Some Cancers Become Contagious
Some Cancers Become Contagious

Some Cancers Become Contagious

So far, six animal species are known to carry transmissible, “parasitic” forms of cancer, but researchers are still mystified as to how cancer can become infectious.

Apr 1, 2019
Katarina Zimmer

Above: © ISTOCK.COM, CraigRJD

The untrained eye likely wouldn’t have noticed, but doctoral student Ruth Pye immediately spotted something unusual about the way the cells were arranged in a tissue sample from a facial tumor of a Tasmanian devil. Tumor cells plucked from the marsupials normally grew and divided more slowly, but these established themselves much faster in culture, and had longer projections extending from their spindle-shaped cell bodies, she recalls.

It was early 2014, and Pye was examining a biopsy taken from a diseased devil on a remote peninsula on the southeast side of Tasmania. Her lab at the Menzies Institute for Medical Research at the University of Tasmania received such samples as part of a government-sponsored monitoring program to study the notorious cancer that had been decimating populations of the island’s namesake marsupial. Known as devil facial tumor disease (DFTD), the cancer differs from most in that it exists as a single transformed cell line, thought to have originated in a devil that lived more than 20 years ago, that is capable of moving between individuals. Each cell genetically resembles the founder devil, and is distinct from the cells of healthy devils.

Such contagious cancers are exceedingly rare: at the time Pye noticed the strange-looking samples, the only other transmissible cancer known was a sexually transmitted oncogenic cell line in dogs, which was by and large harmless to the animals. In Tasmanian devils, DFTD is thought to be transmitted when the animals bite one another, whether it be during battles for mates, during mating itself, or when scrapping over meals of dead animals. Cancerous cells that become lodged in the open wounds of the mouth or face quickly colonize the host tissue, triggering the growth of disfiguring tumors that impair the animals’ ability to feed. Since the first reported case of DFTD in 1996, devil numbers have plummeted by nearly 80 percent in areas affected by the disease.

Once the pro­cess is started, it’s equivalent to the creation of a novel parasitic species, with its own dynamic and its own evolution.

—Frédéric Thomas, French National Center for Scientific Research

Pye made a note of her observation regarding the unusual sheets of cancer cells and put it aside. But when a similar sample came in from a different devil that lived in the same area as the first, she decided to investigate. Neither tumor stained positive for the protein periaxin, a diagnostic marker for DFTD. Karyotyping revealed hallmarks of a transmissible cancer: the tumor cells were different from host cells, but were identical to one another, sharing the same chromosomal aberrations even across devils. This suggested that the cells came from a transmissible cancer that originated, not in the individual carrying the tumor, but from another devil who had passed along its disease, according to Pye, now a research veterinarian at the Menzies Institute.

After puzzling over the findings with her supervisors at the time, Greg Woods and Bruce Lyons, Pye sent some samples to the University of Cambridge in the UK, where Elizabeth Murchison and her team conducted a microsatellite analysis. Sure enough, the results revealed that the fast-growing cells in the two samples represented a distinct transmissible cancer that appeared to have arisen completely independently.1 “I remember when we got the results off the machine, and it was immediately clear that it wasn’t DFTD,” Murchison says. “We just couldn’t believe it.”

The finding challenged the idea that DFTD, now named DFT1, was a one-off. If a second contagious cancer, since named DFT2, could emerge within a few decades of the first, perhaps the phenomenon of cancerous cell lines spreading between individuals was a lot more common than researchers realized, Murchison says. It’s possible, she muses, that researchers only identified DFT2 because the devils were already under such close surveillance due to the original cancer, and that other examples of animals spreading tumors to one another have flown under the radar. “[The discovery] really changed everything we thought we knew” about transmissible cancers, she says.

At the time Pye identified the first unusual samples of Tasmanian devils’ tumors, researchers in the United States were already investigating a leukemia that had been decimating populations of a soft-shelled clam (Mya arenaria) off the coast of New England. In 2015, virologist Stephen Goff and his team at Columbia University reported that leukemia cells across individuals were genetically identical to one another, but genetically different from their host’s healthy cells.“We were forced to the realization at the time that this was a third example of a transmissible tumor happening,” Goff recalls, after the cancers known in devils and dogs. A year later, he and his team found three other species of bivalve that carry their own contagious cancers, raising the number of species affected by naturally occurring transmissible cancers to six.3

The realization that such contagious cancers may be more widespread than previously thought has intensified efforts to understand their biology—not just for the sake of the species they affect, but also to understand how cancer can become an infectious disease. Many questions remain unanswered, including how these diseases emerge and in what populations. But in the last few years, genetic and immunological studies have provided some insight into these cancers’ interactions with their hosts. The findings have led researchers to view them as independent parasites, with the survival of their host species depending on a delicate interplay between the animals’ immune systems and the cancers’ ability to evade them.

