Human Molecular Genetics, 2001, Vol. 10, No. 7 655-656
© 2001 Oxford University Press
The molecular genetics of cancer: down the rabbit hole
Chris Gunter+
Department of Genetics, Case Western Reserve University, 2109 Adelbert Road, BRB 701, Cleveland, OH 44106, USA
Received 6 February 2001; Accepted 6 February 2001.
 |
OVERVIEW
|
|---|
Like Alice in Wonderland, researchers in the molecular genetics
of cancer may feel that each discovery leads to a ‘curiouser
and curiouser’ set of questions. The initial celebration
that follows the discovery of a gene mutation associated with
a cancer soon gives way to perplexity upon learning that mutations
in this gene only account for some (usually small) percentage
of cancer cases. In addition, different mutations in the same
gene do not always cause the same phenotype, or respond in the
same way to therapeutic agents, making patient care increasingly
difficult. Epigenetic phenomena that are potentially reversible
further complicate the matter, as do gene–environment
interactions. Nevertheless, there is a strong motivation to
study these disorders, for they teach us how all cells work
by showing how processes can go wrong, and have the potential
to lead to crucial breakthroughs in care and treatment. This
special review issue first examines new tools that researchers
are using to make these discoveries, followed by epigenetic
phenomena that are common to a number of cancers. Next, this
issue explores the latest developments in two ‘complex’
syndromes, breast cancer and colon cancer. Finally, several
reviews update the new complexities involved in a number of
‘single-gene’ cancers (are they really?), and then
detail the state-of-the-art in gene therapy for one set of cancers,
specifically involving the brain.
|
ENHANCING THE TOOLBOX
|
|---|
Large-scale genomic approaches are most useful to compare and
contrast many variables. Kallioniemi
et al. (this issue, pp. 657–662)
describe a novel technology for looking at DNA, RNA (shown on
the cover) or protein levels in a number of cancerous tissues
at once. These tissue microarrays are similar in theory to oligonucleotide
microarrays being used in many genome-wide studies; however,
these arrays are more versatile and allow examination of similar
tissues from many individuals in one experiment. Similarly,
the mission of the Cancer Genome Anatomy Project [CGAP; detailed
by Riggins
et al. (this issue, pp. 663–667)] is ‘to
decipher the molecular anatomy of the cancerous cell’.
CGAP examines the transcribed sequences from a number of human
and mouse cancer cells, and catalogs both markers in these transcripts
and the expression patterns of these transcripts between cancer
types. Exploring the next step, Resor
et al. (this issue, pp.
669–675)
review the use of the mouse as a model system
for many human cancers, including novel techniques that allow
for tissue- or age-specific expression of genes.
|
COMMON ELEMENTS IN CANCER
|
|---|
In order to grow in an uncontrolled fashion, cancerous cells
must overcome a number of obstacles put in place to determine
normal cell growth. One of these, re-expression of telomerase,
is described by Shay
et al. (this issue, pp. 667–685),
who also suggest that assays for telomerase expression might
be used to diagnose a number of cancers at earlier stages. Similarly,
to attain uncontrolled growth, cells must inactivate tumor suppressor
genes such as
APC,
BRCA1,
Rb and
vHL. Recently Baylin
et al.
(this issue, pp. 687–692) have shown that cells can exploit
epigenetic phenomena to accomplish this task, specifically by
hypermethylation of the gene’s promoter and thus silencing
of transcription. These authors also submit that assays for
hypermethylation can reveal early stages of cancer progression.
DNA methylation works in concert with histone hypoacetylation
to maintain silence at critical sequences for gene expression,
and the review by Wade (this issue, pp. 693–698)
reveals the unfortunate consequences of the inappropriate deacetylation
of regulatory chromatin elements. One class of these histone
deacetylases, including HDAC1 and HDAC2, is affected by the
disruption of the Rb (retinoblastoma) family of proteins. Indeed,
as detailed by Nevins (this issue, pp. 699–703), the Rb/E2F
pathway represents another crucial growth checkpoint that has
been overcome by almost all cancerous cells.
|
COMPLEX SYNDROMES: TWO HITS OR MANY?
|
|---|
The Rb gene has in fact long represented the paradigm of the
‘two-hit’ hypothesis for cancers, but recent data
suggest caveats that must be considered when extending this
hypothesis to other disorders. More research into human cancer
has shown increasing complexity in the manner by which these
‘hits’ can happen and has demonstrated that all
‘hits’ may not be equal. This is true both for ‘complex’
disorders with multiple genes implicated, such as breast or
colon cancer, and ‘single-gene’ cancer disorders.
