Darwin, cancer and the genotype to phenotype mapping
David Basanta (11th November 2008)
2008 commemorates the 150th anniversary of the announcement of the discovery of natural selection in a lecture at the Linnean Society in London. Although the discovery is credited to both Charles Darwin and Russel Wallace, it is mostly Darwin that we associate with evolution and it is unlikely we will see the day in which Wallace is honoured with a portrait back to back with the British Monarch in a piece of paper sponsored by the Bank of England.
Were he alive today, he might have been initially surprised to find his portrait hanging from the door of one Moffitt researcher whose office is near mine. Although the theory of evolution by means of natural selection draws from the experiences of Darwin and Wallace observing ecosystems in exotic (for an Englishman) parts of the world, he would feel right at home in a cancer research institute. He would appreciate how different cellular species competing for ever decreasing resources (nutrients, oxygen) in a complex environment would be a perfect example to complement the many he used to support his case on his book On the Origin of Species.
(Notice that the cover represents evolution as one line of transformations, a powerful visual image but inaccurate).
Sadly, the influence of Darwin’s ideas on medicine in general and in particular upon oncology, is somewhat limited. There have been calls to change this (here, here and here) but the bulk of cancer biology research is based on studying the genes that are altered in tumour cells. As important as genes are to explain tumour growth and progression, doing this in isolation of the environment that tumour cells inhabit makes as much as sense as trying to explain the morphological differences between finches in different parts of the world by means of genetic differences without reference to the ecosystem in which they live (or die).
Of course genes are crucial to have a proper understanding of the potential evolutionary path of tumour progression. Many ecologists and evolutionary biologists study only phenotypes since natural selection only works at the phenotypic level . This approach has limitations: the environment selects for particular traits in a population but not all traits are equally likely to emerge. For exmaple, if it becomes advantageous for a population adapted to land to cross a body of water the individuals could potentially evolve the capability to fly or swim. If those individuals already posses appendices that could easily evolve into wings and have the right aerodynamics then, given the same environment one could single out flying as the most likely evolutionary adaptation.
A integrated understanding of cancer genomics (as favoured by the cancer biology community) and cancer phenotypes (as would be the preference of the evolutionary biology community) could lead to a comprehensive genotype-to-phenotype mapping. Such a mapping would revolutionise how we understand the evolutionary dynamics of a growing tumour with a particular genetic make up within a given (micro)environment. Since the environment would select for specific phenotypic traits where as the genetic make up would make certain adaptations more likely than others. However, it is clear that the genotype-phenotype mapping is an incredibly challenging problem to tackle and can only really be addressed using an integrated framework. Specifically, using computational and mathematical models to build a theoretical bridge between the evolutionary and cancer biology communities.