Cancer: The Role of Genetic Collaboration
Summary: Philip Leder is interested in understanding the genetic interactions that give rise to cancer.
Cancer is a profound disorder of cell growth and migration in which
the delicate balance established by genetically encoded programs of
regulation is disrupted. Instead of reaching an equilibrium, cancerous
cells no longer respond to signals that limit their ability to divide
and migrate beyond their normal confines. In the end, their growth and
invasive potential are out of control, and this loss has profoundly
dangerous consequences for the organism.
Over the past two decades, it has become increasingly clear that
many cancers can be accounted for, at least in part, by damage
occurring to genes that encode the rules for control of cell growth.
Genetic damage, or mutation, can inactivate a gene or cause it to
function at the wrong time or place or to make the wrong product. The
genes in which mutations can give rise to cancer are often those that
normally regulate cell growth. Geneticists refer to the damaged genes
as oncogenes (from the Greek onkos, or tumor).
Undamaged, some of these genes have profound effects on processes that
guide the formation of the early embryo or control the cycle of events
through which a cell passes as it prepares for and undergoes
For some time, my laboratory has been interested in genes that can
contribute to the development of cancer, especially in identifying
genes that specifically collaborate with one another to bring about
malignant transformation. Our work has been considerably advanced by
the technique of introducing active oncogenes into special strains of
laboratory mice. These "transgenic" mice carry oncogenes created in the
laboratory and pass these cancer-causing genes to offspring,
transmitting a strong tendency to develop cancer. Thus, in many ways
transgenic mice become useful models of human malignancy.
Genetically engineered mice that develop cancer of the breast and of
the blood cells (leukemias and lymphomas) have demonstrated that
certain cancers can be caused by specific oncogenes and that many, but
not necessarily all, cancers are the result of a collaboration between
two or more oncogenes. This suggests that cancer is often a "multihit"
process, one that requires several activating events.
One important question concerns our ability to identify sets of
oncogenes that constitute collaborating partners in the process of
tumor induction. One transgenic mouse line that we have constructed
contains a gene for one of the fibroblast growth factors, a gene that
encodes a biologic growth-inducing factor and that is amplified in many
cases of human breast cancer. When this gene is expressed in breast
tissue, it induces the proliferation of mammary epithelial cells, cells
that normally form the functioning mammary gland. The unregulated
expression of this gene, called fgf7, causes breast cancer to
arise in these mice. Such tumors arise when the mice are about 1 year
old (equivalent to a midlife mature female).
Since fgf7 is expressed in all mammary cells during the early
life of the mouse, its expression alone cannot account for the
formation of breast tumors. It is necessary, but not sufficient. Our
hypothesis is that other genes, the collaborating partner oncogenes,
must undergo mutation during this latency period and, when such a
mutation occurs in the precursor cell of the tumor, the precursor cell
becomes malignant, outgrows normal epithelial cells, and grows out as a
If this model is correct, the next goal is to identify the
collaborating partner gene(s). The first required gene is obviously the fgf7 transgene. The second, and perhaps third and fourth, are
genes that have spontaneously undergone mutation over the year or so it
takes for the tumor to develop. We have begun to identify these
additional genes by using a virus known to infect mammary epithelial
cells and integrate among or within genes encoded by such cells. This
mouse mammary tumor virus (MMTV) can "hop" into or near virtually any
gene encoded by the mouse, destroying its function or disrupting its
normal pattern of expression. In either case, the virus can be looked
at as a mutagenic agent, a so-called insertional mutagen. If the
mutation it causes is in a potential oncogene that can collaborate with
the initiation oncogene (in this case, for example, the activated
fgf7 gene), the mammary cell in which this insertional event has
occurred will proliferate and become a malignant tumor.
As expected, transgenic mice that are infected with MMTV develop
tumors faster than the uninfected transgenic mice. In addition, since
we have genetic tools that allow us to find the region of the mouse
genome into which the MMTV has inserted, we can identify the nearby
perturbed gene that must be one of the collaborating partners of the
We have applied this protocol to a large number of tumors and found
that the perturbed partner genes are far from a random assortment.
Rather they are very frequently members of the Wnt family of genes.
These genes play a key role in early proliferation, differentiation,
and development in a number of embryonic organs. Our experiments
suggest, however, that their misexpression can contribute to a loss of
growth control and, in partnership with the misexpression of the
fgf7 gene, can contribute to the development of cancer. These
experiments led to the discovery of a novel member of the Wnt family, Wnt10b, which has been shown to be capable of inducing mammary
gland hyperplasia and, ultimately, mammary carcinoma as well.
Our initial success has led us to extend these studies to another
initiating oncogenic transgene. Overexpression of Her2 is
associated with more than 30 percent of human breast carcinomas and
correlates with a poorer prognosis with respect to recurrence and
survival. The clinical relevance of Her2 and the fact that we
have been able to create excellent transgenic mouse models using its
rat homolog neu make it particularly attractive for further
investigation. The virus-infected transgenic mice develop breast tumors
more rapidly than noninfected transgenic or infected normal mice. These
experiments have led us to the oncogenic collaborators of this
clinically important breast cancer gene, one of which is our original
oncogenic candidate, fgf7.
In addition to identifying collaborating oncogenic partners of
neu, we have also devised a high-throughput, cell-based assay
that allows us to riffle through libraries of thousands of small
chemical compounds to select those that prevent proliferation of tumor
cells that overexpress neu. Using this assay, we have identified
a number of novel compounds that block the growth of neu-based
mammary tumors but have no effect on normal cells. One of these
compounds, effective at low concentrations, unexpectedly blocks growth
by inflicting damage on the cell's energy-producing elements, its
mitochondria. This compound or compounds related to it might someday be
useful in the treatment of breast cancer and other malignancies. For
example, early studies suggest that this compound inhibits tumor growth
in living mice.
Portions of these studies have been supported by funds from the
Harvard University endowment, the American Cancer Society, the Damon
RunyonWalter Winchell Cancer Foundation, the Department of the
Army, and the Helen Hay Whitney Foundation.
Last updated August 01, 2003