The Animals that Catch Cancer

So far, only dogs, Tasmanian devils, and four bivalve species are known to carry transmissible cancers, which have varying effects on their hosts.

Host speciesDomestic dog (Canis lupus familiaris)Tasmanian devil (Sarcophilus harrisii)Bivalves (the soft-shelled clam, Mya
arenaria
, the mussel, Mytilus trossulus, and the cockle species, Cerastoderma edule, and the golden carpet-shell clam, Poltitapes aureus)
CancerCanine transmissible venereal tumor (CTVT)Devil’s facial tumor disease (DFT1 and DFT2)Clam disseminated neoplasia
Outcome
GABRIELE MARINO, UNIVERSITY OF MESSINA, ITALY
CTVT manifests as tumors on dogs’ genitals (seen here at base of penis) that rarely result in death. It  regresses on its own or can be treated easily using chemotherapy drugs such as vincristine.
CAMILA ESPEJO, UNIVERSITY OF TASMANI
DFT1 and DFT2 form tumors around the mouth and face that restrict the devils’ ability to feed. These cancers are typically fatal, and have driven devils to the brink of extinction.
Marisa Yonemitsu and Rachael Giersch, Pacific Northwest Research Institute
These leukemia cells clog bivalves’  internal organs. While the disease has triggered mass die-offs of soft-shelled clams in the past, today it doesn’t seem to cause severe problems.
OriginRecent studies have estimated that CTVT arose in one of the first dog populations to inhabit North America some 8,000 years ago. It is believed to stem from a macrophage or another kind of immune cell.DFT1 is thought to have originated in a devil that lived some 20 years ago. DFT2 is thought to be much younger than DFT1, but its exact age is unknown. Researchers think the cancers stem from Schwann cells, which protect neurons and help repair them after damage.Not much is known about the origin of disseminated neoplasia in clams, but researchers have suggested that the malignant cells stem from hemocyte cells that reside in the animals’ circulatory system.
Mode of transmission
©istock.com, Oleg_0
CTVT spreads through sexual intercourse.
©istock.com, Parfenov yurii
Both devil cancers are thought to spread when the animals bite each other, for instance, during the mating season.
@istock.com, hailshadow
Dead individuals are thought to release cancer cells that drift through seawater, infecting other individuals that likely pick them up through filter feeding.

The birth of a cancer parasite

When North American settlers wandered across the Bering Strait while it was frozen during the last ice age, they brought their dogs with them. Sometime later—some 8,000 years ago—a tumor emerged in one of the canines’ descendants. But something unusual happened. Rather than dying with its canine victim, the cancer began to jump from one pup to another.

That’s according to a 2016 Science study of canine evolution in the Americas. In that paper, Murchison and her collaborators traced back the origins of the oldest known contagious cancer—canine transmissible venereal tumor (CTVT)—through estimations of its mutation rate.4 Currently, the cancer lives on as a cauliflower-like tumor on the genitals of dogs around the world. But because CTVT has accumulated so many mutations over the course of its long history, it’s difficult to identify which ones contributed to its emergence, explains University College London (UCL) immunologist Ariberto Fassati, who together with UCL virologist Robin Weiss conclusively established CTVT’s transmissibility in 2006.5

Murchison and colleagues’ analysis of the more recently arisen DFT1 and DFT2 lines was more successful at uncovering clues to the cancers’ origins.6 Based on the proteins these cancer cells produce, the researchers surmised that both cancer types appear to have emerged in a type of cell associated with the nervous system. DFT1 was already known to produce periaxin and other proteins associated with Schwann cells, which wrap around nerves and rapidly proliferate after nerve injury. DFT2 doesn’t produce periaxin. But like DFT1 it has a mutation in the Hippo pathway, a signaling cascade altered in certain Schwann cell cancers in humans. “Our theory is that they both come from a Schwann cell which has become stuck in this repair mode,” Murchison says.

