The detail is in the devil

Dr Elizabeth Murchison at the Wellcome Trust Sanger Institute. Credit: Wellcome Images.

Two years ago Dr Elizabeth Murchison at the Wellcome Trust Sanger Institute told us about her work on a cancer that spreads between Tasmanian devils through biting. We hear from her again about a new paper that’s shedding light on the genome of this unique tumour.

What is your new paper about?

The transmissible cancer that affects Tasmanian devils is very strange because it’s actually one cancer that has spread throughout the population. The cancer, which occurs on the face, is spread by the direct transfer of living cancer cells when devils bite each other. The focus of this paper¹ is the genome of that cancer – the only cancer we know of that is threatening a whole species with extinction.

It’s a bit like the story behind The Immortal Life of Henrietta Lacks [the book by Rebecca Skloot that tells the story of a woman whose cells were taken, immortalised and used in scientific research for decades without her knowledge]. By sequencing the genome of this cancer, we’re actually sequencing the genome of the original devil that gave rise to this cancer. We’ve been doing a bit of what I call genetic detective work, looking at the genome of this cancer to work out where it came from. Our analysis has led us to conclude that it came from a female devil who probably lived relatively recently. Although she’s now dead, her cells live on in this cancer.

What were you trying to do?

One of the things we’ve been interested in finding out in this study is what’s allowed this cancer to evolve and survive in different hosts. We’re very interested in looking at genes that are normally mutated in cancers, as well as genes that are involved in the immune system.

Of course, we’ve got a bit of a problem with this study because we don’t have any DNA from the normal cells from the original devil. This means that we don’t really know for sure which mutations are somatic (those arising in the cancer, during its development) and which are normal variants in that devil.

The devastating effects of devil facial tumour disease. Credit: Save the Tasmanian Devil Program.

There will be a lot more normal variants than somatic mutations. For example, there are about 3 million normal variants (single nucleotide polymorphisms; SNPs) between two humans. Between a single human and their cancer there are just hundreds or thousands of somatic mutations. Given that we don’t have any normal DNA from the original devil we’ve had to do some quite interesting tricks to try and figure out which are likely to be cancer mutations.The devastating effects of devil facial tumour disease. Credit: Save

How did you get around the lack of normal DNA?

I have fantasies of going to dig up devil remains from the area she’s meant to come from, but obviously that’s not possible, so the best we can do is to search for variants that we find in the cancer genome in normal genomes from other devils. If we don’t find these variants in any other normal devils then there’s a higher chance that they are somatic.

To try and piece together the founder devil’s genome, we sequenced four genomes: one from a normal female, which we used as a reference genome, one from a normal male and two cancer genomes taken from different parts of Tasmania – one from the south, one in the north. We know that the cancers are derived from the same original cancer, so any mutation found in one and not in the other must have arisen by a somatic event, in the time since the two cancers diverged from their most recent common ancestor.

Using that approach we were able to find more than 17 000 variants that we think arose somatically. We also saw that, even though the mutations arose independently, both tumours had a similar, distinctive mutation profile – they have a lot of A to T and T to A mutations. That suggests that there’s either been some kind of environmental carcinogen acting on both of them since their divergence, or some loss of DNA repair processes that would normally fix these errors.

I was expecting that this cancer would have enormous numbers of mutations, just because it’s such a divergent cancer and it’s spread through thousands of individuals. But while the 17 000 or so we found may sound a lot, it is fewer than those found in some human cancers. For example, the melanoma cancer genome published in 2010² had around 30 000 mutations.

What else did the genome tell you?

When we looked at other changes to the genome – deletions, gain of copy number, translocations etc. we found relatively few. This, together with the data on the somatic mutations, indicate that it’s a relatively stable cancer, one whose genome has remained fairly intact in spite of the fact that it’s transmissible and it has passed through a number of individuals. This shows that a cancer does not need to be incredibly unstable to become transmissible. Possibly this cancer has ‘found’ quite a good set of mutations that allow it to survive in different hosts – further instability is not an advantage.

We’ve found more than 400 genes with possible somatic mutations that change the coding region. We found three genes with mutations that are frequently mutated in human cancer. We still don’t know if they’re needed for the cancer to develop, so that’s something for future studies.

Samples of the devil tumours in Elizabeth's lab. Credit: Wellcome Images.

Is the fact that the original devil was female of any significance?

I guess it’s just a bit of trivia really but I think it was quite exciting that we could use genetics to go back in time and find out the identity of the devil that gave rise to the cancer. It’s very strange to imagine that devil was living probably around the time I was born, in the 1980s. She was a normal devil who developed a cancer that somehow managed to spread into other devils and continue to divide even after she died.

Do you have a name for the devil in which the cancer arose?

I call her the “immortal devil” but we need a better name!

What else did you find in the study?

As well as sequencing these two cancers completely we took another 104 cancers from all around Tasmania. We genotyped them [compared their genetic profiles] and used the data to construct what’s called a phylogeny of those 104 tumours to understand how they’re related and how they’ve diverged as they moved through Tasmania.

That’s given us some really interesting insights. There are lots of different genetic types interspersed throughout mainland Tasmania, so it seems like different tumour lineages have spread quite far.

However, on the isolated Forestier Peninsula on the south-east coast of Tasmania, we found that one tumour type has taken over the others there. That opens up the question, what type of genetic advantage has that tumour evolved? That’s also something we want to look at in the future. This is first evidence that this kind of selection is occurring in the devil cancer.

Have you been to Tasmania recently?

Yes, I just got back a couple of weeks ago from doing some fieldwork. Going out there really motivates me. Where we were working we found that 50 per cent of the devils had the disease. Many of them were in a horrible state, and there was nothing we could really do for them.

What’s changed since we last spoke?

Well, the disease has spread towards and actually reached the west cast. Now the only definitely disease-free area is the north-west corner, and the disease is moving at a steady rate towards there.

Researchers in Tasmania have shown that devils are able to reject skin grafts from healthy donors. That’s really important because one idea was that the cancer was able to spread between devils because devils were really inbred and unable to detect non-self tissues. This work has shown that, in fact, that devils have a perfectly good immune response against foreign grafts.

A healthy Tasmanian devil. Credit: Anaspides Photography - Ian D Williams.

So how might the cancer be able to spread?

My guess is that the cancer is actually talking directly to its host’s immune system, perhaps by making cytokines or other soluble factors that directly suppress the immune system. We still don’t know how the cancer is able to spread without eliciting an immune response. I’d like to understand this more, ultimately so that we can work to develop vaccines and therapies that could help with conservation.

The genes are there that allow this cancer to evade the immune system effectively in devils, which are mammals, so I think this kind of thing could happen in other mammals too.

Are the breeding programmes working?

A lot of effort is being put in into captive breeding programmes. Devils are breeding well in captivity and they’ve been put into wildlife parks and free-range enclosures on the mainland of Australia. There’s an island of the east coast of Tasmania that might be used as a kind of sanctuary – but that’s not really a long-term solution because we need to be sure the disease is completely gone before we can reintroduce devils into the wild, and that’s going to take a while. There’s also the risk that the cancer could jump species and even get into one of the other carnivorous marsupials in Tasmania – that’s a very scary thought.

For more, read the 2010 feature on Elizabeth’s work, the accompanying guide to transmissible cancers and the Wellcome Trust’s news story on this paper.

References

1. Murchison EP et al. Genome sequencing and analysis of the Tasmanian devil and its transmissible cancer. Cell 2012;148:780-91.
2. Pleasance ED, Cheetham RK et al. A comprehensive catalogue of somatic mutations from a human cancer genome. Nature 2010;463:191–196.