Page 1
ORIGINAL ARTICLE
Immunohistochemical analysis of heme oxygenase-1 inpreneoplastic and neoplastic lesions during chemicalhepatocarcinogenesis§
Fabiana Caballero*†, Roberto Meiss‡†, Alejandra Gimenez‡, Alcira Batlle* and Elba Vazquez*
*Centro de Investigaciones sobre Porfirinas y Porfirias (CIPYP) CONICET, Departamento de Quımica Biologica, FCEN, UBA – Ciudad
Universitaria, Buenos Aires, Argentina, and ‡Departamento de Patologıa, Instituto de Estudios Oncologicos, Academia Nacional de
Medicina, Buenos Aires, Argentina
I N T E R N AT I O N A LJOURNAL OFE X P E R I M E N TA LPAT H O L O G Y
Summary
Heme oxygenase (HO) breaks down the pro-oxidant heme into carbon monoxide, iron
and the antioxidant biliverdin. The isoform HO-1 plays an effective role to counteract
oxidative damage and to control inflammation. Prolonged cellular damage due to
chronic inflammation is one of the reasons leading to the development of tumours.
The aim of this work was to investigate HO-1 expression and localization along the
different stages of chemically induced hepatocarcinogenesis (HCC) and the occurring
morphological changes. To provoke sustained oxidative stress and chronic inflamma-
tion, CF1 mice received dietary p-dimethylaminoazobenzene (DAB, 0.5%, w/w) during
a whole period of 14 months. HO-1 expression increased along the experimental trial in
morphologically normal hepatocytes in DAB-treated animals. HO-1 expression dimin-
ished in altered hepatic foci (AHF) and oval cells and early preneoplastic lesions.
Otherwise, marked HO-1 overexpression was detected in Kupffer cells and macro-
phages surrounding necrotic and nodular areas. Adenomas showed decreased HO-1
immunostaining. In hepatocellular carcinomas, an inverse relationship was found
between the immunohistochemical expression of HO-1 and the degree of tumour
differentiation, being negative in poorly differentiated tumours. In our experimental
model, down modulation of HO-1 expression correlated with malignancy progression.
Thus, our data point to activation of HO-1 as a potential therapeutic tool.
Keywords
heme oxygenase, hepatocarcinogenesis, inflammation, oxidative stress, p-dimethyl-
aminoazobenzene
Hepatic carcinogenesis (HCC) is a complex process
that histologically progresses from benign precursor lesions
to malignant neoplasms, and it is associated with accumula-
tion of genetic and epigenetic changes, giving rise to the
stages of initiation, promotion and progression (Boone
et al. 1992).
Received for publication:
7 November 2003
Accepted for publication:
17 June 2004
Correspondence:
Prof Dr A. Batlle, Viamonte 1881 10�
‘‘A’’,1056 - Buenos Aires, Argentina.
Fax: +54 11 4811 7447;
E-mail: [email protected]
†The first two authors contributed
equally to the development of this
work.§To the memory of Cesar Polo,
deceased March 9th, 1996.
Int. J. Exp. Path. (2004), 85, 213–221
� 2004 Blackwell Publishing Ltd 213
Page 2
Carcinogenesis can be experimentally induced by exposure
to exogenous agents, and chronic rodent bioassays are the best
models to predict the risks of chemical exposure in humans
(Ames & Gold 1990).
The diagnosis of HCC is based on the occurring architec-
tural and cytological changes. Numerous investigations have
been carried out to elucidate the early hepatocellular lesions
which may represent progenitor populations of HCCs (Ahn
et al. 1999).
The development of phenotypically altered hepatic foci
(AHF) has been a common feature in experimental HCC
and is believed to be caused by mutations in normal
cellular growth control genes and subsequent clonal growth
(Butterworth & Goldsworthy 1991). Subsets (1–5%) of these
foci eventually progress through a process of neoplastic trans-
formation into overt malignancy (Farber 1986; Pitot et al.
1991). Another outstanding finding in early chemical carcino-
genesis is a multipotential cell called the oval cell, described by
Farber and colleagues (Farber 1956). All of these cellular
events are accompanied by increased expression of several
growth factors that enhance cell survival of carcinogen-
activated cells (Dominguez-Malagon & Gaytan-Graham 2001).
The effect of prolonged cellular damage in carcinogenesis
such as that related to chronic inflammation has become
widely recognized (Ohshima & Bartsch 1994). The oxidants
produced by inflammation may stimulate oncogenes and cell
proliferation. Reactive oxygen species (ROS) are known to be
implicated in both the initiation and promotion of tumours.
Multiple sources of ROS may contribute to a permanent oxi-
dative stressing environment leading to pathophysiological
changes and allowing for the selective growth of preneoplastic
initiated cells (Batlle 1993; Klaunig et al. 1998).
