[CANCER RESEARCH 47, 210-220, January 1, 1987]
Elevated Expression and Cell Cycle Deregulation of a Mitosis-associated Target
Polypeptide of a Carcinogen in Hyperplastic and Malignant Rat Hepatocytes1
Sam Sorof2 and R. Philip Custer
Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111
ABSTRACT
A cytoplasmic polypeptide with a molecular weight of 14,000 (pl4)
was previously found to be the principal covalent target protein of the
carcinogen, /V-2-fluorenylacetamide
(2-acetylaminofluorene),
early dur
ing carcinogenesis in rat liver. The level of immunodetected pl4 was
markedly increased specifically during all stages of mitosis of hepatocytes
in normal and regenerating partially hepatectomized livers. In addition,
the polypeptide appeared to be immunologically and behaviorally related
to a polypeptide with a molecular weight of 17,500 that is tightly bound
to nucleosomes of chromatin in hepatocytes.
This report describes the actions of the two polypeptides during
hepatocarcinogenesis
induced by ingestion of 3'-methyl-4-dimethylaminoazobenzene
or Ar-2-fluorenylacetamide.
Both carcinogens
acted simi
larly in bringing about characteristic responses of the two polypeptides,
detected immunohistochemically
with specific rabbit antiserum and peroxidase-antiperoxidase
complex. Four types of discrete hepatocytic le
sions were observed. The earliest and least aberrant were hyperplastic
foci of proliferating hepatoyctes, which generally displayed markedly
higher levels of immunostained pl4 in cytoplasm, compared to levels in
normal diploid hepatocytes. Furthermore, the very high concentrations
of pl4 were continuously present during cell interphase, in contrast to
those in normal and regenerating hepatocytes, in which the elevation is
restricted to the period of mitosis. Later-arising lesions were acidophilic
adenomas of hepatocytes, which were characterized by quiescent mor
phology, fairly uniform and bulky eosinophilic cytoplasm, high levels of
glycogen, and small delicate nuclei. These cells usually displayed little
cytoplasmic pl4. Coexistent hepatocytic lesions, i.e., mixed basophilic
adenomas, exhibited considerable morphological heterogeneity, often
slightly basophilic cytoplasm, and variable immunostain of cytoplasmic
pi 4 during interphase. The fourth lesions were hepatocellular carcino
mas, which continuously demonstrated much higher than normal levels
of cytoplasmic pl4 during cell interphase. During mitosis in the four
types of hepatocytic lesions, the levels of immunostained pl4 were usually
further elevated above those in interphase, regardless of whether they
were already very high during interphase in hyperplasia and malignancy,
or low in acidophilic adenomas. All four kinds of carcinogen-altered
lesions usually displayed little of the detectable M, 17,500 polypeptide
in nuclei. Withdrawal of the carcinogens reversed the abnormal levels of
both immunostained polypeptides only in the mixed basophilic adenomas,
but not in persistent hyperplastic foci, acidophilic adenomas, and hepa
tocellular carcinomas. Considered altogether, the level of discernible pl4
is low when growth activity is low, high only during mitosis in normal
and regenerating hepatocytes, and continuously very high throughout
interphase during proliferation in hyperplastic and malignant hepato
cytes. The findings permit the speculation that focal proliferation brought
about by the two carcinogens may in part be mediated in rat hepatocytes
by the very high and continuous expression of cytoplasmic pl4 during
interphase in hyperplastic and malignant hepatocytes, in contrast to low
levels of the polypeptide in the slowly growing surrounding parenchyma.
INTRODUCTION
Focal cellular proliferation is an important early stage in
chemical carcinogenesis (1,2). Chemical carcinogens generally
inhibit cell multiplication in target tissues. In liver, a few
Received 5/14/86; revised 9/17/86; accepted 9/24/86.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported in part by USPHS Grants CA-05945, CA-30036, CA-06927, and
RR-05539 and an appropriation from the Commonwealth of Pennsylvania.
2To whom requests for reprints should be addressed, at: Institute for Cancer
Research, 7701 Burholme Avenue, Philadelphia, PA 19111.
hepatocytes arise that escape these inhibitory effects, and they
proliferate regeneratively to form hyperplastic foci. Through
successively inherited changes and cell selections, progressively
altered hepatocytes appear within the foci, and those cells give
rise ultimately to invasive carcinomas. The molecular events
that lead to the focal proliferation and subsequent focal malig
nant transformation are largely unknown.
A cytoplasmic polypeptide pl43 is associated with mitosis of
normal and regenerating rat hepatocytes. In the hepatocytes of
late fetal, neonatal, immature, adult, and regenerating livers,
the level of immunostained pi4 in cytoplasm is markedly and
specifically elevated during mitosis, compared to its level in
interphase cells (3, 4).
The pi4 is also associated with the focal hepatocytic prolif
eration brought about by two liver carcinogens. The polypeptide
is the principal covalent protein target of the hepatic carcino
gen, FAA, early during liver carcinogenesis (5-7). During he
patocarcinogenesis by FAA or 3'-Me-DAB, there arise at first
reversible and later irreversible (persistent) foci of proliferating
hyperplastic hepatocytes, in which the concentration of detect
able pl4 is strikingly elevated in cytoplasm, according to initial
study (3). In contrast, the level of demonstrable polypeptide is
low in the surrounding liver parenchyma (3), where responsive
ness to mitotic stimuli is known to be blocked by actions of the
carcinogens (reviewed in Refs. 1 and 2).
Evidence originally indicated that the cytoplasmic pl4 may
be related to a polypeptide with a molecular weight of 17,500
that is tightly bound to nucleosomes of chromatin in hepato
cytes of normal adult rats (8, 9). Unlike the level of cytoplasmic
p 14 immunostain, the level of nuclear pi7 immunostain is
evidently not changed during mitosis of normal and regenerat
ing hepatocytes (3, 4), and is greatly reduced in early and
persistent hyperplastic foci produced by FAA or 3'-Me-DAB
(3).
Directed by the connections of the cytoplasmic target pl4
with mitosis and focal hyperplasia, we examined the levels of
the two polypeptides in rat hepatocytes during carcinogenesis,
progressing from early preneoplasia to neoplasia and ending
with hepatocellular carcinoma. We report here the elevated
expression and cell cycle deregulation of the mitosis-associated
pl4 polypeptide in hyperplastic and malignant hepatocytes.
The level of detectable pl4 correlates with the apparent growth
activities of the different carcinogen-altered hepatocytes.
MATERIALS AND METHODS
Rats and Diets. Rats were raised on commercial stock diet (Prolab
RMH 3500; Agway, Syracuse, NY) and housed in screen-bottomed
cages at 22°Cwith food and acidified tap water available ad libitum.
Lighting was provided from 6 a.m. until 6 p.m. Animals weighed 150225 g at the start of experiments. Male Fischer 344 (CDF) and SpragueDawley (SD) rats (Charles River Breeding Laboratories, Wilmington,
MA) were fed the hepatocarcinogenic aminoazo dye, 3'-Me-DAB, as
0.058% of a synthetic diet (diet 3 of Ref. 10). Similarly, male CDF and
SD rats were fed FAA as 0.02 or 0.03% in a grain diet (11). Female
rats responded more slowly. Their data are not presented. In reversal
3 The abbreviations used are: p 14, 14,000-dalton polypeptide; p 17, 17,500dalton polypeptide; FAA, JV-2-fluorenylacetamide (2-acetylaminofluorene); 3'Me-DAB, 3'-methyl-4-dimethylaminoazobenzene;
PAP, peroxidase-antiperoxidase.
