VTG
Analyt­
ical
Result
on
Homogenate
Samples
for
Each
Concentration
Code
(
1
­
5)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13
Figure
3.
Each
Laboratories
VTG
Concentration
for
A
Given
Concentration
Code
Averaged
Over
Antibody,
Standard,
and
Assay
for
Plasma
Samples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
Figure
4.
Each
Laboratories
VTG
Concentration
for
A
Given
Concentration
Code
Averaged
Over
Antibody,
Standard,
and
Assay
for
Homogenate
Samples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
16
Figure
5.
The
Percentage
Difference
in
VTG
Concentrations
Averaged
Over
Replicates
Between
the
Homologous
and
Purified
fathead
minnow
Standard
Data
in
Plasma
and
Homogenate
Samples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
18
Figure
6.
The
VTG
Concentration
in
Plasma
Samples
for
Each
Standard,
Laboratory,
Antibody,
and
Concentration
Code
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
20
Figure
7.
The
VTG
Concentration
in
Homogenate
Samples
for
Each
Standard,
Laboratory,
Antibody,
and
Concentration
Code
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
21
Figure
8.
The
CVs
of
the
Average
VTG
Concentration
for
Each
Laboratory,
Antibody,
and
Concentration
Code
for
Plasma
Samples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
Figure
9.
The
CVs
of
the
Average
VTG
Concentration
for
Each
Laboratory,
Antibody,
and
Concentration
Code
for
Homogenate
Samples
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
27
Figure
10.
UV
trace
of
equal
amounts
of
BSA
and
VTG
after
purification
for
GC­
MS.
.
.
.
.
.
.
.
.
.
.
.
34
Figure
11.
10
ug
VTG
injected
after
"
spike"
into
219PR33
control
plasma:
Purified
using
microcon
100
(
BSA
added
[
10ug
injected])
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
34
Figure
12.
No
VTG
"
Spiked"
into
219Pk18
Control
Plasma:
Purified
using
microcon
100
with
BSA
added
(
10ug
injected)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
35
Figure
13.
219PU18
Exposed
Plasma:
Purified
using
microcon
100
with
60
ug
BSA
(
6
ug
injected)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
36
Figure
14.
219PU18
Exposed
Plasma:
Diluted
1
to
5
with
additional
BSA
(
20
ug
BSA
injected)
.
.
.
.
36
Figure
15.
Comparison
of
the
peptide
digest
mass
spectra
from
VTG
fractions
collected
from
two
exposed
female
plasma
samples
and
from
the
VTG
standard.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
37
Battelle
Draft
May
2003
iii
ACKNOWLEDGMENTS
This
document
was
prepared
by:

Ann
D.
Skillman,
Valerie
Cullinan,
Michael
L
Blanton,
and
Irvin
R.
Schultz
Battelle's
Pacific
Northwest
Division
1529
West
Sequin
Bay
Rd.
Sequin,
Washington
98382
USA
This
report
would
not
be
possible
with
out
the
participation
of
the
following
laboratories
who
donated
materials
and
time
in
the
application
of
their
methods
to
the
samples
in
this
study.

University
of
Florida,
USA
University
of
Idaho,
USA
Oregon
State
University,
USA
US
EPA
Duluth,
USA
University
of
Exeter,
USA
Brixham
Environmental
Laboratory,
UK
Battelle
Richland,
USA
Battelle
Sequin,
USA
Molecular
Light
Technology,
UK
Biosense,
Norway
INERIS,
France
Cemagref,
France
University
of
Windsor,
Canada
University
of
Southern
Denmark,
Denmark
Finnish
Environmental
Institute,
Finland
The
authors
gratefully
acknowledge
review
comments
from
Les
Touart
(
U.
S.
EPA),
Gary
Ankley
(
U.
S.
EPA­
MED).
Battelle
Draft
May
2003
1
1.0
INTRODUCTION
The
U.
S.
Environmental
Protection
Agency
(
EPA)
is
implementing
an
Endocrine
Disruptor
Screening
Program
(
EDSP)
comprised
of
a
battery
of
Tier
1
screening
assays
and
Tier
2
tests.
An
international
effort
is
also
underway
to
develop
and
coordinate
screens
and
tests
appropriate
for
use
in
investigating
potential
endocrine
disrupting
chemicals.
The
Organization
for
Economic
Cooperation
and
Development
(
OECD)
has
established
an
Endocrine
Disruptor
Testing
and
Assessment
(
EDTA)
task
force
to
oversee
the
coordination
of
this
effort.
One
of
the
Tier
1
assays
under
development
is
a
short­
term
screening
assay
designed
to
detect
substances
that
interact
with
the
estrogen
and
androgen
systems
of
fish.
It
is
thought
that
the
inclusion
of
the
fish
screening
assay
in
Tier
1
is
important,
because
estrogenic
and
androgenic
controls
on
reproduction
and
development
in
fish
may
differ
significantly
from
that
of
higher
vertebrates,
such
that
mammalian
screening
methods
may
not
identify
potential
endocrine
disruptor
chemicals
(
EDCs)
in
this
important
class
of
animals.
The
measurement
of
a
biochemical
marker,
vitellogenin
(
VTG)
in
oviparous
vertebrates
is
generally
agreed
to
be
a
good
indicator
for
estrogenic
and
anti­
estrogenic
effects
and
is
proposed
as
one
of
several
endpoints
in
the
fish
screening
assay.
VTG
is
a
phospholipoglycoprotein
precursor
to
egg
yolk
protein
that
normally
occurs
in
sexually
active
female
oviparous
fishes,
but
can
be
induced
to
occur
in
males
in
response
to
estrogenic
substances.
Different
methods
are
available
to
assess
VTG
induction
in
fishes
including
measurement
of
the
VTG
protein
with
enzyme­
linked
immunosorbant
assays
(
ELISA)
or
gas
chromatography­
mass
spectrometry
(
GC­
MS),
and
mRNA
detection.
Both
plasma
and
whole
body
measurements
have
been
proposed.

The
purpose
of
this
study
was
to
conduct
a
survey
of
existing
VTG
analytical
methods
for
suitability
in
a
routine
screening
program.
This
comparison
was
not
intended
to
be
a
validation
of
a
given
method,
but
an
evaluation
across
methods
to
ascertain
the
qualitative
and/
or
quantitative
comparability
of
the
variety
of
methods
currently
available.

The
specific
objectives
of
the
study
included:

1)
The
preparation
of
a
standard
evaluation
series
of
fish
plasma
and
tissue
taken
from
fathead
minnow
(
Pimephales
promelas)
(
e.
g.
whole
body
homogenate)
to
provide
a
range
of
VTG
and
mRNA
concentrations
produced
in
male
and
female
fish
(
exposed
or
not
exposed
to
an
estrogen
compound).
The
series
was
produced
with
1)
uninduced
male,
2)
uninduced
female,
3)
induced
male
and
4)
induced
female
fathead
minnows.
In
addition
to
the
standard
series,
a
set
of
control
VTG
samples
were
prepared.

2)
The
identification
of
laboratories
to
participate
in
the
analysis
of
the
standard
evaluation
series
and
the
coordinated
transfer
of
the
samples
to
the
participating
laboratories
and
the
collection
of
analytical
results.
Each
laboratory
employed
the
specific
analytical
technique
routinely
used
by
the
laboratory
to
measure
VTG
or
mRNA.
This
resulted
in
a
variety
of
analytical
methods
applied
to
the
standard
series.
Each
laboratory
reported
the
results
of
the
analysis
and
provided
a
detailed
analytical
protocol.
Battelle
Draft
May
2003
2
2.0
SAMPLE
PREPARATION
AND
HANDLING
The
VTG
standard
evaluation
series
was
prepared
from
fathead
minnow
(
Pimephales
promelas)
plasma
and
whole
body
homogenates.
Fish
were
acquired,
exposed,
and
used
to
prepare
plasma
and
whole
body
homogenates
under
an
animal
care
protocol
reviewed
and
approved
by
the
Battelle's
Animal
Care
Committee
(
AAALAC
accreditation).
To
generate
samples
for
the
series
from
uninduced
male,
induced
male,
uninduced
female,
and
induced
female
fish,
sets
of
fish
were
exposed
to
estrogen
in
the
laboratory,
or
remained
unexposed.
The
unexposed
group
of
adult
male
and
female
fish
were
used
to
generate
uninduced
background
concentrations
for
male
and
female
fish.
The
production
of
VTG
was
induced
in
adult
male
and
female
fathead
minnows
via
exposure
to17$­
estradiol
in
the
laboratory
(
Korte
et
al.
2000).
After
a
1­
week
exposure,
when
maximal
VTG
protein
levels
in
plasma
were
anticipated,
the
fish
were
sacrificed,
plasma
was
collected,
and
whole
body
homogenates
were
prepared.
Several
steps
were
employed
to
aid
in
preserving
the
integrity
of
the
samples,
including
the
use
of
an
inhibitor,
cold
processing,
and
quick
freezing
to
stabilize
the
VTG
samples.
To
generate
a
sufficient
amount
of
material
for
the
study,
a
large
number
of
samples
were
collected
and
frozen
upon
collection.
Following
the
processing
of
the
individual
fish,
the
group
of
plasma
or
tissue
samples
were
quickly
thawed
and
combined
to
prepare
pooled
samples
for
the
series.
The
exposed
fish
samples
provided
the
plasma
and
tissue
samples
for
the
induced
male
and
induced
female
for
the
standard
series.
Samples
composited
from
individual
unexposed
fish
provided
the
uninduced
male
and
uninduced
female
sample
in
the
standard
series.
This
approach
resulted
in
four
samples
within
the
series.

In
addition
to
the
samples
representing
four
VTG
levels
found
in
plasma
and
whole
body
homogenate
from
uninduced
and
induced
male
and
female
fish
for
the
standard
series,
two
control
samples
were
generated
to
complete
the
plasma
analytical
series.
These
additional
samples
included
a
positive
and
negative
control,
prepared
with
unexposed
male
plasma.
To
prepare
the
negative
control,
anti­
VTG
antibodies
were
used
to
remove
the
protein
from
the
unexposed
male
plasma
resulting
in
a
sample
with
minimal
levels
of
protein
for
the
analysis.
For
the
positive
control,
VTG
was
purified
from
plasma
from
exposed
male
fathead
minnows
(
Denslow
et
al.
1999)
and
negative
control
plasma
was
spiked
with
a
known
amount
of
the
protein.
This
resulted
in
six
samples
in
the
standard
series,
covering
a
range
of
VTG
concentrations
from
minimal
to
high
levels
of
protein
in
male
and
female
plasma.

With
the
goal
to
provide
fully
replicate
samples
to
the
analytical
laboratories,
multiple
aliquots
of
each
sample
were
prepared
and
stored
in
a
repository
until
coordinated
shipment
to
the
participating
laboratories.
Each
aliquot
from
pooled
samples
represented
an
amount
of
material
(
plasma
or
tissue)
equivalent
to
that
obtained
from
a
single
fish.
The
samples
were
shipped
on
dry
ice
with
appropriate
Chain­
of­
Custody
documentation
and
with
instructions
for
sample
receipt,
storage,
analysis,
and
data
reporting.
The
laboratories
verified
that
the
samples
remained
frozen
during
shipping
and
were
advised
to
store
the
samples
under
conditions
similar
to
those
found
in
the
repository
and
to
avoid
exposing
the
samples
to
freeze­
thaw
cycles.
This
activity
was
thoroughly
documented
to
ensure
sample
integrity
was
not
compromised.
As
a
result
,
2
of
11
shipments
for
ELISA
analysis
were
determined
to
have
been
compromised
during
shipping
(
thawed),
and
replacement
sample
sets
were
provided
for
analysis
following
the
same
procedure.
Battelle
Draft
May
2003
3
A
similar
approach
was
applied
to
the
preparation
of
an
mRNA
standard
evaluation
series.
Adult
male
and
female
fathead
minnows
were
exposed
to
17$­
estradiol,
and
liver
tissue
samples
were
collected
from
fish
after
a
2­
day
static
renewal
exposure.
This
provided
samples
representing
uninduced
male,
induced
male,
uninduced
female,
and
induced
female,
respectively.
The
shipping
of
mRNA
were
analogous
to
those
previously
described
for
ELISA
samples.
All
shipments
were
received
by
testing
laboratories
were
un­
compromised
and
received
in
good
condition.

3.0
SAMPLE
METHODS
Standard
Series
One
hundred
adult
male
fathead
minnows
and
190
adult
females
were
exposed
to
a
nominal
concentration
of
300
nG/
L
of
17ß­
estradiol
in
a
7­
day
static
renewal
exposure,
while
an
additional
400
(
210
male
and
190
female)
fish
remained
unexposed.
On
day
2,
240
exposed
and
unexposed
fish
(
80
male
exposed
/
80
male
unexposed
and
160
female
exposed
and
160
unexposed)
were
sacrificed
and
the
liver
tissue
was
harvested
for
the
mRNA
standard
series.
The
liver
tissue
was
collected
under
cold
conditions
as
rapidly
as
possible
and
placed
in
RNA
stabilizing
buffer
prior
to
being
placed
in
frozen
storage
at
approximately
­
20
°
C.

On
day
7
of
the
exposure,
the
remaining
exposed
fish
were
anesthetized
with
tricaine
methane
sulfonate
(
MS­
222)
and
blood
was
collected
from
the
caudal
vessels
into
heparinized
hematocrit
tubes.
The
hematocrit
tubes
were
centrifuged
and
the
plasma
was
transferred
to
tubes
containing
the
inhibitor
aprotinin,
quick
frozen
in
liquid
nitrogen,
and
stored
at
­
80
°
C.
Plasma
was
collected
from
the
unexposed
fish
in
a
similar
manner.
A
subset
of
fish
from
the
exposed
and
unexposed
groups
was
used
to
prepare
whole
body
homogenates
for
analysis.
Whole
body
tissue
homogenate
was
prepared
by
placing
the
fish
into
ice­
cold
ELISA
assay
buffer
in
a
1:
1
ratio.
The
samples
were
homogenized
on
ice
and
following
homogenization,
the
samples
were
centrifuged
and
the
supernatant
was
harvested
and
frozen
at
­
80
°
C.
Care
was
taken
in
each
of
the
steps
to
collect
and
process
the
samples
in
a
timely
manner
under
cold
conditions
followed
by
a
quick
freeze,
to
limit
the
time
from
collection
to
storage.
The
stability
of
the
protein
vitellogenin
was
addressed
in
the
study
through
controlled
sample
collection,
processing,
storage,
and
shipping.
In
each
of
the
steps
used
to
collect
and
process
the
VTG
samples,
care
was
given
to
performing
the
steps
1)
in
a
timely
manner,
2)
under
cold
conditions,
3)
with
the
use
of
protease
inhibitors,
4)
followed
by
a
quick
freeze,
to
limit
the
time
from
collection
to
storage.
Following
collection,
plasma
and
tissue
homogenate
samples
were
pooled
to
create
the
composite
samples
for
the
series.
From
this
pooled
material
multiple
samples
were
created
for
each
analytical
method
and
given
a
unique
code
,
and
entered
into
a
repository
management
system.

