LETTER
REPORT
ON
PHASE
I
PRE­
OPTIMIZATION
EXPERIMENTS
FOR
SUBSTRATE
CHARACTERIZATION
FOR
HUMAN
RECOMBINANT
AND
HUMAN
PLACENTAL
MICROSOMES
EPA
Contract
Number
68­
W­
01­
023
WA
2­
24
May
6,
2003
PREPARED
FOR
GARY
E.
TIMM
WORK
ASSIGNMENT
MANAGER
U.
S.
ENVIRONMENTAL
PROTECTION
AGENCY
ENDOCRINE
DISRUPTOR
SCREENING
PROGRAM
WASHINGTON,
D.
C.

BATTELLE
505
KING
AVENUE
COLUMBUS,
OH
43201
1
Letter
Report
05
05
03
Work
Assignment
2­
24
Pre­
Validation
of
the
Aromatase
Assay
using
Human,
Bovine,
and
Porcine
Placental
Microsomes,
and
Human
Recombinant
Microsomes
Phase
I
Pre­
Optimization
Experiments
for
Substrate
Characterization
for
Human
Recombinant
and
Human
Placental
Microsomes
1
Introduction
The
pre­
optimization
experiments
were
designed
to
assess
the
chemical
and
biological
properties
of
the
critical
components
that
are
used
in
the
aromatase
assay.
These
experiments
include
characterizing
the
radiolabeled
substrate
and
preparation
of
placental
microsomes.
In
addition,
each
of
the
four
microsomal
preparations
(
human,
bovine,
and
porcine
placental
microsomes
and
the
human
recombinant
microsomes)
will
be
analyzed
for
protein
concentration,
cytochrome
P450
(
P450)
content,
and
aromatase
activity.
The
P450
content
measurement
will
provide
assurance
that
the
enzyme
is
present
(
and
in
what
concentration/
preparation
type)
prior
to
beginning
the
more
elaborate
aromatase
activity
assay.
Finally,
a
single
aromatase
activity
assay
run
using
each
type
of
microsomal
preparation
was
included
as
a
pre­
optimization
experiment
in
order
to
determine
whether
the
preparations
are
of
sufficient
activity
to
conduct
the
definitive
optimization
experiments.

This
report
includes
the
results
of
the
pre­
optimization
experiments
related
to
substrate
characterization
and
human
recombinant
and
human
placental
microsomes.

2
Materials
and
Methods
2.1
Chemicals
Non­
radiolabeled
4­
androstene­
3,17­
dione
(
ASDN)
was
received
through
Battelle
from
Sigma
(
St.
Louis,
MO).
[
1$­
3H(
N)]
Androst­
4­
ene­
3,17­
dione
([
3H]
ASDN)
was
obtained
from
Perkin
Elmer
Life
Science,
Boston,
MA.
NADP,
glucose
6­
phosphate,
glucose
6­
phosphate
dehydrogenase,
glycerol,
niacinamide,
dithiothreitol
and
bovine
serum
albumin
(
BSA)
were
purchased
from
Sigma.
Sodium
phosphate
monobasic,
sodium
phosphate
dibasic,
sucrose
and
propylene
glycol
were
from
JT
Baker.
Human
recombinant
CYP19,
coexpressed
with
P450
reductase,
(
Human
CYP19
+
P450
Reductase
SUPERSOMES
 
)
was
purchased
from
BD
Gentest
(
Woburn,
MA).
Ultima
Gold
scintillation
cocktail
was
purchased
from
Packard
Instruments.
DC
Protein
assay
kit
was
purchased
from
Biorad
(
Hercules,
CA)
.
2
2.2
HPLC
System
The
HPLC
system
consisted
of
a
Waters
2690
Separations
Module,
a
Waters
2487
Dual
8
Absorbance
Detector
and
a
$­
RAM
Model
3
flow­
through
radioactivity
detector
(
IN/
US,
Inc.,
Tampa)
with
a
250
:
L
glass
scintillant
cell.
Data
was
collected
using
Waters
Millennium32
Client/
Server
Chromatography
Data
System
Software,
Version
4.0.

