[CANCER RESEARCH 43, 3335-3347,
July 1983]
Isolation and Characterization of Endoplasmic Reticulum and Golgi
Apparatus from Hepatocyte Nodules in Male Wistar Rats1
Lennart C. Eriksson,2 Ulla-Britta Torndal, and Goran N. Andersson
Department of Pathology, Huddinge University Hospital, S-141 86 Huddinge, Sweden
ABSTRACT
A method is described for the isolation of endoplasmic reticulum and Golgi apparatus from hyperplastic liver nodules pro
duced by discontinuous feeding of 2-acetylaminofluorene to male
Wistar rats. The procedure involves three centrifugation steps
and permits the separation of these cell components and their
subfractions from the same sample of liver tissue with as little
as 1 g, wet weight. The fractions have been characterized by
chemical, enzymatic, and morphological techniques and were
found to be as pure as preparations from normal tissue. Fur
thermore, some of the characteristic histochemical features of
hyperplastic liver nodules have been quantitated by biochemical
methods in the fractions. Glucose-6-phosphatase activity in the
endoplasmic reticulum subfractions of nodules is approximately
15% of the corresponding value in normal livers, whereas the
activity of reduced nicotinamide adenine dinucleotide phosphate:
cytochrome c reducÃ-aseis reduced to 85% of the normal activity.
The amount of cytochrome P-450 in nodular membranes as
measured by differential spectroscopy is 25% of the control,
indicating a decreased Phase I activity in drug metabolism. A 5fold increase in cytosolic glutathione S-transferase activity with
out change in the corresponding microsomal activity was de
tected in hepatocyte nodules in rat liver. The activity of yglutamyltransferase is increased more than 20-fold in all mem
brane fractions prepared from nodular tissue. The cytosolic
activity, which is very low in the normal liver, is similarly increased
more than 20-fold. The membrane-associated 7-glutamyltransferase seems to be an integral membrane protein which cannot
be washed away from the membranes.
Chemically, membranes from nodules have phospholipidrprotein and cholesterol:protein ratios as found in mem
branes from normal liver tissue. However, the composition of
individual phospholipids is changed with a 2-fold increase in
nodular phosphatidylinositol and a slight decrease in phosphatidylcholine content in nodular membranes. The amount of endo
plasmic reticulum membranes is of the same magnitude as in
normal liver, although the smooth-surfaced component consti
tutes almost 60% of the isolated endoplasmic reticulum marker
enzymes in nodules, compared with only 32% in preparations
from normal tissue. The albumin contents of nodular and normal
microsomal and Golgi membrane preparations are similar, indi
cating a normal synthesis of albumin by nodular tissue.
INTRODUCTION
In attempts to understand chemical carcinogenesis, hyperplas
tic nodules appearing in rat liver after treatment with carcinogens
1This work has been supported by grants from the Swedish Medical Research
Council and the Swedish Society of Medical Sciences.
2To whom requests for reprints should be addressed.
Received February 15,1982; accepted April 8,1983.
JULY 1983
have attracted increased attention and have been extensively
reviewed. Until recently, the nodules have been examined mainly
by morphological and histochemical techniques by which various
differences from control tissue have been demonstrated (6, 17,
25, 41, 47).
Efforts to obtain more quantitative information at subcellular
levels have been limited mainly by the lack of adequate methods
for isolating subcellular fractions from nodular liver tissue. Some
quantitative biochemical data have been obtained from liver
homogenates or uncharacterized fractions (12, 31), but more
adequate techniques for membrane preparation are necessary
to examine the role of membranes in carcinogenic and cell
transformation processes.
When fractionating tumor tissue, it is tempting to use methods
developed and characterized for control tissue. With few excep
tions, it is not appropriate to transfer techniques for organelle
preparation from one tissue to another without modification and
extensive characterization.
Differences in cell composition,
stromal content, tissue rigidity, and vascularization as well as
the presence of macrophages, inflammatory cells, and necrosis
in tumors may affect the performance of procedures developed
for isolating various organelles from normal tissues. A thorough
characterization of each fractionation procedure must be con
ducted before fractions are compared with those obtained from
other sources.
In this paper, we present a simple sequential centrifugation
procedure for the preparation of rough and smooth microsomes
and light and heavy Golgi membranes from hyperplastic liver
nodules. The method yields the various fractions with high purity
and adequate recovery from small amounts of nodular tissue.
MATERIALS AND METHODS
Male Wistar rats (Mollegaards Avislaboratorium,
Ejby, Denmark)
weighing 140 to 160 g were used. The basal diet was brood stock food
for rats and mice (R3; Ewos, Sodertalje, Sweden). For production of
hepatocellular nodules, the rats were fed ad libitum with the basal diet
containing 0.05% 2-acetylaminofluorene according to the regimen sug
gested by Epstein ef al, (18) after slight modifications to better fit the
Wistar rats. The drug was bought from Fluka, Buchs, Switzerland, and
mixed into the diet by Ewos.
For 1 week, the rats were allowed to recover from transportation and
adjust themselves to the environment in the animal room where temper
ature, humidity, ventilation, and light were controlled according to inter
national standards. After this week of acclimatization, the rats were fed
the 2-acetylaminofluorene-supplemented
diet for 3 weeks followed by 1
week of basal diet, 2 weeks of 2-acetylaminofluorene diet, and 2 weeks
of basal diet. After this introduction, the rats were fed for 3 periods of 4
weeks, each consisting of 3 weeks of 2-acetylaminofluorene diet and 1
week of basal diet (Chart 1).
Rats with nodules were fed the basal diet for at least 2 weeks and
starved for 48 hr before harvesting of nodules. Control animals were fed
basal diet and were starved for 24 hr before sacrifice.
3335
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
L. C. Eriksson et al.
All animals were anesthetized
with an i.p. injection of Evipan-Natrium
(sodium hexobarbital; Bayer, Wuppertal, West Germany) (20 mg/100 g
body weight), and the livers were perfused in a retrograde direction by
cannulating the inferior vena cava caudal to the liver and clamping the
vena cava cranial to the diaphragm. After the portal vein and hepatic
artery had been cut, the liver was perfused with 0.15 M sodium chloride
at 37°(10 ml/100 g body weight). After the blood has been rinsed away,
perfusion continued with cold 0.25 M sucrose until the tissue was cool
(1 min).
Nodules, homogeneous in structure, color, and size and macroscopically well defined, were picked out with a pair of tweezers (Fig. 1).
Histological examinations of nodules shelled out by this procedure and
cleaned from surrounding tissue as indicated below showed the homo
geneous nodular structure as is well known and described earlier (i.e.,
Ref. 23). Necrotic regions or fibrous regions were not seen. The recovery
of light brown nodules from different rats varied between 0.5 and 4 g of
nodular tissue per liver. In the experiments where small, medium, and
large nodules were collected separately, small nodules measured up to
3 mm, medium nodules were 5 to 8 mm, and large nodules measured
larger than 10 mm in diameter. After careful cleaning from perinodular
tissue, the sample was weighed and minced in cold 0.25 M sucrose. All
preparative work was performed at 4°.The liver pieces (1 g/5 ml 0.25 M
10
weeks of treatment
15
20
Chart 1. Schematic presentation of the diet regimen designed for Wistar rats
for the production of hepatocyte nodules. 2AAF, 2-acetylaminofluorene at a con
centration of 0.05% mixed in basal diet.
Homogenate
sucrose) were homogenized
in a Potter-Elvehjem
glass:Teflon homoge-
nizer and homogenized with 4 strokes (one up and down constitutes
one stroke) at a pestle speed of 440 rpm.