210
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE
AND HEPATOCYTE
experiments, following periods of ingestion of either carcinogen, rats
were fed their respective control diets (no carcinogen) for 4 or 6 weeks.
Rats were sacrificed with minimal prior excitement.
Immunohistochemical Analyses. Immediately after animal sacrifice,
liver sections 1-1.5 mm thick were fixed in 10% formalin in 0.1 M
sodium phosphate buffer, pH 7.4, for 20-24 h. Paraffin blocks and
successive S-^m sections were prepared and stored in a dry atmosphere
at 1-4°C.These conditions were necessary to preserve the detectability
of the two polypeptides. Deparaffinized sections were exposed to spe
cific rabbit antiserum against the cytosolic pl4 (3, 7, 8) and stained
with PAP complex (DAKO; Accurate Chemical and Scientific Corp.,
Westbury, NY) and 3,3'-diaminobenzidine tetrahydrochloride (12).
Rabbit antiserum was previously prepared against p 14 that was
homogeneous according to molecular size and charge (7, 8). The
specificity of the antiserum is supported by six criteria, (a) In reactions
with cytosols of normal livers of rats, mice, and Syrian hamsters, the
antiserum reacts only with pl4 after sodium dodecyl sulfate-gel electrophoresis, followed by western blot immunoanalysis (7, 8). Immunoreaction was not detected in 14 other organ cytosols in these species,
indicative of levels of less than 3-5% of the concentration in normal
rat liver cytosol (7). (o) Immunohistochemical analysis of rat liver
detects the pi4 only in hepatocytes (3, 4, 8). (c) The antiserum discrim
inates perilobular hepatocytes from centrilobular hepatocytes (8). (d)
The antiserum distinguishes mitotic hepatocytes from the greater than
1000-fold more numerous nondividing hepatocytes in normal adult
liver and also does so in regenerating liver as well (3, 4). (e) The
antiserum identifies different types of carcinogen-altered hepatocytes
according to their proliferative activities (Ref. 3; this report). (/) A
single purified protein absorbs out the reactivity of the antiserum (13).
On the other hand, the specificity of the reaction of the antiserum with
the nuclear pi7 is relative, inasmuch as other nuclear proteins appar
ently give weak positive signals when tested at 5-fold higher levels of
proteins in western immunoblot analyses.
Immunohistochemical stain in cytoplasm in hepatic lesions was
considered to be due to pi4. and that in nuclei to be of pl7, based on
our previous electrophoretic immunoblot analyses of cytosolic and
nuclear proteins of normal rat liver (7, 8) and the above evidence.
Comparisons of staining intensities were made between hepatocytes of
adjacent regions of the same sections, in which one region served as
reference for the other. Evaluations of levels of immunostaining in
volved inspections of four adjacent serial sections that were cut and
stained in the following order: those stained with hematoxylin and
eosin; those stained using both immune serum and hematoxylin; those
stained using immune serum only; and finally those processed using
only the corresponding preimmune (control) serum. Preimmune serum
yielded essentially no stain in all histochemical analyses. Counterstaining by hematoxylin was required to disclose nuclei, which would oth
erwise be obscured by heavily immunostained cytoplasm, and to reveal
ductal epithelia, oval cells, vessels, and stroma.
Levels of immunostaining were tentatively interpreted as reflections
of concentrations of the two polypeptides. Observed differences may
also be due in part to changes in accessibility of antibodies to the
polypeptides; e.g., due to the extent of complexation with nucleic acids
(cf. Réf.14). That changes in concentration of pi4 may actually be
involved is supported; (a) by our previous demonstration of the depres
sion in the amount of chemically detected pl4 in liver cytosols of rats
after ingestion of any of three carcinogens (5); (b) by use of polyclonal
antisera, which presumably react with many antigenic determinants of
both polypeptides, making detection less subject to problems of inac
cessibility; and (c) the reproducible observation of intralobular concen
tration gradients of pl4 and pi7 immunostains in normal rat liver.
Noteworthy too is the earlier finding by Gronow and Thackrah (15) of
the disappearance of a chemically detected nonhistone M, 17,500
polypeptide in electrophoretic gels of liver euchromatin of rats given
the hepatocarcinogen, diethylnitrosamine. Nevertheless, if alterations
in accessibility were also involved, the essentials of the findings would
remain the same; only the nature of the underlying changes would in
part be different.
Scoring of Lesions. Ingestion of either 3'-Me-DAB or FAA produced
PROLIFERATION
Relative incidences of these foci were scored on a scale of 0-4. Virtual
absence was scored 0, low incidence 1, moderate 2, high 3, and very
high 4. Table 1 lists ranges of incidences of lesions in which italicized
scores denote relative abundances representative of at least threefourths of the number of lesions, each focus considered as one unit.
Thus, a score of 4 signifies that the particular type of focus was present
in very high relative abundance.
Scoring of Polypeptides. A similar numerical system quantified levels
of immunodetected pl4 in cytoplasm and pi7 in nuclei. Table 1
contains ranges of scores in which italicized values denote relative
concentrations in at least three-fourths of the number of particular
types of lesions, each focus considered as one unit. Emphasis has been
placed on italicized scores as most representative of the lesions. Het
erogeneity was encountered within and between hepatic lesions. Pho
tographs in figures were chosen in part to demonstrate cellular hetero
geneity in foci, sometimes at the sacrifice of more representative immunostainings.
Levels of immunostainings of the two polypeptides in lesions were
also routinely compared with those of simultaneously processed normal
hepatocytes of adult rats. The pi4 in cytoplasm of normal diploid
hepatocytes was previously reported to be at moderate level (score 2)
at lobular peripheries and to decrease steadily to low content (score 1)
near lobular centers (8). A similar concentration gradient of pi 7 exists
in nuclei of normal diploid hepatocytes, starting perilobularly at very
high (score 4) and dropping centrilobularly to low (score 1) (8).
RESULTS
Polypeptides in Hepatocytic Lesions
With continued exposure to either carcinogen, the livers
underwent similar sequential developments of four types of
discrete hepatocytic lesions of increasing morphological and
immunohistochemical
abnormalities in both strains of male
rats. The azocarcinogen, 3'-Me-DAB, acted more rapidly and
aggressively and provided many more biliary ductal derivatives
(oval cells, cholangiofibrosis, cholangiocarinomas)
than did
FAA. These structures rarely immunostained for pi4 or pi7
(Figs. IB, 2B, and 4, B and F).
Hyperplastic Foci
Foci of hyperplasia were the earliest and least morphologi
cally abnormal hepatocytic lesions, appearing mainly after 4-5
weeks of ingestion of either liver carcinogen in both strains of
rats (Table 1). The lesions originated periportally as prolifera
tive cellular buds of larger than normal hepatocytes with little
alteration in ratio of cytoplasmic to nuclear volumes (Figs. 1,
A-D, and 2, A-D). Subsequently, the hyperplastic foci enlarged,
coalesced, and formed nodules separated by bands of com
pressed ductules, oval cells, and stroma.