Sample
sets
of
the
plasma
and
homogenate
standard
series
were
provided
to
11
participating
laboratories
for
VTG
analysis
by
ELISA.
Each
participating
laboratory
was
provided
three
samples
to
be
analyzed
in
triplicate,
packaged
to
limit
the
need
to
freeze­
thaw
during
analysis.
Each
sample
represented
an
individual
fish
equivalent
for
each
of
the
fish
conditions
being
considered
(
i.
e.,
exposed
female,
exposed
male,
unexposed
female,
and
unexposed
male).
Because
the
samples
were
analyzed
fully
blind,
each
vial
contained
enough
Battelle
Draft
May
2003
4
plasma
to
be
analyzed
in
triplicate,
while
representing
the
volume
of
plasma
obtained
from
an
individual
fish.
Messenger
RNA
analysis
was
conducted
on
the
liver
tissue
by
three
participating
laboratories
who
agreed
to
analyze
five
replicate
samples
of
each
sample
type
(
e.
g.,
unexposed
males)
in
triplicate
analysis.
As
for
the
ELISA
samples,
the
integrity
of
the
samples
during
transfer
to
the
participating
laboratories
was
carefully
monitored
and
documented,
and
proper
storage
conditions
were
maintained
prior
to
analysis.

VTG
Purification
For
purification
of
VTG,
adult
male
fish
were
exposed
to
300
ng/
L
of
17$­
estradiol
in
a
7­
day
static
renewal
exposure.
On
day
7,
plasma
was
collected
as
described
for
the
standard
series
and
pooled
for
purification.
The
purified
VTG
was
used
to
create
a
positive
control
sample
within
the
range
of
the
standard
series
and
to
provide
as
a
standard
to
the
analytical
laboratories.
The
VTG
was
purified
from
the
estrogenized
plasma
using
anion
exchange
chromatography
methods
developed
by
Denslow
et
al.
(
1999)
(
Appendix
B).
VTG
was
separated
from
other
plasma
proteins
using
the
BIOCAD
Perfusion
TM
Chromatography
System
and
anion
exchange
media
(
POROS
20HQ).
The
plasma
was
pre­
equilibrated
in
running
buffer
(
20
mM
Bis­
tris
propane,
150
mM
NaCl,
pH
9.0),
loaded
onto
the
column,
rinsed
with
running
buffer
to
elute
non­
binding
proteins,
and
the
VTG
released
using
a
linear
gradient
of
NaCl
(
150­
800
mM).
VTG
was
the
last
protein
to
elute
(
500­
600
mM
NaCl)
the
column.
To
verify
the
identity
of
the
peak,
the
elution
profile
was
compared
to
a
run
using
male
plasma.
After
pooling
the
fractions
containing
VTG,
the
pH
was
adjusted
to
7.0
using
500
mM
Bis­
tris
propane,
the
following
reagents
were
added:
protease
inhibitor
Aprotinin
(
10KIU/
ml),
azide
(
0.02%),
and
cryoprotectant
(
50%
glycerol).
Aliquots
were
stored
at
­
80
º
C
until
needed
and
to
prevent
freeze/
thaw
effects,
once
the
aliquot
was
thawed,
it
was
be
stored
at
­
20
º
C
(
where
it
remains
a
liquid),
with
stability
of
up
to
one
year
(
Kroll
&
Denslow,
unpublished
technique).

To
ensure
that
the
purified
VTG
was
pure
and
of
high
quality,
a
number
of
analyses
were
conducted.
Total
protein
on
the
purified
VTG
was
determined
first
by
Bradford
(
Coomassie
plus
TM,
Pierce)
using
bovine
serum
albumin
as
a
standard
and
then
the
concentration
revalidated
by
amino
acid
analysis.
Purity
was
determined
by
SDS­
PAGE
and
yielded
two
high­
molecularweight
proteins
(
180
&
200
KDa).

Vitellogenin
is
sensitive
to
freeze/
thaw
events
that
can
fracture
the
protein
and
affect
ELISA
results
(
Kroll
&
Denslow,
unpublished
results).
To
control
for
this
variation,
the
purified
vitellogenin
was
frozen
only
once
after
purification.
After
thawing
an
aliquot,
vitellogenin
is
stable
for
1
year
at
 
20
º
C,
and
remains
in
liquid
form
since
it
contains
50%
glycerol.
Stability
of
the
VTG
at
 
20
º
C
has
been
verified
using
positive
controls
and
determining
VTG
concentration
by
ELISA
over
a
period
of
1
year.

4.0
ANALYTICAL
METHODS
OF
THE
PARTICIPATING
LABORATORIES
A
number
of
methods
have
been
developed
for
the
quantification
of
VTG
in
blood
plasma,
liver
tissue,
or
whole­
body
homogenates.
The
various
methods
differ
in
sensitivity,
specificity,
and
technical
difficulty.
Currently,
the
most
popular
approach
to
measure
VTG
is
Battelle
Draft
May
2003
5
some
form
of
an
ELISA.
The
ELISA
employs
enzyme­
linked
antibodies
and
an
adsorbent
surface
to
detect
specific
antigens
in
solution.
The
ELISA
has
been
widely
used
to
quantify
VTG
in
teleosts
due
to
the
ease
in
use
and
unlike
the
radioimmunoassay
(
RIA),
ELISA
does
not
require
the
use
of
radioactive
isotopes.
There
are
a
variety
of
ELISA
designs
that
typically
fall
into
three
general
assay
formats
including
competitive,
sandwich,
and
direct
ELISAs.
Competitive
ELISAs
incorporate
a
step
in
which
the
samples
and
antibody
(
antibody­
capture)
or
labelled
antigen
(
antigen­
capture)
are
incubated
together
prior
to
adding
the
sample
on
the
test
plate.
This
non­
equilibrium
design
is
often
used
to
enhance
sensitivity
and
counteracts
potential
preferential
binding
(
Edmunds
et
al.
2000).
Sandwich
ELISAs
employ
two
antibody
preparations
to
detect
the
antigen.
The
antigens
can
recognize
different
epitopes
on
the
target
analyte,
thereby
providing
a
large
degree
of
specificity
and
sensitivity.
In
a
direct
antibodycapture
ELISA,
the
sample
and
standards
are
adsorbed
directly
on
the
surface
of
the
microwell
plate.
After
incubation,
the
wells
are
blocked
and
anti­
VTG
antibody
is
added
to
bind
to
the
VTG
attached
to
the
well.
As
with
other
ELISAs,
subsequent
steps
culminate
in
the
development
of
color
reflective
of
the
amount
of
antigen
present
in
the
sample.
The
ELISA
protocols
included
in
this
study
are
presented
in
Appendix
A).

Mass
spectrometry
(
MS)
offers
future
possibilities
for
becoming
a
reference
method
for
VTG
and
for
combining
multiple
protein
analysis
from
a
single
tissue
sample.
In
general,
MS
approaches
to
protein
quantification
attempt
to
measure
the
protein
largely
in
its
intact
form
or
rely
on
digestion
procedures
(
chemical
or
enzymatic)
to
reduce
the
size
of
the
protein
into
smaller
fragments.
The
MS
technique
allows
both
the
direct
measurement
of
the
VTG
mass
and
generation
of
peptide­
fingerprinting
data
for
further
identification
(
Wunschel
and
Wahl,
2002).

An
alternative
to
measuring
the
VTG
protein
is
to
quantify
the
messenger
ribonucleic
acid
(
mRNA)
for
VTG
that
codes
for
the
protein.
Two
methods
for
quantifying
fish
VTG
mRNA
have
emerged,
the
ribonuclease
protection
assay
(
RPA)
and
the
quantitative
reverse
transcription­
polymerase
chain
reaction
(
QRT­
PCR),
although
other
methods
exist
(
e.
g.,
Northern
blot,
slot­
blot)
that
have
drawbacks
relative
to
sample
throughput
or
sensitivity.
All
methods
can
be
used
for
absolute
or
relative
quantification
of
mRNA.

Specific
protocols
employed
by
the
participating
laboratories
(
Appendix
A)
were
applied
to
the
samples
in
this
study.
The
analysis
can
be
grouped
into
the
general
categories
of
VTG
and
mRNA
as
the
target
analyte.
Within
these
categories,
multiple
methods
were
applied
to
the
sample
series.
The
participating
labs
received
multiple
aliquots
of
the
standard
series
as
a
contingency
to
prevent
the
need
for
sample
freeze­
thaw
cycles
and
to
limit
the
number
of
shipments
to
each
laboratory.
Each
laboratory
was
asked
to
analyze
the
samples
three
times
within
approximately
4
weeks,
with
a
minimum
of
three
replicates.

5.0
PARTICIPATING
LABORATORIES
The
laboratories
participating
in
this
study
were
selected
based
upon
their
previous
experience
in
the
measurement
of
the
VTG
protein
or
mRNA.
The
laboratories
had
established
protocols
in
routine
use
and
were
willing
to
commit
to
analysis
during
the
study
period.
Two
laboratories
were
provided
with
sample
sets
but
were
unable
to
complete
the
analysis
within
the
Battelle
Draft
May
2003
6
timeline
of
the
study.
The
laboratories
that
analyzed
the
samples
provided
are
presented
in
Table
1.
The
laboratories
that
conduct
ELISA
VTG
measurements
analyzed
the
plasma
and
whole
body
homogenate
standard
series.
The
laboratories
that
measured
mRNA
analyzed
the
liver
tissue
standard
series.
Two
of
these
laboratories,
Oregon
State
University
and
the
Finnish
Environmental
Institute
applied
their
ELISA
assay
to
the
samples,
however,
their
antibodies
(
carp
monoclonal
and
trout)
did
not
interact
with
the
samples.
The
results
from
8
ELISA
laboratories
and
3
mRNA
laboratories
are
summarized
in
this
report.

6.0
DATA
ANALYSIS
Data
analysis
was
intended
to
provide
descriptive
statistics
and
plots
that
allow
a
general
assessment
of
the
objectives
of
the
study.
Statistically,
the
first
objective
was
to
determine
if
an
increasing
concentration
of
VTG
was
produced
by
the
standard
series.
This
series
was
represented,
in
order,
by
1)
uninduced
male,
2)
uninduced
female,
3)
induced
male,
and
4)
induced
female
fathead
minnows.
The
second
statistical
objective
was
to
determine
the
analytical
results
and
variation
for
the
set
of
control
and
spiked
VTG
samples.
The
third
statistical
objective
was
to
compare
the
analytical
results
and
variation
of
each
lab's
analytical
method
including
the
antibody,
standard,
and
assay
used.

Analysis
of
the
data
yielded
descriptive
statistics
including
the
number
of
samples,
means,
standard
deviations,
medians,
first
and
third
quartiles,
and
the
coefficient
of
variation
(
CV).
Simple
linear
regression
of
the
ranked
average
VTG
concentration
(
mean
of
the
with­
in
run
analyses)
and
plots
of
the
analytical
results
against
the
concentration
series
were
used
to
assess
the
strength
of
the
VTG
concentration
trend
(
ignoring
the
positive
control).
Tukey's
Honestly
Significant
Difference
(
HSD)
multiple
comparison
test
was
conducted
on
the
ranked
average
VTG
concentrations
to
specifically
determine
if
neighboring
means
in
the
series
were
significantly
different
(
i.
e.,
the
blank
mean
compared
to
the
uninduced
male
mean,
the
uninduced
male
mean
compared
to
the
uninduced
female
mean,
and
so
on).
Linear
regression
for
each
laboratory
was
also
conducted
on
the
average
VTG
concentrations
observed
for
the
blank
and
the
uninduced
male
data.
The
regression
results
allow
a
test
of
the
null
hypothesis
that
the
slope
equals
0
and
provides
an
measure
of
the
strength
of
the
trend.
The
multiple
comparison
testing
(
which
is
less
powerful
than
the
regression
analysis
due
to
the
smaller
degrees
of
freedom
for
testing)
provides
a
test
of
how
quickly
differences
can
be
detected
in
the
series.
Excel
spreadsheet
software
(
Microsoft
Excel)
and
Minitab
statistical
software
(
Minitab
Inc.)
were
used
for
this
analysis.
Battelle
Draft
May
2003
7
Table
1.
Summary
of
Reporting
Laboratories
Lab
ID
#
Participating
Laboratory
Method(
s)
applied
to
the
standard
series
1
The
University
of
Florida
Protein
Chemistry
and
Biomarkers
Res.
Lab.
Gainesville,
Florida,
USA
mRNA
­
RT­
PCR
ELISA
­
Fathead
minnow
based,
monoclonal
antibody,
direct
ELISA
Biosense
ELISA
kit
2
The
University
of
Idaho
Department
of
Biological
Sciences
Moscow,
Idaho,
USA
mRNA
­
qRT­
PCR
TaqMan
15
Molecular
Light
Technology
Research
Ltd.
Cardiff,
UK
mRNA
­
HPA
(
hybridization
protection
assay)

14
Battelle
Pacific
Northwest
National
Laboratory
Richland,
WA,
USA
GC­
MS
3
Oregon
State
University
Environmental
and
Molecular
Toxicology
Corvallis,
Oregon,
USA
ELISA
­
Trout
based
polyclonal
antibodies
in
a
competitive
ELISA
4
US
EPA
Duluth,
Michigan,
USA
ELISA
­
Fathead
minnow
based
polyclonal
antibodies,
competitive,
antibody­
capture
56
University
of
Exeter
Environmental
and
Molecular
Fish
Biology
Exeter,
United
Kingdom
Brixham
Environmental
Laboratory
AstraZeneca
,
United
Kingdom
ELISA
­
Carp
based
polyclonal
antibodies,
competitive
ELISA
713
Biosense
Laboratories
Bergen,
Norway
Battelle
Pacific
Northwest
National
Laboratory
Sequin,
WA,
USA
ELISA
­
Carp
based
polyclonal
and
monoclonal
antibodies,
sandwich
ELISA
8
INERIS
(
National
Institute
of
Industrial
Environment
and
Risks
­
France)
Verneuil
en
Halatte,
France
ELISA
­
Zebrafish
based
polyclonal
antibodies,
competitive
ELISA
11
University
of
Southern
Denmark
Institute
of
Biology
Odense,
Denmark
ELISA
­
Zebrafish
based
anti­
lipovitellin
direct
non­
competitive
sandwich
ELISA
12
The
Finnish
Environmental
Institute
Helsinki,
Finland
Carp
based
monoclonal
antibody
indirect
ELISA
Battelle
Draft
May
2003
8
7.0
RESULTS
As
noted
in
Section
3.0,
Sample
Methods,
sample
sets
of
the
plasma
and
homogenate
standard
series
were
provided
to
11
participating
laboratories
for
VTG
analysis
by
ELISA.
It
should
be
noted
that
all
laboratories
provided
their
services
without
compensation,
and
that
every
attempt
was
made
to
assist
the
laboratories
in
performing
this
complex
task.
Three
samples
were
provided
to
each
lab
to
be
analyzed
in
triplicate,
each
vial
containing
enough
plasma
to
represent
an
individual
fish.
One
of
the
participating
laboratories
performed
the
analysis
of
the
standard
series
once,
rather
than
in
triplicate,
and
the
results
reflect
this
singular
value
accordingly.

This
study
included
several
techniques
for
the
detection
of
the
induction
of
VTG
in
fathead
minnows.
These
techniques
include
ELISA
(
enzyme­
linked
immunosorbent
assay),
RT­
PCR
(
reverse
transcription­
polymerase
chain
reaction)
and
GC­
MS
(
gas
chromatography
­
mass
spectrometry).
The
advantages
and
disadvantages
of
these
techniques
include
those
of
sensitivity,
reproducibility,
and
cost.
Two
of
the
techniques,
ELISA
and
GC­
MS,
measured
levels
of
the
protein
VTG,
and
RT­
PCR
measured
the
up­
regulation
of
messenger
RNA.
The
advantage
of
measuring
mRNA
compared
with
measuring
the
expression
of
the
protein
VTG
include
very
fast
response
upon
exposure
and
detection
that
can
be
very
sensitive.
However,
the
increased
levels
of
mRNA
are
less
persistent
after
exposure,
and
the
technique
requires
expensive
specialized
equipment.
Alternatively,
ELISA
and
GC­
MS
measure
the
VTG
protein
that
persists
longer
post­
exposure
compared
with
mRNA.
The
immunologically
based
ELISA
relies
on
antigen­
antibody
interactions,
with
associated
antibody
specificity
questions
for
quantification.
GC­
MS
does
not
have
the
problems
of
specificity
associated
with
immunoassay
and
offers
the
potential
to
measure
multiple
proteins
in
a
single
sample.
Although
GC­
MS
requires
very
expensive,
advanced
equipment,
the
technique
can
provide
critical
performance
evaluations
by
providing
a
secondary
means
to
measure
the
level
of
protein
in
standards
and
other
QC
criteria
required
for
a
screening
assay.

