Expression
of
Liver Phenotypes
in
Cultured
Mouse Hepatoma Cells
1.2
Gretchen J. Darlington, 3 Hans Peter Bernhard, 4 Richard
A.
Miller, 5 and Frank
H.
Ruddle 6
ABSTRACT-Mouse
hepatoma cells were established
In
vitro
as
a permanently growing line designated Hepa. The
mall
popula-
tion and a subclone were characterized for their karyotype and
their retention of liver-specific properties.
An
examination of 17
hepatic traits revealed that the cell lines secreted
several
serum
proteins. The actIvities
of
a number of liver-specific enzymes,
however, appeared
to
be
absent
In
these cells. The Identification
of differentiated properties of cultured hepatoma cells permits
the
use
of
these
lines
In
a variety of studies
such
as
cell hybridi-
zation, biochemical
analysis
of tillue-speclfic
gene
products, and
the modulation of expression of genes governing differentiated
phenotypes. This report
presents
the analysis
of
a broad spectrum
of
characteristics
and
thereby describes one
of
the
most
fully
defined hepatoma cell
lines
of murine origin
In
the
IIterature.-
JNCI
64:
809-819, 1980.
Cultivation
of
somatic
ceI1s
in
vitro
is
an
advanta-
geous system for the study of organ-associated func-
tions.
When
serially
propagated
outside
the
body,
somatic
cells
obtained
from
various
tissues often fail to
express specific functions. However, a
number
of
mam-
malian
cell lines have
now
been described
that
retain
characteristics of the tissue of
origin.
Cell lines from
hepatomas
include
several derived from rat
tumors
(1)
and
the
line
reported
by Szpirer
and
Szpirer (2) from
the mouse.
In
addition
to the study of tissue-specific
phenomena
such
as enzyme
induction,
differentiated cell lines
have
been utilized
in
the
production
of
hybrid
cells.
The
interaction
of
differentiated
and
undifferentiated ge-
nomes
and
the expression of
various
organ-associated
traits
in
hybrid
populations
have
been the subjects
of
several
investigations
(3).
It
is
our
purpose
to describe a cell line, isolated from
a
mouse
hepatoma,
that
expresses liver characteristics.
At the
time
of analysis, this cell line,
Hepa,
and
a
cloned
subpopulation,
Hepa-I,
had
TAT
activity
in-
ducible
by corticosteroids.
They
synthesized
and
se-
creted several
serum
proteins
including
albumin,
ceru-
loplasm,
transferrin,
and
AFP.
These
cells
had
activity
for esterase-2
and
pseudocholinesterase,
although
other
enzymes
such
as aldolase
B,
alcohol
dehydrogenase,
and
XOX
were
not
found.
The
chromosome
comple-
ment
was
also
found
to differ from
that
of
the
normal
mouse
karyotype.
MATERIALS AND METHODS
Medium.-Dry
powdered
medium
MAB 87/3 (4) was
obtained
from
GIBCO,
Grand
Island,
New York.
Penicillin
(50,000 U),
streptomycin
(50,000 p.g),
and
FCS (100
ml)
were
added
to 900
ml
of
medium.
FCS
was
obtained
from
twO
sources:
GIBCO
and
Flow
Laboratories
(Rockville, Md.).
809
Isolation
of
Hepa
cells.-Male
C57L/J
mice
carrying
a
subcutaneous
hepatoma
(BW7756) were
obtained
from
The
Jackson
Laboratory,
Bar
Harbor,
Maine.
The
serially
transplantable
BW7756
hepatoma
has been
termed a
"minimal
deviation
tumor"
(5).
The
tumor
was excised
from
a mouse
under
sterile
conditions
when
the
cancer
became a
sphere
approximately
2
em
in
size.
The
tissue was
minced
into
pieces
1-2
mm
at
the
longest
axis
and
then placed
in
a 25-ml Erlen-
meyer flask
containing
15
ml
of
a
pancreatic
enzyme
solution
[tradename,
Viokase (GIBCO)]
and
a
small
magnetic
stirring
bar.