Companion reviews from Welcsh and King (this issue, pp. 705–713)
and Nathanson and Weber (this issue, pp. 715–720) describe
this dilemma for breast cancer. This first review outlines the
role of the
BRCA1 and
BRCA2 genes, which are estimated to account
for only a small percentage of breast and ovarian cancer cases.
In carriers of one mutated allele, a number of mechanisms may
account for the second ‘hit,’ including deletion
of the
BRCA gene or promoter hypermethylation. The BRCA proteins
are involved in repair in response to DNA damage, and haploinsuffiency
may increase the chance of a second ‘hit’ in one
of these genes. Thus, the search for additional genes involved
in breast cancer has focused in part on other genes in the DNA
damage pathway, as reviewed by Nathanson and Weber (this issue,
pp. 715–720). Given that many families do not share linkage
to common genomic areas, it may be that low-penetrance missense
mutations in a number of genes, i.e. in the DNA repair pathway,
can combine to confer a higher risk to breast cancer. In this
case, the number of ‘hits’ may be difficult or impossible
to detail, but examination of common variant alleles in candidate
genes may provide valuable clues.
An analogous situation exists in colon cancer, both in its complexity and in the involvement of DNA repair processes. Fearnhead et al. (this issue, pp. 721–733) describe the large number of mutations in APC that can confer susceptibility to colorectal tumorigenesis, including both lower-penetrance missense mutations and complete loss of function through promoter hypermethylation. There even appears to be a correlation between the ‘hit’ (placement of mutation in the APC protein) and the resultant phenotype (number of polyps developed). Although APC is involved in signaling within the cell, colon cancer can also be caused by mutations in the DNA repair genes MLH1, MSH2 and MSH6, as reviewed by Peltomaki (this issue, pp. 735–740). Still, up to one-third of families with this form of colon cancer have not been accounted for by mismatch repair mutations to date, invoking the possibility of multiple lower-penetrance mutations that have an additive effect, as in breast cancer. Again, the complete genotype of the patient must be taken into account: patients with mismatch repair mutations who also have a variation in the protein cyclin D1 develop cancer at earlier ages. Finally, Mohaghegh and Hickson (this issue, pp. 741–746) explore the effects of mutations in other genes that sense or repair DNA damage, such as BLM (which complexes with BRCA1 and goes to PML bodies), WRN and RECQ4.
|
‘SINGLE-GENE’ CANCER SYNDROMES
|
|---|
The complexity being revealed by ‘single-gene’ cancer
syndromes further suggests a secure future for cancer researchers.
These include neurofibromatosis, reviewed here by Gutmann (this
issue, pp. 747–755);
patched and basal cell carcinoma,
reviewed by Bale (this issue, pp. 757–762); and von Hippel-Lindau
syndrome, reviewed by Friedrich (this issue, pp. 763–767).
Most of these syndromes, as suggested by Gutmann (this issue,
pp. 747–755), effectively represent the results of a mutation
screen for tumor suppressors in the human genome. Therefore,
full understanding of their mechanisms will eventually allow
us to understand the complete growth program of the cell. Similarly,
Pandolfi (this issue, pp. 769–775) describes the partnership
of oncogenes and tumor suppressors formed by chromosome translocations
in acute promyelocytic leukemia. Despite their seemingly simple
explanations, these syndromes are complicated by haploinsuffiency
leading to allele selection, gene–environment interactions
and as yet undescribed modifier genes.
|
FUTURE DIRECTIONS: SOLVING THE RIDDLE
|
|---|
Finally, Lam and Breakefield (this issue, pp. 777–787)
detail the latest research in gene therapy for brain tumors,
an area known for the paucity of effective treatments. Caution
is indicated with many of these proposed therapies, despite
promising results from experimental models. Therapeutic options
for other cancers are discussed in many of the reviews, and
much of the ongoing research into tumor genotypes and mechanisms
of attaining uncontrolled growth will shed light on possible
therapeutic targets.
Our current knowledge of cancer processes is remarkably similar to the smile of the Cheshire cat, and we must make the rest of the cat appear to learn the answer to the cancer riddle.
|
FOOTNOTES
|
|---|
+ Editorial Fellow, Human Molecular Genetics Tel: +1 216 368 3518;
Fax: +1 216 368 3432; Email: cdg3@po.cwru.edu
CiteULike 
Connotea 
Del.icio.us What's this?
TOP
OVERVIEW
ENHANCING THE TOOLBOX