For the preparation of microsomal and Golgi subfractions, the homogenate was centrifugated for 20 min at 10,000 x g (average). The
supernatant was decanted and processed as illustrated in Chart 2 (left)
for the preparation of rough and smooth microsomes as well as of Golgi
I, light Golgi vesicles.
The 10,000 x g pellet was gently resuspended in 0.25 M sucrose and
centrifuged for 10 min at 10,000 x g (average). The supernatant formed
was used for the preparation of Golgi II, heavy Golgi cistemae, as shown
in Chart 2 (fight). Control livers were processed in the same way.
This procedure, suitable for isolation of a series of membranous
organdÃ-es from the same piece of liver, was originally worked out for
control liver (4) and proved useful also for hyperplastic liver tissue.
As a reference fraction for the estimation of recovery, a total paniculate
fraction was prepared by centrifuging the homogenate at 105,000 x g
(average) for 90 min in a Beckman 50 Ti rotor. The pellet was suspended
in 0.25 M sucrose (1 g liver per 20 ml sucrose) and sonicated at 150
watts in ice: salt solution to keep the temperature at 4°.This procedure
did not inactivate enzymes in the fraction but rendered the suspension
homogeneous and the results reproducible.
Chemical Analysis. Protein was measured by the method described
by Lowry ef al. (33) with bovine serum albumin as standard. The amount
of albumin in vesicles of the endoplasmic reticulum and Golgi apparatus
was quantitated by immunoelectrophoresis
on 1% agarose gel plates
containing 0.75 ¿ilrabbit anti-rat albumin per sq cm (United States
Biochemical Corporation) and 1% Triton X-100. The electrophoresis was
performed in a Multiphore (Beckman) using 24 mw Veronal buffer, pH
8.6, also supplemented with Triton X-100 to a concentration of 1% (v/
v). The membrane fractions were dissolved in Triton X-100 (4 mg/mg
membrane protein) before application. The voltage was 2 V/cm, and the
(20% in 0.25M sucrose)
10.000 x g
(20 min)
Supernatant
Pellet
Resuspend
gentlyin
0.25Msucrose(20%)
10.000x g
(10min)
10ML
25MSUCROSEIO
0
10 ML 100009 SUPERNATANT
IN 02SM
SUCROSE
05 ML 0.6» SUCROSE
. 15 mM C»CI
2 OML 1 3M SUCROSE
•
iSmM C«CI
Õ
ML VIC M
SUCROSE'GOLGI
I —i
«l
n'j
1JOMm(SW27TSSUCROSE
I5KL IN
^èZê?
Supernatant
Pelet
(discard)
*<<&?
105.000x g
(90min)
Pellet
Resuspend
(l g liver/ml)
in l.20Msucrose
Supernatant
(discard)
1C ML 0 25 M
SUCROSE
10 ML 1.1SM
SUCROSE
89 000 »g
(SW 27J
PELLETIN I.ÃŽOM
SUCROSE
w
— MICROSOMES D
Chart 2. Schematic representation of the preparation procedure for rough microsomes, smooth microsomes (Sm), and Golgi I and Golgi II fractions.
3336
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Microsomes and Golgi Fractions
electrophoresis was run for 16 hr. Rat albumin (Sigma) was used as a
standard.
Lipids were extracted from membrane fractions by the procedure as
described by Folch ef al. (28). Phospholipids and neutral lipids were
separated by silicic acid column chromatography (10). Neutral lipids were
further separated into cholesterol, triglycérides, and free fatty acids by
thin-layer chromatography on silica gel plates according to the method
of Marinetti (35) with n-heptane:diisobutylketone:acetic
acid (85:15:1) in
the presence of 0.05% butylated hydroxytoluene as an antioxidant. Free
cholesterol and cholesterol esters were extracted with FeSO4-saturated
acetic acid and quantitated as described by Searcy et al. (44). Triglycér
ides were extracted and measured according to the method of Carlsson
(13).
Phospholipids were subfractionated by 2-dimensional thin-layer chro
matography on silica gel plates activated at 110°for 60 min. The solvent
systems were chloroform:methanol:acetic
acid:water (106:50:12:6) for
the first dimension
(40) and chloroform:acetone:methanol:acetic
acid:water (50:20:10:10:5) for the second dimension. The spots were
detected by iodine vapor. After evaporation of the iodine, the spots were
scraped off, and phospholipid phosphorus was determined by the
method described by Marinetti (34).
Enzyme Assays. NADPH:cytochrome c reducÃ-ase,glucose-6-phosphatase, mannose-6-phosphatase,
cytochrome P-450, and cytochrome
bs were measured as described previously (19). Succinatexytochrome
c reducÃ-ase, 0-glycerophosphatase
and AMPase were measured ac
cording to the method of Beaufay ef al. (7). UDP-galactosyltransferase
was measured using desialyated mucin as an exogenous galactose
acceptor in an assay recently developed in our laboratory (1). f-Glutamyltransferase in different subfractions was assayed according to the
method of Tate and Meister (46) using glycylglycine as acceptor. Glutathione transferase activities in serum, cytosol, and membrane fractions
were determined according to the method described by Habig et al. (29).
Electron Microscopy. Aliquots from the isolated microsomal and Golgi
subfractions were fixed in suspension immediately after preparation by
the addition of 3% glutaraldehyde buffer, pH 7.2, for 24 hr and then
sedimented at 12,000 rpm for 60 min in a SW 40 rotor (Beckman).
Following a brief rinse in 0.1 M cacodylate:0.1 M sucrose, the pellets
were cut and postfixed in 2% OsO4 in 0.1 M s-collidine buffer at pH 7.2,
dehydrated in a series of graded alcohols and propylene oxide, and
embedded in Epon 812 epoxy resin. The specimens were stained en
bloc with uranyl acetate. Ultrathin sections were cut with a diamond
knife, mounted on copper grids, and stained with lead citrate. Examina
tion was performed in the JEOL JEM 100 C electron microscope.
RESULTS
The amount of homogeneous hyperplastic liver tissue that
could be easily recovered free from surrounding liver tissue
reached a maximum between the 18th and 22nd weeks of the
dietary regimen. The livers were perfused with 0.15 M sodium
chloride and cold sucrose prior to dissection to avoid contami
nation by serum proteins and lipids and to preserve cell function
during the isolation of nodules.
Epstein ef al. (18) showed that the glycogen content in hyper-
plastic liver cells decreases only very slowly during starvation.
After 48 hr of starvation, the glycogen content reached the same
levels as those of normal rats after 24 hr without food. In our
preparative work, the recovery of microsomal protein increased
with decreasing glycogen in the cells, not shown in the figures
or tables. A maximum amount of microsomal protein was re
covered after 48 hr of starvation from nodular rats. Control rats
were starved for 24 hr to give similar glycogen content.
The centrifugation procedure used utilizes monovalent cations
for selective aggregation of rough endoplasmic reticulum, which
permits a fast separation from the nonaggregated smooth-sur
faced vesicles (14). The induced difference in particle size and
the difference in density between the 2 types of vesicles are the
basis for the separation, which is performed in a Beckman 40.2
fixed-angle rotor with a tube angle of 40°.Two facts make this
procedure particularly advantageous, (a) To increase selectivity
in aggregation and to avoid nonspecific clumping of membrane,
the CsCI was not mixed in the 10,000 x g homogenate. Instead,
an intermediate layer of 0.6 M sucrose containing 15 mw CsCI
was introduced in addition to the 1.3 M sucrose in the bottom of
the tube. This approach proved to be a simple solution to the
well-known problem of nonspecific aggregation ruining the fractionation. (b) The tube angle in the 40.2 rotor is as high as 40°
and permitted a further subfractionation not obtainable with other
rotors of the granulated vesicles with differing amounts of ribosomes on the vesicle surfaces. In Chart 2, these subfractions
have been indicated as R l, R II, and R III. The rough microsomal
subfractions from control liver have been described and charac
terized in earlier publications (19, 22).