A
comprehensive
survey
of
the
literature
and
experts
in
the
field
of
induction
of
VTG
in
fish
(
Battelle
2002)
revealed
that
the
technique
of
ELISA
is
currently
the
most
widely
developed
and
applied
technique,
with
multiple
methods
that
can
be
applied
to
fathead
minnows.
The
ELISA
technique
is
represented
in
this
study
by
assays
developed
in
several
species
of
fish
that
can
be
applied
to
fathead
minnows.
These
various
assays
rely
upon
different
types
of
antibodies
(
e.
g.
monoclonal
and
polyclonal)
and
are
performed
with
multiple
approaches
(
e.
g.
competitive,
sandwich
ELISA).
A
summary
of
the
ELISA
methods
by
general
type
that
were
employed
by
the
laboratories
participating
in
this
study
are
presented
in
Table
2,
with
the
specifics
of
the
ELISA
presented
in
Table
1.
It
should
be
noted
that
one
laboratory
applied
two
ELISA
methods
to
its
set
of
samples,
and
that
this
allowed
the
comparison
of
three
laboratories
employing
a
commercially
available
kit.
Battelle
Draft
May
2003
9
Table
2.
Summary
of
Antibody
Methods
Used
by
Participating
Laboratories
Lab
ID
Lab
Name
ELISA
Method
Antibody
1
University
of
Florida
Fathead
Minnow
1
University
of
Florida
Carp­
sandwich
4
US
EPA
Duluth
Fathead
Minnow
5
University
of
Exeter
Carp­
competitive
6
Brixham
Environmental
Carp­
competitive
7
Biosense
Carp­
sandwich
8
INERIS
Zebrafish
11
University
of
Southern
Denmark
Zebrafish
13
Battelle
Sequin
Carp­
sandwich
The
more
specialized
technique
of
RT­
PCR
to
measure
fathead
minnow
mRNA
is
conducted
in
a
very
limited
number
of
laboratories.
However,
three
laboratories
agreed
to
participate
in
this
study,
thereby
allowing
an
assessment
of
the
variability
of
this
method.
These
laboratories
included
Lab1;
University
of
Florida,
Lab
2;
University
of
Idaho,
Lab
15;
MLT
Research.
The
technique
of
GC­
MS
requires
highly
advanced,
expensive
equipment
and
the
new
application
of
this
method
to
the
measurement
of
fathead
minnow
VTG
was
assessed
by
one
participating
laboratory
(
Lab
14;
Battelle
Richland)
in
this
study.

The
ELISA
methods
that
will
be
applied
to
the
plasma
and
whole
body
homogenate
standard
series
are
immunologically
based.
ELISA
relies
on
antigen­
antibody
interactions
and
the
associated
antibody
specificity
must
be
controlled
for
quantification.
A
variety
of
antibodies
are
used
by
the
various
methods.
In
addition
to
the
use
of
the
standard
homologous
to
each
method,
the
participating
laboratories
were
supplied
with
purified
fathead
minnow
VTG
for
use
as
a
standard.
As
a
result,
six
of
the
laboratories
used
the
VTG
purified
in
this
study
to
analyze
the
samples
in
addition
to
their
homologous
standard
(
Table
3).
In
addition
to
the
samples
within
the
series
from
exposed
and
unexposed
fish,
each
laboratory
received
a
sample
spiked
with
purified
fathead
minnow
VTG.
A
summary
of
the
concentration,
standards,
antibody,
and
assay
codes
are
presented
in
Table
4.
The
concentration
codes
are
based
upon
the
exposure
history
and
sex
of
the
fish
used
to
generate
the
samples
in
the
series.
The
standard
codes
identify
the
results
based
on
the
use
of
the
standard
routinely
employed
by
eh
individual
laboratory,
or
based
upon
the
fathead
minnow
standard
provided
to
each
of
the
participating
laboratories
as
a
part
of
this
study.
The
antibody
codes
group
the
assays
into
4
general
categories
of
antibody
type
used
in
the
various
assays.
The
assay
code
defines
the
use
of
a
commercial
kit
(
Carp
sandwich
ELISA)
vs
the
assay
unique
to
the
individual
laboratories.
These
groupings
were
used
to
analyze
the
variability
of
the
reported
results.
Battelle
Draft
May
2003
10
Table
3.
Summary
of
Standards
Employed
by
Participating
Laboratories
Lab
#
Lab
Name
Homologous
Std
Purified
fathead
minnow
Std
1
University
of
Florida
X
X
4
US
EPA
Duluth
X
X
5
University
of
Exeter
X
­
6
Brixham
Environmental
X
­
7
Biosense
X
X
8
INERIS
X
X
11
University
of
Southern
Denmark
X
X
13
Battelle
X
X
Table
4.
Summary
of
the
Concentration,
Standards,
Antibody,
and
Assay
Codes
Concentration
Conc.
Code
STD
STD
Code
Antibody
Antibody
Code
Assay
Assay
Code
Blank
0
homologous
1
Carp
sandwich
1
kit
1
Unexposed
Male
1
Purified
fathead
minnow
2
Fathead
Minnow
2
Unique
0
Unexposed
Female
2
­­
­­
Carp
competitive
3
­­
­­

Exposed
Male
3
­­
­­
Zebrafish
4
­­
­­
Exposed
Female
4
­­
­­
­­
­­
­­
­­
Positive
Control
5
­­
­­
­­
­­
­­
­

To
assess
the
overall
variability
of
analysis
by
all
of
the
various
methods,
the
reported
ELISA
results
were
analyzed
irrespective
of
method
or
the
standard
employed.
This
reflects
the
variability
encountered
when
a
number
of
methods
for
use
in
the
measurement
of
VTG
in
fathead
minnow
samples
are
applied
to
standard
samples
spanning
a
wide
range
of
concentrations.
When
conducting
the
analysis,
each
of
the
participating
laboratories
evaluated
three
analytical
replicates
for
each
of
three
sample
replicates
(
Appendix
C).
The
three
analytical
replicates
are
a
measure
of
the
within­
run
variability.
When
all
of
the
reported
results
were
used,
the
within­
run
variability
for
plasma
had
a
wide
range
of
coefficients
of
variation
(
CVs),
ranging
from
0%
to
173%
with
a
mean
of
13%
(
Table
5;
Appendix
D).
The
within­
run
variability
for
homogenate
was
similar
with
CVs
ranging
from
0%
to
162%.
Indeed,
most
of
the
large
CV's
were
associated
with
samples
that
had
one
or
two
observations
less
than
detected.
If
all
undetected
values
are
ignored
the
maximum
CV
for
plasma
and
homogenate
become
83%
and
141%,
respectively.
The
quartiles
of
the
CV
distribution
remain
the
same
with
or
without
the
Battelle
Draft
May
2003
11
less
than
detected
values.
For
both
sample
types,
75%
of
the
within­
run
CVs
were
less
than
16%.
The
CVs
for
both
the
plasma
and
homogenate
samples
tended
to
be
less
than
30%
for
all
concentrations
except
for
those
derived
from
the
blank
and
unexposed
males
(
Figures
1
and
2).
With
this
assessment
of
the
within­
run
variability,
all
the
remaining
analyses
were
conducted
on
the
mean
result
of
the
analytical
replicates.

Table
5.
Descriptive
Statistics
of
the
CVs
of
the
Within­
run
Analytical
Results
where
Q1
and
Q3
are
the
1st
and
3rd
Quartiles,
Respectively
Series
Concentration
Mean
Minimum
Maximum
Q1
Q3
N
Plasma
Blank
18%
0%
142%
0%
26%
40
Plasma
Uninduced
Male
23%
0%
173%
0%
23%
45
Plasma
Uninduced
Female
10%
2%
83%
2%
9%
39
Plasma
Induced
Male
9%
0%
74%
3%
10%
40
Plasma
Induced
Female
8%
2%
20%
4%
14%
36
Plasma
Positive
Control
9%
1%
24%
4%
11%
42
Plasma
All
13%
0%
173%
3%
13%
242
Homogenate
Uninduced
Male
26%
0%
162%
0%
39%
45
Homogenate
Uninduced
Female
21%
0%
94%
1%
20%
44
Homogenate
Induced
Male
19%
0%
102%
3%
16%
40
Homogenate
Induced
Female
7%
1%
22%
3%
11%
40
Homogenate
Positive
Control
15%
0%
141%
2%
9%
42
Homogenate
All
18%
0%
162%
2%
15%
211
Battelle
Draft
May
2003
12
Plasma
0
5
10
15
20
25
0%
20%
40%
60%
80%
100%
120%
140%
160%

CV%
of
Within­
run
Analyses
Number
of
Observations
0
1
2
3
4
5
Figure
1.
Distribution
of
the
Within­
run
Coefficient
of
Variation
(
CV)
of
the
VTG
Analytical
Result
on
Plasma
Samples
for
Each
Concentration
Code
(
0
­
5)
Code
0
=
Blank;
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
13
Homogenate
0
5
10
15
20
25
30
0%
20%
40%
60%
80%
100%
120%
140%
160%

CV%
of
Within­
run
Analyses
Number
of
Observations
1
2
3
4
5
Figure
2.
Distribution
of
the
Within­
run
Coefficient
of
Variation
(
CV)
of
the
VTG
Analyt­
ical
Result
on
Homogenate
Samples
for
Each
Concentration
Code
(
1
­
5)
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
In
addition
to
assessing
the
within
run
variability
from
the
analytical
replicates,
the
three
sample
replicates
provided
a
measure
of
intra­
assay
variability.
The
intra­
assay
variability
for
plasma
and
homogenate
had
CVs
ranging
from
0%
to
173%
(
Table
6;
Appendix
E).
For
both
sample
types
75%
of
the
intra­
assay
CVs
were
less
than
51%.
Again,
many
of
the
large
CVS
were
due
to
means
calculated
with
one
or
more
values
less
than
the
detection
limit.
If
all
such
means
are
ignored,
the
maximum
CV
for
plasma
and
homogenate
become
124%
and
125%,
respectively.
Battelle
Draft
May
2003
14
Table
6.
Descriptive
Statistics
of
the
CVs
of
the
Intra­
Assay
Analytical
Results
where
Q1
and
Q3
are
the
1st
and
3rd
Quartiles
Respectively
Series
Concentration
Mean
Minimum
Maximum
Q1
Q3
N
Plasma
Blank
27%
0%
99%
0%
54%
14
Plasma
Uninduced
Male
43%
0%
173%
12%
63%
14
Plasma
Uninduced
Female
36%
2%
111%
12%
57%
14
Plasma
Induced
Male
21%
1%
105%
7%
21%
14
Plasma
Induced
Female
37%
2%
124%
12%
40%
12
Plasma
Positive
Control
27%
6%
92%
12%
36%
14
Plasma
All
32%
0%
173%
9%
44%
82
Homogenate
Uninduced
Male
46%
0%
173%
6%
63%
16
Homogenate
Uninduced
Female
29%
0%
72%
11%
53%
16
Homogenate
Induced
Male
45%
3%
165%
10%
42%
14
Homogenate
Induced
Female
30%
1%
145%
4%
47%
14
Homogenate
Positive
Control
21%
0%
87%
4%
29%
14
Homogenate
All
34%
0%
173%
7%
51%
74
One
goal
of
this
study
was
to
create
a
series
of
analytical
samples
that
included
a
wide
range
of
VTG
concentrations
in
male
and
female
plasma
and
whole
body
homogenates.
It
was
anticipated
that
male
fish
unexposed
to
estrogenic
compounds
would
provide
minimal
levels
of
VTG
in
plasma
and
tissues,
with
unexposed
female
fish,
exposed
male,
and
exposed
female
fish
generating
increasing
levels
of
VTG
in
their
respective
systems.
Inclusive
of
all
reported
results,
the
general
trend
for
the
plasma
samples
observed
for
each
laboratory
averaged
over
antibodies,
standards,
and
assays
was
the
expected
increase
based
on
the
series
(
i.
e.,
uninduced
male
<
uninduced
female
<
induced
male
<
induced
female
fathead
minnows;
Figure
3).
However,
for
several
of
the
average
laboratory
results,
the
uninduced
male
(
Code
1)
results
were
equal
or
only
slightly
greater
than
the
prepared
blank
(
Code
0).
For
half
of
the
average
laboratory
homogenate
results
the
uninduced
male
(
Code
1)
VTG
concentrations
were
approximately
equal
to
or
greater
than
that
of
uninduced
female
results
(
Code
2;
Figure
4).
All
but
Laboratory
13
showed
an
increase
in
average
VTG
concentration
between
uninduced
female
results
and
induced
males
(
Code
3).
Further,
two
of
the
average
laboratory
results
(
laboratories
5
and
6)
showed
a
decrease
in
average
VTG
concentration
between
the
induced
male
samples
and
the
induced
female
samples
(
Code
4).
Descriptive
Statistics
for
the
Mean
VTG
Results
Averaged
Over
Laboratory,
Antibody,
Standard,
and
Assay
are
provided
Table
7.
Battelle
Draft
May
2003
15
Plasma
0.00
0.01
0.10
1.00
10.00
100.00
1000.00
10000.00
100000.00
1000000.00
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL
Lab
1
Lab
4
Lab
5
Lab
6
Lab
7
Lab
11
Lab
13
Max
detection
Limit
Figure
3.
Each
Laboratories
VTG
Concentration
for
A
Given
Concentration
Code
Averaged
Over
Antibody,
Standard,
and
Assay
for
Plasma
Samples
Code
0
=
Blank;
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
16
Homogenate
0.00
0.01
0.10
1.00
10.00
100.00
1000.00
10000.00
100000.00
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL)

Lab
1
Lab
4
Lab
5
Lab
6
Lab
7
Lab
8
Lab
11
Lab
13
Max
detection
Limit
Figure
4.
Each
Laboratories
VTG
Concentration
for
A
Given
Concentration
Code
Averaged
Over
Antibody,
Standard,
and
Assay
for
Homogenate
Samples
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
17
Table
7.
Descriptive
Statistics
for
the
Mean
VTG
Results
Averaged
Over
Laboratory,
Antibody,
Standard,
and
Assay
Series
Concentration
Code
N
Average
Standard
Deviation
CV%

Plasma
0
43
0.4
0.7
161%

1
48
6
38.3
635%

2
41
10560
19574
185%

3
43
43847
37612
86%

4
38
74071
60909
82%

5
45
1258
2113
168%

Homogenate
0
0
­
­
­
1
48
7.2
42.6
588%

2
47
2
3.1
160%

3
43
1452
1588
109%

4
43
4881
4492
92%

5
45
251
237
95%

A
number
of
the
participating
laboratories
were
able
to
perform
their
analysis
using
both
their
homologous
standard
(
i.
e.
unique
carp,
zebrafish
or
fathead
minnow)
and
the
fathead
minnow
standard
purified
for
this
study.
This
offers
the
comparison
of
method
specificity
and
standardization
in
a
screening
assay.
To
examine
the
results
based
upon
the
standard
used
within
the
assay,
the
percentage
difference
between
the
average
replicate
VTG
concentrations
obtained
with
the
homologous
standard
(
H)
and
Purified
fathead
minnow
standard
(
B)
was
calculated
as
(
H­
B)/
B
(
100%).
Thus,
negative
values
represent
greater
VTG
concentrations
obtained
with
the
Purified
fathead
minnow
standard.
The
results
of
this
analysis
are
presented
in
Figure
5
(
Appendix
E)
by
analytical
laboratory.
Note,
for
small
concentrations,
small
absolute
differences
between
the
results
of
each
standard
may
still
be
a
large
proportion
of
the
Purified
fathead
minnow
standard
result.
The
intent
in
highlighting
these
small
differences
is
due
to
the
need
for
great
precision
at
small
doses.