The
tissue
chl!nks
were
further
dispersed by
agitation
of
the
spinning
bar
for 20
minutes.
The
solution
containing
Viokase,
mono-
dispersed cells,
and
small
tissue
fragments
was
then
transferred from the Erlenmeyer flask to a tube
and
centrifuged. Pelle ted cells were resuspended
in
medium
MAB 87/3,
plated,
and
allowed
to
grow
for 1 week.
After
that
time a 0.5-ml
aliquot
containing
approxi-
mately
10
6
cells was
inoculated
ip
into
each of 2
C57L/J
males.
Tumors
appeared
in
2-3
weeks
and
were
not
encapsulated.
When
the
tumors
reached the
appropriate
size (2
cm
in
diameter), they were
minced
and
cultured
as above.
The
cells were passaged from
animal
to
in
vitro
culture
six times
over
a
period
of
6
months.
This
procedure
for cell selection was described
by Buonasissi et al. (6). After passage
6,
the cells,
designated
Hepa,
grew
in
culture
with
a
generation
time
of
about
30
hours.
We
generated
colonies
by
plating
50 cells
per
60-mm
2
Falcon
culture
petri dish.
One
colony,
Hepa-I,
was selected for
further
study.
Cell
lines.-Other
cell lines were used as
controls
and
are
referred to
in
various
experiments.
WI-38
is
a
human
diploid
cell
line
obtained
from the
American
Type
Culture
Collection,
Rockville,
Maryland.
RAG
is
derived from a
BALB/c
murine
renal
adenocarcinoma
ABBREVIATIONS
tiSEO,
AChE
= acetylcholinesterase.
AFP
= alpha-feto-
protein;
AHH=aryl
hydrocarbon
hydroxylase;
FCS=fetal
calf
serum;
TAT=
tyrosine
aminotransferase;
XOX
=
xanthine
oxidase.
I Received May
14,
1979; accepted
September
19,
1979.
2
Animals
were
maintained
under
the
guidelines
set forth
by
Yale
University, New
Haven,
Conn.
1
Department
of Medicine,
The
New York
Hospital-Cornell
Uni-
versity Medical Center, 525 East 68th St., New York, N.Y. 10021.
4
Department
of Cell Biology, Biocenter, University of Basel,
Basel, Switzerland.
l Sydney
Farber
Cancer
Center,
44
Binney St., Boston, Mass. 02115.
6
Department
of Biology, Yale University, New
Haven,
Conn.
06520.
JNCI,
VOL.
64.
NO.4,
APRIL
19HO
810
Darlington, Bernhard, Miller,
and
Ruddle
(7).
LM(TK-)
is derived
from
the
mouse
L-cell
and
lacks
thymidine
kinase activity (8).
KOP
is a
human
diploid
fibroblast
line
(9). H
4
AzC
2
,
a
subclone
of a
rat
hepatoma
line
originally
isolated
by
Richardson
et
al.
(10), was
supplied
by Dr.
Armen
Tashjian,
Harvard
University,
Cambridge,
Massachusetts.
Strain
501-1 is
a
mouse
fibroblast
line
which,
like
LM(TK-),
was de-
rived
from
L-cells.
Testing
for
mycoplasma
is
routinely
done
with
the use of
Hoechst
33258.
Histochemical
analysis for intracellular
glycogen.-
Hepa
cells were
grown
on
microscope slides
and
stained
with
periodic
acid-Schiff
following
the proce-
dures
outlined
by
Merchant
et
al. (11). Slides were
examined
under
bright-field
and
phase
illumination.
Karyotype
studies.-
The
general
methods
for
chro-
mosome
preparation
from
cultured
cells were previ-
ously
described (12).
Starch gel
electrophoresis.-
The
general procedures
for
starch
gel electrophoresis were described by Brewer
(13). Cell samples were frozen
prior
to
their
use
and
stored as pellets. Shortly before electrophoresis the
samples
were
thawed
and
homogenized
in
distilled
water
at
a
concentration
of 50Xl0
6
cells/ml
unless
noted
otherwise.