Golgi membranes were floated from the smooth microsomal
fraction in a separate step using the techniques outlined by
Ehrenreich er al. (16). By choosing a 2-step sequential procedure
for subfractionation of rough and smooth microsomes and prep
aration of Golgi membranes, centrifugation conditions could be
selected to give a high purity of the different vesiculated mem
brane populations. Recovery of Golgi profiles was increased
significantly by resuspension of the 10,000 x g pellet as illus
trated in Chart 2. This fraction represented the heavier part of
the Golgi apparatus. The recovery of the endoplasmic reticulum
membranes both from nodules or control livers could be almost
doubled by repeating centrifugation in the discontinuous-cationcontaining sucrose gradient using supernatant from the resuspended 10,000 x g pellet as starting material. This procedure
has been described earlier for normal liver (20).
Chemical Characterization.
In Table 1 is shown the recovery
of protein in the different subfractions from nodular and control
livers. Sodium chloride (0.15 M) and sucrose perfusion increased
the liver weight by 20 to 25% on a protein basis, which makes
recovery look low in comparison to nonperfused livers (4), but
did not influence the isolation of subcellular fractions in any
Table 1
Recovery of protein and RNA in microsomal and Golgi membranesin control and nodular liver tissue
Protein
(mg/g
liver)Total
paniculate fraction
Rough microsomes
Smooth microsomes
Golgi I
Golgi IIControl"132.1
±12.1"
±14.3
6.10 ±1.21
9.10± 1.32
6.90 + 1.46
4.20 ± 0.61
0.41 ± 0.08
0.39 ±0.09
0.49 ±0.12RNA
0.49 ± 0.11Nodules131.0
liver)Control7.92
(mg/g
±1.18
±1.25
±0.004
±0.003
1.82 ±0.21
1.29 ±0.22
0.20 ±0.01
0.21 ±0.02
0.34 ±0.06Nodules7.30
0.34 ±0.05RNA:proteinControl0.060
0.070 ±0.009
0.080 ±0.007Nodules0.070
. of 6 experiments.
JULY 1983
3337
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
L. C. Eriksson et al.
respect. A striking feature in Table 1 is the relatively high amount
of smooth microsomal membranes recovered from nodular tis
sue. The RNA:protein ratios illustrate the distribution of ribosomes between the 2 fractions derived from endoplasmic reticulum. The values were not corrected for the amount of free
ribosomes which have been shown previously to be equally
distributed between the subfractions (19).
The relationship between rough and smooth microsomes in
preparations from hyperplastic liver tissue is further illustrated in
Table 2. Repeated resuspensions of 10,000 x g pellets were
used for the preparation of the 2 membrane fractions. In contrast
with the control tissue, where rough microsomal protein was
twice that of smooth microsomes, nodular smooth microsomal
protein exceeded the rough counterpart by 30%. The absolute
increase of smooth microsomal protein was from 5.9 to 9.3 mg/
g liver (57%). This result is in agreement with morphological
studies describing a proliferation of smooth-surfaced endo
plasmic reticulum in the nodular cells (for review, see Ref. 24).
The lipid composition of the endoplasmic reticulum membranes
is given in Table 3. Only minor variations in the amounts of total
phospho- and neutral lipids between control and nodular tissue
could be detected. These results are in agreement with results
of Merrit ef al. (36) but differ from those reported by Floyd ef al.
(27), where a substantial decrease in phospholipid content and
phospholipidiprotein and phospholipid:cholesterol ratios has
been suggested. The relative distribution of individual phospholipids is illustrated in Table 4, where an increase in phosphatidylinositol at the expense of the phosphatidylcholine content in
nodular tissue is obvious.
Enzymatic Characterization. Enzymeactivitiestogetherwith
electron microscopic appearance are widely used to describe
the composition of fractions and to calculate purity and contam
ination. The calculations often presuppose the existence of
marker enzymes which in the proper sense of the word do not
exist. On the contrary, more and more data indicate that most
enzymes are localized in more than one organelle. Nevertheless,
the distribution pattern and the combination of several enzyme
activities or amounts can be used to describe a membrane
fraction in functional terms.
In Table 5 are shown the activities of NADPH:cytochrome c
reducÃ-ase,glucose-6-phosphatase, and UDP-galactosyltransferase in preparations from perfused control livers and nodular
tissue prepared as described in Chart 2. The results indicate a
20% decrease in NADPH:cylochrome c reducÃ-aseactiviÃ-yand
8
8
J)
If)
9-c
Ö œ
ts
o
Table 2
Amount of microsomal protein recovered by repeated resuspension of 10,000 x g
pellet of nodular homogenate and the specific activities of NADPH:cytochrome c
reducÃ-ase
Preparations were performed as illustrated in Chart 1 for rough and smooth
microsomes. The resuspended 10,000 x g pellets were centrifuged at 10,000 x g
for 20 min to obtain Supernatants II, III, and IV. The values from one representative
experiment are shown.
Protein (mg/g liver)
No. of 10,000 x g
supernatants
Rough
microsomes
Smooth
microsomes
NADPHicytochrome c
reductase (fimol/min/mg
protein)
Rough
microsomes
Smooth
microsomes
IIIIIIIVI
111
+
+ II (control)5.401.701.601.287.1012.806.472.851.280.509.325.920.080.090.080.080.110.120.130.12
3338
o
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Microsomes and Golgi Fractions
Table4
Phospholipid composition of microsomes from control and nodular tissue
Fractions were prepared as described in "Materials and Methods."
phosphohpidsControl
% of total
liver)5.1
(mg/g
ethanolamine28.7
0± 0.56a
±0.24
microsomes
±3.1
±0.37
±0.38
±1.9
3.60 ±0.31Phosphatidylcholine59.0
Nodular microsomesPhospholipids
5.40 ±0.68Sphingomyelin3.50
53.2 ±3.0Phosphatidylinositol3.40
8.4 ±1.1Phosphatidylserine5.40
6.10 ±0.43Phosphatidyl 28.9 ±2.0
a Mean ±S.D. of 4 experiments.
as much as an 85% decrease in glucose-6-phosphatase activity
in nodular microsomes in comparison with control values. The
activity of UDP-galactosylasialomucin transferase measured by
the method developed in our laboratory (1) showed a slight
decrease in specific activity, although this is not significant. The
specific activities of "marker" enzymes in the different subtrac
tions are consistent with a high fraction purity, especially consid
ering the possibilities of indigenous NADPH:cytochrome c reduc
Ã-aseand glucose-6-phosphatase in Golgi vesicles (30).
Table 5 also permits calculations of recovery of the different
membrane subfractions from the total particulate fraction which
is defined as the 105,000 x g pellet of the liver homogenate.
When membrane-bound markers are used, the total particulate
fraction, in which the enzymes can easily be measured, is an
adequate reference preparation. If the total particulate fraction is
diluted to 5% with sucrose and sonicated gently, the preparation
is quite homogeneous and can be easily pipeted even with
micropipets. The substrate and/or protein amounts used in the
enzyme assays must be adjusted to obtain optimal assay con
ditions using homogenate or total particulate fractions. This is of
special importance when measuring galactosyltransferase where
the substrate is degraded by pyrophosphatases. Our results
from the work with hepatocyte nodules point to a slight but
measurable reduction in the enzyme activities recovered from
total particulate fraction in both microsomes and Golgi fractions.