The immunostained pl4 in cytoplasm of hyperplastic hepa
tocytes reached very high levels at 4-8 weeks of ingestion of
3'-Me-DAB or FAA in both strains of rats (Table 1, main
scores 4 and 3). The levels were considerably above the low to
moderate concentrations previously encountered with normal
diploid hepatocytes (scores 1-2) (8) and the surrounding par
enchyma.
The hepatocytes in hyperplastic foci displayed their high and
very high levels of cytoplasmic pl4 immunostain during cell
interphase (Figs. I, A-D, and 2, A-F). Table 1 documents and
the cited figures demonstrate that the hyperplastic foci exhibit
in general the intense immunostain throughout the lesions. The
vast majority of the hepatocytes were in interphase. This con
tinuous enhancement in pl4 in response to the carcinogens
markedly contrasts with the situation in hepatocytes of normal
and partially hepatectomized regenerating livers, in which con
siderable elevation of immunostained pl4 in cytoplasm is re
four types of focal hepatocellular lesions: hyperplastic foci; acidophilic
stricted to the period of mitosis (cf. Refs. 3 and 4).
adenomas; mixed basophilic adenomas; and hepatocellular carcinomas.
Their characteristics are described in "Results" and in Figs. 1-5.
Conversely, the level of pi7 immunostain dropped in nuclei
211
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE
AND HEPATOCYTE
PROLIFERATION
O
O
h
pt
acá
H
Fig. 1. A-D, early hyperplastic foci of hepatocytes in liver of SD male rat fed 0.058% 3'-Me-DAB for 4 weeks. In A, hyperplastic buds (A)of hepatocytes radiating
from portal triads (pt), separated by compressed septa of ductules and oval cells (o) (H & E, x 100). B, strong PAP immunostain of cytoplasmic pl4 in hyperplastic
hepatocytes (A) in serial section adjacent to A. Scattered white dots indicate little or no immunoreaction of pi7 in nuclei. Portal triads (pt) and intervening septa of
compressed ductules and oval cells (o) are virtually unstained (PAP only, x 100). C, margin of hyperplastic bud (A) in ./. Hepatocytes are larger than those of normal
parenchyma, but with undisturbed nucleocytoplasmic ratio. Ondular twigs and oval cells (o) are compressed between hyperplastic foci (H & E, x 400). />. intense
immunostain of pl4 in most cytoplasm and near absence of nuclear pi 7 in hyperplastic hepatocytes (A) in section adjacent to C. Oval cells (o) are unstained (PAP
only, x 400). E-H, acidophilic adenoma of hepatocytes located between hyperplastic foci in liver of SD male rat fed 0.058% 3'-Me-DAB for 4 weeks. In E, hepatocytes
of acidophilic adenoma (ara) are relatively pallidly stained with eosin. They retain same morphological relationships to portal triads (pt) as do their ill-defined
hyperplastic neighbors (A) (H & E, x 100). /•'.
minimal immunostaining of pi 4 and pi 7 in an acidophilic adenoma (sea) and none in portal triads (/;() in serial section
adjacent to F (PAP only, x 100). G, acidophilic adenomatous hepatocytes (oca) affare large with bulky, foamy (glycogen), eosinophilic cytoplasm and small nuclei
(H & E, x 400). In //. acidophilic adenomatous hepatocytes (oca) show minimal immunoreactivity for both polypeptides in the section adjacent to G (PAP only, x
400).
of hyperplastic foci produced by either carcinogen. At 4 weeks
and thereafter, immunostained pi7 was characteristically ab
sent or low (Table 1, main scores 0 and 1; Figs. 1, A-F, and 2,
A-F).
Irreversibility of Aberrant Levels of Immunostained Polypep
tides in Persistent Hyperplastic Foci. Early-appearing hyper
plastic lesions are known to be mainly reversible and later are
less so (termed irreversible or persistent) following removal of
carcinogens (1, 2). After long term omission of carcinogens
from the diets, only a few persistent hyperplastic foci were
discernible morphologically. These irreversible foci usually dis
played the abnormal levels of the two polypeptides evident in
212
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE AND HEPATOCYTE PROLIFERATION
acá
o
acá
.ti
H2«
aca
Fig. 2. /f-ß,early hyperplastic foci of hepatocytes in liver of CDF male rat fed 0.02% FAA for 8 weeks. A, periportal outgrowths of hyperplastic hepatocytes (A)
radiating from portal triads tin). Hyperplastic foci are separated by peripheral zone of oval cells (o) and ductular twigs (H & E, x 100). In B, immunostain of pI4
distinguishes hyperplastic buds by virtue of high cytoplasmic level, in serial section adjacent to I. Lighter and white dots indicates nuclei with low or absent pl7
reactions. Oval cells (o) and ductules are not stained, apart from slight hepatocytic admixture (PAP only, x 100). C detail of hyperplaslic buds (A) in A, similar to
those induced by 3'-Me-DAB, shown in Fig. 1C(H & E, x 400). In /'. hyperplastic hepatocytes (H) show marked immunostain of cytoplasmic pl4 and little or none
of nuclear pi 7 (PAP only, x 400). E-H, acidophilic adenoma of heptocytes among hyperplastic hepatocytes in liver of CDF male rat fed 0.02% FAA for 8 weeks. In
/;', line of cleavage between acidophilic adenoma (aca) and hyperplastic (A) foci is indicated by open arrow. Dark compressed bands are mostly oval cells (o) (H & E,
x 100). In /•',
immunostained serial section adjacent to E shows usual weak immunodetection of both polypeptides in acidophilic adenoma (ara) and strong
immunoreaction of cytoplasmic pl4 in hyperplastic foci (A). Neither polypeptide is detected in portal triad (pi) (PAP only, x 100). In G, detail of £showing acidophilic
adenomatous hepatocytes (aca) and many coarsely vacuolated cells laden with glycogen (H & E, x 400). //. minimal immunodetection of both polypeptides in
acidophilic adenomatous hepatocytes (aca) in serial section adjacent to G (PAP only, x 400).
the early foci. The concentration of immunostained pl4 was
generally high or moderate (main scores 3 and 2) and pi7
virtually absent (main scores 0), after ingestion of carcinogen
for 4-20 weeks followed by control diet for 4 or 6 weeks (Table
1; Fig. 5, C and D). Thus, the abnormal levels of immunostain
of the two polypeptides were usually fixed irreversibly at high
pl4 and virtually absent pi7 in the persistent hyperplastic
foci.