XOX
(EC
1.2.3.2).-Activity
for
XOX
was observed
by
histochemical
staining
of starch gels
according
to
the
method
of Yen
and
Glassman
(14).
The
staining
solution
was
modified
by the
omission
of NAD from
the
reaction
mixture.
The
stock gel buffer
(Tris-
ethylenedinitrilo
tetra-acetic
acid-borate
buffer)
has
been described (15).
Alcohol
dehydrogenase (EC
1.1.2.1).-Alcohol
dehy-
drogenase
isoenzymes were separated
with
the use of
the
electrophoretic
conditions
described above for
XOX.
The
staining
mixture
contained
3
ml
of
70%
ethanol,
0.6 mM NAD,
0.1
mM
phenazine
methosulfate,
0.43
mM
nitro
blue
tetrazolium, 0.012 N
potassium
cyanide,
and
0.035 M
Tris-HCl
(pH
8.0).
Esterase-2.-
The
procedures
for
this
assay
have
been
previously
reported (16).
Aldolase
(EC
4.1.2.13).-Following
the
procedure
of
Omenn
and
Cohen
(17),
Hepa
and
Hepa-l
were
examined
for the presence of aldolase activity.
TAT
activity (EC
2.6.1.5).-TAT
activity was mea-
sured
by the
method
of
Granner
et
al. (18).
Protein
determinations
were
performed
as described by
Lowry
et
al. (19). Specific
TAT
activity was
examined
in
Hepa-l
cells
in
the
presence of varying
amounts
of
dexamethasone
(Sigma
Chemical
Co., St.
Louis,
Mo.)
to
determine
the
most
effective dose for enzyme
induc-
tion.
Cells were
plated
in
25-cm
2
flasks
at
a
density
of
2.6XlO5
per
bottle.
The
cells were
allowed
to
attach
and
grow
for
72
hours.
At
that
time
flasks were divided
into
groups
of
three,
and
each
group
received
medium
containing
one
of the
following
concentrations
of
dexamethasone:
0,
lO-5
M,
lO-6
M,
lO-7
M,
lO-8
M,
or
lO-9
M.
Twenty-six
hours
after
exposure
to
the
test
medium,
the cells were harvested
and
the
contents
of
each
bottle were. assayed for specific
TAT
activity.
Induction
of
TAT
over a 48-hour
period.-
To
test
JNCI.
VOL. 64.
NO.4.
APRIL
1980
the effects of cell
density
on
the
inducibility
of
TAT
by
dexamethasone,
Hepa-l
cells were assayed periodically
for 48
hours.
Replicate
flasks were
plated
with
4XlO
5
cells
in
MAB
87/3
medium
without
dexamethasone.
After
24
hours,
3 flasks were
sampled
for base-line
TAT
activity,
and
the
remainder
was fed
dexametha-
sone (10-
5
M)
in
fresh
medium.
Antisera.-Monospecific
antimouse
AFP
is a
goat
antiserum
prepared
against
electrophoretically
purified
fetoproteins 1-3 derived from
mouse
amniotic
fluid
and
was
the
gift
of
Dr. Ernest
Zimmerman,
University
of
Cincinnati,
Cincinnati,
Ohio.
The
antiserum
gives a
single
precipitin
on
immunoelectrophoresis
of
amnio-
tic fluid
and
does
not
react
with
adult
mouse
serum.
Commercially
prepared
rabbit
antisera
to
mouse
fibrogen. C3
(third
component
of
complement),
and
transferrin were
purchased
from
Cappel
Laboratories,
Inc.
(Cochranville,
Pa).
Antigens.-Amniotic
fluid from
14l7-day
mouse
em-
bryos was
also
sent
to us by Dr.
Zimmerman.
Super-
natant,
serum-free, tissue
culture
medium
containing
the secreted
proteins
of H
4
AzC2,
KOP,
501-1,
and
Hepa-l
was collected
in
the
following
way.
Confluent
monolayers
were washed twice
with
serum-free MAB
87/3,
then
overlaid
with
lO
ml
of
this
medium.