Some enzyme activities are shown in Table 6 to give an
estimate of fraction purity and indication of the contamination by
mitochondria, lysosomes, and plasma membrane. The degree of
contamination shown in this way was low in all fractions and
practically identical in control and nodular preparations. It should
be stressed again that these enzymes are not true markers. For
instance, 0-glycerophosphatase and AMPase have been shown
to be endogenous components on the endoplasmic reticulum,
as well as in Golgi membranes (8, 26). From the values in Table
6, it can be seen that Golgi II is a special fraction with high
specific activities of both galactosyltransferase and AMPase.
These findings are hard to explain and might reflect plasma
membrane contamination, which would not be surprising since
the isolation procedure resembles that for preparation of blood
sinusoidal plasma membranes used by Evans (23, 48), but a
specific part of Golgi membranes with a high AMPase activity
cannot be excluded. Intermediate Golgi plasma membrane vesi
cles have recently been postulated by Evans.3
Electron Microscopic Characterization. Electron microscopic
examination of the rough microsomal fractions from nodules (Fig.
2a) revealed almost exclusively rough-surfaced vesicles. Other
profiles are very rare; occasionally, a lysosome-like vesicle can
be seen. In Fig. 2b, the appearance of the nodular smooth
3 H. Evans, personal communication.
JULY 1983
fraction which has been cleared from Golgi profiles by use of
flotation is shown. Even in this fraction, an occasional lysosomelike structure was present. Between the membrane vesicles are
ribosomes and ribosome clusters which the centrifugation pro
cedure did not exclude from the fraction. The fractions were
fixed in suspension, and when the aggregated materials were
spun down no layering of the pellet could be detected. The
photographs are representative of the whole fraction.
The 2 Golgi fractions from nodular tissue depicted in Fig. 3
are morphologically much more heterogeneous. In the Golgi I
fraction (Fig. 3a), lipoprotein-filled, rounded Golgi vesicles ap
peared together with empty-looking smooth vesicles and vesicles
filled with membranous material, some of them with a structure
resembling lysosomes or autophagic vacuoles. Some vesicles
appeared to be opened. Golgi vesicles are very fragile, as shown
by experiments where the Golgi fractions were spun down and
resuspended before fixation and morphological examination. Un
der such treatment, the vesicles seen are completely empty, and
most appear fragmented or opened (3).
The nodular Golgi II fraction (Fig. 30) which was floated on
1.15 M sucrose is dominated morphologically by cisternal lipidfilled Golgi profiles. Some open vesicles and sheets are also
present, together with small lipoprotein-filled spherical vesicles.
This fraction is very similar with respect to both preparation
procedure and density to Evans' fraction of hepatocyte plasma
membrane originating from the surface facing the space of Disse
(48). However, in our Golgi II fraction, where homogenization
and preparation have been performed in isotonic sucrose instead
of hypotonie carbonate buffer, few profiles are identified to be of
plasma membrane origin.
The electron microscopic appearance of the fractions obtained
from preneoplastic liver nodules fully corresponds to the appear
ance of membrane fractions prepared from control livers (2).
Examination of the Latent Activities of Mannose-6-phosphatase and Glucose-6-phosphatase. In an attempt to study
the intactness of the microsomal membranes, the latencies of
mannose-6-phosphatase and glucose-6-phosphatase activities
were determined (Table 7). Latency is expressed as the differ
ence between the specific activity of the enzymes using maximal
detergent stimulation and that without detergent, divided by the
specific activity at maximal detergent stimulation. Latency is
often used as an expression of the membrane permeability
barrier (11) such as, in our experiments, the permeability barrier
for the enzyme substrates. In spite of the difference in absolute
and specific activities between nodular and control membranes,
latency remains constant at 0.45 for glucose-6-phosphatase and
0.90 for mannose-6-phosphatase,
indicating no differences in
membrane permeability between control and nodular mem
branes.
Activities of Some Drug-metabolizing Enzymes. Nodular
liver cells are resistant to the cytotoxic effect of some mitoinhib3339
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
L C. Eriksson et al.
Table
Activities of microsomal and Golgi
Fractions were prepared as described in "Materials and Methods."
NADPH:cyctochrome
Glucose-6-
c reducÃ-ase
ControlTotal
Activity"7.65
±1.039
activity"5.89
Activity60.048
paniculate fraction
0.10
1.05
±0.11
Rough microsomes
0.17
0.860 ±0.091
Smooth microsomes
0.01 6
0.0080 ±0.001 3
Golgi 1
0.010 ±0.008Specific 0.018
Golgi IIAbsolute
a (imol cytochrome c reduced per min per g liver.
" //mol cytochrome c reduced per min per mg protein.
0 /imol P, released per 20 min per g liver.
d Mmol P, released per 20 min per mg protein.
e nmol galactose transferred per 30 min per g liver.
' nmol galactose transferred per 30 min per mg protein.
9 Mean ±S.D. of 7 experiments.
±0.003
±0.006
0.56
±0.002
0.59
±0.002
0.008
0.01 2
±0.002NodulesAbsolute
activity"0.045
±0.83
±0.062
±0.071
±0.001 9
±0.0023Specific
0.08
0.12
0.021
0.026
activity0413.4
±0.006
±0.01
70.80
±0.02
28.0
±0.003
0.40
±0.003ControlAbsolute
0.56
activity"2.60
±53.0
± 6.20
± 3.70
± 0.038
± 0.069Specific
7.04
5.60
0.80
1.08
±0.15
±0.38
±0.41
±0.10
±0.12
Table
Distribution of some enzymes in
Fractions were prepared as described in
reducÃ-aseControlTotal
Succinate:cytochrome
activity60.025
activity"3.93
c
activity*4.97
activity60.038
activity0286.6
activity"1.80
±0.46"
±0.0018
±0.68
±0.0041
paniculate fraclion
±34.0
0.014 + 0.0021
0.055 ±0.0081
0.0050 ±0.00036
0.0020 ±0.0001 8
Rough microsomes
9.91 ± 1.07
0.99
0.0030 ±0.00064Specific 0.0006 ±0.00008NodulesAbsolute
0.004 ±0.0006Specific 0.0008 ±0.000100-Glycerc-ControlAbsolute
5.14 ± 0.64
Smoolh microsomes
1.02
Golgi 1
2.31 ± 0.18
4.53
0.54 ± 0.07Speäfic0.94
Golgi IIAbsolute
a »¿mol
cytochrome c reduced per min per g liver.
0 /irnol cytochrome c reduced per min per mg protein.
P, released
P, released
P, released
'/imol P, released
±0.13
±0.11
±0.14
±0.76
±0.12
in 30 min per g liver.
in 30 min per mg protein.
in 20 min per g liver,
in 20 min per mg protein.
9 Mean ±S.D. of 4 experiments.
Table 7
Effects of detergents on the activities of glucose-6-phosphatase
and mannose-6-phosphatase
The assays were performed in the presence of 0.05% deoxycholate or without detergent. Latency, expressed as
Activity with detergent - activity without detergent
Activity with detergent
is 0.45 for glucose-6-phosphatase
and 0.91 for mannose-6-phosphatase
in control and nodular tissue.