Acidophilic Adenomas
Hepatocytes of acidophilic adenomas had fairly uniform di
ameters, relatively bulky eosinophilic cytoplasm, high levels of
glycogen, and unusually small nuclei that stained lightly with
hematoxylin. Mitoses were exceedingly rare. Altogether, the
hepatocytes exhibited the appearance of cells at near arrest or
slow growth (Figs. 1, E-H, 2, E-H, and 3, A-D). Apparently
the acidophilic adenomatous hepatocytes had previously under213
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE
AND HEPATOCYTE
PROLIFERATION
if, '
Fig. 3. Acidophilic adenoma, mixed basophilic adenoma, and hepatocellular carcinoma within the same large hepatic nodule from CDF male rat fed 0.058% 3'Me-DAB for 16 weeks. All cells are counterstained with hematoxylin. In A. sharp boundary (dark arrow) divides acidophilic adenoma (oca) from mixed basophilic
adenoma (mba). Small ill-defined area within the acidophilic adenomatous portion in lower left is hepatocellular carcinoma (hca). Foamy cells in upper left are
glycogen laden: vacuolated cells in lower right contain lipid (H&E, x 100). In //, immunostaining for cytoplasmic p 14 conforms to the above morphologic differences,
revealed in the serial section adjacent to A. Acidophilic adenomatous area (oca) is relatively poorly immunoreactive in contrast to focus of mixed basophilic adenoma
(mim), while hepatocellular carcinomatous focus (hca) is very densely immunostained (PAP + hematoxylin, X 100). C, acidophilic adenomatous hepatocytes (oca) in
detail from . I. similar to Fig. 1C (H & E, x 400). In D, little immunodetection of cytoplasmic pI4 in acidophilic adenomatous hepatocytes (aca) in serial section
adjacent to C. Hematoxylin counterstains nuclei, but only few dark ones in lower right contain detectable pl7 (not fully evident in black and white) (PAP +
hematoxylin. x 400). In E, hepatocytes of mixed basophilic adenoma (mba) from A are characterized by disorderly arrangement of cells of varied size, shape, and
staining quality. Cytoplasm is slightly basophilic, and nuclei are generally larger, are coarser, and have greater than normal nucleoli (H & E, x 400). In F, immunostains
of pi 4 and pi 1 are exceedingly variable among mixed basophilic adenomatous cells (mba). in section adjacent to /•..
Example shown is immunostained intensely (PAP
+ hematoxylin, x 400). In G, hepatocellular carcinoma (hca) of A shows bizarre cytological abnormalities, including a multipolar mitosis (H & E, x 400). In II.
hepatocellular carcinomatous cells (hca) immunostain moderately to intensely for pl4 in the serial section adjacent to that of G. Heterogeneously stained region is
intentionally shown. Overall reaction is strong. Nuclear pl7 is obscured by hematoxylin counterstain (PAP + hematoxylin. x 400).
gone limited growth to form foci with visible compression
capsules at their perimeters. Similar irreversible neoplastic le
sions, termed clear and acidophilic cell foci, have previously
been described and reviewed by Bannasch et al. (16, 17). Foci
of acidophilic adenomas were evident starting at 4 and 8 weeks
of ingestion of 3'-Me-DAB and FAA, respectively (Table 1).
The lesions were usually rare. After longer term ingestion of
carcinogen, acidophilic adenomas usually contained inter
spersed basophilic hepatocytes (below), suggestive of possible
sequential relationships between them, as has been suggested
previously (16,17). In some instances, even small hepatocellular
carcinomas resided within acidophilic adenomas (Fig. 3, A, B,
G, and H).
Hepatocytes of acidophilic adenomas displayed little or no
214
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE AND HEPATOCYTE PROLIFERATION
Fig. 4. Mixed basophilic adenoma, hepatocellular carcinoma, and cholangiocarcinoma in liver of male SD rat fed 0.03% FAA for 32 weeks. In . I. heterogeneous
cellular mass of mixed basophilic adenomatous hepatocytes (mba), hepatocellular carcinoma (Ara), and cholangiocarcinoma (<7;ra),all richly laden with lipids. Details
of each region are described in C to H (H & E, x 100). In lì,adjacent serial section immunostained for p 14 in cytoplasm and pl7 in nuclei shows predominantly
nonreactive cholangiocarcinoma (chea) sharply delineated from highly immunoreactive hepatocellular carcinoma (hail Mixed basophilic adenomatous hepatocytes
(mila} are partially reactive (PAP only, x 100). C, detail of mixed basophilic adenomatous hepatocytes (mba) of A and B, resembling those induced by 3'-Me-DAB
demonstrated in Fig. IE (H & E, x 400). In l>, mixed basophilic adenomatous hepatocytes (mhu) in section adjacent to C show moderate immunoreactivity in both
cytoplasm and nuclei. (PAP only, x 400). £,poorly differentiated cholangiocarcinoma (chea) of A with pseudocolumnar cells. Ductular structures were abundant, and
mucin secretion is evident (not shown) (H & E, x 400). /'. typical lack of immunostains of p 14 and pi 7 in cholangiocarcinoma (chea) shown in . I and B, in serial
section adjacent to / (PAP only, x 400). G, hepatocellular carcinoma (hca) from A. The cells form a near syncytium, have coarse nuclear membranes, and have large
nucleoli (H & E, x 400). //, general strong immunoreaction of pi4 in cytoplasm and frequent absence of pi7 in nuclei of hepatocellular carcinoma (hca), in serial
section adjacent to G (PAP only, x 400).
p 14 immunostain, consistent with a morphological appearance
of quiescence or slow growth (Table 1, usual scores 0 and 7).
Figs. 1, E-H, 2, E-H, and 3, A-D show the uniformly very low
immunostain of cytoplasmic p 14, below that in normal liver
(cf. Refs. 3, 4, and 8).
Accompanying the depressed level of immunostain of pl4 in
cytoplasm of acidophilic adenomatous hepatocytes was the near
absence of evident pi7 in nuclei (Table 1, usual score 0; see
Figs. 1, E-H, 2, E-H, and 3, A-D). The virtual absence of
detected pi 7 contrasts with the high levels ofthat polypeptide
in perilobular regions in normal and regenerating livers (cf.
Refs. 3, 4, and 8).
Irreversibility of Low Levels of Immunostained Polypeptides
in Acidophilic Adenomas. Ingestion of the carcinogens for 1220 weeks, followed by their control diets for 4 or 6 weeks,
caused no apparent disappearance of the acidophilic adenomas
215
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
mba
acá
••<
o
~ •-•-,•
V ;i.if
w
;¿V
Fig. 5. A and A, immunoreactivity of pl4 and pl7 in liver of male CDF rat fed 0.058% 3'-Me-DAB for 12 weeks and then control diet (no carcinogen) for 6
weeks. A: left, acidophilic adenoma (ara); top, a mixed basophilic adenoma (mba); right, a hepatocarcinomatous focus (Ara); bottom, a cholangiofibrotic tongue (chf).
Septa of oval cells (o) divide the four foci (H & E, x 63). B, strong immunostain of pl4 in mixed basophilic adenomatous hepatocytes (mba). The acidophilic adenoma
(ara) is interspersed with heavily immunostained basophilic adenomatous hepatocytes. Right, an intensely immunostained hepatocarcinoma (Ara). Cholangiofibrotic
cells (chf) and oval cells (o) are not immunostained (PAP + hematoxylin, x 63). C and D, immunostain of p 14 in liver of male CDF rat fed 0.02% FAA for 20 weeks
and then control diet (no carcinogen) for 6 weeks. C, hyperplastic foci (h) adjoining distorted portal triad (pt) and predominantly acidophilic (ora) and mixed
basophilic (mba) adenomas (H & E, x 63). D, strong immunostain of hyperplastic foci (h) that emanate from nonreactive portal triad (pt). (The darkening in the
compacted portal triad region is due to hematoxylin counterstain, rather than immunostain.) Individual mixed basophilic adenomatous hepatocytes (mba) immunostain
intensely amidst poorly reactive acidophilic adenomatous hepactocytes (ara) (PAP + hematoxylin, x 63). E-J, elevated levels of pl4 immunostains in cytoplasm
during mitosis in normal, hyperplastic. acidophilic adenomatous, mixed basophilic adenomatous, and malignant hepatocytes. All sections are counterstained with
hematoxylin. Photographs were of low exposure, in order to display the mitotic nuclei which would otherwise be obscured by the heavy pl4 immunostain of the
dividing hepatocytes
PAP + hepatocyte
hematoxylin,
12SO).