Heme oxygenase (HO) is the rate-limiting enzyme in the
degradation of heme. It breaks down the pro-oxidant heme
into the vasodilator carbon monoxide, free iron and the anti-
oxidant biliverdin (Maines 1997; Foresti & Motterlini 1999;
Tomaro & Batlle 2002). Recently, interest has been intensively
focused on the biological effects of carbon monoxide, bili-
verdin and bilirubin. Not so long ago, these heme-derivative
metabolites were considered toxic waste products; however,
accumulating data suggest that they have antioxidative, anti-
inflammatory, antiapoptotic and signalling properties and
possibly also some immune modulatory function (Wagener
et al. 2001). It was proposed that the inducible isoform HO-1
could play an effective role to counteract the damage caused
by oxidative and nitrosative stress (Maines 1997; Naughton
et al. 2002). HO-1 is induced by host oxidative stress stimuli,
and activation of HO-1 gene expression is considered to be an
adaptive cellular response to different types of chemicals and
mediators that change the redox status of the cell (Motterlini
et al. 2002). Other suggested functions for HO-1 are related to
the regulation of blood pressure (Motterlini et al. 1998;
Abraham et al. 2002) and to neovascularization (Dulak et al.
2002). Macrophages are key participants in angiogenesis, and
its infiltration is correlated to the appearance of histologic
malignancies. Nishie et al. (1999) proposed HO-1 as a marker
for activated macrophage infiltration and for neovasculariza-
tion in human gliomas.
It was also suggested that in oral carcinoma, HO-1 expres-
sion can be used to identify patients with low risk of lymph
nodule metastases (Tsuji et al. 1999). Other authors have
detected HO-1 expression only in tumour cells, and it has
been found that intraarterial zinc protoporphyrin administra-
tion in rat hepatomas considerably suppressed tumour growth
(Doi et al. 1999).
We have developed an experimental mouse model to study
the onset of hepatocarcinogenesis induced by the administra-
tion of p-dimethylaminoazobenzene (DAB) (Gerez et al. 1997;
Caballero et al. 2001).
In our model, DAB metabolization generates free radicals
and triggers the activation of oxidative processes as indicated
by an important and sustained P450 level increase, enhanced
lipid peroxidation and diminished levels of the natural anti-
oxidant defence system (Gerez et al. 1997; Gerez et al. 1998).
We have proposed a possible cytoprotective role of HO-1
during the carcinogenic process. We have shown that HO-1
mRNA levels were greatly induced in the liver of DAB-treated
animals, and that this level of induction persisted as long as the
oxidative damage lasted. However, HO activity would be
modulated by post-transcriptional events (Vazquez et al.
2002).
Based on own previous results about the characterization of
mice with chemically induced HCC by chronic DAB adminis-
tration, we have designed the intoxication protocol used in the
present study to provoke continuous and permanent oxidative
stress, leading to inflammation and consequent liver injury, as
essential components of the carcinogenic process. The aim of
this work was to investigate HO-1 expression and localization
along the different morphological stages occurring during
experimental HCC.
Materials and methods
Reagents
Chemicals were reagent grade and were purchased from Sigma
(St Louis, MO, USA). The following primary polyclonal anti-
bodies, rabbit anti-HO-1 (Stressgen Biotech, Victoria, BC
Canada) and goat anti-actin (I-19) (Santa Cruz Biotech,
Santa Cruz, CA, USA) were used. The secondary antibodies
214 F. Caballero et al.
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 3
used for HO-1 Western blot analysis were antigoat IgGHRP or
antirabbit IgGHRP (Santa Cruz Biotech). The secondary
antibody used for HO-1 immunochemical detection was a
biotinylated linking antirabbit IgG followed by alkaline
phosphatase – labelled streptavidin (Biogenex, San Ramon,
CA, USA).
Animals and treatment
Male CF1 mice weighing 30 g were employed. A group of
animals (n5 100) was placed on dietary DAB (0.5%, w/w)
during a whole period of 14 months. Control animals (n5 40)
were fed with a standard laboratory diet (Purina 3, Asociacion
de Cooperativas Argentinas, San Nicolas, Buenos Aires,
Argentina) for the same period. All animals received food
and water ad libitum. Throughout the study, all animals
were inspected at least twice daily. Body weight and food
consumption were measured at intervals throughout the
study. Food was removed from animals 16 h before killing.
Similar groups of animals (at least three animals per group) on
the same laboratory standard diet, with and without the car-
cinogen, were killed under ether anaesthesia. Animals were
scheduled for killing at 54, 74, 89, 104 and 164 days and at
8, 10, 13 and 14 months.
All animals received humane care and were treated in accord-
ance with the guidelines established by the Animal Care and
Use Committee of the Argentine Association of Specialists in
Laboratory Animals (AADEALC) and in accordance with the
UKCCCR Guidelines for the Welfare of Animals in Experi-
mental Neoplasia (Workman et al. 1998).