After
24-48
hours,
the
supernatant
medium
was
concentrated
in
a
Schleicher-Schuell
membrane
(Arthur
H.
Thomas
Co.,
Philadelphia,
Pa).
These
concentrates
were frozen
until
use
in
polyacrylamide
electrophoresis
and
im-
munologic
procedures.
Secreted cell
proteins
were
radiolabeled
with
[14C]leu-
cine as follows. Serum-free
medium
minus
leucine
to
which
10
/-LCi
[14C]leucine/ml
had
been
added
was
placed
over
plateau-phase
cells
and
incubated
at
37°
C
for 24
hours.
The
supernatant
medium
was collected,
added
to
carrier
protein,
and
tested by the
immunoelec-
trophoresis
method
of
Laurell
(20).
Autoradiography
of
the
precipitin
peaks
permitted
visualization
of the
radio
labeled cell
products.
Aerylamide
gel
electrophoresis.-Electrophoretic
an-
alysis was
done
on
Hepa-l-secreted
proteins
with
the
use of
ORTEC
(Ortec, Inc.,
Oak
Ridge,
Tenn.)
slab
acrylamide
gels by the
procedure
of
Tischfield
et al.
(2I)
with
a
7.65%
separating
gel
and
a
4.5%
spacer gel.
The
voltage levels were 280 V for 1 gel
or
320 V if 2
gels were
being
run
simultaneously.
The
acrylamide
slabs were removed from the gel
mold
and
fixed for 30
minutes
in
12%
trichloroacetic
acid
if the gel was to be
stained
for
protein.
If
the gel
was
to
be
stained
for
ceruloplasmin
or
cholinesterase
(as described below),
no
prior
fixation
was done.
Serum
proteins.-Protein
bands
were
stained
with
Coomassie
brilliant
blue.
Prior
to
electrophoretic
anal-
ysis, some
samples
were pretreated
with
neuraminidase
(Sigma
Chemical
Co.), as described by
Gustine
and
Zimmerman
(22).
One
volume
of
purified
neuramini-
dase,
1.1
U/mg,
200 /-Lg/ml
in
35
mM
sodium
acetate
(pH
5.0), was added to
one
volume
of
Hepa-l-secreted
protein
mixture
or
to
mouse
serum
or
amniotic
fl
uid
as described below.
The
samples
were
incubated
at
37°
C for 40
minutes,
and
the
reaction
was
stopped
by
immersion
of
the
samples
in
ice
and
addition
of
0.5
volumes
of
1.25 M
Tris
base.
Ceruloplasmin.-Ceruloplasmin
was detected by the
method
of
McCombs
and
Bowman
(23)
with
the use
of
o-dianisidine
(Sigma
Chemical
Co.)
in
acetic acid.
The
orange
bands
of
activity were visible
within
1-2
hours.
Cholinesterase.-After
electrophoresis, cholinesterase
isoenzymes were identified by
the
direct
coloring
method
of
Karnovsky
and
Roots
(24)
with
the use
of
acetylthio-
choline
iodine
or
butyrylthiocholine
iodide (Sigma
Chemical
Co.) as substrates.
The
cholinesterase
inhibi-
tor eserine (Schwarz-Mann,
Orangeburg,
N.Y.) was
used
at
10-
4
M,
and
the
AChE
inhibitor
BW284C51 (a
gift
of
Burroughs
Wellcome, Inc., Beckenham, Kent,
England)
was used
at
10-
5
M;
when
inhibitors
were
used,
the
gel was
incubated
with
inhibitor
for 30
minutes
before the substrate was included.
Double
diJfusion.-Double-diffusion
analysis was
performed
on
microscope slides coated
with
1%
agarose
(Seakem,
Rockland,
Maine)
in
barbital
buffer
(pH
8.6)
with
an
ionic
strength
of
0.025 M. Each well
contained
about
5
1-'1
of
antigen
or
antiserum.
RESULTS
Cell Morphology and Growth Characteristics
Populations
of
Hepa-I
cells
kept
in
logarithmic
phase
doubled
every 24
hours.
The
generation
time
of
Hepa
was 30
hours.