E, normal
adult hepatocytes hyperplastic
in mitosis invariably
pl4 moreimmunostains
intensely thanfordopl4
adjacent
interphase
hepatocytes.(allNormal
(n) xis in
telophase.
/•'.
carcinogen-induced
hepatocyteimmunostain
h in mitosisfor(prophase)
more
strongly than even the above-normal levels of interphase neighbors. </. acidophilic adenomatous hepatocytes generally immunostain weakly for p 14. Nevertheless,
rare mitoses stand out by greater immunoreaction for p 14. An acidophilic adenomatous hepatocyte (ara) is shown in prophase. H, mixed basophilic adenomatous
hepatocytes vary in degrees of immunoreactivity, usually strongly staining. Two mixed basophilic adenomatous hepatocytes (mba) in metaphase and an adjacent high
polyploid hepatocyte immunoreact more strongly than do interphase neighbors. /, in contrast, another mixed basophilic adenomatous hepatocyte (mini) in mitosis
(anaphase) is not immunoreactive. J. two metaphases in a hepatocellular carcinoma (hca), each with opposite intensities of immunoreaction.
216
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE
AND HEPATOCYTE
PROLIFERATION
Table 1 Levels ofp¡4 and pl7 polypeptides in carcinogen-induced focal lesions in rat liver
fociRatCarcinogen*
Relative incidences and concentrations in
basophilic
adenomasIncidence"01-21-312-01-03-\1-42001-23-\00221100-202-113fociRats26553333634242133342333333Incidence"34-32-Õ0-10400001-30-10-12-3422001010001pl4"3447-34-33-12-33-43-44-343-42-33-4PIT"2110-10-20-2
carcinomasRats00121012530120000000300000pl4"4
adenomasIncidence"012-31-32-403-\3-43003-11-200333-43-400-202-32-32-44pl4"2strain3'-Me-DAB
Wk
SD4812164
3e
CDFg121620Reversal
4SD8
+ 4
612+
CDFFAA0.02%
+ 6
CDF581216200.03% 4
CDF32
3.5
SDReversal,
CDF0.02%8
4+ 6
612+
616+
11pl4"1-21-21-31-41-20-10-421-21-32-11-21-21-221-31-21-3Ì-3PIT"0-10-20-30-30-20-10-41-20-20-21-00-2
620+
+ 6Hyperplastic
" Relative incidences of morphological foci and their apparent relative concentrations of pi4 and pi7 are indicated on a scale of 0-4. In listed ranges of scores,
italicized first values denote predominant scores of greater than 75% of the number of foci.
'' Male rats of the indicated strains were fed either of two hepatic carcinogens, 3'-Me-DAB at 0.058% in a synthetic diet or FAA at 0.02 or 0.03% in a grain diet.
Reversals indicate that rats were fed the carcinogenic diet for the first number of weeks shown, followed by the control diet (no carcinogen) for the second number of
weeks (e.g., 12 + 6). Details are presented in "Materials and Methods."
' Ingestion of 3'-Me-DAB diet or 0.03% FAA diet for 1 or 2 weeks resulted in no hepatocytic foci.
(Table 1). Acidophilic adenomas were thus irreversible lesions.
Omission of the carcinogens for 4 or 6 weeks did not alter
the low and very low levels of both immunostained polypeptides
in acidophilic adenomas. The apparent scarcity of the polypep
tides is seen in Fig. 5, A-D. The very low concentrations of
both immunostained polypeptides in acidophilic adenomatous
hepatocytes thus appeared to be irreversibly fixed.
Mixed Basophilic Adenomas
Foci of mixed basophilic adenomas coexisted and often in
termingled with acidophilic adenomas. Also clearly marginated,
hepatocytes in mixed basophilic adenomas showed morpholog
ical aberrancies (Figs. 3, A, B, E, and F\ and 4, A-D), but not
the bizarre forms characteristic of malignant cells (below).
Hepatocytes in mixed basophilic adenomas varied in size,
shape, and nuclear detail, frequently in atypical forms. The
cytoplasm exhibited little glycogen and slight basophilia in
hematoxylin- and eosin-stained sections. Nuclei were larger,
coarser, and more hyperchromatic, and mitoses were much
more common than those in acidophilic adenomas. Similar
neoplastic lesions, termed mixed basophilic foci and adenomas,
were previously described and reviewed by Bannasch et al. (16,
17). Pure basophilic adenomas referred to by those investigators
were not observed in the present study. Mixed basophilic ade
nomas were most numerous at 8-16 weeks of ingestion of 3'Me-DAB or FAA (Table 1, usual score 3). Foci of hepatocellular
carcinomas were often discernible within these lesions starting
at 8 weeks (data not shown).
Consistent with their heterogeneous morphological charac
ter, hepatocytes in mixed basophilic adenomas exhibited differ
ent cell-to-cell levels of immunostained cytoplasmic pl4 (see
Fig. 3, A, B, E, and F and Fig. 4, A-D). Table 1 lists levels of
the immunostained polypeptide. Uncertainties pertain due to
dispersion of concentrations within and between these lesions.
The main scores varied from 0 to 3. No consistent trend was
evident with either carcinogen. Nevertheless, it is of interest
that high levels of pi4 immunostain in the cytoplasm of indi
vidual mixed basophilic hepatocytes were present at interphase.
There was marked loss of immunostained pi 7 in most nuclei
of mixed basophilic adenomatous hepatocytes (see Fig. 3, A, B,
E, and F and Fig. 4, A-D). However, there was a notable lack
of uniformity of the pi7 immunostain within individual ade
nomas. The pi7 immunostain was generally absent or low
during all periods of ingestion of either of the two agents (Table
1, usual scores 0 and 7). These values are much below those of
pi7 in perilobular diploid hepatocytes of normal adult rats (cf.
Refs. 3, 4, and 8).
Reversibility of Abnormal Levels of Immunostained Polypep
tides in Mixed Basophilic Adenomas. We ascertained whether
the highly varied concentrations of cytoplasmic pl4 immuno
stain and the generally low levels of nuclear pi7 in mixed
basophilic adenomas could be reversed. Rats were fed the azo
dye diet for 4 or 8 weeks or FAA for 8 to 20 weeks, followed
by their control diets without carcinogen for 6 weeks (Table 1).
The pi4 in all groups of rats rebounded to high levels (main
score 3). Unlike the original variations of p 14, the later levels
were remarkably constant (Fig. 5, A-D). These high concentra-
217
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE AND HEPATOCYTE PROLIFERATION
lions of p 14 exceeded those of normal liver (cf. Refs. 3, 4, and
8).