2.3
Substrate
Characterization
The
nonradiolabeled
ASDN
was
dissolved
in
ethanol
(
0.01
mg/
mL)
and
analyzed
by
HPLC
using
a
Zorbax
SB­
C18
column
(
4.6
x
250
mm).
The
mobile
phase
was
55:
15:
30
(
v:
v:
v)
distilled,
deionized
water:
tetrahydrofuran:
methanol
with
a
flow
rate
of
1
mL/
min.
The
eluant
was
monitored
by
UV
absorbance
at
240
nm.
Under
these
conditions,
the
ASDN
had
a
retention
time
of
ca.
15
min.

The
purity
of
the
[
3H]
androstenedione
([
3H]
ASDN)
was
determined
by
HPLC
using
the
conditions
described
above,
with
the
addition
of
monitoring
eluant
with
the
radiochemical
detector.
Eluant
fractions
were
collected
manually
into
vials
containing
ca.
10
mL
Ultima
Gold
and
assayed
for
radiochemical
content
by
liquid
scintillation
spectrometry
(
LSS).
The
retention
time
of
[
3H]
ASDN
was
ca.
15
min.

2.4
Specific
Activity
Determination
A
sample
containing
1
mg
ASDN/
mL
ethanol
was
prepared.
Dilutions
containing
0.54
to
4.28
ng
ASDN/
20
:
L
were
prepared
in
distilled,
deionized
water
and
were
analyzed
by
HPLC
in
duplicate
using
the
conditions
described
above.
A
standard
curve
was
constructed
relating
peak
height
to
ASDN
concentration.
Samples
of
[
3H]
ASDN
were
analyzed
by
HPLC
using
the
same
conditions
and
fractions
were
collected
and
assayed
for
radiochemical
content
by
LSS.
The
specific
activity
of
the
[
3H]
ASDN
was
determined
by
the
relationship
between
the
height
of
the
ASDN
peak
and
the
amount
of
radioactivity
contained
in
the
peak.

2.5
Placental
Microsome
Preparation
A
human
placenta
was
received
from
a
local
hospital
and
placed
on
ice
within
10
min
of
delivery.
The
tissue
was
placed
on
a
ice­
chilled
board
and
the
soft
tissue
was
dissected
away
from
the
membrane.
The
soft
tissue
was
placed
in
ice­
cold
buffer
(
2:
1
tissue
weight:
buffer;
0.25
M
sucrose,
0.05
M
sodium
phosphate
(
pH
7.0),
0.04
M
niacinamide),
minced
with
scissors
and
then
was
homogenized
in
portions
using
a
Polytron
homogenizer.
The
homogenate
was
transferred
to
centrifuge
tubes
and
centrifuged
at
a
setting
of
10,000g
for
30
min
at
4
NC
in
an
IEC
B­
22M
centrifuge.
The
supernatant
was
transferred
to
ultracentrifuge
tubes
and
was
centrifuged
at
a
setting
of
35,000
rpm
(
which
is
equivalent
to
approximately
100,000g)
in
a
refrigerated
Beckman
L5­
50B
Ultracentrifuge
for
1
h
to
obtain
the
crude
microsomal
pellet.
The
supernatant
was
decanted
and
discarded
and
the
microsomal
pellet
was
resuspended
in
a
chilled
buffer
containing
0.1
M
sodium
phosphate
buffer,
pH
7.4.
The
sample
was
centrifuged
again
at
a
setting
of
35,000
rpm
in
the
Beckman
L5­
50B
for
1
h
to
wash
the
microsomes.
This
washing
3
procedure
was
repeated
one
additional
time.
The
twice­
washed
microsomal
pellet
was
resuspended
in
chilled
0.1
M
sodium
phosphate
buffer
(
pH
7.4)
containing
0.25
M
sucrose,
20%
glycerol
and
0.05
mM
dithiothreitol.
The
microsomal
suspension
(
total
volume
ca.
20.2
mL)
was
divided
among
30
vials,
was
flash
frozen
in
liquid
nitrogen
and
was
stored
at
ca.
­
70
NC.