Specific activity
Glucose-6-phosphatase"
Nodules
Control
detergent3.94
±0.62"
±0.38
Rough micro
somes
3.1
2
±0.81Detergent7.04
5.60
±0.41No
Smooth micro
somesNo
a nmol P, released in 20 min per mg protein.
6 »¿mol
P, released in 5 min per mg protein.
c Mean ±S.D. of 4 experiments.
Mannose-6-phosphatase
Control
0.0810.44
±
±0.11
±0.0062
±0.073Detergent1.06
0.80 ±0.10No
itors and grow in the presence of 2-acetylaminofluorene. In
recent publications (5, 21) and in the Tables 8 to 13, we have
presented data from measurements of some drug-metabolizing
enzymes in an attempt to explain the mechanism of this resist
ance.
The amount of cytochrome P-450 as determined by differential
spectroscopy in nodular membranes was markedly decreased.
It appears from Table 8 that only 25% of the specific P-450
amount was left in rough microsomes as well as in smooth
membranes. The corresponding value for cytochrome b5 was
3340
Nodules
detergent60.041
detergent0.58
detergent0.024
±0.033
0.030 ±0.0068Detergent0.48
0.37 ±0.032No
±0.0038
±0.019
0.01 9 ±0.0041Detergent0.26
0.20 ±0.021
about 85% of the control amount.
The activity of glutathione S-transferase has been shown by
Morgenstern ef al. (37, 38) to be an endogenous constituent in
microsomal membranes, as well as an abundant cytosolic pro
tein. In Table 9, the distribution of glutathione S-transferase in
microsomal and Golgi membranes is shown, together with the
activity in the cytosol. The cytosolic activity represented 96% of
the total enzyme activity in the cell. The remaining activity was
associated with membrane structures even after repeated TrisHCI washing, most of it in the microsomal fractions, but a certain
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL.
43
Microsomes and Golgi Fractions
membrane marker enzymes
UDP-galactosyltransferase
phosphatase
Nodules
Control
Specific activity1*
Absolute activity0
62.88
8.707.42
+
1.403.92
+
0.440.05
±
0.0080.08
±
+ 0.0160.48
Nodules
Absolute activity6
0.031.06
±
0.110.80
+
0.100.14
+
0.020.17
±
±0.021298
Absolute activity6
Specific activit/
13129.6+
±
3.152.1
4.8108.4±
±
12.490.7
± 10.810.20
Specific activit/
15631
+
3.975.8
.3 ±
8.985.2
±
9.180.4
+
± 8.912.20
±1.104.71
0.5117.42
+
±1.71387
42335 ±
± 401339
±1.205.13
±0.8215.54
±2.20368
47321 +
± 46
membranesphosphatasaeAMPaseNodulesAbsolute
m/crosomal and Golgi
activity0162
activity"1
activity8164.63
+ 21.0
.24 ±0.091
0.70 ±0.060
4.84 ± 0.69
0.71 + 0.073
4.83 ± 0.73
1.84+ 0.24
1.53 ±0.12
2.01 ±0.16ControlAbsolute 6.0 ± 0.81Specific
±26
4.92 ± 0.58
3.48 ± 0.47
0.55 ± 0.076
0.98 ± 0.12Specific
activity8131
activit/1.04
+ 0.081
±22
0.48 ±0.043
2.73 ± 0.34
6.71
+
1.03
0.96 ±0.10
3.60 ±0.31
1.45± 0.21
10.20 ±1.70NodulesAbsolute3.85 ± 0.59Specific
activity'1.0
±0.093
0.39 ±0.043
1.37 ±0.12
3.73 ±0.56
7.85 + 1.31
Tables
Amounts of cytochrome P-450 and cytochrome D5
Fractions were prepared as described in "Materials and Methods."
Cytochrome P-450
activity85.98
Cytochrome 05
Nodules
Control
activity"0.63
activity0.92
Control
activity0.1
activity4.62
Nodules
activity0.52
activity3.09
activity0.44
±0.83C
+ 0.064
+ 0.042
±0.14
8 ±0.020
±0.73
±0.38
±0.034
Rough microsomes
0.21 + 0.024Absolute
2.26 ±0.34Specific
2.35 ±0.31Specific
0.48 ±0.036
3.77 ±0.49Specific0.76 ±0.086Absolute1.04 ±0.18Specific
0.50 ±0.067Absolute
Smooth microsomesAbsolute
8 nmol/g liver.
" nmol/mg protein.
c Mean ±S.D. of 5 experiments.
Table 9
Activities oÃ-glutathione S-transferase in different subtractions from control and nodular liver tissue
The membrane fractions are Tris-washed twice.
Glutathione S-transferase (/imol/mm)
Control
liver7.24
g
Total paniculate fraction
Rough microsomes
Smooth microsomes
Golgi I
Golgi II
105,000 x g supernatantPer
± 0.54s
1.02 + 0.092
0.63 ± 0.061
0.008 ± 0.0010
0.040+ 0.0043
174
±21Per
Nodules
protein0.090
mg
0.24
0.23
0.08
0.14
1.54
+ 0.0083
+0.021
+0.017
+0.012
±0.017
+0.19Per
liver28.84
g
± 3.64
0.91 ± 0.14
0.80 ± 0.12
0.032 ± 0.0051
0.15 ± 0.024
432
±51Per
protein0.31
mg
±0.033
0.27 + 0.031
1.1
0.31 ±0.028
1.2
0.25 + 0.027
3.1
0.47 ±0.051
3.3
7.21 ±0.64Nodules:control3.5
4.7
a Mean + S.D. of 3 experiments.
activity could be detected in the Golgi fractions. In nodular tissue,
the cytosolic activity was markedly increased, as was the activity
in Golgi membranes, while the activity of glutathione S-transfer
ase in microsomes was unchanged. There is a possibility that
the increased activity of glutathione S-transferase in Golgi mem
branes may represent cytosolic contamination which has not
been eliminated by alkaline buffer washings, but an indigenous
"cytosolic" form of the enzyme in Golgi apparatus cannot be
excluded and could represent a step in the biogenesis of the
JULY 1983
cytosolic enzyme.
Distribution of 7-Glutamyltransferase.
The -,-glutamyltrans-
ferase activity was clearly increased in hepatocyte liver nodules.
In order to study the localization of this increase in activity, the
subcellular distribution of -y-glutamyltransferase in control and
nodular liver tissue was monitored as was the total activity in
pooled nodules of different sizes (Table 10). The large majority
of the enzyme activity was membrane-associated as is shown in
Table 11. Repeated washings and hypotonie incubations did not
3341
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
L. C. Eriksson et al.
TabtelO
activities in different liver cell subfractions and in homogenates from hyperplastic liver nodules of different
sizes
Fractions were prepared as illustrated in Chart 1. Data from small, medium, and large nodules represent one experiment using pooled nodules.