Likewise, the loss of immunostained nuclear pl7 in mixed
basophilic adenomas brought about by the two carcinogens was
reversed by omission of the carcinogens from the diets. From
the original levels of 0 and 1 in mixed basophilic adenomatous
foci, removal of the carcinogens for 6 weeks resulted in re
bounds comparable to the high concentrations present perilobularly in normal liver. The main scores were 3 in all groups of
rats (Table 1). Thus, in contrast to all other hepatocytic lesions
(above and below), mixed basophilic adenomas responded to
removal of the carcinogens by reverting the concentrations of
the two polypeptides to uniformly high levels that matched or
exceeded those in normal liver.
Hepatocellular Carcinomas
Hepatocellular carcinomas were the most aberrant of the
hepatocytic lesions. Occasional well-differentiated tumors re
tained cytological and topologica! features of liver. Malignant
hepatocytes stained with hematoxylin and eosin had hyperchromatic nuclei, prominent nucleoli, and high ratios of nuclear to
cytoplasmic volumes. The carcinomas demonstrated a variety
of morphological atypias, many giant cells, strange mitotic
figures, and vascular invasion.
The azocarcinogen produced mixtures of liver tumors in both
the Fischer and Sprague-Dawley rats. Partially to poorly differ
entiated heptocellular carcinomas appeared in 2 of 8 rats fed
the azocarcinogen for 8 weeks (Table 1), 4 developed cholangiocarcinomas, and 1 had both. In 8 rats fed the azo dye for 12
weeks, 4 had hepatocellular carcinomas, 4 had cholangiocarcinomas, and 3 had both. At 16 weeks, 6 of 9 rats had hepato
cellular carcinomas (Fig. 3, A, B, G, and //), 5 had cholangiocarcinomas, and 5 had both. Finally, at 20 weeks 3 developed
both kinds of cancers.
Hepatocellular carcinomas arose in 3 Sprague-Dawley rats
fed 0.03% FAA for 32 weeks (Table 1). One rat had a well to
poorly differentiated trabecular carcinoma. The second had a
mixed hepatocellular and biliary carcinoma, in which hepato
cellular components immunostained intensely, and ductal con
stituents very little or not at all (Fig. 4, A, B, E-H). The third
had a conglomerate of acidophilic and mixed basophilic ade
nomas and hepatocellular carcinoma.
Virtually all hepatocellular carcinomas exhibited very high
levels of immunostained pi 4 in cytoplasm (Table 1, usual score
4). The exceptions were one group that scored high and another
moderately (3 and 2, respectively). Otherwise, all were very high
in pi4 and stained remarkably alike. These evaluations of the
levels of pl4 ininninosi ain were aided by comparisons with
adjacent liver parenchyma, hyperplastic and acidophilic lesions,
and normal liver sections, including those simultaneously proc
essed and stained. Cholangiocarcinomas were not immunoreactive (Fig. 4, A, B, E, and F ) like their nonneoplastic counter
parts (Figs. 1, A, B, E, and F, and 2, A, B, E, and F).
Furthermore, the very high levels of pl4 immunostain
throughout the hepatocellular carcinomas were in interphase
cells. Table 1 documents and Figs. 3, A, B, G, and //. and 4, A,
B, G, and //. demonstrate the intense p 14 immunostain
throughout these lesions. The presence of these very high
cytoplasmic levels of pl4 immunostain during interphase in the
hepatocellular carcinomas matched the similar behavior of the
polypeptide in hyperplastic foci (above). The elevation of pi4
during interphase in these two hepatic lesions in response to
the carcinogens was in marked contrast to the behavior of
normal and regenerating hepatocytes, in which the considerable
elevation of p 14 immunostain in cytoplasm is confined to the
interval of mitosis (cf. Refs. 3, 4, and 8).
The pi7 was not detected in most, but not all, nuclei in all
groups of hepatocellular carcinomas produced by the two car
cinogens (Table 1). Fig. 4, A, B, G, and H, shows immunoreacted sections of FAA-induced liver tumors, in which nuclei
were not counterstained with hematoxylin.
Irreversibility of Abnormal Levels of Immunodetected Poly
peptides in Hepatocellular Carcinomas. Two Fischer CDF rats
were fed the azo carcinogen for 8 or 12 weeks, followed by
control diets for an additional 6 weeks. The resultant hepato
cellular carcinomas still exhibited very high levels of immuno
stained pl4 (Table 1; main score 4 and 3; Fig. 5, A and B).
Further, the apparent accumulations of pi 4 were present during
interphase.
Likewise, the general near absence of immunostained pi7 in
nuclei persisted in azo dye-induced hepatocellular carcinomas
after ingestion of control diet for 6 weeks (Table 1, main scores
0 and 1). No hepatocellular carcinomas were found in rats that
were fed FAA and then control diet.
Polypeptides
during Mitosis of Carcinogen-altered
Hepatocytes
We asked whether the concentration of detectable cyto
plasmic pi4 increases during mitosis in the carcinogen-altered
hepatocytes, as it does in normal and regenerating hepatocytes.
Such elevations would need to be above the high levels in
interphase hyperplastic and malignant hepatocytes, and above
the low concentrations in interphase hepatocytes in acidophilic
adenomas.
Mitosis was usually accompanied by enhancement of cyto
plasmic immunostained pi4 in all carcinogen-induced lesions.
Further, the increase occurred during all stages of mitosis. Fig.
5£illustrates the typical increase in cytoplasm during mitosis
(anaphase) in a normal adult hepatocyte, compared to its interphase neighbors. Likewise, Fig. 5F shows a representative ele
vation during mitosis (metaphase) of a hepatocyte in a hyper
plastic focus. The greater level of p 14 immunostain in cyto
plasm is evident above the high concentrations in adjacent
interphase cells. Even the rarely dividing hepatocytes in acido
philic adenomas exhibited a rise of detectable pl4 (Fig. 56'). In
hepatocytes of mixed basophilic adenomas most divisions were
accompanied by an enhancement (Fig. 5H). However, consist
ent with their cellular heterogeneity, a few multiplying mixed
basophilic hepatocytes showed equivalent or less p 14 than did
their neighboring interphase cells (Fig. 5/). Similarly, while
most dividing hepatocytes in hepatocellular carcinomas showed
more pl4 above already very high levels in adjacent interphase
carcinomatous cells, a few had less. Both types of dividing cells
are pictured in Fig. 5J. In summary, the level of immunostained
pi 4 was greater during mitosis of most, but not all, hepatocytes
in all types of carcinogen-induced lesions, regardless of whether
interphase levels were very low in acidophilic adenomas, or
very high during hyperplasia and malignancy.
DISCUSSION
Cytoplasmic pl4 appears to be connected with the prolifera
tion of rat hepatocytes (Fig. 6). First, the level of immuno
stained polypeptide is elevated specifically during all stages of
mitosis in late fetal, neonatal, immature, adult, and regenerat
ing partially hepatectomized livers, and also in carcinogenaltered and malignant hepatocytes (Refs. 3 and 4; this report).
Secondly, the concentration of immunostained pi4 is consid
erably enhanced in the cytoplasm of proliferating hepatocytes
in early and persistent hyperplastic foci brought about by FAA
and 3'-Me-DAB (Refs. 3; this report). Conversely, there is little
218
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE AND HEPATOCYTE PROLIFERATION
p 14 evident in liver parenchyma that surrounds the early hyperplastic foci (3), where cell multiplication has been amply
reported to be inhibited by the actions of carcinogens (reviewed
in Refs. 1 and 2). Next, detectable p 14 is low in quiescentappearing hepatocytes of acidophilic adenomas (this report).