2.6
Protein
determination
The
protein
concentration
of
the
human
placental
microsome
preparation,
and
the
human
recombinant
microsomal
preparation,
was
determined.
A
6­
point
standard
curve
was
prepared
using
BSA,
ranging
from
0.13
to
1.5
mg
protein/
mL.
Protein
was
determined
by
using
a
DC
Protein
Assay
kit.
To
a
25
:
L
aliquot
of
unknown
or
standard,
125
:
L
of
BioRad
DC
Protein
Kit
Reagent
A
was
added
and
mixed.
Next,
1
mL
of
BioRad
DC
Protein
Kit
Reagent
B
was
added
to
each
standard
or
unknown
and
the
samples
were
vortex
mixed.
The
samples
were
allowed
to
sit
at
room
temperature
for
at
least
15
min
to
allow
for
color
development.
The
absorbances
are
stable
for
about
1
h.
Each
sample
(
unknowns
and
standards)
was
transferred
to
disposable
polystyrene
cuvettes
and
the
visible
absorbance
(@
750
nm)
was
measured
using
a
spectrophotometer.
The
protein
concentration
of
the
microsomal
sample
was
determined
by
extrapolation
of
the
absorbance
value
using
the
curve
developed
from
the
absorbance
of
the
protein
standards.

2.7
P450
Content
P450
content
was
determined
for
the
human
placental
microsome
preparation
and
the
human
recombinant
microsomal
preparation.
Using
the
carbon
monoxide
(
CO)
spectrum
assay
of
Omura
and
Sato
(
1964),
a
single
experiment
using
each
of
the
preparations
was
conducted
as
described
below.

A
sample
of
each
microsomal
preparation
was
diluted
1:
20
in
0.1
M
phosphate
buffer
(
pH
7.4).
The
diluted
sample
was
gently
bubbled
with
carbon
monoxide
for
approximately
10
s
and
then
was
divided
between
a
pair
of
matched
cuvettes
(
1
mL/
cuvette).
Next,
a
few
grains
of
solid
sodium
dithionite
was
added
to
the
sample
cuvette
with
gentle
mixing.
The
visible
spectrum
was
then
recorded
from
400
to
500
nm
using
a
split­
beam
spectrophotometer.

The
concentration
(
nmol/
mL)
of
P450
was
calculated
according
to
Beer's
Law
using
an
extinction
coefficient
value
for
P450
of
100
mM­
1
cm­
1.
The
specific
content
(
nmol/
mg
protein)
was
calculated
by
multiplying
the
P450
concentration
(
nmol/
mL)
times
the
dilution
factor
and
dividing
this
product
by
the
protein
content
(
mg/
mL)
of
the
original
sample.

2.8
Aromatase
Activity
Aromatase
activity
was
determined
for
the
human
placental
microsome
preparation
and
the
humna
recombinant
CYP19.
A
single
experiment
was
conducted
using
only
the
substrate
([
3H]
ASDN/
ASDN)
with
each
of
the
microsomal
preparations.
The
assay
was
conducted
as
described
in
the
following
paragraph.

The
[
3H]
ASDN/
ASDN
substrate
solution
was
prepared
by
combining
solutions
of
4
[
3H]
ASDN
and
ASDN.
A
1
mg/
mL
solution
of
ASDN
was
prepared
in
ethanol.
Serial
dilutions
of
this
solution
were
prepared
in
assay
buffer
to
yield
a
solution
containing
ca.
1
:
g
ASDN/
mL.
The
[
3H]
ASDN
stock
was
diluted
1:
100
in
assay
buffer
to
yield
a
solution
containing
ca.
10
:
Ci/
mL.
The
substrate
solution
was
prepared
by
combining
275
:
L
of
the
1
:
g
ASDN/
mL
solution,
100
:
L
of
the
10
:
Ci
[
3H]
ASDN/
mL
solution
and
625
:
L
buffer.