Why would this [transmissible] cancer not be rejected like any regular tissue transplant would be?

—Bruce Lyons, University of Tasmania

Still, it’s not clear what made the cancer cells first able to jump hosts. Biologists have proposed that transmissibility results from a perfect storm of four key factors. First, tumor cells need a route of transmission to a new host, and second, they must be able to maintain survival during that transition. Third, the species must have an immune environment that facilitates invasion by foreign cells. And finally, the tumors must have mechanisms to evade immune attacks by their new host.

The sustained physical contact during canine intercourse and the devils’ habit of biting other individuals during the mating season set the stage for transmission to occur. Then, low genetic diversity in both species might have allowed the cancers to flourish in new but similar environments.

Under normal circumstances, foreign cells are rejected through a vertebrate’s self/nonself recognition immune system. Individual-specific cell surface antigens known as the major histocompatibility complex (MHC) classes I and II alert the animal’s T cells to the presence of foreign material, the former triggering killer T cells to attack. In Tasmanian devils, however, MHC diversity is very low, as a result of major population bottlenecks caused by extreme climatic events throughout the species’ history. And the CTVT founder dog is believed to have been a member of a population with low genetic diversity: “It looks like it came from quite an isolated population,” says Máire Ní Leathlobhair, a doctoral student in Murchison’s lab. The role of low host genetic diversity in the emergence of transmissible cancers is also supported by experiments from the 1960s showing that a sarcoma could be transmitted between individuals of an inbred laboratory population of Syrian hamsters by the injection of tumor cells.7 Despite the fact that the cells were of nonself origin, the hamsters’ immune systems did not reject them.

In bivalves, cancerous cells are released from infected animals and drift through the water to other individuals, which are thought to pick them up during filter feeding, explains Michael Metzger, a former postdoc in Goff’s lab, who now heads a lab at the Pacific Northwest Research Institute. And the mollusks lack the adaptive immune system that includes the self/nonself recognition system in vertebrates, he adds. “That seems to lower the bar for transmissible cancers.”

Once these initial hurdles are cleared, the tumors can adapt to their new environment to evade immune attacks by their hosts, says Frédéric Thomas, an evolutionary biologist at the French National Center for Scientific Research who with colleagues formalized the “perfect storm” hypothesis in 2016.8 “Once the process is started, it’s equivalent to the creation of a novel parasitic species, with its own dynamic and its own evolution.”

Immune hide and seek

For a while, researchers believed that the devils’ precariously low MHC diversity was enough to explain why DFT1 could thrive in different individuals. But an experiment in 2011 turned that idea on its head. When researchers at the University of Tasmania cut small skin slices from five healthy devils and transplanted them to other individuals, all of the allografts were rejected within 14 days.9 This suggested that MHC diversity in devils is sufficient for them to reject tissue from other individuals, puzzling immunologists as to how a cancer could be so easily transmitted. “Why would this [transmissible] cancer not be rejected like any regular tissue transplant would be?” asks the University of Tasmania’s Lyons.

A clue surfaced in 2013, when British researchers demonstrated that DFT1 cells don’t express any MHC antigens.10 It turns out that the tumor cells epigenetically downregulate several genes that encode proteins required for building MHC antigens. Murchison’s group reported last year that DFT1 lacks a copy of β2 microglobulin (B2M), which encodes a component of MHC class I molecules.6 “That’s probably at least part of the reason why the immune system isn’t able to detect [the cancer cells] as being foreign,” Murchison says. Researchers have similarly shown that CTVT cells downregulate MHC class I expression in dogs.5

The Devil Is in the Details

In one of the most extensive studies of devil facial tumor disease (DFT1) to date, an international team of researchers has uncovered a mechanism that drives the cancer’s metastasis and helps it to evade the Tasmanian devils’ immune system.