Protein extraction and Western blotting
Expression of HO-1 protein was studied by Western blot
analysis. Protein from liver tissues (100 mg chopped in small
pieces) was extracted by using lysis buffer [50 mM Tris–HCl,
pH 6.8; 10% sodium dodecyl sulfate (SDS)] and homogenized.
After 30 min of incubation at 4 �C, the lysates were heated at
100 �C during 5 min and were centrifuged at 10 000 · g for
30 min at 4 �C. Lysates containing equal amounts of proteins
(100 mg) were resolved on 8.5% SDS-polyacrylamide gel elec-
trophoresis. Rainbow coloured protein molecular weight stand-
ards obtained from Amersham were used for the estimation of
molecular size. The proteins were blotted to a Hybond-
Enzyme Chemio Luminiscence (ECL) nitrocellulose membrane
that was probed and washed according to the instructions for
the enhanced chemiluminescence Western blotting detection
system (Amersham Pharmacia Biotech, Little Chalfont, UK),
with transfer buffer (pH 8.3) containing 20% methanol (v/v)
using an Hoefer miniEV electrotransfer unit (Amersham
Pharmacia Biotech). The membrane with transferred proteins
was blocked with 5% serum albumin in Tris-buffered saline
(TBS) containing 0.1% Tween 20 (TBST) for 1 h at room
temperature and incubated with the first antibody diluted in
TBST for 1 h at room temperature. After washing in TBST, the
membrane was incubated with horseradish peroxidase-
labelled secondary antibody for 1 h at room temperature.
For quantification of immunoblots, relative intensities of bands
were quantified by densitometry using IMAGE MASTER image
analysis software (Amersham Pharmacia Biotech). Control for
loading and transfer was obtained by probing with anti-b-actin.
Histology
For histological study, a random section, about 4 mm thick,
was removed from the median, left lateral and right lateral
lobes of the liver. This operation was repeated to examine six
samples from each liver. Additional sections were taken of
grossly visible lesions not included within the above sections.
When gross tumours were observed, sections were invariably
prepared from them for histological characterization. Sections
from other organs were also prepared.
All sections were cut 3–5 mm thick from buffered formalin-
fixed, paraffin-embedded tissue. After deparaffinization, sec-
tions were stained with haematoxylin–eosin (H&E), periodic
acid–Schiff (PAS) reagent, Gomori’s silver impregnation for
reticulin and Prussian blue for iron methods.
Liver foci as well as tumour histology were identified
according to published criteria (Frith et al. 1994).
Immunohistochemistry
Sections obtained from the paraffin blocks used for the histo-
logical diagnosis were deparaffinized with three changes of
xylene for 10 min each; they were hydrated in decreasing con-
centrations of ethanol and rinsed in water. Prior to staining,
the slides were placed in 10 mM sodium citrate buffer pH 6 in a
microwave oven (5· 2 min at 750 W, with a cooling period of
6 min after each treatment).
Sections were pretreated with horse normal serum, diluted
at 1 : 20, for 20 min. Then, they were incubated with: 1) a
rabbit polyclonal antibody against HO-1, diluted at 1 : 500,
overnight; 2) biotinylated antirabbit immunoglobulins serum
diluted at 1 : 15 for 30 min and 3) alkaline phosphatase-
labelled streptavidin diluted at 1 : 15 for 30 min. Alkaline
phosphatase activity was developed with the Fast-Red system
(DAKO). Slides were lightly counterstained with H&E. All
incubations were performed at room temperature, and all
washings were performed with phosphate-buffered saline
buffer (pH 7.5).
Immunohistochemical analysis of HO-1 in hepatocarcinogenesis 215
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 4
Control sections used for determination of antibody reac-
tion specificity included: (a) positive controls (sections of
mouse spleen) and (b) negative controls (serial sections of
each sample omitting the primary antibody).
The HO-1 immunostaining was qualitatively evaluated by
the presence of positive or negative staining, in addition to the
type of cells staining positive.
Statistics
Six to 8 animals per treatment group were used in all experi-
ments. Data were analysed using the Student’s t-test. Results
were considered statistically significant when P < 0.05.
Results
Western blot analysis
Liver samples of animals treated with DAB during a whole
period of 14 months were analysed by Western blotting. HO-1
expression increased significantly along the treatment. In
whole liver, maximal induction (sixfold) was detected after
13 months of DAB administration, whereas in nodules sam-
ples, increase was only 1.7-fold (Figure 1).
Histology – immunohistochemistry
Normal hepatocytes. In the liver of all mice treated with DAB,
HO-1-positive morphological normal hepatocytes were seen;
the immunostaining became more intense with prolonged
treatment (Figure 2a). No increased immunostaining for HO-1
was observed in control animal hepatocytes (data not shown).
Macrophages. Proliferation and enlargement of sinusoidal
lining Kupffer cells was seen after day 54 and became more
pronounced with prolonged treatment. Liver macrophages
surrounding necrotic foci and nodular lesions were seen.