When
the
type
of
assay
is
examined,
the
Carp
sandwich
ELISA
followed
by
the
zebra
fish
assay
produced
the
greatest
variation
between
standard
results
for
plasma
(­
82%
and
­
75%
respectively),
and
the
fathead
minnow
followed
by
the
zebra
fish
antibody
based
assays
produced
the
greatest
variation
between
standards
for
the
homogenate
samples
(
200%
and
99%
respectively;
Figure
5).
The
large
difference
observed
with
the
homogenate
results
is
due
to
averaging
small
numbers
associated
with
three
less
than
detected
values
in
the
homologous
standard
and
seven
less
than
detected
values
with
the
Purified
fathead
minnow
standard.
In
contrast
to
the
plasma
samples,
nearly
all
of
the
fathead
minnow
antibody
homogenate
data
had
greater
VTG
concentrations
with
the
homologous
standard.
Battelle
Draft
May
2003
18
Plasma
­
100%
­
80%
­
60%
­
40%
­
20%
0%
20%
40%
60%

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Lab
Id
%
Difference
Between
Standards
Carp­
sandwich
Fathead
Minnow
Zebra
Fish
Homogenate
­
100%
­
50%
0%
50%
100%
150%
200%
250%

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Lab
Id
%
Difference
Between
Standards
Carp­
sandwich
Fathead
Minnow
Zebra
Fish
Figure
5.
The
Percentage
Difference
in
VTG
Concentrations
Averaged
Over
Replicates
Between
the
Homologous
and
Purified
fathead
minnow
Standard
Data
in
Plasma
and
Homogenate
Samples
Battelle
Draft
May
2003
19
In
the
remaining
types
of
assays
(
zebrafish
and
the
carp­
based
sandwich),
the
concentrations
based
upon
the
purified
fathead
minnow
standard
were
equal
to
or
lower
than
the
homologous
standard.
This
reflects
a
reduced
specificity
of
the
zebrafish
and
carp­
based
antibodies
to
the
fathead
minnow
vitellogenin,
resulting
in
lower
values
reported
with
the
homologous
standard.
Because
of
the
large
variation
in
the
results
between
the
two
standards,
data
based
on
each
standard
was
evaluated
separately
for
all
remaining
analyses.

When
the
concentration
of
VTG
in
plasma
and
homogenate
samples
is
examined
with
the
assays
performed
in
the
manner
routine
to
the
participating
laboratories
(
i.
e.
homologous
standard),
the
carp
antibody
based
competitive
ELISAs
tended
to
produce
the
lowest
VTG
concentrations
in
the
plasma
standard
samples
(
Figure
6).
Further,
these
ELISAs
did
not
distinguish
between
the
exposed
male
and
unexposed
female
samples
(
concentration
codes
2
and
3
respectively).
It
should
be
noted
that
the
carp
antibody
competitive
ELISA
were
not
reported
with
the
Purified
fathead
minnow
standard.
The
carp
antibody
based
competitive
ELISAs
with
homogenate
samples
still
produced
low
VTG
concentrations,
but
not
the
lowest
reported
values
(
Figure
7).
There
was
slightly
less
variation
in
VTG
concentrations
with
the
Purified
fathead
minnow
standard
for
both
sample
types.
For
the
data
produced
using
the
purified
fathead
minnow
standard,
the
carp
sandwich
ELISA
and
the
fathead
minnow
antibody
tended
to
produce
greater
variation
between
laboratories
for
the
blank
and
uninduced
male
plasma
samples
(
Figure
6).
The
fathead
minnow
antibody
produce
greater
variation
between
laboratories
for
the
uninduced
male
and
female
samples
for
homogenate
samples
(
Figure
7).
Battelle
Draft
May
2003
20
Plasma
Samples
with
Battelle
Std
0.0001
0.01
1
100
10000
1000000
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL)

Carp­
sandwich­
1
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Zebra
Fish­
11
Maximum
Detection
Limit
Plasma
Samples
with
Homologous
Std
0.0001
0.01
1
100
10000
1000000
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL)

Carp­
sandwich­
1
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Carp­
5
Carp­
6
Zebra
Fish­
11
Maximum
Detection
Limit
Figure
6.
The
VTG
Concentration
in
Plasma
Samples
for
Each
Standard,
Laboratory,
Antibody,
and
Concentration
Code
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
21
Homogenate
Samples
with
Homologous
Std
0.001
0.01
0.1
1
10
100
1000
10000
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL)

Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Carp­
5
Carp­
6
Zebra
Fish­
8
Zebra
Fish­
11
Maximum
Detection
Limit
Homogenate
Samples
with
Battelle
Std
0.001
0.01
0.1
1
10
100
1000
10000
100000
0
1
2
3
4
5
6
Concentration
Code
VTG
Concentration
(
ug/
mL)

Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Zebra
Fish­
8
Zebra
Fish­
11
Maximum
Detection
Limit
Figure
7.
The
VTG
Concentration
in
Homogenate
Samples
for
Each
Standard,
Laboratory,
Antibody,
and
Concentration
Code
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
22
Simple
linear
regression
analysis
of
the
ranked
VTG
concentration
within
sample
type,
laboratory,
standard,
and
antibody
confirmed
significant
positive
slopes
(
p
<=
0.003)
across
the
concentration
codes
0
­
4
(
blank
through
induced
female
series)
for
all
combinations
except
for
Laboratory
5
using
the
homogenate
samples,
homologous
standard,
and
carp
antibody
data
(
p
=
0.128,
n
=
12;
Table
9).
The
lack
of
significance
for
this
set
of
observations
was
due
to
a
greater
average
concentration
of
VTG
observed
in
the
uninduced
males
than
in
the
uninduced
females
(
Figure
7).
All
regressions
had
a
minimum
of
8
degrees
of
freedom
for
error.
The
positive
control
(
concentration
code
5)
was
not
used
for
this
analysis.
With
the
one
exception
for
homogenate
samples,
this
demonstrates
the
successful
measurement
of
the
increasing
concentrations
of
VTG
in
the
samples
among
multiple
assays,
sample
type,
standard,
antibody
and
laboratory.

Tuckey's
HSD
was
used
to
compare
the
mean
ranked
concentrations
between
successive
pairs
in
the
series
using
laboratories
as
replicates
(
Table
10).
For
this
comparison,
observations
were
ranked
within
sample
type,
standard,
and
antibody.
There
were
only
2
to
3
replicates
for
each
concentration
and
4
to
9
degrees
of
freedom
for
error.
None
of
the
successive
concentrations
were
found
to
be
significantly
different
for
any
of
the
sample
type,
standard,
and
antibody
combinations
(
family­
wise
error
rate
=
0.05).
However,
most
comparisons
separated
by
two
concentrations
were
significantly
different
for
plasma
samples
used
with
the
Carp
sandwich
ELISA
antibody.
The
remainder
of
the
comparisons
had
too
little
replication
and
too
large
of
a
variation
within
classes
to
achieve
significance
other
than
between
the
lowest
and
highest
concentrations.
Homogenate
samples
analyzed
with
homologous
standards
and
fathead
minnow
and
carp
antibodies
had
no
detected
significant
differences
between
any
of
the
concentrations.
Therefore,
even
though
the
regressions
across
the
series
from
blank
to
induced
female
were
significant,
the
variability
between
laboratories
to
reproduce
the
ranks
for
a
given
concentration
code
was
too
large
to
detect
differences
between
successive
concentration
pairs
in
the
series.

When
the
variability
of
results
by
laboratory
and
standard
were
examined,
the
C'S
associated
with
the
plasma
zebrafish
antibody
results
were
lower
with
the
Purified
fathead
minnow
standard
than
with
the
homologous
standard
(
Figure
8).
The
Carp
sandwich
ELISA
CVs
tended
to
be
equal
or
greater
with
the
Purified
fathead
minnow
standard
than
with
the
homologous
standard.
This
trend
was
repeated
in
the
homogenate
samples
(
Figure
9).
The
fathead
minnow
CVs
tended
to
be
greater
in
the
homogenate
samples
than
in
the
plasma
samples
for
both
the
homologous
and
Purified
fathead
minnow
standard
results.
Greater
than
50%
of
the
CVs
obtained
with
the
carp
antibody
were
greater
than
30%.

Three
laboratories
applied
the
same
commercial
carp
sandwich
ELISA
to
the
sample
series,
including
the
low
level
samples
and
positive
control.
This
allows
for
some
measure
of
the
application
of
one
of
the
methods
in
this
study
by
multiple
laboratories;
however,
similar
assessments
of
the
other
methods
in
this
study
would
be
required
to
draw
general
conclusions
based
on
this
very
limited
data
set.
For
those
laboratories
that
used
this
method
in
this
study,
a
simple
linear
regression
of
the
average
VTG
concentration
(
within­
run
average)
between
the
blank
and
the
uninduced
male
concentrations
was
conducted
to
determine
whether
a
significant
slope
was
obtained.
For
the
plasma
samples
(
Laboratories
1,
7,
and
13),
only
Laboratory
13
produced
a
significant
slope
between
these
two
concentrations
using
both
standards
(
p
=
0.006,
4
Battelle
Draft
May
2003
23
d.
f.
for
error).
For
the
homogenate
samples
(
Laboratories
7
and
13),
both
Laboratories
7
and
13
produced
significant
slopes
using
the
homologous
standard
(
p
<
0.05,
4
d.
f.
for
error),
and
Laboratory
7
produced
a
significant
slope
using
the
purified
fathead
minnow
standard
(
p
=
0.005,
4
d.
f.
for
error).
The
means
and
standard
deviations
for
each
of
the
concentrations
is
presented
in
Table
11.
Although
these
results
indicate
that
the
detection
of
background
or
low
levels
of
VTG
in
fathead
minnow
samples
can
be
achieved
with
the
application
of
this
ELISA,
the
lack
of
consistent
concentrations
and
significant
differences
between
samples
among
laboratories
requires
additional
study
for
method
assessment.

The
positive
control
samples
for
both
the
plasma
and
homogenate
were
spiked
to
500
:
g/
ml
VTG
with
the
fathead
minnow
VTG
purified
for
this
study.
With
three
laboratories
conducting
the
carp
sandwich
ELISA,
these
results
can
be
examined
for
statistical
significance.
The
positive
control
concentrations
was
not
achieved
with
the
carp
sandwich
ELISA
using
laboratories
as
replicates
(
Table
12).
The
results
of
the
positive
control
were
again
highly
variable
both
within
and
between
laboratories.
The
concentrations
detected
in
plasma
ranged
from
337
to
8834
:
g/
ml
VTG
in
plasma
and
from
235
to
745
:
g/
ml
VTG
in
the
homogenate
control
sample.
In
addition
to
the
variation
found
in
the
Carp
sandwich
ELISA
method,
a
similar
wide
range
of
concentrations
for
the
control
samples
were
detected
among
all
of
the
assays
(
Table
8;
Figure
6
and
7)
Battelle
Draft
May
2003
24
Table
8.
Results
of
the
Linear
Regression
on
Within­
Laboratory,
Sample
Type,
Standard,
and
Antibody
Ranked
VTG
Concentrations
Observed
in
the
Series
Codes
0
Through
4
Series
Standard
Antibody
Lab
Slope
Std
Error
d.
f.
p­
value
Plasma
1
1
1
2.629
0.2175
8
<
0.001
Plasma
1
1
7
2.9333
0.2368
13
<
0.001
Plasma
1
1
13
2.7
0.3965
13
<
0.001
Plasma
1
2
1
3.4
0.2736
12
<
0.001
Plasma
1
2
4
2.6099
0.2962
12
<
0.001
Plasma
1
3
5
2.7944
0.5344
12
<
0.001
Plasma
1
3
6
2.485
0.3272
11
<
0.001
Plasma
1
4
11
4.7
0.6595
27
<
0.001
Plasma
2
1
1
2.629
0.2175
8
<
0.001
Plasma
2
1
7
2.9333
0.2368
13
<
0.001
Plasma
2
1
13
2.4
0.5243
13
0.001
Plasma
2
2
1
3.4
0.2736
12
<
0.001
Plasma
2
2
4
2.6074
0.3509
12
<
0.001
Plasma
2
4
11
4.0104
0.3722
19
<
0.001
Homogenate
1
1
7
3.2333
0.2993
10
<
0.001
Homogenate
1
1
13
2.5333
0.5582
10
0.001
Homogenate
1
2
1
3.5
0.6455
10
<
0.001
Homogenate
1
2
4
2.6667
0.4922
10
<
0.001
Homogenate
1
3
5
1.5667
0.9432
10
0.128
Homogenate
1
3
6
2.8
0.4115
10
<
0.001
Homogenate
1
4
8
2.4
0.6143
10
0.003
Homogenate
1
4
11
4.7521
0.5216
18
<
0.001
Homogenate
2
1
7
3.2333
0.2993
10
<
0.001
Homogenate
2
1
13
2.5333
0.5582
10
0.001
Homogenate
2
2
1
3.5
0.6455
10
<
0.001
Homogenate
2
2
4
2.8667
0.3627
10
<
0.001
Homogenate
2
4
8
2.4
0.6143
10
0.003
Homogenate
2
4
11
4.0057
0.5089
15
<
0.001
Battelle
Draft
May
2003
25
Table
9.
Results
of
Tukey's
HSD
Multiple
Comparison
Test
On
the
Ranked
VTG
Concentrations
Using
Laboratories
as
Replicates
Series
Standard
Antibody
0
to
1
0
to
2
0
to
3
0
to
4
1
to
2
1
to
3
1
to
4
2
to
3
2
to
4
3
to
4
Error
d.
f.
Plasma
1
1
NS
*
*
*
NS
*
*
NS
NS
NS
9
Plasma
1
2
NS
NS
NS
*
NS
*
*
NS
NS
NS
5
Plasma
1
3
NS
NS
NS
*
NS
NS
NS
NS
NS
NS
5
Plasma
1
4
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
Plasma
2
1
NS
*
*
*
NS
*
*
NS
NS
NS
9
Plasma
2
2
NS
NS
NS
*
NS
NS
*
NS
NS
NS
5
Plasma
2
4
­
­
­
­
NA
NA
NA
NA
NA
NA
0
Homogenate
1
1
­
­
­
­
NS
NS
*
NS
NS
NS
4
Homogenate
1
2
­
­
­
­
NS
NS
NS
NS
NS
NS
4
Homogenate
1
3
­
­
­
­
NS
NS
NS
NS
NS
NS
4
Homogenate
1
4
­
­
­
­
NS
NS
*
NS
NS
NS
6
Homogenate
2
1
­
­
­
­
NS
NS
*
NS
NS
NS
4
Homogenate
2
2
­
­
­
­
NS
NS
NS
NS
NS
NS
4
Homogenate
2
4
­
­
­
­
NS
NS
*
NS
*
NS
6
NS
=
Not
Significant
*
=
Significant
at
a
Family­
Wise
Error
Rate
of
0.05
NA
=
Not
Applicable
­
=
No
Data
Battelle
Draft
May
2003
26
Plasma
Samples
with
Homologous
Std
0%
20%
40%
60%
80%
100%
120%
140%
160%
180%
200%

