Response to Expert
Commentaries on Epigenetic Mechanisms of Chemical Carcinogenesis
James E. Klaunig, Ph.D.
Division of Toxicology, Department of Pharmacology and Toxicology, Indiana
University School of Medicine, 635 Barnhill Drive - MS-1021, Indianapolis, IN 46202-5120
Overall, we found it difficult to respond to this critique. We disagree
with the comment that "if pigs had wings, could they fly, . indicating that
research conducted by academia. .has little relevance to the real world".
Academic institutions train toxicologists with an understanding of the concepts
involved in the toxicological process including carcinogenesis. How
mechanistically based data produced in the academic setting is utilized is many times out of
the control of the academician. Unfortunately, for a variety of reasons,
information that is produced in a pure scientific and investigative format is
misinterpreted when used to understand the potential safety or risk of a chemical agent
or drug. Hopefully, a well-trained toxicologist will be able to understand the
differences in quality and application of the data produced, thus, leading to
a logical and scientifically-based assessment of the risk or safety.
With regard to this reviewer's concern about the paradigm of cancer as
a 1 to 3 stage process, the utilization of three stages to describe the
development of neoplasia has been historically applied, and thus was the foundation
by which we discussed the mechanistic processes that impact on these stages. It
is not irrelevant to the human setting, since the multistage nature of cancer
appears to also occur in humans.
We take exception to this responder's comment of our discussion
of cell foci and the role of replication in cell foci. While we respect
this reviewer's personal opinion it lacks proper citation and critical evaluation
of literature, modification of cell growth in preneoplastic lesions is important
in the stepwise sequence of chemical carcinogenesis. Similarly, the
dismissing of gap-junctional communication and altered methylation as important
factors in the cancer process is not justified. An extensive experimental
database showing these two cell functions to be important in carcinogenesis
exists.
We agree with Dr. Goodman's assessment that the role of epigenetic
regulation in gene expression is an important component of the carcinogenesis
process. As Dr. Goodman and his co-workers have shown during the past
ten years, hyper- or hypo-methylation of important genes in the
carcinogenesis process may facilitate the expression, or suppression, of cell growth
and apoptosis. With regard to chemical carcinogenesis, while Dr.
Goodman's premise requires further substantiation with additional carcinogens, the
available data with the model compounds examined by Goodman and
co-workers provides a strong foundation for the role of epigenetic events in general,
and DNA methylation in particular, in modulating changes from a normal
to preneoplastic to neoplastic state.
Dr. Preston's response is, for the most part, in line with our presentation.
The article discussed by Dr. Pressman by Hanahan and Weinberg is not in
conflict with the review we presented. More specifically, the Hanahan
and Weinberg model details the possible cellular and molecular mechanisms
that occur during the change from a preneoplastic to neoplastic state. Our
review, while simple in context, should not be taken to imply that the development
of human cancer is a simple mechanism, but rather involves multistages and
multiple mechanisms. The proposal by Hanahan and Weinberg, as eloquently
summarized by Dr. Preston, fits within the context of epigenetic mechanisms
of action. Again, while DNA mutation may be necessary at points in the
cancer process, changes in the cellular environment that in turn modify gene
expression, cell proliferation, and/or cell death, are important in transferring a
modified preneoplastic cell to the neoplastic and eventually the malignant state.
The response by Dr. De Angelo is consistent with our presentation on
the importance of the balance between cell proliferation and cell death in
chemical carcinogenesis. As noted by Dr. De Angelo, the development of cancer, in
particular liver cancer, can occur through multiple mechanisms. Our review
was, by nature, limited to an overview, As Dr. De Angelo noted his own studies
with dichloroacetic acid has shown the importance of modification of altered
gene expression, effects on cell-to-cell communication, and production of
oxidative damage to the cell, in leading to modification of cell proliferation and/or
cell death. While, as rightly pointed out by Dr. De Angelo, liver cancer may not
be predicted strictly by an increase cell proliferation, we believe that alteration
of cell growth (changes in the balance of apoptosis and cell division)
in preneoplastic, focal hepatocytes in the liver, resulting in an increase
in preneoplastic cell number correlates with the transference from a
preneoplastic to neoplastic state. This modification in cell growth is dependent on changes
in gene expression, as well as possible changes in the cellular and molecular
machinery of the cell. As Dr. DeAngelo notes in his final paragraph,
"mechanistic research must be conducted and related to the cancer bioassay.", to
allow for clear understanding of the mode and mechanism of action of each
chemical on neoplasms produced in the target species.
If a mechanism, such as oxidative stress, is established as the means by
which an agent may produce its carcinogenic effect, then given that the normal
reactive oxygen species "load" on the cell is detoxified by antioxidants, an
overload, either through the reduction of antioxidants, or an increase in reactive
oxygen species, could result in oxidative lipid, protein, or DNA damage. Based on
a number of reported studies, it is apparent that an overload of antioxidant
systems in a cell must occur for the oxidative damage to be manifested. This,
by our definition, is a threshold event, indicating no effect at doses of the
chemical that do not overload the system, while producing an effect by doses of the
compound that overload the system. As rightly pointed out by Dr. Lutz, the
prediction of a population-derived threshold may be difficult to make, because
of differences in genetic, metabolic and dietary/environmental influences.
This does not preclude however, the fact that the mode, or mechanism, by which
an agent produces a chemical response, may not exhibit a threshold. Thus
the application, of a threshold concept to the population at large, falls under
the province of dose response and risk assessment.
Tel: 317-274-7824,
Fax: 317-274-7787
E-mail:jklauni@iupui.edu
Response to Dr. Carl Alden:
Response to Dr. Jay Goodman:
Response to Dr. Julian Preston:
Response to Dr. Anthony DeAngelo:
Response to Dr. Werner Lutz:
Dr. Lutz's response that "because of genetic and lifestyle-dependent
susceptibility differences, each animal or human has its individual threshold dose.
And, therefore, for a group no single threshold dose can be defined,
irrespective of mode of action of carcinogen action". While we agree that differences
in response by animals and humans to the carcinogenic effect of a specific
chemical is influenced by individual animal or human genetics, diet, co-exposure
to other chemicals (i.e. tobacco smoke, alcohol) thus, making individual
susceptibilities to a carcinogen different, this does not take from the fact that the
development and understanding of the mechanisms, or mode of action, of a
carcinogen may exhibit a threshold for a response. We believe this is true of not
only epigenetic or non-genotoxic chemicals, but is also seen with DNA
reactive agents. In fact, a threshold response is readily observed in an individual
animal by the differences in the neoplastic response of various tissues.