Both types of cells showed golden brown granular and diffuse
brownish yellow pigments, Prussian blue staining and a PAS-
positive reaction.
A positive, intense immunostaining for HO-1, in all liver
macrophages was observed (Figure 2a,b,c,d).
Necrosis. Foci of liver necrosis was first found after 67 days of
DAB feeding; from then onwards, all animals showed foci of
hepatic necrosis of different size and shape at any time up to
14 months.
A marked decreased immunostaining for HO-1 mainly in
hepatocytes from necrotic foci, with an increased staining
in surrounding normal and regenerative hepatocytes and in
macrophages was observed in all cases (Figure 2b).
Proliferation of ductular epithelial cells (‘oval cells’). In DAB-
feeding animals, the ductular epithelial cell (DEC) prolifer-
ation was first visible in portal areas at day 89, then increasing
with time. Duct profiles with well-defined lumina were seen at
month 13.
The proliferated DECs were negative for HO-1 immuno-
staining (Figure 2c) as were biliary epithelial cells in normal
nontreated animals (data not shown).
Nodular lesions. AHF: Hepatic foci of cellular alterations
were a common histopathological finding in this study and
included acidophilic and vacuolated AHF. These lesions were
seen after day 54 up to the end of the assay. There was a
decreased HO-1 immunostaining in AHF cells when compared
with surrounding normal hepatocytes at the different stages
(Figure 2d).
[_________________]Months
Indu
ctio
n of
HO
-1 p
rote
in (
fold
)
0
2
4
6
8
*
**
*
*
Months
CC
HO-1
1.5 5 8 13
1.5 5 8 13
HCC
HCCDAB C DAB C DAB C DAB
[___________________________]
Figure 1 Effect of p-dimethylaminoazobenzene treatment on
heme oxygenase (HO)-1 protein expression. Western blotting was
performed by standard procedures with polyclonal antibody to
HO-1. b-actin was used as a loading control. Other experimental
conditions are as described under Materials and methods. One
typical gel is shown. These results are representative of two
independents experiments. The densitometric data are mean6 SE
of six animals. ‘*’ indicates data significantly different (P < 0.05)
from that of control.
216 F. Caballero et al.
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 5
(a)
(c) (d)
(f)
(h)
(e)
(g)
(b)
(i) (j)
Figure 2 Representative findings of
heme oxygenase (HO-1) immunoreac-
tivity. (a) HO-1 expression in non-
atypical hepatocytes and in hyperplastic
Kupffer cells (day 104) (original mag-
nification ·250). (b) Negative HO-1
expression in necrotic hepatocytes
(arrow) with intense HO-1 expression
in marginal regenerative hepatocytes
and hyperplastic Kupffer cells (original
magnification ·250). (c) Isolated
ductular epithelial cell proliferation and
poor defined duct profiles, HO-1
immunostaining negative (original
magnification ·250). (d) Decreased
HO-1 expression in vacuolated AHF
surrounded by intense positive HO-1
macrophages (original magnification
·250). (e) Marked decreased HO-1
expression in adenomatous cells with
intense positive macrophages (original
magnification ·400). (f) Intense diffuse
HO-1 expression in trabecular hepato-
cellular carcinoma (original magnifica-
tion ·250). (g) Negative HO-1
expression in atypical cells from an
infiltrating undifferentiated hepatocar-
cinoma with weak positive comprised
hepatocytes (original magnification
·400). (h) Increased HO-1 expression
in nontumour hepatic tissue (bottom
half) at edge of carcinoma (arrow)
(original magnification ·250). Com-
parative heme oxygenase (HO-1)
expression according to carcinoma dif-
ferentiation degree. (i) Heterogeneous,
weak to negative immunostaining in
poorly differentiated trabecular hepa-
tocarcinoma (original magnification
·250). (j) Homogeneous, intense
immunostaining in well-differentiated
trabecular hepatocarcinoma (original
magnification ·250).
Immunohistochemical analysis of HO-1 in hepatocarcinogenesis 217
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 6
Neoplasms: The neoplasms found, in agreement with the
proposed classification, were identified as hepatocellular
adenoma and hepatocellular carcinoma.
The adenomas, mainly eosinophilic and vacuolated, exam-
ined at day 104 and later, were small and existed as distinct
nodules with well-differentiated hepatocytes. A marked
decreased or negative HO-1 immunostaining in adenomatous
tissue as compared to surrounding nonadenomatous tissue
was observed (Figure 2e).
Hepatocellular carcinomas were not obvious until 10
months. They were usually considerably larger and more irregu-
lar nodules. Several histological varieties of the trabecular
hepatocellular carcinoma were found; some of them showed
a well-differentiated pattern with a glandular appearance, and
others were well demarcated but poorly differentiated
tumours. Hepatoblastomas or cholangiomas were not found.