See full infographic: WEB | PDF
© MESA SCHUMACHER


1Cancer cells that form DFT1 tumors  overproduce transmembrane enzymes known as ERBB receptors.
2When stimulated by specific proteins—likely EGF and NRG1, which are also overproduced in DFT1 cells—ERBB receptors induce production of a signaling protein and transcription factor called Stat3 and drive its activation.
3In the nucleus, Stat3 drives the production of genes such as MMP2 that are known to trigger metastasis in humans.
4Stat3 also binds to and inhibits another transcription factor, Stat1, which normally drives the expression of genes necessary for the generation of MHC class I molecules.
5MHC class I molecules normally interact with receptors on cytotoxic T cells to discriminate self from foreign cells. By downregulating the production of MHC proteins, DFT1 cells are able to evade detection by the animals’ immune system.  
6Under normal circumstances, cells lacking MHC markers would be detected by natural killer cells (NK), but for reasons that are unclear, devil NK cells don’t react to DFT1.

Interference with the MHC system is a mechanism also exploited by human cancers to hide from the immune system. Several large-scale studies have established a frequent occurrence of downregulation among genes encoding human leukocyte antigens (HLA)—the human variant of MHC—in head, neck, lung, and other tumors. And in a rare case of transmission of cancer cells from a pregnant mother to her developing fetus in 2009, genetic analysis indicated that the fetus’s maternally derived leukemic cells carried a deletion of a particular HLA allele.

Earlier this year, an international group of researchers uncovered a mechanism that explains how the devils’ tumors manipulate the production of MHC proteins. The team found that two proteins, ERBB and Stat3, known to boost the expression of metastasis-driving genes in human cancers, are overactive in DFT1 cells.11 These proteins are part of “a very well-described signaling cascade in human cancer,” explains senior author and immunologist Andreas Bergthaler of the Research Center for Molecular Medicine of the Austrian Academy of Sciences. But in devils, the pathway has an additional effect: it also regulates MHC class I production by stifling the expression of genes required for the creation of MHC proteins, such as B2M. Blocking the pathway in immunodeficient mice not only stunted the growth of DFT1 cells by crippling the expression of genes that drive metastasis, it also revived MHC class I expression on the surface of DFT1 cells. In mice with functioning immune systems, this could further slow tumor growth by triggering T cells to attack the cancer cells, explains Bergthaler.

We’re getting an increas­ing number of animals that are actually surviving, in which the tumors regress and in some cases disappear.

—Menna Jones, University of Tasmania

But stunted MHC production alone doesn’t explain why DFT1 is not rejected by new hosts. Vertebrate immune systems have other mechanisms to ward off pathogens. For instance, natural killer (NK) cells that fight infections would normally detect and attack cells that don’t produce MHC proteins. NK cells function in devils, yet for reasons that are unclear, they don’t seem to attack DFT1,12 suggesting that the cancers exploit additional mechanisms to evade them, such as by secreting immunosuppressive cytokines, says Pye.

DFT2 cells, however, do present MHC class I molecules—at least in cell culture, Murchison’s group established last year.13 This finding indicates that the loss of these MHC antigens may not be necessary for the emergence of a contagious cancer. Perhaps for now, DFT2 can spread thanks to extremely low genetic diversity of the southeastern devil population, the researchers speculate. Indeed, Murchison and colleagues found that the MHC class I molecules of DFT2 are very similar to those of devils infected by the cancer. DFT2 may also eventually begin to mimic DFT1 and lose its MHC expression if it spreads to less closely related individuals. 

“It’s quite possible that DFT2 is just a precursor, so to speak, to DFT1,” notes Fassati.

Prayer for the devil

The devil as a species is safe for now. Years ago, biologists gathered a number of uninfected animals and shuttled them to an uninhabited island off the coast of Tasmania for safekeeping, and numerous breeding populations live in zoos around the world. But the race is still on to protect the 25,000 or fewer wild animals that remain in Tasmania. For them, DFT2 is not yet a threat, as the disease hasn’t spread beyond a small area in southern Tasmania. But DFT1 is still killing off devils at an alarming rate, worrying researchers about the fate of the charismatic marsupials. Many fear that the demise of the island’s only remaining native carnivore will bring major disruption to the Tasmanian ecosystem.

One approach to protecting the devils is to vaccinate uninfected animals kept in captivity or on the uninhabited island against the contagious cancers before returning them to the wild. Vaccines produced from heat-treated or otherwise damaged DFT1 cells, for example, were used on devils released in groups of 19 or more in different parts of Tasmania between 2015 and 2018. But trapping the animals again to see whether the vaccine had any protective effect has proven difficult, as devils disperse widely across the island and often become victims of vehicular collisions, explains Carolyn Hogg, a conservation biologist at the University of Sydney who is working with the Tasmanian government–sponsored Save the Tasmanian Devil Program.