0
1
2
3
4
5
6
Concentration
Code
CV%
of
Intra­
assay
VTG
Concentrations
Carp­
sandwich­
1
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Carp­
5
Carp­
6
Zebra
Fish­
11
Plasma
Samples
with
Battelle
Std
0%
20%
40%
60%
80%
100%
120%

0
1
2
3
4
5
6
Concentration
Code
CV%
of
Intra­
assay
VTG
Concentrations
Carp­
sandwich­
1
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Zebra
Fish­
11
Figure
8.
The
CVs
of
the
Average
VTG
Concentration
for
Each
Laboratory,
Antibody,
and
Concentration
Code
for
Plasma
Samples
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
27
Homogenate
Samples
with
Homologous
Std
0%
20%
40%
60%
80%
100%
120%
140%
160%
180%
200%

0
1
2
3
4
5
6
Concentration
Code
CV%
of
Intra­
assay
VTG
Concentrations
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Carp­
5
Carp­
6
Zebra
Fish­
8
Zebra
Fish­
11
Homogenate
Samples
with
Battelle
Std
0%
20%
40%
60%
80%
100%
120%
140%
160%
180%

0
1
2
3
4
5
6
Concentration
Code
CV%
of
Intra­
assay
VTG
Concentrations
Carp­
sandwich­
7
Carp­
sandwich­
13
Fathead
Minnow­
1
Fathead
Minnow­
4
Zebra
Fish­
8
Zebra
Fish­
11
Figure
9.
The
CVs
of
the
Average
VTG
Concentration
for
Each
Laboratory,
Antibody,
and
Concentration
Code
for
Homogenate
Samples
Code
1
=
Uninduced
Male;
Code
2
=
Uninduced
Female;
Code
3
=
Induced
male;
Code
4
=
Induced
Female;
Code
5
=
Positive
Control
Battelle
Draft
May
2003
28
Table
10.
Descriptive
Statistics
for
the
VTG
Concentrations
Obtained
from
the
Carp
sandwich
ELISA
Each
Sample
Type,
Standard,
Laboratory,
and
Concentration
Code
Series
Std
Lab
Id
Concentration
Code
Mean
Qualifier
Stdev
Plasma
Homologous
1
0
0.0005
U
0
Plasma
Homologous
7
0
0.24
0.03
Plasma
Homologous
13
0
0.06
0.01
Plasma
Homologous
1
1
0.67
0.33
Plasma
Homologous
7
1
0.30
0.04
Plasma
Homologous
13
1
0.10
0.01
Plasma
Homologous
1
2
5056
96
Plasma
Homologous
7
2
6611
194
Plasma
Homologous
13
2
11197
8227
Plasma
Homologous
1
3
25584
8410
Plasma
Homologous
7
3
25252
5227
Plasma
Homologous
13
3
105878
13282
Plasma
Homologous
1
4
NA
NA
Plasma
Homologous
7
4
70513
14022
Plasma
Homologous
13
4
71344
11469
Plasma
Homologous
1
5
930
335
Plasma
Homologous
7
5
337
177
Plasma
Homologous
13
5
1573
207
Plasma
Purified
fathead
minnow
1
0
0.0005
U
0
Plasma
Purified
fathead
minnow
7
0
0.42
0.04
Plasma
Purified
fathead
minnow
13
0
0.11
0.03
Plasma
Purified
fathead
minnow
1
1
0.89
0.69
Plasma
Purified
fathead
minnow
7
1
0.50
0.06
Plasma
Purified
fathead
minnow
13
1
0.25
0.03
Plasma
Purified
fathead
minnow
1
2
6130
2459
Plasma
Purified
fathead
minnow
7
2
10727
296
Plasma
Purified
fathead
minnow
13
2
56311
62311
Plasma
Purified
fathead
minnow
1
3
31277
10442
Plasma
Purified
fathead
minnow
7
3
34673
7224
Plasma
Purified
fathead
minnow
13
3
104978
13163
Plasma
Purified
fathead
minnow
1
4
NA
NA
Plasma
Purified
fathead
minnow
7
4
99859
19822
Plasma
Purified
fathead
minnow
13
4
70144
11469
Plasma
Purified
fathead
minnow
1
5
1117
411
Plasma
Purified
fathead
minnow
7
5
572
204
Series
Std
Lab
Id
Concentration
Code
Mean
Qualifier
Stdev
Battelle
Draft
May
2003
29
Plasma
Purified
fathead
minnow
13
5
8834
804
Homogenate
Homologous
7
1
0.21
U
0
Homogenate
Homologous
13
1
0.3
0.11
Homogenate
Homologous
7
2
0.61
0.13
Homogenate
Homologous
13
2
4.48
2.51
Homogenate
Homologous
7
3
2212
334
Homogenate
Homologous
13
3
2.87
0.28
Homogenate
Homologous
7
4
2946
649
Homogenate
Homologous
13
4
5281
60.1
Homogenate
Homologous
7
5
235
8.34
Homogenate
Homologous
13
5
275
91.4
Homogenate
Purified
fathead
minnow
7
1
0.36
U
0
Homogenate
Purified
fathead
minnow
13
1
0.57
0.31
Homogenate
Purified
fathead
minnow
7
2
1.13
0.23
Homogenate
Purified
fathead
minnow
13
2
10.3
7.36
Homogenate
Purified
fathead
minnow
7
3
3987
566
Homogenate
Purified
fathead
minnow
13
3
5.67
0.86
Homogenate
Purified
fathead
minnow
7
4
5219
1079
Homogenate
Purified
fathead
minnow
13
4
16222
201
Homogenate
Purified
fathead
minnow
7
5
252
218
Homogenate
Purified
fathead
minnow
13
5
745
278
U
=
Value
less
than
detected
Battelle
Draft
May
2003
30
Table
11.
Upper
and
Lower
95%
Confidence
Limits
for
the
Mean
VTG
Concentrations
Obtained
from
the
Carp
Sandwich
ELISA
Using
Laboratories
as
Replicates
Series
Std
Lab
Id
Concentration
Code
Mean
Average
Stdev
LCL
UCL
Plasma
Homologous
1
5
929.8889
946.6546
618.4238
­
589.596
2482.905
Plasma
Homologous
7
5
336.7841
Plasma
Homologous
13
5
1573.291
Plasma
Purified
fathead
minnow
1
5
1117.333
3508.06
4620.824
­
7970.71
14986.83
Plasma
Purified
fathead
minnow
7
5
572.4025
Plasma
Purified
fathead
minnow
13
5
8834.444
Homogenate
Homologous
7
5
235.2678
255.0291
27.94667
3.939306
506.1188
Homogenate
Homologous
13
5
274.7903
Homogenate
Purified
fathead
minnow
7
5
252.0114
498.6467
348.7949
­
2635.14
3632.431
Homogenate
Purified
fathead
minnow
13
5
745.2819
mRNA
Results
Laboratory
1
and
2
Protocols;
Taqman
theory.
The
Taqman
real­
time
PCR
technique
provides
a
gene
specific
assay
for
the
measurement
of
VTG
mRNA.
The
Taqman
is
very
similar
in
action
to
regular
PCR,
except
that
in
addition
to
amplifying
target
DNA,
it
also
has
the
ability
to
calculate
the
amount
of
material
present
at
each
cycle
by
fluorescent
signal
allowing
for
relative
comparison
of
starting
mRNA
material.
The
fluorescent
signal
is
attached
to
a
probe
that
binds
between
the
forward
and
reverse
primers
of
the
target
sequence.
The
signal
is
quenched
when
intact
and
is
only
released
when
the
target
sequence
is
amplified
creating
a
direct
correlation
between
amount
of
fluorescent
signal
and
amount
of
amplified
material.
To
calculate
the
amount
of
mRNA
in
samples,
a
standard
curve
is
used
with
known
amounts
of
mRNA.
Based
on
amplification
data
relative
to
the
standard
curve,
it
is
possible
to
calculate
the
amount
of
starting
material
in
unknown
samples.
It
is
necessary
to
know
a
portion
of
the
sequence
of
the
gene
to
be
measured
in
order
to
design
Taqman
specific
PCR
primers
and
probe.
The
specific
sequence
for
FHM
VTG
has
been
published
in
the
database
and
is
available.

Laboratory
1.
The
FHM
VTG
specific
sequences
were
used
to
develop
Taqman
specific
primers
and
probe
using
Primer
Express
software
(
Applied
Biosystems).
Gene
specific
plasmids
containing
the
VTG
fragment
were
used
to
make
the
standards
for
the
standard
curve.
To
develop
the
standard
curve,
plasmid
containing
the
cloned
gene
fragment,
was
diluted
to
1ng/
ul
(
2.56x108
ER"
copies
and
2.68x108
VTG
copies
respectively).
Performing
serial
dilutions,
a
standard
curve
range
of
268
to
2.68x107
copies
was
obtained
for
VTG.
18S
rRNA
was
used
as
a
normalizing
gene
to
accommodate
for
error
in
preparation.
Once
a
sample
was
run
for
ER"
and/
or
VTG
mRNA
and
18S
rRNA,
mRNA
amounts
could
be
calculated.
Raw
data
was
Battelle
Draft
May
2003
31
normalized
based
on
average
vs.
individual
18S
rRNA
values
and
then
calculated
to
determine
amount
of
mRNA
in
copies
per
ug
of
total
RNA.

Laboratory
2.
Total
RNA
was
extracted
from
a
portion
of
the
liver
tissue
sample
supplied
using
a
commercially
available
kit
(
QIAGEN).
Total
RNA
was
used
in
a
real­
time
quantitative
reverse
transcription­
polymerase
chain
reaction
(
qRT­
PCR)
on
an
Applied
Biosystems
(
ABI)
Model
7900
real­
time
PCR
machine.
The
ABI
TaqMan
reagent
system
was
used
with
primers
and
probe
designed
to
a
specific
region
of
the
fathead
minnow
VTG
DNA
sequence.
Primers
and
probe
sequence
were
selected
using
ABI
Primer
Express
software,
and
sufficient
reagents
were
synthesized
by
ABI.
RT
reactions
were
performed
with
oligo­
dT,
followed
by
PCR
using
TaqMan
Universal
PCR
Master
Mix
and
the
VTG
specific
probe
(
fluorescent­
labeled)
and
primers.
As
an
internal
standard
the
fathead
minnow
12S
ribosomal
gene
was
used
to
allow
post­
PCR
normalization
of
starting
RNA
amounts.
Specific
primers
and
probe
for
the
12S
ribosomal
gene
were
designed
and
synthesized
similar
to
the
VTG
reagents
described
above.
A
standard
curve
developed
by
serially
diluting
a
known
reference
sample
was
used
in
each
assay,
to
which
unknown
samples
were
compared.

Laboratory
15.
A
hybridization
protection
assay
(
HPA)
was
applied
to
RNA
extracted
from
the
liver
tissue
of
the
fathead
minnows.
An
oligonucleotide
probe,
labeled
with
a
chemiluminescent
acridinium
ester,
was
introduced
into
a
sample
of
the
extracted
RNA
to
hybridize
with
any
complimentary
target
present.
Hybridization
is
followed
by
a
selection
step,
in
which
label
attached
to
free
probe
is
hydrolyzed
to
a
non­
chemiluminescent
derivative,
while
label
attached
to
hybridized
probe
is
protected
from
such
hydrolysis.
Following
this
selection
step,
chemiluminescence
is
measured
using
a
luminometer.
The
light
intensity
is
proportional
to
the
concentration
of
VTG
mRNA
in
the
sample.

The
reported
results
from
the
three
laboratories
that
applied
their
method
to
the
detection
of
mRNA
in
the
fathead
minnow
samples
are
summarized
in
Table
12.
As
described
previously,
Laboratory
1
and
2
applied
RT­
PCR
and
the
third
Laboratory
(
Lab15)
applied
a
hybridization
protection
assay
(
HPA)
to
the
samples.
All
three
laboratories
were
provided
with
liver
tissue
and
performed
all
analytical
steps,
including
isolation
of
RNA
from
the
tissue,
as
would
be
applied
to
analytical
samples
collected
as
a
part
of
a
screening
assay,
representing
individual
fish
units.
Although
not
directly
comparable
due
to
method
differences
and
approach,
all
three
methods
distinguished
between
exposed
and
unexposed
fish.
In
the
unexposed
fish,
the
levels
were
variable
with
each
lab
reporting
generally
higher,
similar,
or
lower
levels
in
the
unexposed
males
vs.
the
levels
found
in
unexposed
females.
Battelle
Draft
May
2003
32
Table
12.
Summary
of
the
reported
mRNA
results
for
the
liver
samples
from
unexposed
and
exposed
male
and
female
fathead
minnows.

Sample
Code
Sample
Type
Lab
1
Lab
2
Lab
15
1
unexposed
male
3380
0.81
<
0.1
1
unexposed
male
430
0.61
<
0.1
1
unexposed
male
500
0.75
<
0.1
1
unexposed
male
43
0.48
<
0.1
1
unexposed
male
36
0.47
<
0.1
1
unexposed
male
875
+/­
1415*
0.62
+/­
0.20*
<
0.1
+/­
0*

2
unexposed
female
26
1.56
<
0.1
2
unexposed
female
23
1.22
<
0.1
2
unexposed
female
700
0.32
0.22
2
unexposed
female
100
0.27
<
0.1
2
unexposed
female
15
0.50
NA
2
unexposed
female
180
+/­
307*
0.78
+/­.
60*
0.22
+/­
0*
3
exposed
male
216000
1470
19.9
3
exposed
male
186000
1300
13.8
3
exposed
male
145000
851
3.38
3
exposed
male
134000
878
13.2
3
exposed
male
198109.4
2150
NA
3
exposed
male
176000
+/­
35000*
1330
+/­
533*
13
+/­
7.0*
4
exposed
female
180000
1950
16.6
4
exposed
female
179000
3260
18.8
4
exposed
female
154000
3040
13.3
4
exposed
female
160000
2170
6.36
4
exposed
female
176000
2500
NA
4
exposed
female
170000
+/­
12000*
2600
+/­
560
14
+/­
5.0*

*
pg
VTG
mRNA
/
:
g
total
RNA;
*
mRNA
VTG
/
total
RNA
;
*
fmol
VTG
mRNA
/
:
g
total
RNA
GC­
MS
Results
Mass
spectrometric
(
MS)
analysis
of
VTG
and
other
large
glycosylated
biomolecules
has
allowed
for
a
general
approach
to
be
utilized
for
specifically
identifying
proteins
both
in
purified
forms
and
from
within
mixtures.
A
common
type
of
ionization
technique
for
MS
that
allows
analysis
of
large
biomolecules
without
fragmentation
is
matrix
assisted
laser
desorption/
ionization
mass
spectrometry
(
MALDI­
MS).

The
technique
of
peptide
mapping
using
MALDI­
MS
has
been
used
for
nearly
a
decade
to
identify
proteins.
This
technique
relies
on
the
mass
measurement
of
peptides
produced
by
proteolytic
digestion
and
comparing
them
with
the
predicted
peptide
masses
from
each
protein
in
the
database.
Algorithms
are
then
used
to
compare
and
determine
a
probability
score
to
match
the
experimental
data
with
candidates
in
the
database.
The
mathematical
tools
for
identifying
proteins
using
this
approach
are
available
publicly
at
<
http://
prowl.
rockefeller.
edu/
cgibin
ProFound>
through
an
easily
navigable
web
interface.
Protein
identification
using
the
Battelle
Draft
May
2003
33
peptide
mapping
approach
is
limited
to
a
small
(
mixtures
of
four
or
fewer)
number
of
proteins,
so
pre­
separation
of
proteins
from
complex
mixtures
is
often
required.