The
cell
morphology
of
Hepa
and
Hepa-I
(fig. I)
was
epithelioid.
The
nuclei
contained
many
nucleoli,
and
the
cytoplasms
were
granular,
making
the
nuclear
outlines
very
apparent.
Histochemical
staining
of
Hepa
cells for stored
glycogen was negative.
Karyotype Analysis
Table
I
compares
the
mean
chromosome
numbers
of
Hepa
and
Hepa-1.
In
general,
populations
of
Hepa-I
had
less
variability
in
the total
number
of
chromo-
somes
per
cell.
The
distribution
in
the
Hepa-I
subpop-
ulation
was clustered
around
the
mode
of
67, whereas
the
numbers
in
the
parental
Hepa
population
ranged
from
43
to
151
with
a very weak
mode
of
66 (text-fig.
I).
Two
to
four
biarmed
chromosomes
appeared
in
both
Hepa
and
Hepa-I
cells.
TABLE
I.-Comparison
of
the
mean chromosome numbers
of
Hepa and Hepa-l
No.
of chromosomes
Cell
No.
of
line
Mean Range Mode
Biarmed
cells counted
chromosomes
Hepa 81.0
43-151
66
2
69
Hepa-1 65.9 49-129
67
5
70
Liver
Phenotypes
In
Mouse Hepatoma Cells
811
HEPA
'Of
NU::ER
: 1
III
••
I~'
••
i
iMi'~~
....
,~
...
1 • I
"I
I I I I
P"
I
I~
I
••
'I
I I
I~
Pi
I
"1
CELLS
I
iol
t
HEPA-'
:
,.
"~.IIII,,.
"'"'0'''
.,,,
••
,,, •. ,
"'~'"
.,,, .. ,.
".,
~
00
~
ro
~
~
00
I~
IW
OOI~
100
NUMBER
OF
CHROMOSOMES
TEXT·FIGURE
I.-Histogram
of
chromosomes in
Hepa
and
Hepa·1.
Secreted Proteins
Serum
proteins.-Serum-free
medium
collected from
cultures
of
Hepa-I
and
analyzed by
polyacrylamide
gel
electrophoresis
contained
at
least 20
electrophoretically
distinguishable
protein
components.
In
addition
to
albumin
[ef. (25)], we have been
able
to identify AFP,
ceruloplasmin,
and
transferrin
among
the secreted
proteins.
Our
identifications
were based
on
immuno-
logic,
electrophoretic,
and
biochemical
techniques. Fig-
ure 2 shows the
positions
of
AFP
and
transferrin
secreted by
Hepa-I
after
polyacrylamide
slab gel elec-
trophoresis.
Gustine
and
Zimmerman
(22) showed
that
amniotic
fluid from
14l-2-day
mouse
embryos
contained
three
principal
regions
of
protein:
an
albumin
region
moving
just
behind
the dye front, a set of five
AFP
bands
behind
albumin,
and
a set
of
three transferrin
bands
that
migrated
least
anodally.
They
also
demon-
strated
that
the
faster members of each set were derived
in
vivo by the
addition
of sialic acid residues to the
slower
bands
and
that
neuraminidase
treatment
of
amniotic
fluid
prior
to electrophoresis
partially
con-
verted the
fast-migrating
bands
into
the slower species.
Figure
2 shows the results of
neuraminidase
treatment
of
supernatant
medium
from
Hepa-I
(channel
4).
On
the basis of
mobility
and
response to removal by sialic
acid residues, we were able to identify the
bands
indicated
as
AFP
(fetoproteins 3,
4,
and
5)
and
trans-
ferrin.
No
AFP-like
protein
was
present
in
samples of
adult
mouse
serum
(channels
I
and
2) regardless of
neuramidase
pretreatment.
The
predominant
transfer-
rin
species of
adult
mouse
serum
migrated
to the same
electrophoretic
position
as
did
the fastest species seen
in
Hepa-I
and
in
amniotic
fluid.
