|
|
|
Functional genomics to study cancer progression and metastasis
|
|
|
|
|
While extensive analysis over the last two decades led to a deep insight into the
control of cell proliferation and survival, and their alterations during cancer
onset, still much remains to be clarified about the genetic lesions and alterations
of cell signalling that lead to aberrant activation of invasive growth, cancer progression
and metastasis. To this aim, great advantages may come from the historical changes
in perspectives and modality of gaining information that biomedical research is
currently facing. Genome-sequencing projects have been completed for many organisms,
including Homo sapiens
(
http://www.ncbi.nlm.nih.gov/genome/guide/human/
)
and Mus musculus
(
http://www.ncbi.nlm.nih.gov/genome/guide/mouse/
).
This reversed the conventional approach to biomedical discovery, in which understanding
a certain biological function required identification of one or more genes involved
in that function. The current situation is that thousands of genes have been sequenced
but still wait for any functional information to be assigned to them. The fact that
genes of unknown function represent over 70% of all genes suggests that current
comprehension of most biological and pathological processes is by far incomplete.
This is particularly true in the case of cancer progression, where systematic exploration
of gene function is likely to yield a huge amount of information in the next years.
There are several ways of obtaining information about gene function, some of which
have been evolving at an incredibly high pace. For example, the relatively recent
development of the DNA microarray technology currently enables mRNA expression analysis
in parallel for thousands of genes. Indeed, being expressed (at least at the RNA
level) is an essential prerequisite for a gene to exert its function, and by studying
the sites and pathways of regulation of a certain gene expression it is possible
to putatively assign it to a broad functional group. In this view, genes with restricted,
tissue-specific expression are likely to play key roles in the biochemical and biological
processes specifically occurring at the expression sites.
Another powerful approach to gene functional characterization is exploration of
the consequences of gene loss-of-function in various model organisms, ranging from
unicellular microorganisms to invertebrates, vertebrates and mammals. In particular,
generation of mutations in murine ES cells by targeted and random approaches offers
a powerful tool for loss-of-function studies in the mouse. ES cells can be grown
in vitro as a continuous cell line, genetically modified and subsequently returned
to the embryo, where they can generate chimeric mice and eventually contribute to
the germ line. Mouse ES cells are now widely used for gene disruption by homologous
recombination or chemically induced mutagenesis, to create mutant mice that lack
or express an altered form of a specific gene. Recently, an International Mouse
Mutagenesis Consortium has been established, with the long-term goal of producing
at least one heritable mutation, in either ES cells or mice, in every gene in the
mouse genome. In many cases however, functional redundancy or subtle phenotypes
may impair functional characterization of the targeted genes. Moreover, this approach
is aimed at defining gene function in the context of the organism, but is hard to
direct at exploring basic biological and biochemical functions at the cellular level.
This latter type of information can be achieved by systematic screenings exploring
features of the gene protein product, like subcellular localization, biochemical
activity, interactions, and others. Recently, development of small interfering RNA
(siRNA)-based approaches rendered loss-of-function studies more easily practicable
in cell lines and higher organisms (Science 296:550-553, 2002). Finally, genes can
be characterized by gain-of-function approaches, relying on overexpression of cloned
genes in cells and organisms (Cancer Res. 61:5861-5868, 2001) or on random activation
of gene expression (RAGE; Nat Biotechnol. 19:440-445, 2001).
From this brief outline of the major strategies for gene functional characterization
it clearly emerges that a crucial issue in functional genomics is the development
of technologies for high-throughput functional analysis. Towards this aim, development
of large-scale functional screens focussed on cancer will require a coordinated
approach involving complementary competences and establishment of dedicated facilities,
for which TRANSFOG intends to provide an optimal organizational and financial framework.
The TRANSFOG project consists of seven research components that will synergistically
enable streamlined translation of large-scale genomic screenings into high-impact
contributions to cancer diagnosis and therapy. A wide variety of cancer-oriented
genomic screenings will be carried out by 14 partners and finally merged with the
particular aim of identifying and prioritising novel genes (ESTs or poorly characterised
genes) with a clear potential role in cancer metastasis, the candidate genes. Recent
works have shown that it is possible to exploit gene expression profiling of tumour
samples to define sets of genes (signatures) whose expression correlates, positively
or negatively, with metastasis-free survival, e.g. in breast cancer (Nature 415:530-536,
2002; N Engl J Med. 347:1999-2009, 2002.). It has also been found that a general
signature associated with metastatic behaviour can be shared between solid tumours
of different organs (Nat Genet. 33:49-54, 2003), which indicates that common alterations
of basic cellular functions and signalling pathways trigger metastatic progression
of cancer. The TRANSFOG screenings will concentrate on breast, lung and colon cancer,
which altogether account for the majority of cancer deaths in the general population.
Selected tumour samples and cell-based models will be analysed for the identification
of molecular signatures for metastasis, through three main approaches:
- Paired tumour/metastasis samples. In this case, the aim is to identify genes consistently
showing differential expression between the primary tumour and its metastasis.
- Primary tumour samples with no synchronous metastasis, having adequate followup
data to assess the clinical history. In this case, the aim is to identify genes
whose expression in a primary, non-metastatic tumour is associated with its tendency
to give rise to secondary lesions after surgical removal.
- Organ-specific metastases. The aim is to identify genes whose expression in the
primary tumour is associated with metastasis homing to a specific target organ.
As a target organ for metastasis, the initial focus will be posed on the liver,
the most frequent metastatic target of colon cancer.
Apart from tumours, screenings will also include cancer-oriented experimental models,
like serine and tyrosine kinase receptor-driven transcriptional responses, ligand-induced
in vitro epithelial morphogenesis and invasive growth, in vitro angiogenesis of
endothelial cells. The aim is to obtain a genome-wide exploration of the basic mechanisms
of cancer progression. By merging the results of the screenings, we therefore expect
to find "common" genes, i.e. genes emerging from more than one screening as associated
to invasion and metastasis, and "specific" genes, whose expression is only altered
in small subgroups or subtypes of tumours/metastases or cellular models. The relevant
genes will have to be ranked for priority towards functional characterization and/or
diagnostic validation, with the main priority criterion being their emergence in
more than one screening.
Two other activities will produce, respectively, the full-length coding sequences
and the siRNA sequences for these genes, enabling their functional, "mirrored" characterization
by gain- and loss-of-function in cell lines and model organisms. Up-or down-regulation
of these genes in models of tumour aggressiveness will provide proof-of-concept
validation of their potential usefulness as new molecular targets for drug screenings.
The coding sequences will also be used for systematic protein-protein interaction
studies, aimed at defining the signal transduction properties of the identified
candidates. The diagnostic potential of the candidate cancer genes will be evaluated
and validated by systematic analysis of tumour samples with different techniques,
from DNA microarrays to tissue microarrays, to quantitative real-time PCR and finally
to diagnostic-grade antibodies.
The huge amount of experimental data generated by the Consortium will require establishment
of a shared platform for data handling and gene functional annotation.
|
|
|
|
|
|
|