In hepatocellular carcinoma, various patterns of HO-1
immunostaining were observed (Figure 2f,g). In all nodular
areas (AHF, adenomas and carcinomas), the HO-1-positive
immunostaining was always of lower intensity than in the
surrounding normal hepatocytes (Figure 2h).
A decreased or negative relationship was found in poorly
differentiated tumours between the immunohistochemical
expression of HO-1 and tumour differentiation degrees;
instead, more differentiated tumours exhibited varying and
increased positive immunostaining patterns (Figure 2i,j).
No pertinent changes occurred in any organ other than liver
and the spleen, which at late stages showed marked enlarge-
ment with hypertrophy, mainly of the white splenic pulp.
Nodular lesions were not identified in sections from livers of
control animals at any time up to 14 months after laboratory
standard diet feeding. A minimal fatty change but no liver
necrosis or DEC proliferation was seen. Proliferated and pigment
containing macrophages were not observed at any stage.
Discussion
The hepatocarcinogenic protocol used here was designed to
induce tumours as a consequence of prolonged inflammation
and cell damage after oral administration of a carcinogenic
azo dye to mice. The model was very effective in terms of
tumour incidence (100%) and in reproducing all the steps in
the carcinogenic process. The continuous feeding with DAB
was useful to study the morphological changes that take place
during HCC. It was possible to identify early preneoplastic
lesions, which were characterized by proliferation of pheno-
typically AHF and oval cells. The emergence of such cells
generally precedes the development of HCC (Dominguez-
Malagon & Gaytan-Graham 2001). Oval cells proliferate in
response to several carcinogens and toxic agents (Saeter &
Seglen 1990). Although oval cells may be facultative liver
stem cells, they have the potential of evolving into tumours,
histologically identified as hepatocellular carcinoma (Braun
et al. 1989); recently, it was shown that preneoplastic foci
can originate from mature hepatocytes; these findings are
consistent with the hypothesis that dedifferentiation of mature
hepatocytes may occur during the course of a carcinogenic
regimen (Gourny et al. 2002).
It is well known that three stages are identified in HCC:
initiation, promotion and progression (Pitot et al. 1996).
Initiation of cellular transformation involves an irreversible
genetic change, occurring either spontaneously or as a result
of exposure to chemical or physical agents. During the promo-
tion stage which depends on the continuous presence of a
promoter agent (Dominguez-Malagon & Gaytan-Graham 2001),
selected cells clonally expand into AHF which grow progres-
sively as compared with surrounding normal tissue (Seo et al.
1988). The progression stage may be characterized by the
development of malignant neoplasm (Pitot & Dragan 1991).
The present study indicates that chronic exposure to DAB
promotes AHF at day 54 and DEC proliferation in portal areas
at day 89, both of these events are characteristic features of
early preneoplastic hepatic lesions. A significant finding was
foci of liver necrosis after 67 days of DAB feeding. In this in
vivo bioassay, the alterations provoked by the carcinogen
would be fixed through cell proliferation, reflected by hepato-
megaly and in response to necrosis, as we have previously
reported (Caballero et al. 2001; Vazquez et al. 2002). Under
the experimental conditions described here, we demonstrated
the presence of adenomas at day 104 and hepatocellular car-
cinomas at 10 months of DAB feeding. In a recent study, we
employed an experimental design administrating tamoxifen
citrate to promote hepatocarcinogenesis initiated with DAB
in mice. Animals also developed trabecular hepatocellular
carcinoma (Caballero et al. 2001; Vazquez et al. 2002).
One important component observed in the histological char-
acterization of this model was the proliferation and enlarge-
ment of Kupffer cells from day 54 as well as the presence of
macrophages surrounding foci of necrosis and nodular lesions,
indicative of the inflammatory response occurring under the
current protocol.
Oxidative stresses such as oxidant stimuli, inflammation,
exposure to xenobiotics and ionizing irradiation elicit various
tissue injuries and provoke cellular responses, principally
involving transcriptional activation of genes encoding proteins
which participate in the defence reactions (Camhi et al. 1995).
Chronic inflammation is leading to cancer in animals, and it is
therefore a risk factor of human cancer (Dominguez-Malagon
& Gaytan-Graham 2001). The incidence in liver cancer is low
218 F. Caballero et al.
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 7
in humans (but not in some strains of mice) unless the liver is
chronically damaged (Ames & Gold 1990).
HO-1 plays a major protective role against oxidant stimuli
(Clark et al. 2000; Tanaka et al. 2003). The strong adaptative
response of HO-1 expression on the diverse array of stress
stimuli suggests an important role for HO other than heme
degradation (Maines 1997; Maines 2000). In fact, the over-
expression of HO-1 is associated with the resolution of inflam-
mation and may act as a feedback mechanism (Wagener et al.