In
addition,
transferrin
and
AFP
could
each be
identified by
double-diffusion
analysis
in
agarose. Fig-
ure 3A depicts a reaction of
immunologic
identity
among
antigens
common
to
adult
mouse
serum,
Hepa
and
Hepa-I
secretions,
precipitated
by a
rabbit
anti-
mouse
transferrin
serum
that
gave a
single
band
when
tested by
immunoelectrophoresis
against
adult
mouse
serum. FCS
and
an
aliquot
of
concentrated
medium
were
included
as negative
controls
to
demonstrate
that
the
antiserum
would
not
detect
either
FCS
contamina-
tion
or
some
artifact
of
our
concentration
technique.
lNCI.
VOL. 64.
NO.4.
APRIL
1980
812
Darlington, Bernhard, Miller,
and
Ruddle
Figure
3B shows a
reaction
of
immunologic
identity
between
antigens
present
in
mouse
amniotic
fluid
and
in
the
supernatant
medium
from
Hepa
and
Hepa-l
cells
that
were
precipitated
by
anti
mouse
AFP
sera.
Albumin,
AFP,
and
transferrin
were synthesized
and
secreted by
the
Hepa
and
Hepa-l
cells.
Figure
4
illustrates
autoradiographs
of
Laurell
immunoelectro-
phoresis
plates
on
which
was tested
supernatant
me-
dium
containing
secreted
proteins
labeled
with
[14C]leucine.
Precipitin
peaks
contained
labeled
pro-
teins for each
serum
protein
when
media
from
Hepa
and
Hepa-l
were used (wells 1
and
2).
No
peak
was
seen
on
the
autoradiograph
above well
3,
which
contained
supernatant
medium
from
RAG
cells for
albumin
and
AFP. Howeve.r,
RAG
cells
produced
a
small
amount
of transferrin.
Ceruloplasmin.-Ceruloplasmin
was
identified
on
the basis
of
its characteristic
orange-staining
reaction
with
a-dianisidine.
Figure
5
demonstrates
that
the
stained
protein
in
Hepa-l
supernatant
was
similar
in
mobility
to
a
protein
found
in
adult
mouse
serum.
The
slightly
slower
mobility
of the
material
produced
by
Hepa-l
was
reproducible
from
run
to
run.
The
lot
of
FCS
used for cell
growth
did
not
contain
histochemi-
cally detectable
ceruloplasmin.
Attempts
to
demonstrate
the presence of
fibrinogen
or
of
C3
in
freshly
concentrated
supernatant
from
Hepa-l
cultures
by
immunoelectrophoresis
were unsuc-
cessful.
Cholinesterase.-Mammalian
sera
contained
several
isoenzymes of
pseudocholinesterase
(EC 3.1.1.8)
capable
of
hydrolyzing
acetylcholine
or
butyrylcholine
in
addi-
tion
to
other
esters. Pseudocholinesterase (EC 3.1.1.8)
was synthesized
principally
in
the liver
and
was distin-
guishable
from the
AChE
(EC
3.1.1.
7)
that
is character-
istic of
cholinergic
synapse by the resistance of pseudo-
cholinesterace
to
the
inhibitor
BW284C51
and
by its
more
rapid
hydrolysis
of
butyrylcholine
esters. We
examined
Hepa-l-conditioned
medium
for its
ability
to
hydrolize esters of
thiocholine;
the results are presented
in
figure
6.
Normal
mouse
serum,
channels
4,
6,
8,
and
10,
had
several
bands
of cholinesterase activity
when
acetyl-
thiocholine
was used
as
substrate.
Of
these,
two
bands
(labeled
with
an
asterisk
in
channel
4,
fig. 6)
appeared
to be
AChE
inasmuch
as
both
disappeared
in
the
presence of the
AChE
inhibitor
BW284C51
(channel
6)
and
failed to utilize
butyrylcholine
as substrate (chan-
nel
8).
The
remaining
bands
were classified as pseudo-
cholinesterases;
channel
10
shows
that
all
were eserine-
sensitive.
Hepa-l,
channels
3, 5,
7,
and
9,
produced
at
least
four
electrophoretically
distinguishable
regions of
cho-
linesterase activity
with
acethylthiocholine
as substrate.