The
assays
were
performed
in
13x100
mm
test
tubes
(
two
for
each
microsomal
preparation)
maintained
at
37
±
1
°
C
in
a
shaking
water
bath.
An
aliquot
(
100
:
L)
of
propylene
glycol
was
added
to
the
tubes
to
serve
as
a
co­
solvent.
The
substrate,
[
1$­
3H]­
androstenedione
(
0.1
:
Ci,
50
nM),
was
added
to
the
tubes.
An
NADPH­
generating
system
comprised
of
NADP+
(
1.7
mM),
glucose­
6­
phosphate
(
2.8
mM)
and
glucose­
6­
phosphate
dehydrogenase
(
1.0
units)
was
added
to
each
tube.
The
tubes
were
placed
at
37
±
1
°
C
in
the
water
bath
for
5
min
prior
to
initiation
of
the
assay
by
the
addition
of
the
diluted
microsomal
suspension
(~
0.1
mg
microsomal
protein/
mL).
The
total
volume
was
2.0
mL,
and
the
tubes
were
incubated
for
30
min.
The
incubations
were
stopped
by
the
addition
of
methylene
chloride
(
2.0
mL);
the
tubes
were
vortexmixed
for
about
30
s.
The
tubes
were
then
centrifuged
using
a
Beckman
GS­
6R
centrifuge
with
a
GH­
3.8
rotor
for
10
min
at
a
setting
of
1000
rpm
(
which
is
approximately
equivalent
to
230g).
The
methylene
chloride
layer
was
removed
to
a
vial
and
weighed;
the
aqueous
layers
were
extracted
again
with
methylene
chloride
(
2.0
mL).
This
extraction
procedure
was
performed
one
additional
time,
each
time
reserving
and
weighing
the
methylene
chloride
layer
in
a
separate
vial.
The
aqueous
layers
was
transferred
to
vials,
weighed,
and
duplicate
aliquots
(
0.5
mL)
were
weighed
into
20­
mL
liquid
scintillation
counting
vials.
Duplicate
aliquots
of
each
methylene
chloride
fraction
were
weighed
into
scintillation
vials.
Liquid
scintillation
cocktail
(
Ultima
Gold,
Packard,
10
mL)
was
added
to
each
counting
vial
and
shaken
to
mix
the
solution.

The
radiochemical
content
of
the
substrate
solution
was
determined
by
analyzing
5
weighed
aliquots
by
LSS.
The
substrate
solution
specific
activity
was
determined
by
dividing
the
radiochemical
content
of
the
substrate
solution
(
dpm/
g)
by
the
total
concentration
of
ASDN
in
the
solution
(
ASDN
+
[
3H]
ASDN;
nmol/
g
solution).

Analysis
of
the
samples
was
performed
using
LSS
as
described
in
SOP
METAB­
610.
Radiolabel
found
in
the
aqueous
fractions
represents
3H2O
formed,
and
that
in
the
methylene
chloride
fractions
represents
unreacted
substrate.

The
amount
of
estrogen
product
formed
was
determined
by
dividing
the
total
amount
of
3H2O
formed
by
the
specific
activity
of
the
[
3H]
ASDN
substrate
solution
(
expressed
in
dpm/
nmol).
The
activity
of
the
enzyme
reaction
was
expressed
in
nmol
(
mg
protein)­
1
min­
1
and
was
calculated
by
dividing
the
amount
of
estrogen
formed
by
the
product
of
mg
microsomal
protein
used
times
the
incubation
time,
e.
g.
30
min.
5
3
Results
and
Discussion
3.1
Substrate
Analysis
4­
Androstene­
3,17­
dione
(
ASDN),
lot
number
072K1134,
had
a
stated
purity
of
99%
(
Figure
1).
The
[
3H]
ASDN
coeluted
with
the
nonradiolabeled
ASDN
on
HPLC
(
Figure
2).
Information
provided
by
the
supplier
of
[
3H]
ASDN
regarding
its
purity
and
specific
activity
is
presented
in
Figure
3.
The
radiochemical
purity
of
the
[
3H]
ASDN
was
determined
by
HPLC
at
RTI
to
be
98%
(
Figure
4).
Samples
of
known
concentration
of
ASDN
were
analyzed
by
HPLC,
and
a
standard
curve
relating
peak
height
to
concentration
of
ASDN
was
generated.
A
sample
of
[
3H]
ASDN
was
analyzed
by
HPLC
in
triplicate
and
eluant
fractions
were
collected
and
assayed
for
radiochemical
content
by
LSS.
The
specific
activity
of
the
[
3H]
ASDN
stock
was
determined
by
dividing
the
dpm
in
the
peak
fractions
by
the
amount
of
ASDN
in
the
peak
(
calculated
using
the
peak
height
and
the
standard
curve
parameters)
.
The
data
are
presented
in
Table
1.
The
calculated
specific
activity
is
26.4
Ci/
mmol.
This
figure
is
within
5%
of
the
specific
activity
value
(
25.3
Ci/
mmol)
provided
by
the
supplier,
therefore,
25.3
Ci/
mmol
will
be
used
as
the
specific
activity
of
the
stock
[
3H]
ASDN
for
this
study.
The
data
presented
confirm
that
the
ASDN
and
[
3H]
ASDN
are
suitable
for
use
in
these
studies.