This
approach
was
used
to
directly
analyze
VTG
from
fathead
minnow
plasma
using
a
combination
of
liquid
chromatography
and
mass
spectrometry.
A
simple
membrane
filtration
pre­
purification
step
was
coupled
to
an
analytical
scale
anion
exchange
separation.
This
approach
to
MALDI­
MS
analysis
requires
a
relatively
small
plasma
sample
(<
10
µ
l)
and
is
suitable
for
use
with
plasma
volumes
typically
obtained
from
individual
fathead
minnows.
Correction
for
incomplete
recovery
of
the
VTG
was
achieved
through
the
use
of
an
internal
standard.
Peak
identity
as
VTG
was
confirmed
through
automated
fraction
collection,
trypsin
digestion
and
MALDI­
MS
analysis.

VTG
was
relatively
easy
to
purify
although
performing
analytical
separation
on
it
proved
challenging.
VTG
is
well
known
to
degrade
quickly
and
so
precautions
to
add
protease
inhibitors
(
PMSF)
and
keep
the
sample
on
ice
need
to
be
made.
Even
with
these
precautions,
analysis
of
the
sample
needs
to
occur
immediately
after
purification,
as
VTG
apparently
degrades
in
a
matter
of
hours
after
the
microcon
purification.
Both
the
large
size
and
dimeric
nature
can
lead
to
peak
broadening,
so
urea
was
added
as
a
denaturant.
Once
unfolded,
large
proteins
are
able
to
take
on
different
conformations
and
so
the
pore
size
of
the
resin
used
in
the
separation
must
be
consid­
ered
to
avoid
loses.
For
this
reason,
a
non­
porous
resin
was
selected
for
this
work.
A
smaller
inner
diameter
(
ID)
column
was
also
used
to
minimize
dead
volume,
flow
rate
and
increase
sensitivity
for
smaller
quantities
of
material.
The
drawback
to
using
a
small
ID
column
with
a
non­
porous
resin
is
the
decreased
sample
capacity.
Precautions
must
be
taken
to
not
overload
the
column
and
reduce
the
resolution
of
the
separation.
If
a
large
amount
of
protein
is
apparent
(
often
from
a
small
precipitate
visible
on
the
microcon
100
filter)
then
a
larger
volume
of
buffer
can
be
used
when
retrieving
the
sample
from
the
membrane.
A
detailed
protocol
is
presented
in
Appendix
F.

Figure
10
contains
the
UV
trace
of
equal
amounts
of
BSA
and
VTG
after
going
through
the
microcon
100
cleanup
step
and
directly
injection
on
the
AX
column.
Twenty
µ
g
of
each
protein
was
used
in
the
cleanup
(
the
BSA
added
just
prior
to
retrieval)
and
10
µ
g
of
each
injected.
The
peaks
are
well
resolved
and
of
equal
area.
This
sample
represents
roughly
the
detection
limit
of
VTG
that
can
be
injected,
with
a
concentration
of
250
ng/
ul
producing
a
peak
visible
in
the
chromatogram.
Figure
11
shows
the
chromatogram
of
a
sample
similar
to
shown
in
figure
ten,
however
in
this
sample
the
20
µ
g
of
VTG
has
been
added
to
10
µ
l
of
plasma
taken
from
an
untreated
fish
219PR33.
In
figure
12,
the
chromatogram
for
plasma
from
an
untreated
fish
(
219PK18)
has
been
cleaned
up
in
a
similar
fashion
without
the
addition
of
VTG.
Only
the
peak
for
BSA
is
apparent
with
no
peak
appearing
at
the
retention
time
of
VTG.
This
indicates
that
there
is
a
nearly
linear
relationship
between
the
peak
areas
for
BSA
and
the
VTG
standard
using
10
or
20
µ
g
of
BSA
as
an
internal
standard
when
compared
to
5
µ
g
to
40
µ
g
of
VTG
injected.
50
µ
g
is
nearing
the
upper
limit
of
column
capacity
and
so
amounts
of
VTG
beyond
were
not
investigated
in
this
study.
Battelle
Draft
May
2003
34
10
ug
each
VTG
and
BSA
injected
­
0
.005
­
0
.003
­
0
.001
0.001
0.003
0.005
0.007
0.009
1
583
1165
1747
2329
2911
3493
4075
4657
5239
5821
6403
6985
7567
8149
8731
9313
9895
Time
(
600/
min)
Series1
BSA
Vtg
10
ug
VTG
injected
after
"
spike"
into
219PR33
control
plasma:
Purified
using
microcon
100
(
BSA
added
[
10ug
injected])

­
0
.005
­
0
.003
­
0
.001
0.001
0.003
0.005
0.007
0.009
1
579
1157
1735
2313
2891
3469
4047
4625
5203
5781
6359
6937
7515
8093
8671
9249
9827
Time
(
600/
min
)
Series1
BSA
Vtg
Figure
10.
UV
trace
of
equal
amounts
of
BSA
and
VTG
after
purification
for
GCMS

Figure
11.
10
ug
VTG
injected
after
"
spike"
into
219PR33
control
plasma:
Purified
using
microcon
100
(
BSA
added
[
10ug
injected])
Battelle
Draft
May
2003
35
No
VTG
"
Spiked"
into
219Pk18
Control
Plasma:
Purified
using
microcon
100
with
BSA
added
(
10ug
injected)

­
0
.005
­
0
.003
­
0
.001
0
.001
0
.003
0
.005
0
.007
0
.009
1
569
1137
1705
2273
2841
3409
3977
4545
5113
5681
6249
6817
7385
7953
8521
9089
9657
Time
(
600/
min)
Series1
BSA
Figure
12.
No
VTG
"
Spiked"
into
219Pk18
Control
Plasma:
Purified
using
microcon
100
with
BSA
added
(
10ug
injected)

The
exposed
plasma
samples
had
a
large
amount
of
protein
remaining
on
top
of
the
microcon
membranes
in
some
cases,
and
so
the
volume
of
buffer
added
before
the
sample
was
collected
from
the
membrane
was
larger
(
100
µ
l).
Figure
13
is
the
chromatogram
for
treated
plasma
sample
219PU18
where
10
µ
l
of
the
100
µ
l
total
volume
was
injected.
The
six
µ
g
of
BSA
is
barely
visible
in
comparison
to
the
VTG
peak.
Note
the
scale
difference
between
this
figure
and
previous
figures
making
the
BSA
peak
appear
almost
non­
existent.
A
one
to
five
dilution
of
the
same
sample
was
run
with
additional
BSA
added
(
for
a
final
of
20
µ
g
injected)
and
the
chromatogram
is
shown
in
Figure
14.
Battelle
Draft
May
2003
36
219PU18
Exposed
Plasma:
Diluted
1
to
5
with
additional
BSA
(
20
ug
BSA
injected)

­
0
.005
0
0.005
0.01
0.015
0.02
0.025
1
572
1143
1714
2285
2856
3427
3998
4569
5140
5711
6282
6853
7424
7995
8566
9137
9708
Time
(
600/
min)
Series1
BSA
Vtg
219PU18
Exposed
Plasma:
Purified
using
microcon
100
with
60
ug
BSA
(
6
ug
injected)

­
0
.005
0.005
0.015
0.025
0.035
0.045
1
575
1149
1723
2297
2871
3445
4019
4593
5167
5741
6315
6889
7463
8037
8611
9185
9759
Time
(
600/
min)
Series1
BSA
Vtg
Figure
13.
219PU18
Exposed
Plasma:
Purified
using
microcon
100
with
60
ug
BSA
(
6
ug
injected)

Figure
14.
219PU18
Exposed
Plasma:
Diluted
1
to
5
with
additional
BSA
(
20
ug
BSA
injected)
Battelle
Draft
May
2003
37
1000
1500
2000
2500
3000
219PU18
VTG
AX
fraction
Digest
1000
1500
2000
2500
3000
Mass
(
m/
z)
219PS18
VTG
AX
fraction
Digest
1000
1500
2000
2500
3000
Standard
curve
20ug
Each
Vtg
&
BSA:
VTG
AX
fraction
digest
Relative
abundance
Comparison
of
MALDI­
MS
spectra
from
Vtg
Digests
vitellogenin
precursor
[
Pimephales
promelas]
Rank:
1
Probability
1.0
e
+
000
Z
score:
1.98
%
coverage:
17
vitellogenin
precursor
[
Pimephales
promelas]

Rank:
1
Probability
1.0
e
+
000
Z
score:
1.67
%
coverage:
25
vitellogenin
precursor
[
Pimephales
promelas]
Rank:
1
Probability
1.0
e
+
000
Z
score:
1.51
%
coverage:
15
Figure
15
is
a
comparison
of
the
peptide
digest
mass
spectra
from
VTG
fractions
collected
from
two
"
treated"
plasma
samples
and
from
the
VTG
standard.
The
associated
statistics
from
the
Profound
search
engine
for
the
protein
hits
for
each
data
set
are
given.
The
search
was
conducted
against
all
chordate
proteins
in
the
"
non
redundant"
database
(
combination
of
swissprot,
NCBI,
OWL
and
others)
using
a
1.5
Da
mass
error
for
the
average
peptide
mass
(
M+
H+)
and
a
protein
size
range
of
100
to
300
kDa.
The
output
provides
what
rank
each
database
hit
received
(
top
50
listed),
the
probability
score
associated
with
it
(
with
1.0
being
a
perfect
match).
Also
included
is
a
percentage
of
the
protein
sequence
represented
as
well
as
a
"
Z
score"
for
the
top
match
which
is
a
measure
of
distance
between
the
top
match
and
nearest
matches
(>
1.2
=
90th
percentile
of
correct
matches
if
compared
to
a
random
population
of
sequences,
>
1.6
=
95th
percentile
and
>
2.0
=
99th
percentile
respectively).
Table
13
is
a
list
of
the
top
sequence
matches
for
the
VTG
peak
from
treated
plasma
sample
219PT18.
The
top
two
hits
receiving
perfect
probability
scores
are
the
VTG
sequences
for
both
fathead
minnow
and
the
common
carp.
A
match
with
the
zebra
fish
was
also
made
at
a
much
lower
rank
and
probability.

Figure
15.
Comparison
of
the
peptide
digest
mass
spectra
from
VTG
fractions
collected
from
two
exposed
female
plasma
samples
and
from
the
VTG
standard.
Battelle
Draft
May
2003
38
Rank
Probability
Z
score
Protein
Information
and
Sequence
Analyse
Tools
(
T)
%
pI
+
1
1.0e+
000
2.08
gi|
4572552|
gb|
AAD23878.1|
AF130354_
1
vitellogenin
precursor
[
Pimephales
promelas]
17
9.0
­
­
gi|
15778562|
gb|
AAL07472.1|
AF414432_
1
vitellogenin
[
Cyprinus
carpio]
13
9.1
+
2
2.8e­
005
0.14
T
gi|
6006011|
ref|
NP_
005492.1|
(
NM_
005501)
integrin
alpha
3,
isoform
b,
[
Homo
sapiens]
14
6.5
+
3
6.4e­
007
­
T
gi|
21362287|
ref|
NP_
653099.1|
RIKEN
cDNA
2210402G22
[
Mus
musculus]
25
9.1
4
2.3e­
007
­
T
gi|
20344336|
ref|
XP_
111772.1|
similar
to
put.
gag
and
pol
gene
(
aa
1­
814)
[
Mus
musculus]
19
9.6
+
5
9.8e­
008
­
T
gi|
16550881|
dbj|
BAB71072.1|
(
AK056006)
unnamed
protein
product
[
Homo
sapiens]
27
9.8
6
1.4e­
008
­
T
gi|
13385164|
ref|
NP_
079982.1|
RIKEN
cDNA
4432405K22
[
Mus
musculus]
17
5.2
+
7
1.1e­
008
­
T
gi|
14735371|
ref|
XP_
027054.1|
(
XM_
027054)
KIAA0674
protein
[
Homo
sapiens]
19
5.0
+
8
7.6e­
009
­
T
gi|
6005944|
ref|
NP_
009058.1|
(
NM_
007127)
villin
1;
Villin­
1
[
Homo
sapiens]
15
6.0
9
6.8e­
009
­
T
gi|
21391472|
gb|
AAK58480.1|
vitellogenin
1
[
Danio
rerio]
12
8.9
kDa
146.26
148.24
118.74
66.13
85.00
64.01
76.51
92.49
92.68
128.02
Vtg
Anion
Exchange
Fraction
from
treated
plasma
219PT18:
database
search
with
peptide
masses
Table
13.
The
top
sequence
matches
for
the
VTG
peak
from
exposed
female
plasma
sample
219PT18.

Table
14
lists
the
results
from
all
of
the
treated
plasma
samples
including
whether
any
VTG
could
be
detected,
if
so
what
was
the
estimated
amount
present.
All
VTG
concentrations
are
calculated
for
the
10
µ
l
total
Plasma
volume
received.
Two
of
the
samples
(
219PS18
and
219
PG18
unexposed
female)
were
analyzed
early
in
the
refinement
of
the
method
were
initially
diluted
too
much.
The
undiluted
samples
were
analyzed
later
after
some
degradation
had
occurred
due
to
the
presence
of
many
peaks
eluting
earlier
than
expected.
Therefore,
no
estimates
of
VTG
concentration
were
calculated
for
those
samples
although
a
definite
peak
appeared
at
the
correct
retention
time
and
a
fraction
for
219PS18
was
collected,
digested
and
analyzed
by
MALDI­
MS.
The
results
for
that
are
given
in
Table
14.
Two
samples
had
no
detectable
VTG
in
them,
219PI18
and
219PM18
(
unexposed
male).
Sample
219PD18
(
unexposed
male)
did
appear
to
have
a
very
small
peak
at
the
correct
retention
time,
but
it
is
very
small
and
is
difficult
to
definitively
say
that
VTG
is
present.
Therefore
it
is
given
a
tentative
assignment.
Not
all
samples
containing
VTG
had
the
fractions
collected
for
MALDI­
MS
analysis.
For
those
fractions
analyzed
by
MALDI­
MS
fingerprinting,
the
rank,
probability
score
and
Z
score
are
given
for
each.
Z
score
values
above
1.6
are
considered
to
be
a
confident
match
in
the
95th
percentile
(
or
95%
chance
that
it
is
a
correct
hit).
Battelle
Draft
May
2003
39
Table
14.
Summary
of
the
results
of
GC­
MS
analysis
of
unexposed
and
exposed
male
and
female
fathead
minnow
plasma
samples.

Sample
ID
Sample
type
VTG
detected
Probability
Z
score
%
coverage
Original
Concentration
219PU18
exposed
female
Yes
1
1.67
25
100
ug/
ul
219PT18
exposed
female
Yes
1
2.08
17
75
ug/
ul
219PS18
unexposed
female
Yes
1
1.98
17
NC
219PG18
unexposed
female
Yes
NC
NC
NC
NC
219PI18
unexposed
male
No
##
##
##
##

219PM18
unexposed
male
No
##
##
##
##

219PD18
unexposed
male
Yes**
NC
NC
NC
<
0.5
ug/
ul
*

**
=
Tentative
assignment
NC
=
not
calculated
##
=
Not
applicable
*
=
estimate
of
original
concentration,
not
concentration
injected.