One
band
(identified by
an
asterisk, fig. 6)
could
be
classified as
an
AChE
on
the
grounds
that
it
was
much
less active
either
when
butyrylthiocholine
was
provided
as substrate
(column
7)
or
when
BW284C51 was
included
with
an
acetylthiocholine
substrate
(column
5).
The
other
three
bands
appeared
to
be
pseudocho-
JNCI.
VOL. 64.
NO.4.
APRIL
1980
linesterases.
These
three
bands
exhibited
electropho-
retic
mobilities
similar
but
not
identical
to those of
pseudocholinesterases
found
in
adult
mouse
serum
(column
4). In
addition,
the
more
rapidly
migrating
bands
of
normal
mouse
serum
pseudocholinesterase
activity were
not
observed
in
the
Hepa-l
supernatant.
The
cholinesterases of FCS are
shown
in
column
10
of figure 6.
Inasmuch
as
at
least
two
of these
bands
did
not
occur
in
the
Hepa-l
concentrated
medium,
we
believe
that
there
had
been
no
significant
contamina-
tion
by FCS
and
thus
that
all
four
cholesterase
bands
in
the
Hepa-l
supernatant
had
been synthesized by the
cultured
cells.
The
Hepa-l
AChE
migrated
slightly
less
rapidly
than
did
the
corresponding
band
of
serum
AChE
from
normal
mouse
serum,
and
the
Hepa
pseudocholinester-
ase
migrated
slightly
less
rapidly
than
did
the strongest
serum
pseudocholinesterase.
These
relationships
were
reproducible
from
experiment
to
experiment.
TAT Activity
In
Hepa-1 Cells
TAT
activity was
measured
in
the absence of dexa-
methasone
during
lag,
logarithmic,
and
plateau
phases
of the
growth
of the
culture.
The
greatest specific
TAT
activity
in
the absence of
dexamethasone
was observed
immediately
after
attachment
while
the cells were still
in
the
lag
phase
of
growth.
During
logarithmic
growth
and
plateau,
the activity declined
somewhat.
The
induction
of
TAT
by
dexamethasone
was exam-
ined
shortly
after
attachment,
i.e.,
in
lag
and
early log
phase
rather
than
when
the cells were
approaching
confluence.
TAT
activity
in
Hepa-l
cells
during
that
DEXAMETHASONE
TEXT-fIGURE
2.-
TAT
activity
in
cells
grown
in
dexamethasone
at
various
concentrations.
Specific activity (S.A.) is expressed as
nmol
a-ketoglutarate
converted
per
milligram
cell
protein/min
at
25°
C.
35
25
15
10
5
o
6
12
18
24
HOURS
30
36
42
48
TEXT·FIGURE
3.-Kinetics
of
TAT
induction
with
or
without
dexa·
methasone.
Hepa·)
cells
in
the
presence of 10-' M
dexamethasone
(e);
Hepa·)
cells
in
the absence
of
dexamethasone
(0).
period
was increased by the
addition
of
this
synthetic
corticosteroid
fourfold to sixfold over the basal activity.
For
the
determination
of
the
most
effective
concen-
tration
of
inducer,
Hepa-I
cells were
grown
in
varying
doses of
dexamethasone
as described
in
"Materials
and
Methods."
Text-figure
2
shows
that
10-
6
dexametha-
sone
and
10-
5
M
dexamethasone
were
equally
effective
in
induction,
both
concentrations
evoking
a fourfold
elevation
of
TAT
activity above the
control
level.
Dexamethasone
at
a
10-
9
M
concentration
had
little
or
no
effect
on
enzyme activity, whereas
at
10-
8
and
10-
7
M,
TAT
reached
partially
induced
levels.
The
increase
in
TAT
activity for
48
hours
following
the
addition
of
10-
5
M
inducer
to the
medium
is
illustrated
in
text-figure
3.
The
cells
in
the
control
medium
and
in
the
medium
containing
dexamethasone
both
showed
an
increase
in
TAT
activity
at
6
hours.