Distribution of y-glutamyltransferase
•y-Glutamyltransferase
Control
liver321
56s10.21
±
Nodules
protein2.74
liver7020
protein56
±0.181.13
965211
±
7.847 ±
±0.0823.26
2.3416.14
±
35291
±
6.985 ±
2.731.42±
±
±0.216.17
4826±
±10185
0.646.23
±0.7018.54
7.4110±
±
29552 ±
±2.650.12
1.237.0
±
28162
682.7±
± 0.53nmol/min/mg
±0.0110.030
34485062055832nmol/min/mg
±
0.360.50
±
±0.0026nmol/min/g
0.11263332control2143Z730292317
±
Total
fractionRough
paniculate
microsomesSmooth
microsomesGolgi
IGolgi
II105,000
supernatantSerumSmall
x g
mm)Medium
nodules (<2
mm)Largenodules (4-8
nodules (>10 mm)nmol/min/g
3 Mean ±S.D. of 3 experiments.
Table 11
Effect of Tris:water:Tris washing on the specific activity of i-glutamyltransferase
in nodular membrane fractions
protein)FractionsRough
Activity8 (nmol substrate consumed/min/mg
microsomes
Smooth microsomes
Golgi 1
Golgi IINo
treatment47.0
± 6.9"
85
185
552
±10
±29
±68Tris
wash65.0
139
238
672
wash84.0
± 7.1
±11
168
±27
260
±71Tris:water:Tris
731
protein)Rough
± 7.8
±14
±29
±76
The total activities in the different subfractions remained unchanged through
the washing procedure.
b Mean ±S.D. of 3 experiments.
release the enzyme, indicating that the membrane-bound
form
of the enzyme is an integral membrane protein. Only a minor
part of the total activity could be recovered in the microsomal
and Golgi fractions. Calculations based on the hypothetical situ
ation of 100% recovery of endoplasmic reticulum and Golgi
membranes, less than one-half of the -y-glutamyltransferase was
found in those membranes. The residual activity might be asso
ciated with the plasma membranes as shown in histochemical
studies, which indicates an extremely high specific activity in the
plasmalemma. This suggestion was supported by the very high
values in the Golgi II fraction, which might be contaminated by
plasma membrane fragments. In nodular tissue, the distribution
pattern was largely unchanged in spite of a tremendous increase
in activity. The differences in the degree of induction between
the different subfractions were small, with the possible exception
of the rough endoplasmic reticulum fraction.
As is also shown in Table 10, pooled nodules of different size
classes have been analyzed with respect to 7-glutamyltransferase activity. Nodules larger than 4 mm in diameter are essentially
identical in enzyme activity. Lesions less than 3 mm in diameter
show a slightly lower activity, which may be due to technical
difficulties in harvesting small nodules free from perinodular
tissue. These measurements indicated that very small nodules
should be excluded in this preparative work. Other enzyme
activities and amounts have also been analyzed in individual
nodules uniform in size with essentially the same conclusion.
Although the examination of small foci in the treated livers is of
great interest for studies of early steps in carcinogenesis, this
model system is not appropriate for their isolation.
Albumin Content in Normal Liver and Hepatocyte Nodules.
The major protein synthesized in the normal liver for transport
3342
Table 12
Amount of albumin in membrane preparations from control and nodular tissue
Fractions were prepared as illustrated in Chart 1 and washed twice in 0.15 M
Tris-HCI, pH 8.O. Albumin was measured by rocket immunoelectrophoresis with
Triton X-100 in gel and membrane suspension. Values are the means of 2
experiments. Deviation between the 2 experiments less than 15%.
microsomes
Smooth microsomes
Golgi I
Golgi IIControl2.4
Albumin (ng/mg total fraction
2.9
25.8
30.4Nodules3.1
3.4
1.2
24.9
1.00.9
26.5Nodulesicontrol1.3
to the blood is albumin. The treated livers consisted mostly of
nodular liver tissue. In an attempt to determine the source of the
albumin, its content was measured in normal and nodular
subfractions.
The amount of intravesicular albumin in microsomal and Golgi
membrane preparations is shown in Table 12. In spite of the
changes in cell and tissue structure and function with, among
other things, a significant production of a-fetoprotein (39), the
nodular fractions contained albumin to the same extent as do
the control. This indicates that albumin synthesis and transport
take place in the nodule cells. However, for the quantification of
albumin secretion kinetic experiments must be performed.
Perinodular or Surrounding Liver Tissue Concept. In studies
of the carcinogenic processes using this model system, the ideal
control tissue is not the liver from an untreated control rat but
rather the so-called perinodular tissue that surrounds the nod
ules. Histological examination of a liver from an animal treated
with carcinogens for 20 weeks reveals very small amounts of
normal liver tissue. Most of the liver is occupied by nodules of
different sizes. Even the tissue that appears free of nodules to
the naked eye contains small nodules. The perinodular tissue is
therefore extremely difficult to isolate for routine use as control
tissue.
In an attempt to determine whether the properties of the
nodules were unique for this tissue or if they showed similarities
with all hepatic cells, we performed experiments in which nodules
and perinodular tissue were dissected from the same liver. The
perinodular tissue could not be isolated in pure form due to
technical difficulties in eliminating all small nodules even when
early stages of the process (12 to 15 weeks) were used. The
activities of different nodular marker enzymes in nodules and
perinodular tissue are illustrated in Table 13. The activities in
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Microsomes and Golgi Fractions
Table 13
Activities of some nodular marker enzymes in total paniculate fraction from control, nodular and "perinodular"
liver tissues
The "perinodular tissue" was prepared by careful dissection but was not completely free from small nodules
due to technical difficulties. Nodules and "perinodular tissue" are from the same liver.
Stransferase" (nmol
CDNB6 conjugated/
Control
Perinodular tissue
Treated animal with
few nodules
Nodulesl-Glutamyl-
transferase (nmol/
min/g
liver)299
min/mg soluble
protein)1.54
±31
2776 ±460
221
5738390
1 ±
±0.21
2.81 ±0.34
2.1
±0.358.50
2
±586Glutathione
P-450
(nmol/g
liver)37.2
phosphate (iimol P,
released/20 min/g
liver)413
±4.3
23.5 ±3.2
5.39.40
35.3 ±
±0.93Cytochrome
±1.32Glucose-6-
+ 44
228 ±34
261
38143+
±
16.8
Activity was measured in the 105,000 x g supernatant.
6CDNB, 1-chloro^^-dinitrobenzene.
c Mean ±S.D. of 3 experiments.
normal liver and in livers from 2-acetylaminofluorene-treated
rats
with extremely few nodules are given for comparison.
The data clearly show distinct differences between nodules
and perinodular tissue, with the values for the surrounding tissue
in between the nodular and the normal liver. However, these
results show that the changes in the nodules were characteristic
for this hyperplastic nodular tissue and not a common reaction
of all the liver cells to 2-acetylaminofluorene in the diet.
DISCUSSION
A series of events which occur in the sequential process of
tumor development is associated with pericellular or intracellular
membrane structures. In studies of these mechanisms, there is
an obvious need for an adequate fractionation procedure of
different subcellular membranes. In this paper, we have de
scribed a fractionation technique which permits the preparation
of subfractions of endoplasmic reticulum and subfractions of the
Golgi apparatus from hyperplastic liver tissue with a high degree
of purity and a reasonable recovery. The fractions are sequen
tially prepared from the same tissue homogenate with a 3-step
centrifugation procedure which is fast (less than 6 hr) and repro
ducible. The method, which is adaptable to small amounts of
starting material (less than 1 g), may be used to subfractionate
single large nodules.
The isolated fractions could be characterized enzymatically
and morphologically and compared to control fractions prepared
in the same manner. By using a combination of several mem
brane-bound enzymes and their distribution pattern between the
different subfractions, the membrane preparations have been
described, and the degrees of fraction purity and recovery have
been determined.
In comparison with control tissue, the enzyme patterns for
nodular subcellular fractions are almost identical with few excep
tions. The recovery of endoplasmic reticulum and Golgi mem
branes are, however, slightly reduced.
Electron microscopic examination indicates a high fraction
purity and intactness which is also illustrated by studies of
membrane permeability.