2001). Enhancement of HO-1 expression by gene transfer
provides cellular resistance against haemoglobin-heme toxicity
(Abraham et al. 1995; Yang et al. 1999). Recent observations
suggest that the mechanism of HO-1-mediated control of
inflammation may originate from the modulation of adhesion
molecule expression (Willis et al. 1996; Wagener et al. 1997).
Our results reveal and underline the importance of HO-1
expression in normal tissue, and its up-regulation suggests a
cell-protective response to the stress provoked by the gen-
eration of ROS during the carcinogen metabolism (Caballero
et al. 2001; Caballero et al. 2002). Similarly, HO-1 overexpression
in Kupffer cells and macrophages surrounding necrotic and
nodular areas very likely reflects the protective role of this
protein to the stress reaction triggered by inflammation.
Our findings are consistent with recent data reported
about the induction of HO-1 in isolated hepatocytes and
macrophages during acetaminophen-induced hepatotoxicity
(Chiu et al. 2002). Similarly, transgenic mice lacking func-
tional HO-1 are more sensitive to endotoxin-induced tox-
icity and highly susceptible to the development of chronic
liver inflammation (Poss & Tonegawa 1997).
Western blot analysis of whole liver samples confirmed the
induction of HO-1 expression along the experimental trial
with DAB.
The in vivo bioassay used here turned out to be a valuable
model to follow the complex cascade of morphological events
and the associated cellular response of HO-1. Within this
context, an important observation was the decrease of HO-1
immunostaining in early preneoplastic lesions (AHF) and the
marked decrease of HO-1 immunoreactivity in nodular lesions
compared to surrounding nontumoural tissue. An attenuated
response of HO-1 in nodules was also documented by employ-
ing Western blot analysis.
Furthermore, the relationship between HO-1 immuno-
reactivity and tumour differentiation was a relevant finding.
To the best of our knowledge, this is the first report on the
association between diminution of HO-1 expression and the
neoplastic transformation, and these findings give further sup-
port to the assigned protective role of HO-1 to oxidative stress
(Yang et al. 1999; Motterlini et al. 2000; Otterbein & Choi
2000; Vazquez et al. 2002).
Previous reports have demonstrated that HO-1 overexpression
is cytoprotective by attenuating the oxidative stress-mediated
pro-inflammatory reaction cascade. Within this context, HO-1
overexpression constitutes an alternative therapeutic approach to
disrupt inflammatory tissue deterioration occurring in a wide
range of diseases (Wang et al. 1998).
More promising is the observation that down modulation of
HO-1 expression correlated with malignancy progression in
chemical experimental carcinogenesis. Our data emphasize the
protective role of HO-1 in this model of chronic liver damage
and point to activation of HO-1 as a potential therapeutic tool
in liver cancer.
Acknowledgements
We are very grateful to Dr Nora Navone, Department of
Genitourinary Oncology, MD Anderson Cancer Center,
University of Texas, Houston, Texas, USA for generously
providing some reagents. We also thank Mrs B. Corvalan and
M. Binaghi for valuable technical assistance. A. Batlle and
E. Vazquez are members of the Career of Scientific Researcher
at the Argentine National Research Council (CONICET).
F. Caballero holds the post of Research Assistant at the
CONICET. R. P. Meiss is Chief and A. M. Gimenez is staff
member of Department of Pathology at the National Academy
of Medicine, Argentina. This work has been supported by grants
from the CONICET; the University of Buenos Aires, Argentina;
AICR, UK; and the Science and Technology Argentine Agency.
References
Abraham N., Botros F., Rezzani R., Rodella L., Bianchi R.,
Goodman A. (2002) Differential effect of cobalt protoporphyrin
on distributions of heme oxygenase in renal structure zand on
blood pressure SHR. Cell Mol. Biol. 48, 895–902.
Abraham N., Lavrovsky Y., Schwartzman M. et al. (1995)
Transfection of the human heme oxygenase gene into rabbit
coronary microvessel endothelial cells: protective effect against
heme and hemoglobin toxicity. Proc. Natl. Acad. Sci. USA 92,
6798–6802.
Ahn B., Han B., Kim D., Ohshima H. (1999) Immunohistochem-
ical localization of inducible nitric oxide synthase and
3-nitrotyrosine in rat liver tumours induced by N-nitroso-
diethylamine. Carcinogenesis 20, 1337–1344.
Ames B. & Gold L. (1990) Chemical carcinogenesis: too many
rodent carcinogens. Proc. Natl. Acad. Sci. USA 87, 7772–7776.
Batlle A. (1993) Porphyrins, porphyrias, cancer and photo-
dynamic therapy – a model of carcinogenesis. J. Photochem.
Photobiol. B. 20, 5–22.
Immunohistochemical analysis of HO-1 in hepatocarcinogenesis 219
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 8
Boone C., Kelloff G., Steele V. (1992) Natural history of
intraepithelial neoplasia in humans with implications for
cancer chemoprevention strategy. Cancer Res. 52, 1651–1659.