Further, the level of detected pl4 varies markedly among the
morphologically heterogeneous hepatocytes of mixed basophilic adenomas (this report). Lastly, very high concentrations
of discernible polypeptide are present in the subsequent hepatocellular carcinomas (this report). Considered altogether, the
level of pl4 in rat hepatocytes appears generally to be low when
growth activity is low, high during mitosis in hepatocytes of
normal and regenerating livers, and very high during prolifer
ation in hyperplastic and malignant hepatocytes brought about
by the two carcinogens.
The cell cycle-dependent expression of pl4 appears to be
altered during carcinogenesis by FAA and 3'-Me-DAB. In late
fetal, neonatal, immature, adult, and regenerating hepatocytes,
the content of detectable p 14 in cytoplasm is elevated only
during mitosis (3, 4). In contrast, the immunostained polypep
tide is highly enriched continuously during interphase in early
and persistent hyperplastic hepatocytes, individual mixed basophilic adenomatous hepatocytes, and hepatocellular carcino
mas produced by the two carcinogens (this report). This appar
ent change in cell cycle regulation of p 14, considered together
with its altered levels in hepatocytes in various normal, regen
erative, and carcinogen-induced abnormal states, permit spec
ulation that the high levels of pl4 in cytoplasm during interphase may be involved in mediating the proliferation of hyper
plastic and malignant hepatocytes in these rat livers.
The conversion of normal hepatocytes to malignancy is
marked by lesions with characteristic growth properties and
contents of immunostained p 14. Which lesions lie directly in
the oncogenic process, rather than as epiphenomenal offshoots
thereof, are unknown. A hypothetical sequence is depicted in
Fig. 6. Reversible hyperplastic foci of growing hepatocytes
represent an early morphological lesion induced by the carcin
ogens. Continued exposure to the carcinogens produces irre
versible hyperplastic heptocytes from which carcinomas have
been inferred to arise (1, 2). It seems reasonable to speculate
that proliferation in the early and persistent hyperplastic he
patocytes is directly connected with high interphase levels of
p 14. Some lineages of carcinogen-altered hepatocytes appear
to undergo a limited number of irreversible changes, and slowly
divide thereafter. Presumably they differentiate to the acido
philic adenomatous state with its low concentration of detect
able p 14, possibly in mimicry of normal hepatocyte differentia
tion. Other lineages of hepatocytes may undergo other or ad
ditional combinations of fixed alterations to form mixed basophilic adenomas with characteristic diverse morphologies and
reversible contents of p 14. Compared to acidophilic adenomas,
mixed basophilic adenomas exhibit greater mitotic activity.
Those hepatocytes that show high levels of pl4 during interphase are assumed to be proliferative. Finally, hepatocellular
carcinomas contain very high, irreversible concentrations of
pl4 during cell interphase. It is important to note that acido
philic adenomas are neoplasms with low levels of p 14. Hence,
the neoplastic state, itself, does not require that high concen
trations of pl4 be present. Instead, the levels of p 14 in cyto
plasm appear to reflect the growth activities that are character
istic of the various carcinogen-altered states of the hepatocytes.
The levels of detectable pl7 were depressed generally in all
types of the carcinogen-induced hepatic lesions, regardless of
their proliferative activities. These very low concentrations
continued as such in the persistent hyperplastic foci, acidophilic
adenomas, and hepatocellular carcinomas after discontinuation
of the carcinogens. Thus, high levels of detectable pi 7 in nuclei
appear to be characteristic of normal hepatocytes, and that
property is much reduced in hepatocytes altered by the carcin
ogens.
Various lines of evidence that seemingly connected the cytoplasmic pi4 to the nuclear pi7 were previously summarized (4,
8, 9). The present findings indicate that the two polypeptides
also act alike with regard to fixation of their changes in concen
trations induced by the two carcinogens. When the levels of
p 14 are not reversible, neither are those of pi 7 (as in persistent
hyperplastic foci, acidophilic adenomas, and hepatocellular car
cinomas). Conversely, when the abnormal levels of pl4 are
reversible, those of pi7 are too (as in mixed basophilic adeno
mas). These similar responses to the cessation of carcinogen
ingestion further support an unknown relationship between the
two polypeptides. Noteworthy, whereas purified pl4 can com
pletely absorb out the western immunoblot reaction of the
antiserum with pi 4 in cytosol, the same amount of p 14 fails to
/••••'Û2V':/.
I «t Lf
MORPHOLOGICALLY
NON-TRANSFORMED
PARENCHYMA
.
' .
J
REVERSED
PARENCHYMA
NO
CARCINOGEN
Fig. 6. Hypothetical succession of focal le
sions of rat hepatocytes with observed levels of
cytoplasmic p 14 and nuclear pi7 during liver
carcinogenesis. The unknown relevance and
positions of acidophilic and mixed basophilic
adenomas in oncogenesis are indicated by
question marks. Relative concentrations of the
two polypeptides are depicted by relative shad
ing intensities. K, molecular weight in thou
sands.
NORMAL
INTERPHASE
MITOSIS
PERSISTENT FOCAL
HYPERPLASIA
FOCAL HYPERPLASIA
CARCINOMA
MIXED BASOPHILIC
ADENOMA
ACIDOPHILIC
ADENOMA
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
POLYPEPTIDE
AND HEPATOCYTE
PROLIFERATION
do so with pi7 in nuclear extract." Further experimentation is ACKNOWLEDGMENTS
required to determine the nature of the apparent connection
We thank Bernice B. Althouse, Grace Kroetz, Susan K. Wilkinson,
between the two polypeptides.
Marian J. Mastrangelo, and Michael G. Jacobs for dedicated technical
The p 14 acts in a manner that relates to the focal nature of assistance.
proliferation of rat hepatocytes during oncogenesis by the two REFERENCES
carcinogens. Focal proliferation results from the growth of few
1. Farber, E., Cameron, R. G., Laishes, B., Lin, J-C., Mediine, A., Ogawa, K.,
carcinogen-altered hepatocytes in the midst of multiplicationand Soit, D. B. Physiological and molecular markers during carcinogenesis.
inhibited parenchyma (1, 2, 18). Fig. 6 depicts relevant rela
In: A. C. Griffin and C. R. Shaw (eds.), Carcinogens: Identification and
Mechanisms of Actions, pp. 319-335. New York: Raven Press, 1979.
tionships of the p 14. In hepatocytes of normal and regenerating
E., and Cameron, R. The sequential analysis of cancer development.
partially hepatectomized livers, the p 14 appears to be highly 2. Farber,
Adv. Cancer Res., 31: 125-226, 1980.
elevated in cytoplasm specifically during all phases of mitosis 3. Custer, R. P., and Sorof, S. Target polypeptide of a carcinogen is associated
with normal mitosis and carcinogen-induced hyperplasias in adult hepato
(3, 4). However, in foci of early and persistent hyperplastic
cytes. Proc. Nati. Acad. Sci. USA, 81:6738-6742, 1984.
hepatocytes brought about by FAA and 3'-Me-DAB, the con
4. Custer, R. P., and Sorof, S. Mitosis in hepatocytes is generally associated
centration of immunostained p 14 is even further increased
with elevated levels of the target polypeptide of a liver carcinogen. Differen
tiation, 30: 176-181, 1985.
during cell interphase (Réf.