8.0
DISCUSSION
The
purpose
of
this
study
was
to
survey
the
existing
ELISA
methods
that
are
currently
available
for
use
in
detecting
the
protein
vitellogenin
in
fathead
minnow
plasma
and
whole
body
homogenate
samples.
A
variety
of
methods
routinely
performed
by
the
participating
laboratories
were
applied
to
a
standard
series
of
samples.
All
of
the
samples
were
provided
blind
coded
and
each
laboratory
received
three
replicates
of
each
sample
within
the
coded
sample
set,
and
each
sample
was
assayed
in
triplicate
(
3
wells
per
sample).
For
example,
the
standard
series
consisted
of
4
samples
derived
from
induced
and
uninduced
male
and
female
fish,
and
a
positive
and
negative
control.
This
resulted
in
6
samples
within
the
plasma
series,
spanning
a
range
of
concentrations.
Each
of
the
participating
laboratories
were
provided
with
these
six
samples
in
triplicate,
for
a
total
of
18
samples
for
analysis
(
blind
coded).
Upon
analysis,
each
laboratory
analyzed
each
of
the
18
samples
in
triplicate
wells
(
with
the
exception
of
one
laboratory,
as
noted
in
the
results,
which
provided
results
for
one
analysis
rather
than
in
triplicate).
Because
of
the
blind
nature
of
the
samples
and
the
wide
range
of
concentrations
present
in
the
samples,
this
required
multiple
dilutions
of
the
sample
to
ensure
a
response
within
the
working
range
of
the
assays.
This
required
a
significant
investment
of
time
and
resources
that
was
donated
by
the
participating
laboratories
to
aid
in
reaching
the
goals
of
this
study,
which
is
gratefully
acknowledged.
In
addition
the
intent
of
study
was
not
to
validate
a
given
method,
protocol,
system,
or
technique
but
rather
a
survey
of
methods
that
are
currently
used
in
attempt
to
discern
the
relative
variability
among
those
methods.
Also,
the
use
of
trade
names,
identification
of
Battelle
Draft
May
2003
40
laboratories,
and
methods
described
in
this
do
not
constitute
endorsement
by
the
U.
S.
Environmental
Protection
Agency
or
Battelle
Memorial
Institute.

To
generate
the
standard
series
of
samples,
wet
laboratory
exposures
to
estrogen
were
conducted
with
male
and
female
fathead
minnows,
and
the
plasma
and
whole
body
homogenates
were
prepared
from
pooled
fish.
This
resulted
in
four
sample
types
within
the
series
(
unexposed
male
and
female,
and
exposed
male
and
female).
In
addition
to
the
4
samples
in
the
series,
antibodies
were
used
to
remove
the
VTG
signal
from
unexposed
male
plasma
to
create
a
blank
sample.
A
known
amount
of
purified
fathead
minnow
VTG
was
added
to
a
portion
of
this
plasma
to
create
a
positive
control
sample,
resulting
in
six
samples
in
the
plasma
series.
Multiple
aliquots
of
each
sample
in
the
series
were
created
and
entered
into
a
repository
system.
The
samples
were
processed
with
care
to
limit
the
degradation
of
the
protein
during
collection
and
processing
and
were
maintained
at
­
80
°
C
in
the
repository.
Within
a
two
week
period,
the
samples
were
sent
to
all
of
the
participating
laboratories,
while
maintaining
the
integrity
of
the
samples
(
samples
remained
frozen
throughout
the
transfer
and
were
stored
at
­
80
°
C
by
the
participating
laboratories).

Twelve
laboratories
received
sets
of
the
homogenate
and
plasma
sample
series
for
analysis.
Two
laboratories
were
unable
to
analyze
the
samples
within
the
time
frame
of
the
study,
and
two
laboratories
conducted
the
analysis
but
found
that
their
antibodies
did
not
react
with
the
fathead
minnow
samples
(
carp
monoclonal
and
trout).
Eight
laboratories
reported
results
of
the
plasma
and
homogenate
analysis,
and
six
of
these
also
used
the
fathead
minnow
VTG
that
was
purified
for
this
study
(
and
used
to
create
the
positive
control
sample
in
the
series)
in
their
assay
(
Table
2).
Of
the
eight
laboratories
reporting
results,
2
applied
a
fathead
minnow
ELISA
(
one
monoclonal
antibody­
based
and
one
polyclonal
based),
two
applied
a
carp­
based
competitive
ELISA,
two
applied
zebrafish­
based
ELISAs
and
three
applied
a
commercially
available
carp­
based
sandwich
ELISA
kit
(
Biosense)
to
the
samples
(
note:
one
laboratory
applied
both
a
fathead
ELISA
and
the
commercial
kit).
It
should
be
noted
that
all
of
the
participating
laboratories
have
significant
experience
in
conducting
ELISAs
and
their
method
was
applied
to
the
samples
in
the
routine
manner
employed
by
each
laboratory.

Please
note
that
the
major
goal
of
this
study
was
to
conduct
a
survey
of
existing
ELISA
methods
that
might
be
applied
to
fathead
minnow
samples.
In
the
process
of
conducting
the
study,
three
laboratories
used
the
same
method
during
the
study,
thereby
allowing
for
some
statistical
comparison
of
the
analysis
of
the
samples
in
the
series
by
one
method.
This
was
not
an
attempt
to
validate
a
particular
method,
and
the
results
obtained
from
the
use
of
this
method,
by
circumstance,
by
a
statistically
valid
number
of
laboratories
should
not
be
used
to
assess
the
strength
or
weakness
of
this
method
compared
to
other
methods.
Rather,
it
should
be
assumed
that
the
encountered
variability
in
results
would
be
found
with
the
application
of
any
one
of
the
methods
in
this
study
by
multiple
laboratories.

All
of
the
reported
results
in
this
study
can
be
pooled
to
examine
the
inter­
and
intraassay
variation
in
the
analysis,
and
the
trend
of
concentration
in
the
standard
series.
The
results
of
the
study
can
also
be
examined
based
upon
the
type
of
ELISA
(
e.
g.
fathead
minnow),
the
standard
used
(
the
one
provided
(
purified
fathead
minnow
VTG)
or
the
one
typically
used
in
the
assay
(
homologous))
and
the
laboratory
performing
the
analysis
(
e.
g.
Lab
1).
These
categories
Battelle
Draft
May
2003
41
can
be
used
to
assess
the
trend
in
concentration
of
the
series,
the
ability
to
measure
the
concentration
in
the
positive
control,
and
the
ability
to
distinguish
between
the
concentration
of
VTG
in
individual
samples
(
e.
g.
blank
from
unexposed
male).
Each
of
these
factors
has
direct
implications
to
the
application
of
an
ELISA
method
for
use
in
a
routine
fish
screening
assay.

The
analyses
in
this
report
address
the
within­
run
variability,
the
intra­
assay
variability
(
based
on
the
mean
triplicate
result),
and
the
general
trend
of
the
ELISA
VTG
results
associated
with
the
standard
evaluation
series
of
fish
plasma
and
tissue
(
e.
g.,
whole
body
homogenate).
This
series
was
represented
by
1)
uninduced
male,
2)
uninduced
female,
3)
induced
male,
and
4)
induced
female
fathead
minnows,
respectively.
In
addition
to
the
standard
series,
a
set
of
positive
and
negative
control
VTG
results
are
summarized.
The
distribution
of
CVs
of
the
resulting
triplicate
mean
VTG
concentrations
are
summarized
for
a
given
concentration,
laboratory,
antibody,
and
standard;
across
laboratories,
antibodies,
standards,
and
assays
for
a
given
concentration;
and
by
antibody
for
a
given
laboratory,
standard,
and
concentration.

All
reported
results:

Within­
run
variability
­
Each
laboratory
analysed
each
sample
in
three
replicate
wells
(
within­
run)
during
analysis.
When
all
of
the
reported
results
are
examined,
the
within­
run
variability
for
plasma
had
CVs
ranging
from
0%
to
173%
with
a
mean
of
13%
(
Table
6;
Appendix
D).
The
within­
run
variability
for
homogenate
was
similar,
with
CVs
ranging
from
0%
to
162%.
The
distribution
of
these
CVs
can
be
seen
in
Figure
1
and
Figure
2,
revealing
that
the
results
from
replicate
wells
per
sample
within
the
ELISA
assays
resulted
in
CVs
less
than
16%
for
75%
of
the
samples
(
when
all
of
the
reported
results
for
a
sample,
in
both
the
homogenate
and
plasma
series
(
1­
5),
among
all
laboratories
are
pooled).
The
CVs
for
both
the
plasma
and
homogenate
samples
tended
to
be
less
than
30%
for
all
concentrations
except
for
those
derived
from
the
blank
and
unexposed
males
(
Figure
1)
which
are
near
the
detection
limit
of
the
assays.
The
range
of
CV's
for
replicate
ELISA
wells
per
samples
indicates
that
when
all
of
the
methods
are
applied
to
the
range
of
samples,
low
variation
is
typically
achieved
except
when
measuring
the
background
to
low
level
response
(
pre­
absorbed
plasma
blank
and
unexposed
male
assumed
to
be
devoid
of
VTG).
The
detection
of
low­
level
VTG
is
a
critical
component
of
a
method
for
use
in
a
screening
assay
to
detect
the
induction
of
the
VTG
protein.
It
should
be
noted
that
several
laboratories
reported
the
results
of
triplicate
analysis
noting
that
one
of
the
value
in
the
series
of
three
was
an
outlier
that
they
would
normally
not
include
in
their
analytical
reporting,
but
that
were
included
for
the
purposes
of
this
study.

Intra­
assay
­
Each
of
the
six
samples
in
the
series
(
e.
g.
unexposed
male,
positive
control)
were
provided
to
the
laboratories
in
triplicate
(
blind­
coded).
These
triplicate
samples
for
analysis
provided
a
measure
of
intra­
assay
variability.
The
intra­
assay
variability
for
plasma
and
homogenate
had
CVs
ranging
from
0%
to
173%
(
Table
7;
Appendix
D).
For
both
sample
types,
75%
of
the
intra­
assay
CVs
were
less
than
51%.
This
high
level
of
intra­
assay
variability
indicates
that
when
a
sample
is
provided
to
multiple
laboratories
employing
a
variety
of
methods
(
the
results
using
multiple
standards
are
also
included
in
this
sample
set)
the
methods
provide
a
high
degree
of
variability
when
replicate
samples
are
analyzed.
To
further
examine
this
variability
that
is
critical
to
the
application
of
ELISA
to
a
screening
assay,
the
data
was
further
examined
by
individual
laboratory.
Battelle
Draft
May
2003
42
Trend
of
standard
series
­
A
goal
of
the
study
was
to
generate
a
range
of
concentrations
of
VTG
in
male
and
female
plasma
and
whole
body
homogenates.
Continuing
the
examination
of
all
of
the
reported
data,
the
general
trend
for
the
plasma
samples
observed
for
each
laboratory
averaged
over
antibodies,
standards,
and
assays
was
an
increasing
concentration
of
VTG
in
the
series
(
i.
e.,
uninduced
male
<
uninduced
female
<
induced
male
<
induced
female
fathead
minnows;
Figure
3).
The
homogenate
samples,
however,
tended
to
suggest
that
uninduced
male
VTG
concentration
was
approximately
equal
or
slightly
greater
than
that
of
uninduced
females,
which
were
both
less
than
that
of
the
induced
males
(
Figure
4).
Induced
female
fathead
minnows
produced
the
greatest
concentrations
of
VTG
in
both
the
plasma
and
the
homogenate.
The
CVs
for
each
concentration
are
all
greater
than
80%;
thus
the
variability
between
laboratories,
antibodies,
and
standards
is
large.
This
high
degree
of
variability
indicates
that
there
are
significant
differences
in
the
results
among
reporting
laboratories,
indicating
that
the
methods
currently
in
use
provide
a
variety
of
results
and
that
for
the
application
of
these
methods
to
fathead
minnow
samples
must
be
further
examined.
This
analysis
included
all
reported
results
and
additional
analysis
was
conducted
based
upon
laboratory
and
type
of
method
to
examine
this
variability.

Difference
in
Standards
A
critical
aspect
of
the
performance
of
an
analytical
method
is
the
use
of
standard
and
controls
in
an
assay.
For
this
study
we
purified
fathead
minnow
VTG
and
used
it
to
create
a
positive
control
sample
and
for
use
as
a
standard
in
the
assay.
Six
of
the
participating
laboratories
reported
results
based
on
this
standard,
in
addition
to
their
own
standard.
The
percentage
difference
in
results
between
standards
calculated
as
(
H­
B)/
B
(
100%)
was
large
ranging
from
­
99%
to
200%.
This
is
a
significant
difference,
although
anticipated
based
upon
the
nature
of
the
antibody­
based
ELISA
to
VTG
from
various
species
and
from
various
purifications.
Thus,
standards
were
analyzed
separately
in
the
subsequent
analysis
in
this
report.

Ranking
of
VTG
concentrations
in
the
series.
When
the
results
from
each
of
the
laboratories
is
analyzed
by
standard
employed
in
the
assay,
all
produced
a
significant
regression
of
the
ranked
VTG
concentrations
against
the
series
code
0
though
4,
with
the
exception
of
Laboratory
5
using
the
homogenate
samples,
homologous
standard,
and
carp
antibodies.
However,
none
of
the
successive
pairs
in
the
series
were
found
to
be
significantly
different
using
Tuckey's
HSD
when
using
laboratories
as
replicates.
Therefore,
even
though
the
regressions
across
the
series
from
blank
to
induced
female
were
significant,
the
variability
between
laboratories
to
reproduce
the
ranks
for
a
given
concentration
code
was
too
large
to
detect
differences
between
successive
concentration
pairs
in
the
series.
This
indicates
that
with
or
without
the
inclusion
of
a
standard
for
use
in
the
analysis,
the
survey
of
methods
in
this
study
shows
that
the
variation
in
results
among
laboratories
and
methods
does
not
allow
for
the
determination
of
the
concentration
of
samples
within
the
standard
series
of
samples,
although
the
general
trend
of
the
samples
can
be
observed
when
all
results
are
examined.

Standard
homologous
to
method.
Eight
laboratories
reported
results
based
upon
their
homologous
standard
routinely
run
in
their
assay.
When
examined
by
type
of
antibody
and
form
of
ELISA,
the
carp
antibody
based
competitive
ELISAs
tended
to
produce
the
lowest
VTG
concentrations
in
the
plasma
homologous
standard
samples
(
Figure
6).
(
As
previously
noted,
the
Battelle
Draft
May
2003
43
Biosense
kit
and
the
"
carp­
based"
ELISA
use
antibodies
generated
in
carp.
However,
the
Biosense
kit
uses
multiple
antibodies
(
monoclonal
and
polyclonal)
in
a
form
of
ELISA
known
as
a
sandwich
ELISA,
which
differs
from
the
carp­
based
competitive
from
of
the
ELISA
employed
by
two
of
the
participating
laboratories)
Further,
in
contrast
to
the
other
methods,
the
carp­
based
competitive
ELISAs
did
not
distinguish
between
the
exposed
male
and
unexposed
female
samples
(
concentration
codes
2
and
3
respectively).
The
overall
low
measured
concentrations
and
the
inability
to
distinguish
between
exposed
and
unexposed
fish
limit
the
application
of
the
carp­
based
competitive
ELISA
to
the
fathead
minnow
samples
for
use
in
a
routine
screening
assay.
The
Biosense
kit,
fathead
minnow
and
zebrafish
methods
distinguish
among
the
samples
in
the
plasma
series
with
similar
reported
concentrations.
The
Biosense
and
fathead
minnow
antibodies
generally
produced
the
greatest
plasma
concentrations,
the
zebra
fish
antibody
the
next
greatest,
and
the
carp
the
least
(
Figure
6).
This
pattern
was
not
maintained
for
the
homogenate
samples
which
had
the
zebra
fish
antibodies
with
the
least
VTG
concentrations
in
the
induced
male
and
female
samples.
There
was
no
consistent
pattern
for
the
CVs
(
Figures
8
and
9),
indicating
that
the
among
the
methods
in
the
study
greater
variation
can
not
be
attributed
to
a
type
of
antibody­
based
assay.
Large
CVS
are
present
in
the
both
the
plasma
and
the
homogenate
series,
showing
analytical
variation
irrespective
of
sample
type.