The
enzyme activity of
the
cells
growing
in
the
presence
of
dexamethasone
continued
to rise to a
peak
of
30
nmol/mg
cell
protein/minute
at
48
hours,
at
which
time
the
experiment
was
terminated.
In
the
absence
of
dexamethasone,
the
TAT
activity leveled off
after 6
hours
and
increased
only
slightly.
The
treated
culture
had
a final increase
of
fivefold above the
untreated
cells.
AHH Activity
The
activity of AHI;I was
measured
in
Hepa-I
cells
by Dr.
Daniel
Nebert
(National
Institute
of
Child
Liver Phenotypes
In
Mouse Hepatoma Cells 813
Health
and
Human
Development, Bethesda, Md.)
and
co-workers
and
was presented
in
a
separate
report
(26).
AHH
activity
could
be elevated
in
Hepa-l
by the
addition
of
phenobarbital,
aromatic
hydrocarbons,
and
biogenic
amines.
Inducibility
by the first
compound
appeared
to be a liver-specific trait.
Starch Gel Electrophoresis
Activity for
aldolase
A (the form characteristic of
muscle) was observed
in
Hepa,
Hepa-l,
and
the BW7756
tumor.
Mouse
aldolase
A
migrated
anodally,
whereas
the B-form
of
aldolase, seen
in
the liver, moved
toward
the cathode.
Neither
Hepa
nor
Hepa-l
had
aldolase
activity
corresponding
to
the
B-form
of
the
mouse
enzyme. An
additional
band,
the C-form, was
most
prominent
in
brain.
One
heteropolymeric
band
of
A-
and
C-subunits
was observed
in
the
cultured
cells (data
not
shown).
Other
enzyme activities tested by
starch
gel
techniques
were
XOX
and
alcohol
dehydrogenase.
Neither
of these
enzymes was detected
in
Hepa
or
Hepa~l
cells. Ester-
ase-2,
an
anodally
migrating
esterase
band
controlled
by the Es-2 locus, was
found
to be
present
in
both
Hepa
and
Hepa-l.
DISCUSSION
The
chromosome
numbers
of
Hepa
cells
extended
over a
wider
range
than
did
those of the
subpopulation
Hepa-l,
although
variation
in
the total
chromosome
number
was
apparent
in
Hepa-l
as
well.
The
mean
number
of
chromosomes
in
Hepa-I
cells was less
than
that
in
Hepa
cells, whereas the
number
of
biarmed
chromosomes
increased. Possibly,
some
of
the
biarmed
chromosomes
resulted from
centric
fusion of telocen-
trics; those
of
Hepa
appeared
from
their
banding
patterns
to
be isochromosomes.
The
production
of
albumin
by
rat
hepatoma
cells
in
culture
has
been observed by several investigators
(27,
28).
Our
results
showed
that
Hepa
and
Hepa-I
secreted
ceruloplasmin,
transferrin,
and
AFP
in
addition
to
albumin.
In
contrast
to
the Fu-5 cells
of
Bertolotti
and
Weiss (29), the enzymatic traits of
Hepa
and
Hepa-I
(alcohol dehydrogenase, esterase-I,
TAT
inducibility,
aldolase
B,
and
XOX)
were
diminished
or
absent,
whereas the
production
of
serum
proteins
was
more
extensive
and
has
remained
stable over several years
in
culture.
Pseudocholinesterase
acUvItles
in
mouse
serum
showed
some differences from the
electrophoretic
pat-
terns
found
in
human
serum
on
starch gel (30). In
our
acrylamide
system, we resolved
10
bands
of
pseudocho-
linesterase activity
in
normal
mouse
serum
as
opposed
to 7
bands
in
man.
These
additional
bands
may reflect
species differences
and/or
increased
resolution
by
poly-
acrylamide
electrophoresis.
The
hepatoma
cells
did
not
produce
the
entire
spectrum
of
cholinesterases
that
were observed
in
serum; however,
both
AChE
and
pseudocholinesterase
JNCI,
VOL
64,
NO.4,
APRIL
1980