3; this report). In contrast, in the
Blackburn, G. R., Andrews, J. P., Rao, K. V. K., and Sorof, S. An early event
surrounding liver parenchyma, where responsiveness to mitotic 5. associated
with liver carcinogenesis involving loss of a polypeptide that binds
carcinogen. Cancer Res., ¥0/4688-4693, 1980.
stimuli is inhibited (reviewed in Refs. 1 and 2), biochemical and
immunohistochemical evidence indicates that the polypeptide 6. Blackburn, G. R., Andrews, J. P., Custer, R. P., and Sorof, S. Early events
during liver carcinogenesis involving two carcinogen:protein complexes. Can
is depressed (3, 5). The pi4 is the principal early covalent
cer Res., 41: 4039-4049, 1981.
protein target of metabolites of FAA in vivo(5-7). That rela
7. Blackburn, G. R., Schnabel, S. J., Danley, J. M., Hogue-Angeletti, R. A.,
and Sorof, S. Principal polypeptide target of carcinogen at the beginning of
tionship may contribute additionally to the lack of functional
liver carcinogenesis by three carcinogens. Cancer Res., 42:4664-4672,1982.
pl4 in the surrounding liver parenchyma, possibly resulting in 8. Vinores, S. A., Churey, J. J., Haller, J. M., Schnabel, S. J., Custer, R. P.,
and Sorof, S. Normal liver chromatin contains a firmly bound and larger
even greater mitotic inhibition by FAA there. It is supportive
protein related to the principal cytosolic target polypeptide of a hepatic
that FAA is especially able to serve as mitotic inhibitor of liver
carcinogen. Proc. Nati. Acad. Sci. USA, 81: 2092-2096, 1984.
parenchyma in the rapid production of hyperplastic foci by the 9. Bassuk, J. A., and Sorof, S. A new nucleosoma) protein in normal liver
related to the cytoplasmic polypeptide target of a carcinogen. Mol. Cell.
protocol of Farber et al. (1, 2) and Soit et al. (18). Next, it is
Biochem., 68: 49-57, 1985.
unknown whether and how acidophilic and mixed basophilic 10. Miller, E. C, Miller, J. A., Kline, B. E., and Rusch, H. P. Correlation of the
adenomas may relate to focalization in oncogenesis by the two
level of hepatic riboflavin with the appearance of liver tumors in rats fed
aminoazo dyes. J. Exp. Med., 88:89-98, 1948.
carcinogens. Lastly, hepatocellular carcinomas produced by
E. C., Miller. J. A., Sandin, R. B., and Brown, R. K. The carcinogenic
these carcinogens continuously display very high concentrations 11. Miller,
activities of certain analogs of 2-acetylaminofluorene in rat. Cancer Res., 9:
504-509, 1949.
of pi4 during cell interphase (this report). The collected find
T. (ed.), Reference Guide Series 1: PAP/Immunoperoxidase,
pp.
ings are consistent with a hypothetical model in which pi4 12. Boenisch,
3-18. Santa Barbara, CA: DAKO Corp., 1980.
mediates the focal proliferation of rat hepatocytes elicited by 13. Bassuk, J. A., Tsichlis, P. V, and Sorof, S. Fatty acid binding protein is
the actions of the carcinogens. The focalization acts presumably
closely related to the polypeptide associated with normal mitosis and prolif
eration in hyperplastic and malignant hepatocytes. Abstract. J. Cell Biol., in
to render the carcinogenic process more effective by funneling
press, 1986.
cell growth to relatively few hepatocytes that consequently 14. Rodman,
T. C., l'r usiin. F. H., and Allfrey, V. G. Protamine-DNA associa
tion in mammalian spermatozoa. Exp. Cell Res., 750: 269-281, 1984.
undergo greater numbers of successive divisions and additive
M., and Thackrah, T. Changes in the composition of rat liver
deviations. As a result, early converted cells in the proliferative 15. Gronow,
chromatin fractions during nitrosamine carcinogenesis. Eur. J. Cancer, 10:
foci have enhanced likelihood of progressing to independent
21-25, 1974.
16. Bannasch, P., Hacker, H. J., Klimek, F., and Mayer, D. Hepatocellular
neoplasms.
glycogenolysis and related pattern of enzymatic changes during hepatocarciIt seems attractive to speculate that oncogenesis may be
nogenesis. Adv. Enzyme Regul., 22: 97-121, 1984.
additionally facilitated by the actions of p 14 in rat hepatocytes 17. Bannasch, P., and Zerban, H. Pathogenesis of primary liver tumors induced
by chemicals. Recent Results Cancer Res., 100: 1-15, 1986.
altered by the two carcinogens. The pl4-mediated proliferation
D. B., Mediine, A., and Farber, E. Rapid emergence of carcinogenmay bring about elevated expression of cellular oncogenes. 18. Solt,
induced hyperplastic lesions in a new model for sequential analysis of liver
carcinogenesis. Am. J. Pathol., 88: 595-610, 1977.
Expressions of ras, myc, and fas protooncogenes have been
reported to be increased during cellular proliferation (19-23).
19. Goyette, M., Petropoulos, C. J., Shank, P. R., and Fausto, N. Expression of
a cellular oncogene during liver regeneration. Science (Wash. DC), 219:510The pl4 may facilitate growth of hepatocytes by providing
512, 1983.
constituents that promote their cell division. Ten types of 20. Kelly, K., Cochran, B. H., Stiles, C. D., and Leder, P. Cell-specific regulation
of the c-myc gene by lymphocyte mitogens and platelet-derived growth factor.
evidence, including close sequence homology of cloned cDNA,
Cell, 35:603-610, 1983.
point to the identity of the pi4 as being liver fatty acid binding 21. Greenberg, M. E., and Ziff, E. B. Stimulation of 3T3 cells induces transcrip
tion of the c-fos proto-oncogene. Nature (I mid.). 311:433-438, 1984.
protein (13). We accordingly speculate that increased levels of
R., Bravo, R., Burckhardt, J., and Curran, T. Induction of c-fos gene
liver fatty acid binding protein, a proposed intracellular carrier 22. Muller,
and protein by growth factors precedes activation of c-myc. Nature (Lond.),
polypeptide (24), may be intimately connected with normal
312: 716-720, 1984.
mitosis and the carcinogen-induced proliferation of rat hepa 23. Muicahy, L. S., Smith, M. R., and Stacy, D. W. Requirement for ras protooncogene function during serum-stimulated growth of NIH 3T3 cells. Nature
tocytes.
(Lond.), 313: 241-243, 1985.
24. Glatz, J. F. C., and Veerkamp, J. H. Intracellular fatty acid-binding proteins.
Int. J. Biochem., 17: 13-22, 1985.
4 K. M. Munir and S. Sorof, unpublished data.
220
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1987 American Association for Cancer Research.
Elevated Expression and Cell Cycle Deregulation of a
Mitosis-associated Target Polypeptide of a Carcinogen in
Hyperplastic and Malignant Rat Hepatocytes
Sam Sorof and R. Philip Custer
Cancer Res 1987;47:210-220.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/47/1/210
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 18, 2017. © 1987 American Association for Cancer Research.