‘Drivers’ and ‘Passengers’: who’s in charge?

Genome sequencing has revealed that a plethora of gene mutations can co-exist in individual cancers: thousands in some cases 1,2. Based on Darwinian theory, we assume that whilst most are irrelevant, buried in the background is a modest number of mutations (perhaps counted in single figures) that are functionally active in a way that contributes to cancer clonal development. The ‘offspring’ of the cells with these mutations will be more successful than the cells that surround them. Mutations in the genes in a cancer are commonly referred to as ‘drivers’ and ‘passengers’ implying that we know what they do or don’t do. This is important because the study of these mutations informs our understanding of cancer biology. Mutated genes, or their products, can be candidates for prognostic markers or targets for treatments.

‘Drivers’ are, by definition, mutations that give a fitness advantage to the cells that carry them, via the modifications in phenotype that they cause. Improved fitness can come from any of the ‘hallmark’ features described by Hanahan and Weinberg 3 but ultimately translates, in evolutionary currency, to enhanced or competitive survival and reproduction of cells.

When mutations in the same gene are found in many patients (recurrency) ‘driver’ status is implied. This is endorsed if the mutated gene can contribute to cancer phenotypes in model systems 4.

‘Passenger’ mutations are assumed to have neutral impact on cell phenotype. For each individual, or patient, they can arise at any time in the antecedent life history of any cell 5, including a cancer cell lineage. Or they may be a product of genomic instability and/or aetiological exposures causally linked to the cancer 6. They usually register in genomic screens as idiosyncratic, patient and clone-specific mutations and the fact that they can show up in most cells or in dominant clones is no reflection of their impact.

That much is, in principle, clear and reasonable enough. The trouble is these definitions can encourage us to adopt a gene-centric perspective – appearing to imply that having a particular function is an intrinsic or autonomous property of a mutation when this is far from certain. Rather we should assume that mutations arise more or less randomly with respect to any functional effect they may have on a target gene or cellular process. For those that have the potential to alter gene regulation or exome encoded structure, the functional impact they actually have will depend on several factors:

• The cell type involved and its phenotypic response to the mutation. Some oncogenic mutations such as RAS are broad spectrum, impacting on many cellular processes or cell types. Others (e.g. fusion genes) are more cell-type restricted, possibly because they are selectively expressed and often encode transcription factors impacting on cell lineage-specific differentiation programmes 7.

• Cancers develop within a dynamic local ecosystem. In this context, the local adaptive landscape and presence or absence of particular selective pressures may determine whether or not a mutation actually exercises a fitness advantage, i.e. that it is adaptive. An obvious example of this would be a mutation that confers drug resistance. As with antibiotic resistance in bacteria 8, such mutations will arise independently to, and prior to, drug exposure but are only functionally relevant when and if the clone carrying the mutation is exposed to the relevant drug. Even the most frequent oncogene mutations – in p53, will be active and functionally relevant only in the presence of stress signals that co-opt p53 function. Conversely, a ‘driver’ mutation, though persistently present, could become functionally redundant or insignificant (vestigial) under altered selective conditions.

It might then be argued that ecosystem selective pressures are the real ‘drivers’ of cancer clone evolution. But the reality is that cancer cells and components of their ecosystem engage in a vigorous, reciprocal dialogue, such that, in certain circumstances, cancer cells can re-model their microenvironments. As in evolution in general (though Richard Dawkins might argue otherwise), the gene is not omnipotent; a complex interplay exists (see Figure 1).

Theodosius Dobzhansky

M Escher, 1948

• In evolution in general, many and, possibly most potential innovations arising via individual mutations are dependent upon genetic background and functional interaction with other mutations that are not simply additive – a phenomenon known as epistasis 9. So too in cancer clone evolution. A potential ‘driver’ could be deleterious, neutral or of minor impact in the absence of ‘co-operative’ mutations. In the absence of other mutations, many single oncogenic mutations will signal cell senescence or apoptosis rather than proliferative drive and clonal advantage 10. Thus they will be deleterious to the fitness of the cell unless other mutations overriding (abrogating) the default options of senescence or cell death are present or are acquired. This complex networking arrangement is considered to be an essential and ancient evolutionary development, both in sustaining the integrity of multi-cellularity and in restraining cancer 10. Other synergistic interactions between oncogenic mutations and tumour suppressor gene deletion can collectively re-set signal pathways in a way that increases the fitness of cells for survival and/or proliferation 11.

Overall, these complications highlight the question of what, when and how mutations, arising by chance, acquire selective currency in cancer clone development. What makes them ‘passengers’ or ‘drivers’? As so often in biology, context is everything.



1. Stratton MR (2011) Exploring the genomes of cancer cells: progress and promise. Science 331(6024):1553-1558.

2. Vogelstein B, et al. (2013) Cancer genome landscapes. Science 339:1546-1558.

3. Hanahan D & Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144(5):646-674.

4. Fröhling S, et al. (2007) Identification of driver and passenger mutations of FLT3 by high-throughput DNA sequence analysis and functional assessment of candidate alleles. Cancer Cell 12(6):501-513.

5. Welch JS, et al. (2012) The origin and evolution of mutations in acute myeloid leukemia. Cell 150:264-278.

6. Pleasance ED, et al. (2009) A small-cell lung cancer genome with complex signatures of tobacco exposure. Nature 463:184-190.

7. Mitelman F, Johansson B, & Mertens F (2007) The impact of translocations and gene fusions on cancer causation. Nat Rev Cancer 7(4):233-245.

8. Lambert G, et al. (2011) An analogy between the evolution of drug resistance in bacterial communities and malignant tissues. Nat Rev Cancer 11:375-382.

9. Breen MS, Kemena C, Vlasov PK, Notredame C, & Kondrashov FA (2012) Epistasis as the primary factor in molecular evolution. Nature 490:535-538.

10. Lowe SW, Cepero E, & Evan G (2004) Intrinsic tumour suppression. Nature 432(7015):307-315.

11. De Raedt T, et al. (2014) PRC2 loss amplifies Ras-driven transcription and confers sensitivity to BRD4-based therapies. Nature 514(7521):247-251.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s