Braun L., Mikumo R., Fauston N. (1989) Production of
hepatocellular carcinoma by oval cells: cell cycle expression
of c-myc and p53 at different stages of oval cell transformation.
Cancer Res. 49, 1554–1561.
Butterworth B. & Goldsworthy T. (1991) The role of cell
proliferation in multistage carcinogenesis. Proc. Soc. Exp. Biol.
Med. 198, 683–687.
Caballero F., Gerez E., Batlle A., Vazquez E. (2002) Interaction of
cimetidine with P450 in a mouse model of hepatocarcinogen-
esis initiation. Br. J. Cancer 86, 630–635.
Caballero F., Gerez E., Oliveri L., Falcoff N., Batlle A., Vazquez E.
(2001) On the promoting action of tamoxifen in a model of
hepatocarcinogenesis induced by p-dimethylaminoazobenzene in
CF1 mice. Int. J. Biochem. Cell Biol. 33, 681–690.
Camhi S., Lee P., Choi A. (1995) The oxidative stress response.
New Horiz. 3, 170–182.
Chiu H., Brittingham J., Laskin D. (2002) Differential induction
of heme oxygenase-1 in macrophages and hepatocytes during
acetaminophen-induced hepatotoxicity in the rat: effects of
hemin and biliverdin. Toxicol. Appl. Pharmacol. 181,
106–115.
Clark J., Foresti R., Green C., Motterlini R. (2000) Dynamics of
haem oxygenase-1 expression and bilirubin production in
cellular protection against oxidative stress. Biochem. J. 348,
615–619.
Doi K., Akaike T., Fujii S. et al. (1999) Induction of haem
oxygenase-1 nitric oxide and ischaemia in experimental solid
tumours and implications for tumour growth. Br. J. Cancer 80,
945–954.
Dominguez-Malagon H. & Gaytan-Graham M. (2001) Hepato-
cellular carcinoma: An update. Ultrastruct. Pathol. 25,
497–516.
Dulak J., Jozkowicz A., Foresti R. et al. (2002) Heme oxygenase
activity modulates vascular endothelial growth factor synthesis
in vascular smooth muscle cells. Antioxid. Redox Signal. 4,
229–240.
Farber E. (1956) Similarities in the sequence of early histologic
changes induced in the livers of rats by ethionine,
2-acetylaminofluorene and 3´methyl-4-dimethylaminoazoben-
zene. Cancer Res. 16, 142–148.
Farber E. (1986) Some emerging general principles in the
pathogenesis of hepatocellular carcinoma. Cancer Surv. 5,
698–718.
Foresti R. & Motterlini R. (1999) The heme oxygenase pathway
and its interaction with nitric oxide in the control of cellular
homeostasis. Free Radic. Res. 31, 459–475.
Frith C., Ward J., Turusov V. (1994) Tumours of the liver:
tumours of the mouse. In: Pathology of Tumours in Laboratory
Animals, Vol. 2 (eds. V. Turusov, U. Mohr). Lyon: IARC
Scientific Publications No. 111, International Agency for
Research on Cancer, pp. 223–269.
Gerez E., Caballero F., Vazquez E., Polo C., Batlle A. (1998)
Hepatic enzymatic metabolism alterations and oxidative stress
during the onset of carcinogenesis: protective role of alpha-
tocopherol. Eur. J. Cancer Prev. 7, 69–76.
Gerez E., Vazquez E., Caballero F., Polo C., Batlle A. (1997)
Altered heme pathway regulation and drug metabolizing
enzyme system in a mouse model of hepatocarcinogenesis:
effect of veronal. Gen. Pharmacol. 29, 569–573.
Gourny J., Auvigne I., Pichard V., Ligeza C., Bralet M., Ferry N.
(2002) In vivo cell lineage analysis during chemical hepato-
carcinogenesis in rats using retroviral-mediated gene transfer:
evidence for dedifferentiation of mature hepatocytes. Lab.
Invest. 82, 781–788.
Klaunig J., Xu Y., Isenberg J. et al. (1998) The role of oxidative
stress in chemical carcinogenesis. Environ. Health Perspect.
106, 289–295.
Maines M. (1997) The heme oxygenase system: a regulator of
second messenger gases. Annu. Rev. Pharmacol. Toxicol. 37,
517–554.
Maines M. (2000) The heme oxygenase system and its functions
in the brain. Cell Mol. Biol. 46, 573–585.
Motterlini R., Foresti R., Bassi R., Green C. (2000) Curcumin,
an antioxidant and anti-inflammatory agent, induces heme
oxygenase-1 and protects endothelial cells against oxidative
stress. Free Radic. Biol. Med. 28, 1303–1312.