Battelle
standard.
For
the
data
produced
using
the
Battelle
provided
standard,
the
zebrafish
antibody
tended
to
produce
lower
plasma
results
than
the
Biosense
and
fathead
minnow
antibodies
(
Figure
6).
This
is
consistent
with
the
use
of
the
homologous
standard.
In
general,
the
Battelle
standard
resulted
in
higher
concentrations
derived
for
the
samples.
This
is
consistent
with
the
reduced
specificity
of
the
various
antibodies
to
fathead
minnow
VTG.
However,
as
for
the
homologous
standard,
the
CVS
are
still
generally
large
for
both
the
homogenate
and
the
plasma
samples
limiting
the
interpretation
of
the
results.
The
CVS
associated
with
the
plasma
zebrafish
antibody
results
were
lower
for
higher
concentration
codes
(
Figure
8).
This
trend
is
not
repeated
in
the
homogenate
samples
(
Figure
9).

Multiple
Laboratories
Applying
Same
Method
The
design
of
this
study
does
not
permit
the
validation
of
any
of
the
methods
employed
to
measure
VTG
or
mRNA.
Rather
this
study
is
designed
to
be
a
comparative
survey
of
the
existing
methods
and
evaluation
of
how
similar
or
dissimilar
results
obtained
among
a
variety
of
methods
vary.
Although
the
assessment
of
the
application
of
one
method
by
multiple
laboratories
was
not
a
major
goal
of
this
study,
in
the
process
of
conducting
the
study,
three
laboratories
applied
the
same
carp­
based
sandwich....
to
the
sample
series,
including
the
low
level
samples
and
positive
control.
This
is
a
very
limited
data
set,
and
the
authors
of
this
report
urge
readers
to
please
note
that
the
results
of
this
study
should
not
be
used
as
a
validation
of
this
method.

The
carp­
based
sandwich
ELISA
which
was
utilized
by
multiple
laboratories
in
this
study
during
the
process
of
sample
preparation
and
analysis.
As
presented
previously,
the
CVs
of
the
results
of
the
application
of
the
various
methods
to
standard
series
limit
the
interpretation
of
the
results
of
this
study.
The
triplicate
analysis
of
the
series
by
one
method
allows
for
some
measure
of
the
application
of
one
of
the
methods
in
this
study
by
multiple
laboratories,
however,
similar
assessments
of
the
other
methods
in
this
study
would
be
required
to
draw
general
conclusions.
Battelle
Draft
May
2003
44
Blank
and
unexposed
male
samples.
For
those
laboratories
that
used
the
Biosense
kit
in
this
study,
a
simple
linear
regression
of
the
average
VTG
concentration
(
within­
run
average)
between
the
blank
and
the
uninduced
male
concentrations
was
conducted
to
determine
if
a
significant
slope
was
obtained.
For
the
plasma
samples,
only
Laboratory
13
produced
a
significant
slope
between
these
two
concentrations
using
both
standards
(
p
=
0.006,
4
d.
f.
for
error).
For
the
homogenate
samples,
both
laboratories
7
and
13
produced
significant
slopes
using
the
homologous
standard
(
p
<
0.05,
4
d.
f.
for
error),
and
Laboratory
7
produced
a
significant
slope
using
the
Battelle
standard
(
p
=
0.005,
4
d.
f.
for
error)
(
Table
11).
With
the
desire
to
measure
the
induction
of
VTG
in
male
fish
in
a
screening
assay,
the
variable
results
of
the
application
of
an
established
assay
by
several
skilled
laboratories
when
applied
to
the
same
set
of
samples
is
significant.
Although
these
results
indicate
that
the
detection
of
background
or
low
levels
of
VTG
in
fathead
minnow
samples
can
be
achieved
with
the
application
of
an
ELISA,
the
lack
of
consistent
concentrations
and
significant
differences
between
samples
among
laboratories
requires
additional
study
for
method
assessment.

Positive
control.
One
additional
analysis
was
conducted
on
the
results
from
the
commercial
ELISA
kit
as
applied
to
fathead
minnow
samples.
The
positive
control
samples
for
both
the
plasma
and
homogenate
were
spiked
to
500
:
g/
ml
VTG
with
the
fathead
minnow
VTG
purified
for
this
study.
With
three
laboratories
conducting
the
Biosense
ELISA,
these
results
can
be
examined
for
statistical
significance.
It
should
first
be
noted
that
the
results
were
not
unique
to
the
use
of
the
Biosense
kit,
and
similar
results
were
found
with
the
less
replicated
methods
(
e.
g.
fathead
minnow).
The
positive
control
concentrations
was
not
achieved
with
the
Biosense
kit
using
laboratories
as
replicates
(
Table
12).
The
results
of
the
positive
control
were
again
highly
variable
both
within
and
between
laboratories.
The
concentrations
detected
in
plasma
ranged
from
337
to
8834
:
g/
ml
VTG
in
plasma
and
from
235
to
745
:
g/
ml
VTG
in
the
homogenate
control
sample.
As
previously
noted,
In
addition
to
the
variation
found
in
the
Biosense
assay,
a
similar
wide
range
of
concentrations
for
the
control
samples
were
detected
among
all
of
the
assays
(
Table
8;
Figure
6
and
7)

9.0
SUMMARY
A
standard
series
of
plasma
and
homogenate
samples
representing
a
range
of
VTG
concentrations
in
male
and
female
fathead
minnows
was
generated
for
this
study.
In
addition
to
the
samples
in
the
series,
a
set
of
positive
and
negative
controls
were
generated
for
the
study.
Fathead
minnow
VTG
was
purified
for
use
in
the
positive
control
and
as
a
standard
for
use
by
the
participating
laboratories
in
their
method.
A
repository
of
samples
was
created
and
maintained,
and
12
laboratories
agreed
to
participate
in
the
study.
Sets
of
the
homogenate
and
plasma
series
(
blind
coded)
were
shipped
in
coordination
to
the
participating
laboratories
while
maintaining
the
integrity
of
the
samples.
Two
of
the
participating
laboratories
were
unable
to
complete
their
analysis
in
the
time
frame
of
the
study,
and
two
additional
laboratories
reported
that
their
antibodies
did
not
react
with
the
fathead
minnow
samples.
Of
the
eight
remaining
laboratories,
six
completed
the
analysis
of
their
sample
sets
with
their
homologous
standard
and
with
the
standard
provided
in
the
study.
Two
laboratories
employed
fathead
minnow­
based
ELISA,
two
were
zebrafish
based,
two
were
competitive
carp
based.
Three
of
these
used
a
commercially
available
kit.
Battelle
Draft
May
2003
45
There
was
significant
variation
in
the
reported
ELISA
results
in
this
study.
When
all
of
the
reported
results
were
analyzed,
the
within­
run
variability
ranged
from
0%
to
173%,
with
75%
of
the
CVs
less
than
16%.
This
indicates
a
wide
range
of
intra­
assay
variability
among
the
methods
in
this
study.
The
CVs
for
both
the
plasma
and
homogenate
samples
tended
to
be
less
than
30%
for
all
concentrations
except
for
those
derived
from
the
blank
and
unexposed
males
(
Figure
1)
which
are
near
the
detection
limit
of
the
assays.
With
the
need
to
measure
the
induction
of
VTG
in
male
fish
in
a
screening
assay,
the
ability
to
detect
VTG
in
unexposed
males
was
examined
further
in
the
results
from
three
laboratories
conducting
the
same
assay.
The
ability
to
detect
VTG
and
to
discern
between
the
blank
and
the
unexposed
male
samples
was
present,
but
inconsistent
among
laboratories.

When
the
replicate
samples
were
examined
for
all
reported
results,
the
intra­
assay
variability
for
plasma
and
homogenate
had
CVs
ranging
from
0%
to
173%
(
Table
7;
Appendix
D).
For
both
sample
types,
75%
of
the
intra­
assay
CVs
were
less
than
51%.
This
demonstrates
that
for
a
sample
provided
to
multiple
laboratories
employing
a
variety
of
methods,
the
results
will
be
subject
to
a
high
degree
of
variability
among
replicates,
limiting
the
usefulness
of
the
results.
A
similar
high
degree
of
variability
was
shown
when
the
control
samples
were
analyzed
by
one
method
by
multiple
laboratories.
Determining
the
source
of
this
variability
will
be
critical
in
the
application
of
ELISA
in
a
screening
assay.

Although
affected
by
large
variations
in
reported
results
by
the
various
methods,
the
trend
of
the
standard
series
revealed
increasing
levels
of
VTG
(
uninduced
male
<
uninduced
female
<
induced
male
<
induced
female
fathead
minnows;
Figure
3).
This
trend
was
demonstrated
irrespective
of
the
standard
used
in
the
assay.
However,
the
carp­
based
competitive
ELISAs
tended
to
produce
the
lowest
VTG
concentrations
in
the
plasma
homologous
standard
samples
(
Figure
6)
and
the
carp­
based
competitive
ELISAs
did
not
distinguish
between
the
exposed
male
and
unexposed
female
samples.
The
significantly
lower
reported
concentrations
and
the
inability
to
distinguish
between
exposed
and
unexposed
fish
may
limit
the
application
of
the
carp­
based
competitive
ELISA
to
the
fathead
minnow
samples
for
use
in
a
routine
screening
assay.

The
use
of
the
provided
fathead
minnow
VTG
standard
in
the
assays
typically
resulted
in
higher
concentration
reported
for
the
samples.
This
is
generally
reflective
of
the
differences
in
specificity
of
the
various
antibodies
in
the
methods.
The
use
of
the
standard
did
not
result
in
a
normalization
of
reported
concentrations,
nor
in
achieving
the
concentration
in
the
positive
control
(
spiked
with
the
same
VTG
as
in
the
standard).
The
variation
in
the
reported
concentration
for
the
positive
control
varied
significantly
among
laboratories
and
among
assays.

The
application
of
mRNA
analysis
to
the
detection
of
mRNA
in
the
fathead
minnow
samples
was
performed
by
three
laboratories
in
this
study
(
Table
12).
Two
laboratories
(
Lab
1
and
Lab
2)
applied
RT­
PCR
and
the
third
Laboratory
(
Lab15)
applied
a
hybridization
protection
assay
(
HPA)
to
the
samples.
All
three
laboratories
performed
all
analytical
steps,
including
isolation
of
RNA
from
the
liver
tissue,
as
would
be
applied
to
analytical
samples
resulting
from
a
fish
screening
assay.
All
three
methods
distinguished
between
exposed
and
unexposed
fish,
but
direct
comparison
is
limited
due
to
differences
in
methods
and
analytical
approach.
The
range
of
mRNA
levels
encompassed
a
wider
range
in
the
RT­
PCR
analysis
than
in
those
found
in
the
Battelle
Draft
May
2003
46
series
through
the
application
of
the
HPA
method.
In
the
unexposed
fish,
the
detection
of
mRNA
levels
were
variable,
with
the
labs
reporting
generally
higher,
similar,
or
lower
levels
in
the
unexposed
males
vs.
the
levels
found
in
unexposed
females.

For
GC­
MS
analysis,
a
purification
and
separation
step
was
implemented
prior
to
MALDI­
MS
analysis
that
allowed
detection
of
VTG
on
an
analytical
scale
suitable
for
smaller
sample
sizes.
MALDI­
MS
provided
a
general
means
of
confidently
identifying
VTG
by
matching
experimental
data
with
sequences
in
protein
databases.
However,
problems
with
sample
degradation
after
thawing
prevented
quantitative
estimates
of
VTG
in
plasma
samples
harvested
from
estradiol
exposed
and
unexposed
minnows.
Further
insight
into
the
true
detection
limits
and
reproducibility
of
quantitation
for
this
method
are
required.

10.0
REFERENCES
Battelle,
2002.
Draft
Detailed
Review
Paper,
Fish
Screening
Assays
for
Endocrine
Disruption.
Prepared
for
United
States
Environmental
Protection
Agency
by
Battelle
Pacific
Northwest
Division.

Denslow
N.
D.,
Chow
M.
C.,
Kroll
K.
J.,
and
Green
L.
(
1999).
Vitellogenin
as
a
biomarker
for
estrogen
or
estrogen
mimics.
Ecotox.
8,
385­
398.

Edmunds,
J.
S.,
McCarthy,
R.
A.,
and
Ramsdell,
J.
S.
2000,
"
Permanent
and
Functional
Maleto
female
Sex
Reversal
in
d­
rR
Strain
Medaka
(
Oryzias
latipes)
Following
Egg
Microinjection
of
o,
p'­
DDT,"
Environmental
Health
Perspectives,
Vol.
108,
No.
3,
pp
219­
224.

Kishida,
M.
&
Callard,
G.,
2001,
"
Distinct
Cytochrome
P450
Aromatase
Isoforms
in
Zebrafish
(
Danio
Rerio)
Brain
and
Ovary
Are
Differentially
Programmed
and
Estrogen
Regulated
During
Early
Development,"
Endocrinology,
Vol.
142,
No.
2,
pp.
740­
750.

Korte,
J.
J.,
Kahl,
M.,
Jensen,
K.,
Pasha,
M.,
Parks.
L.,
LeBlanc,
G.,
and
Ankley,
G.,
2000,
"
Fathead
Minnow
Vitellogenin:
Complementary
DNA
Sequence
and
Messenger
RNA
and
Protein
Expression
after
17$­
Estradiol
Treatment,"
Environmental
Toxicology
and
Chemistry,
Vol.
19,
No.
4,
pp.
972­
981.

Spitsbergen,
J.
M.,
Tsai,
H.,
Reddy,
A.,
Miller,
T.,
Arbogast
Dan,
Hendricks,
J.
D.,
&
Bailey,
G.
S.
2000,
"
Neoplasia
in
Zebrafish
(
Danio
Rerio)
Treated
with
N­
methyl­
n'­
nitro­
nnitrosoguanidine
by
Three
Exposure
Routes
at
Different
Developmental
Stages,"
Toxicologic
Pathology,
Vol.
28,
No.
5,
pp.
716­
725.

Van
Den,
H.
U.
R.
K.,
Schoonen,
W.
G.
E.
J.,
Van
Zoelen,
G.
A.,
&
Lambert,
J.
G.
D.
1987,
"
The
Biosynthesis
of
Steroid
Glucuronides
in
the
Testis
of
the
Zebrafish
Brachydanio­
rerio
and
Their
Pheromonal
Function
as
Ovulation
Inducers,"
Gen
Comp
Endocrinol,
Vol.
68
(
2).
1987.
179­
188.
Battelle
Draft
May
2003
47
Wunschel,
D.
S.,
and
Wahl,
K.
L.
2002,
"
Deglycosylation
and
Analysis
of
Vitellogenin
and
Ovalbumin
by
Mass
Spectrometry."
Unpublished
studies.
Advanced
Analytical
and
Organic
Methods
Group,
Pacific
Northwest
National
Laboratory,
Richland,
WA.

Yasuda,
T.,
Aoki,
K.,
Matsumoto,
A.,
&
Ishikawa,
Y.
2000,
"
Heritable
Malformations
Induced
by
X­
irradiation
in
the
Medaka,"
Journal
of
Radiation
Research,
Vol.
41,
No.
4,
p.
484.