Histochemically demonstrated markers characteristic for no
dular tissue (6,17, 24, 41, 47) have in this paper been confirmed
biochemically and quantitated. These markers include the in
crease in smooth endoplasmic reticulum, the decrease in glucose-6-phosphatases and cytochrome P-450 and the dramatic
increase of 7-glutamyltransferase. In addition, we have found a
JULY 1983
2-fold increase in the amount of phosphatidylinositol
in nodular
membranes. In a series of recent publications, phosphatidylino
sitol has been described as a regulatory phospholipid, the con
tent of which is clearly associated with membrane structure and
function (32,42,43). The significance of this substantial increase
in the relative phosphatidylinositol amount in nodular tissue with
respect to membrane function is not yet known. The slight
decrease of phosphatidylcholine in the microsomal membranes
is also of interest, especially in view of the fact that a cholinedeficient diet stimulates the selective growth of the carcinogenaltered cells in the nodules (45).
By careful dissection of nodular livers early in the regimen, it
is possible to collect so-called perinodular or surrounding liver
tissue. From the results presented in this work, we have drawn
2 conclusions: (a) the cells in the nodules have unique properties
different from those of the perinodular cells which have been
treated in the same way; (o) it is very difficult to isolate pure
perinodular tissue, inasmuch as even the most careful prepara
tion will yield a mixture of nodular and perinodular tissue. We
believe that it is preferable to use identically treated nonnodular
liver cells as control, but with this technique it is very time
consuming and inexact. Such a control could, however, be used
for some specific questions.
In a recent publication, De Gerlache ef al. (15) have shown
that cell suspensions from nonnodular parts of nodular livers are
intermediate in all respects between cells from nodules and cells
from normal livers. One interpretation of their results is that the
perinodular cell suspension contains contaminating nodular cells;
the possibility that the perinodular cells really differ from normal
cells is of course obvious and cannot be excluded. Still, their
technique is useful and important for analysis of the carcinogenic
potential of nodular cells.
According to the theory formulated by Farber (25), one of the
most crucial properties for the clonal growth of the initiated cells
and for the development of nodules is their acquired resistance
to the cytotoxic effects of the mitoinhibitors used as selectors in
the model system. This resistance may theoretically be due to a
decreased uptake of the drug and/or a decreased formation
intracellularly of toxic metabolites of the mitoinhibitor. In nodular
tissue, the amount of cytochrome P-450 was observed to be
markedly decreased. Furthermore, the activity of glutathione Stransferase was increased 5-fold over control values. In experi
ments performed in collaboration with Aström and DePierre (5)
and with Bock eÃ-al. (9), we have also shown an increase in
epoxide hydrolases and UDP-glucuronyltransferase,
respec3343
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
L. C. Eriksson et al.
tively. These findings strongly support the idea that the selective
growth of initiated cells may be due to the low monooxygenase
activity and the very high capacity for degradation and conjuga
tion of toxic metabolites within the nodular cells. These findings
do not exclude the possibility of specific Phase I reactions
producing carcinogenic metabolites. Whether or not the bioactivation process is involved in the increased sensitivity of this cell
population for the carcinogenic effect of 2-acetylaminofluorene is not known.
ACKNOWLEDGMENTS
The authors wish to thank Professor Anders Bergstrand and Liv Grontoft for
preparing the electron micrographs.
REFERENCES
1. Andersson, G. N., and Eriksson, L. C. Characterization of UDP-galactosyl
asialo-mucin transferase activity in the Golgi system of rat liver. Biochim.
Biophys. Acta, 570. 239-247. 1979.
2. Andersson. G. N., and Eriksson, L. C. Endogenous localization of UDPgalactose-asialomucin galactosyltransferase activity in rat liver endoplasmic
reticulum and Golgi apparatus. J. Biol. Chem., 256: 9633-9639, 1981.
3. Andersson, G. N., Eriksson, L. C., and Bergstrand, A. F. Membrane damage:
an electronmicroscopical and biochemical study. J. Ultrastruct. Res., 76: 341342,1981.
4. Andersson, G. N., Torndal, U-B., and Eriksson, L. C. Sequential preparation of
rat liver microsomal and Golgi membranes. Biochim. Biophys. Acta, 572: 539549, 1978.
5. Äström,A., DePierre, J. W., and Eriksson, L. C. A preliminary characterization
of drug-metabolizing systems in preneoplastic nodules from the livers of rats
receiving 2-acetylaminofluorene in their diet. Acta Chem. Scand. Ser. B, 219220,1981.
6. Bannasch, P., Mayer, D., and Hacker, H. J. Hepatocellular glycogenesis and
hepatocarcinogenesis. Biochim. Biophys. Acta, 605: 217-245, 1980.
7. Beaufay, H.. Amar-Costesec, A., Feytmans, E., Thines-Sempoux, D., Wibo,
M., Robbi, M., and Berthet, J. Analytical study of microsomes and isolated
subcellular membranes from rat liver. I. Biochemical methods. J. Cell Biol., 67:
188-200,1974.
8. Bergeron, J. J. M., Ehrenreich, J. W., Siekevitz, P., and Palade, G. E. Golgi
fractions prepared from rat liver homogenates. II. Biochemical characterization.
J. Cell Biol., 59:73-88, 1973.
9. Bock, K. W., Lilienblum, W., Pfeil, H., and Eriksson, L. C. Increase in UDPglucuronyl transferase activity in hyperplastic liver nodules and Morris hepatoma. Cancer Res., 42: 3747-3752, 1982.
10. Borgstrom, B. Investigation on lipid separation methods. Separation of phospholipids from neutral lipids and fatty acids. Acta Physiol. Scand., 25: 101110,1952.
11. Bretz, R., and Stäubli,W. Detergent influence on rat-liver galactosyltransferase
activities towards different acceptors. Eur. J. Biochem., 77: 181-192,1977.
12. Cameron, R., Sweeney, G. D., Jones, K., Lee, G., and Farber, E. A relative
deficiency of cytochrome P-450 and aryl hydrocarbon [benzo(a)pyrene] hydroxylase in hyperplastic nodules induced by 2-acetylaminofluorene in rat liver.
Cancer Res., 38: 3888-3893, 1976.
13. Carlsson, L. A. Determination of serum triglycérides. J. Atheroscler. Res., 3:
334-336, 1963.
14. Dallner, G. Studies on the structural and enzymic organization of the membra
nous elements of liver microsomes. Acta Pathol. Microbiol. Scand., Suppl.
166,1-94,1963.
15. De Gerlache, J., Lans, M., Taper, H., and Roberfroid, M. Separate isolation of
cells from nodules and surrounding parenchyma of the same precancerous rat
liver: biochemical and cytochemical characterization. Toxicology, 78: 225-232,
1980.
16. Ehrenreich, J. H., Bergeron, J. J. M., Siekevitz, P., and Palade, G. E. Golgi
fractions prepared from rat liver homogenates. I. Isolation procedure and
morphological characterization. J. Cell Biol., 59: 45-72, 1973.
17. Emmelot, P., and Scherer, E. The first relevant cell stage in rat liver carcinogenesis. A quantitative approach. Biochim. Biophys. Acta, 605: 217-304,
1980.
18. Epstein, S., Ito, N., Merkow, L., and Farber E. Cellular analysis of liver
carcinogenesis: the induction of large hyperplastic nodules in the liver with 2fluorenylacetamide or ethionine and some aspects of their morphology and
glycogen metabolism. Cancer Res., 27:1702-1711,1967.
19. Eriksson. L. C. Studies on the biogenesis of endoplasmic reticulum in the liver
3344
cell. Acta Pathol. Microbiol. Scand. Sect. A Pathol. Suppl. 239, 1-72, 1973.