Motterlini R., Gonzales A., Foresti R., Clark J., Green C.,
Winslow R. (1998) Heme oxygenase-1-derived carbon mon-
oxide contributes to the suppression of acute hypertensive
responses in vivo. Circ. Res. 83, 568–577.
Motterlini R., Green C., Foresti R. (2002) Regulation of heme
oxygenase-1 by redox signals involving nitric oxide. Antioxid.
Redox Signal. 4, 615–624.
Naughton P., Hoque M., Green C., Foresti R., Motterlini R.
(2002) Interaction of heme with nitroxyl or nitric oxide
amplifies heme oxygenase-1 induction: involvement of the
transcription factor Nrf2. Cell Mol. Biol. 48, 885–894.
Nishie A., Ono M., Shono T. et al. (1999) Macrophage
infiltration and heme oxygenase-1 expression correlate with
angiogenesis in human gliomas. Clin. Cancer Res. 5,
1107–1113.
Ohshima H. & Bartsch H. (1994) Chronic infections
and inflammatory processes as cancer risk factors: possible role
of nitric oxide in carcinogenesis. Mutat. Res. 305, 253–264.
Otterbein L. & Choi A. (2000) Heme oxygenase: colors of
defense against cellular stress. [review]. Am. J. Physiol. Lung
Cell Mol. Physiol. 279, L1029.
Pitot H. & Dragan Y. (1991) Facts and theories concerning the
mechanisms of carcinogenesis. FASEB J. 5, 2280–2286.
Pitot H., Dragan Y., Sargent L., Xu Y. (1991) Biochemical markers
associated with the stages of promotion and progression during
220 F. Caballero et al.
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221
Page 9
hepatocarcinogenesis in the rat. Environ. Health Perspect. 93,
181–189.
Pitot H., Dragan Y., Teeguarden J., Hsia S., Campbell H. (1996)
Quantitation of multistage carcinogenesis in rat liver. Toxicol.
Pathol. 24, 119–128.
Poss K. & Tonegawa S. (1997) Heme oxygenase 1 is required for
mammalian iron reutilization. Proc. Natl. Acad. Sci. USA 94,
10919–10924.
Saeter G. & Seglen P. (1990) Cell biology of hepatocarcinogen-
esis. Crit. Rev. Oncog. 1, 437–466.
Seo M., Lynch K., Podolsky D. (1988) Multiplicity of transform-
ing growth factors in human malignant effusions. Cancer Res.
48, 1792–1797.
Tanaka S., Akaike T., Fang J. et al. (2003) Antiapoptotic effect of
haem oxygenase-1 induced by nitric oxide in experimental solid
tumour. Br. J. Cancer 88, 902–909.
Tomaro M. & Batlle A. (2002) Bilirubin: its role cytoprotection
against oxidative stress. Int. J. Biochem. Cell Biol. 34,
216–220.
Tsuji M., Yanagawa T., Iwasa S. et al. (1999) Heme oxygenase-1
expression in oral squamous cell carcinoma is involved in
lymph node metastasis. Cancer Lett. 138, 53–59.
Vazquez E., Gerez E., Caballero F. et al. (2002) On the promot-
ing action of tamoxifen in the p-dimethylaminoazobenzene
induced hepatocarcinogenesis CF1 mice model and the cyto-
protective role of heme oxygenase. In: Heme Oxygenase
in Biology and Medicine. (eds. N. Abraham, J. Alam, K.
Nath). New York: Kluwer Academic Plenum Publishers,
pp. 469–479.
Wagener F., Eggert A., Boerman O. et al. (2001) Heme is a potent
inducer of inflammation in mice and is counteracted by heme
oxygenase. Blood 98, 1802–1811.
Wagener F., Feldman E., de Witte T., Abraham N. (1997) Heme
induces the expression of adhesion molecules ICAM-1, VCAM-1,
and E selectin in vascular endothelial cells. Proc. Soc. Exp. Biol.
Med. 216, 456–463.
Wang L., Lee T., Lee F., Pai R., Chau L. (1998) Expression of
heme oxygenase 1 in atherosclerotic lesions. Am. J. Pathol.
152, 711–720.
Willis D., Moore A., Frederick R., Willoughby D. (1996) Heme
oxygenase: a novel target for the modulation of the inflamma-
tory response. Nat. Med. 2, 87–90.
Workman P., Twentyman P., Balkwill F. et al. (1998)
United Kingdom Co-Ordinating Committee on Cancer
Research (UKCCCR) Guidelines for the Welfare of Animals
in Experimental Neoplasia (Second Edition). Br. J. Cancer 77,
1–10.
Yang L., Quan S., Abraham N. (1999) Retrovirus-mediated HO
gene transfer into endothelial cells protects against oxidant-
induced injury. Am. J. Physiol. 277, L127.
Immunohistochemical analysis of HO-1 in hepatocarcinogenesis 221
� 2004 Blackwell Publishing Ltd, International Journal of Experimental Pathology, 85, 213–221