3.2
Human
Recombinant
CYP19
The
data
sheet
for
the
Human
Recombinant
CYP19
is
presented
in
Figure
5.
One
tube
of
this
product
was
thawed
rapidly
at
37
NC
and
the
contents
were
rehomogenized
and
analyzed
for
protein
and
P450
content
and
aromatase
activity.
The
protein
content
was
found
to
be
3.5
mg/
mL,
compared
with
the
4.2
mg/
mL
stated
on
the
data
sheet.
The
P450
content
was
calculated
to
be
0.38
nmol/
mg
protein.
This
value
is
similar
to
the
0.24
nmol/
mg
value
calculated
from
the
data
sheet
information.
This
microsomal
preparation
had
aromatase
activity
of
0.022
nmol
estrogen
formed/
mg
protein/
min
under
the
conditions
of
the
assay
as
described
above.
The
data
sheet
reported
an
aromatase
activity
value
(
1.38
nmol/
mg
protein/
min)
obtained
using
a
different
substrate
at
a
significantly
higher
concentration.
It
is
unclear
whether
the
activities
determined
under
such
different
conditions
should
be
similar.

3.3
Human
Placental
Microsomes
A
human
placenta
from
a
28
year
old
nonsmoker
with
a
full
term
Caesarean­
section
delivery
was
obtained
from
local
hospital
and
microsomes
were
prepared.
A
sample
of
the
microsomes
was
thawed
rapidly
in
a
water
bath
and
rehomogenized
prior
to
assay
for
protein
and
P450
content
and
aromatase
activity.
The
protein
content
of
the
human
placental
microsomes
was
determined
to
be
ca.
44
mg/
mL.
The
total
protein
yield
for
the
preparation
was
calculated
to
be
ca.
900
mg.
This
exceeds
the
250
mg
of
protein
criteria
set
in
the
protocol.
P450
content
of
the
human
placental
microsomes
was
determined
to
be
ca.
0.048
nmol/
mg
protein,
which
exceeds
the
criteria
of
0.005
nmol
P450/
mg
protein
set
in
the
protocol.
The
aromatase
activity
of
the
human
placental
microsomes
was
ca.
0.015
nmol
estrogen
formed/
mg
protein/
min;
this
exceeds
the
5
pmol
estrogen
formed/
mg
protein/
min
acceptance
criteria
for
this
parameter.
6
4
Conclusion
The
ASDN
and
[
3H]
ASDN
substrates
are
of
sufficient
purity
for
use
in
these
studies.
The
specific
activity
stated
by
the
supplier
of
the
[
3H]
ASDN
was
confirmed
experimentally.
Both
the
human
recombinant
CYP19
and
the
human
placental
microsomes
had
sufficient
protein
and
P450
content
for
the
conduct
of
these
studies.
The
aromatase
activity
for
the
two
microsome
preparations
was
similar
and
sufficient
to
proceed
with
the
optimization
phase
of
the
study.

5
References
Omura,
T.;
Sato,
R.
The
carbon
monoxide
binding
pigment
of
liver
microsomes.
I.
Evidence
for
its
hemoprotein
nature.
J.
Biol.
Chem.
1964,
239,
2370­
2378.
7
Figure
1
Data
Sheet
for
ASDN
8
Figure
2
HPLC
Radiochromatogram
of
ASDN
and
[
3H]
ASDN
9
Figure
3
Data
Sheet
for
[
3H]
ASDN
10
Figure
4
HPLC
Purity
Analysis
of
[
3H]
ASDN
11
Figure
5
Data
Sheet
for
Recombinant
Human
CYP19
12
Table
1
Determination
of
Specific
Activity
of
[
3H]
ASDN
Stock
HPLC
Run
#
ng
ASDN
in
peak
DPM
in
peak
DPM/
ng
ASDN
Ci/
mmol
Average
Ci/
mmol
1
2.008
408691
203531
26.3
26.4
2
2.000
411805
205903
26.6
3
1.994
406758
203991
26.3