20. Eriksson, L. C. Preparation of liver microsomes with high recovery of endo
plasmic reticulum and a low grade of contamination. Biochim. Biophys. Acta,
508:155-164,19/8.
21. Eriksson, L. C., Äström,A., DePierre, J. W., and Bock, K. W. Induction of
drug-metabolizing enzymes in preneoplastic liver nodules. Abstr. Biochem.
Soc., 9. 271, 1981.
22. Eriksson, L. C., and Dallner, G. Isolation of ribose-poor rough microsomes
from rat liver. FEBS Lett., 79: 163-165, 1971.
23. Evans, W. H. Fractionation of liver plasma membranes prepared by zonal
centrifugation. Biochem. J., 776: 833-842, 1970.
24. Farber, E. Hyperplastic liver nodules. Methods Cancer Res., 7:345-375,1973.
25. Farber, E. The sequential analysis of liver cancer induction. Biochim. Biophys.
Acta, 605: 149-166. 1980.
26. Farquhar, M. G., Bergeron, J. J. M., and Palade, G. E. Cytochemistry of Golgi
fractions prepared from rat liver. J. Cell Biol., 60: 8-25, 1974.
27. Floyd, R. A., Steward, J., Hampton, M. J., Wisemann, B. B.. and Griffin, M. J.
Membrane changes during development of hyperplastic nodules during hepa
tocarcinogenesis. Abstracts of the Eleventh International Congress of Bio
chemistry, Toronto, 05-4-R116, p. 391, 1979.
28. Folch, J., Lees, M., and Stanley, G. H. S. A simple method for the isolation
and purification of total lipids from animal tissue. J. Biol. Chem., 226: 497509, 1957.
29. Habig, W. H., Pabst, M. I., and Jakoby, W. B. Glutathione^S-transferase: the
first enzymatic step in mercapturic acid formation. J. Biol. Chem., 249: 71307139,1975.
30. Howell, K. E., Ito, A., and Palade, G. E. Endoplasmic reticulum marker enzymes
in Golgi fractions—what does this mean. J. Cell Biol., 79: 581-589, 1978.
31. Lin, J-C., Hiasa, Y., and Farber, E. Preneoplastic antigen as a marker for
endoplasmic reticulum of putative premalignant hepatocytes during liver car
cinogenesis. Cancer Res., 37: 1972-1981,1977.
32. Low, M. G., and Finean, B. J. Specific release of plasma membrane enzymes
by a phosphatidylinositol-specific
phospholipase C. Biochim. Biophys. Acta,
508: 565-570. 1978.
33. Lowry. O. H., Rosebrough, N. J.. Farr, A. L.. and Randall, R. J. Protein
measurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275,
1951.
34. Marinetti. G. V. Chromatographie separation, identification and analysis of
phosphatides. J. Lipid Res., 3: 1-20, 1962.
35. Marinetti, G. V. Chromatography of lipids on commercial silica loaded paper.
J. Lipid Res., 6: 315-317, 1965.
36. Merrit, W. D., Keenan, T. W., and Morré,D. J. Gangliosides and other lipids of
hyperplastic liver nodules induced by W-2-fluorenylacetamide: Cancer Biochem.
Biophys., 7: 179-185, 1976.
37. Morgenstern, R., DePierre, J. W., and Ernster, L. Activation of microsomal
glutathione-S-transferase
activity by sulfhydryl reagents. Biochem. Biophys.
Res. Commun., 87: 657-663, 1979.
38. Morgenstern, R., Meijer, J., DePierre, J. W., and Ernster, L. Characterization
of rat-liver microsomal glutathione-S-transferase
activity. Eur. J. Biochem.,
704: 167-174, 1980.
39. Okita, K., Gruenstern, M., Klaiber, M., and Farber, E. Localization of afetoprotein by immunofluorescence in hyperplastic nodules during hepatocar
cinogenesis induced by 2-acetylaminofluorene. Cancer Res., 34: 2758-2763,
1974.
40. Parker, F., and Peterson, N. F. Quantitative analysis of phospholipids and
phospholipid fatty acids from silica gel thin layer chromatograms. J. Lipid Res.,
6: 455-459, 1965.
41. Pilot, H. C., and Sirica, A. E. The stages of initiation and promotion in
hepatocarcinogenesis. Biochim. Biophys. Acta, 605:191-215,
1980.
42. Quist, E. E.. and Reece, K. L. The role of diphosphatidylinositol in erythrocyte
membrane shape regulation. Biochem. Biophys. Res. Commun., 95: 10231030, 1980.
43. Ratman, S., Fraser, I. H., and Mookerjea, S. Effect of phosphatidylinositol and
phosphatidylserine
on membrane-bound
galactosyl-transferase.
Can. J.
Biochem.. 58: 58-66, 1980.
44. Searcy, R. L., Bergquist, L. M., and Jung, R. C. Rapid ultramicro estimation
of serum total cholesterol. J. Lipid Res., 7: 349-351, 1960.
45. Shinozuka, H., Sells, M. A., Katyal, S. L., Sell, S., and Lombardi, B. Effects of
a choline-devoid diet on the emergence of 7-glutamyltranspeptidase-positive
foci in the liver of carcinogen-treated rats. Cancer Res., 39: 2515-2521,1979.
46. Täte,S. S., and Meister, A. Interaction of -y-glutamyltranspeptidase with amino
acids, dipeptides and derivatives and analogs of glutathione. J. Biol. Chem.,
249:7593-7602,1975.
47. Williams, G. M. The pathogenesis of rat liver cancer caused by chemical
carcinogens. Biochim. Biophys. Acta, 605: 167-189, 1980.
48. Wisher, M. H., and Evans, W. H. Functional polarity of the rat hepatocyte
surface membrane. Isolation and characterization
of plasma-membrane
subfractions from the blood sinusoidal, bile canalicular and contiguous surface
of the hepatocyte. Biochem. J., 746: 375-383, 1975.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Microsomes and Golgi Fractions
1d
Fig. 1. Pieces of liver and isolated nodules from a rat treated for 20 weeks with the 2-acetylaminofluorene
0, fresh liver tissue; c and d, hematoxylin-eosin-stained
JULY
diet regimen described in "Materials and Methods." a and
sections of formalin-fixed liver pieces and isolated nodules, a, x 1.5; b, x 2; c and d, x 7.5.
1983
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
3345
L. C. Eriksson et al.
5l'
Fig. 2. Electron microscopic appearance of isolated rough microsomes (a) and smooth microsomes (b). The fractions were fixed in glutaraldehyde, postfixed in osmium
tetroxide. and embedded in Epon, and sections were stained with uranyl acetate as described in "Materials and Methods.' a, x 10,000; b. x 6,600.
3346
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Microsomes and Golgi Fractions
>"*-•*;>'_•.
fw
i T»'
Ã--t*
Ã-X°^^"
JÄ'^fiöiw
Fig. 3. Electron microscopic appearance of isolated Golgi I (a) and Golgi II (o) membranes. Fractions were isolated and processed for electron microscopy as described
in the legend to Fig. 2. x 6,600.
JULY
1983
3347
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.
Isolation and Characterization of Endoplasmic Reticulum and
Golgi Apparatus from Hepatocyte Nodules in Male Wistar Rats
Lennart C. Eriksson, Ulla-Britta Torndal and Göran N. Andersson
Cancer Res 1983;43:3335-3347.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/43/7/3335
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1983 American Association for Cancer Research.