Main Menu

Why some childhood cancers are examples of ‘big bang’ evolution


Normally cancer occurs through the accumulation of mutations – but in some childhood cancers, it seems that a single mutation can be enough.

Posted on 28 April, 2015 by Liz Burtally

Thankfully acute leukaemias – a type of blood cancer – in infants are rare, but when they do happen, they are aggressive and extremely hard to treat.

Here at the Institute of Cancer Research in London, Professor Mel Greaves has devoted much of his research to investigating what triggers leukaemia in children by examining the genetic influences and biological pathways that lead to the disease. In the 1990s, Professor Greaves and his team made a major discovery in identifying mutations that initiate leukaemia before birth, in utero.

In March this year, Dr Anna Andersson and colleagues at St Jude, Memphis and Washington University Pediatric Cancer Genome Project, and researchers from elsewhere in the USA, Sweden and Australia published an interesting study in Nature Genetics on the founder mutations in childhood leukaemia. They analysed 22 cases of childhood acute lymphoblastic leukaemia (ALL) all of which had chromosomal translocations – where a piece of chromosome has broken off and attached itself to another chromosome – within the Mixed Lineage Leukaemia (MLL) gene.

They found evidence that this MLL gene fusion is the only detectable abnormality within the whole genome that was present in all cells. Other mutations were present, particularly in a signalling pathway called P13K-RAS that is associated with cancer development. But these appear to be less important and not always found in patients who have relapsed, unlike the MLL gene fusion.

The findings suggest that although additional ‘driver’ mutations occur in childhood ALL, the initiating MLL fusion is probably sufficient enough to spawn a large malignant clone. These in-depth findings fit with earlier research at the ICR – published in Genes, Chromosome and Cancer – in which Professor Richard Houlston and Professor Greaves sequenced the genomes of three cases of infant ALL, and found them all to contain the MLL gene fusion.

The possibility that there is a single driver mutation causing an aggressive leukaemia in children conflicts with the usual cancer story, in that multiple and sequential mutations are needed in order for cancer to develop. So Professor Greaves has provided his thoughts on this intriguing question in a paper published in Cancer Cell.

So what might be bending this convention of cancer?

It turns out that this concept of a single driver mutation is also found in other aggressive childhood cancers. In rhabdoid tumours (a type of kidney cancer) there are mutations or deletions leading to loss of function of the SMARCB1 gene, and in brain ependymomas (cancers that develop in the passageways of the brain and the spinal cord) there is a C11orf95-RELA gene fusion.

But how is it possible for a single mutation to spawn a fully-fledged malignant clone and in such a short time frame?

Professor Greaves suggests that the answer may reside in jobs that different gene products do during embryonic development and the cellular targets that are involved. MLL proteins are global chromatin re-modellers, meaning that they cause disruption in the way the DNA is packaged in a cell. MLL gene fusions have an influence on multiple genes, which may have the same impact on cellular fitness as several independent mutations in other genes.

Interestingly, MLL fusions are also found in adult cancers but appear with additional driver mutations. So there must be something special about the paediatric setting, or else something specific to the type of cells in which the mutations occur.

One plausible explanation is that there is a developmental window during embryonic or foetal development, where early stem cells are dividing and expanding, during which particular mutations could cause cancer on their own. Mutations in genes that decide and regulate cell fate at this critical stage, such as MLL / C11orf95-RELA fusions, or SMARCB1 deletions, might then trap these cells in a self-renewing mode that triggers the development of cancer.

Professor Greaves concludes by stating that having normal unassuming cells that can be converted to highly cancerous cells by a ‘big bang’ in a single step seems a very dangerous liability. The fortunate rarity of these types of cancers may then reflect the very low probability of those mutations occurring during a very transient time window.


leukaemia childhood cancer
comments powered by Disqus