BE FREE: Camila Espejo, a PhD student at the University of Tasmania, releases a diseased devil back into the wild after checking its health status.
MANUEL RUIZ

Another tack is to treat infected individuals. A combined vaccine and immunotherapy trial with six devils in 2017 proved promising, with half of the animals’ tumors vanishing entirely.14 “It was essentially a cure,” for those devils, says Lyons, who led the study. For him, the trial is a proof of principle that the devils’ immune system is capable of fighting off the cancer, and it could help the team build better vaccines in the future. But capturing wild devils and treating them with immunotherapy isn’t logistically feasible on a large scale, he notes.

Fortunately, devils seem to be getting better at battling the disease without medical intervention. “We’re getting an increasing number of animals that are actually surviving, [in which] the tumors regress and in some cases disappear,” says Menna Jones, a zoologist at the University of Tasmania. “And some of those individuals will live to quite a ripe old age.” Such resilient devils appear to be genetically different from devils that succumb to the tumor, she and others found last year. For example, alleles thought to underlie cancer risk and immune responses to tumors in humans were overrepresented among the survivors. This suggests that some populations are adapting to the disease, Jones says.

On the flip side, the cancer itself also appears to be evolving. In a 2013 study, researchers found that large areas of the DFT1 genome have become demethylated over time, possibly resulting in a change in gene expression.16 “We think that the tumors are adapting to the immune response and tumor suppressor response from the host via epigenetic modification,” explains lead author Beata Ujvari, an evolutionary biologist then at the University of Sydney.

Contagious Cancer Cases in Humans

The “elephant in the room” of transmissible cancer research is the question of whether one could arise in humans, says immunologist Andreas Bergthaler of Vienna’s Research Center for Molecular Medicine of the Austrian Academy of Sciences. But so far, there have only been “freak cases” of cancers jumping from one person to another, he says. In some two dozen cases over the past several decades, for instance, doctors have reported mother-to-child transfer of cancer cells during pregnancy. It is unclear how most of these occurred, but genetic analyses suggest that, in some cases, mutations arise in the cancer cells that allow them to escape immune detection by their new host.

One of the first well-documented examples of cancer transmission between two adults came in 1996, when a surgeon was reported to have developed a tumor-like swelling after he injured his palm during an operation on a cancer patient. The tumor was excised, and didn’t grow back. Clinicians aren’t sure why the malignant cells weren’t initially rejected by the surgeon’s immune system, but suggest they could have escaped immune detection by tinkering with their expression of MHC molecules that help the body discern self from foreign cells—the same mechanisms employed by transmissible cancers in Tasmanian devils.

In other cases, cancer appears to have made the leap between humans when a person’s immune system is not functioning properly. Organ transplant recipients, for instance, can be vulnerable to contracting cancer from organ donors, as they are administered immunosuppressive drugs to prevent rejection of foreign tissue. Just last year, for example, clinicians documented four patients who developed breast cancer after receiving kidneys, lungs, heart, and liver from a 53-year-old donor who had died due to stroke. The cancer cells did not match the patients’ own cells, but those of the donor, who at the time of the transplant was not known to have the disease. In an exceptionally unusual case that also involved an immunocompromised individual, a 41-year-old HIV patient was reported to have died of cancer contracted from a tapeworm living inside his gut.

Unless the immune system is compromised, there’s very little chance a contagious cancer could arise in humans, Bergthaler says, and “there’s no hard evidence” that anything akin to the devils’ or dogs’ cancer is making the rounds in human populations. Then again, he adds, there was a time when scientists only knew of retroviruses that could infect animals other than humans. While it’s highly unlikely the story could turn the same way for transmissible cancers, “it goes to show we might still be unaware of certain diseases in humans. . . . We just don’t know.”

Ujvari, now at Deakin University near Melbourne, thinks there’s also a possibility that DFT1 cells are facing a selective pressure to become less virulent. From a parasite’s perspective, it might be advantageous to allow its devil host to survive as long as possible, in order to increase its chances of transmission. This could explain why dogs rarely die from CTVT, which could have been a much more aggressive cancer in the past, she speculates. “The host and the cancer cell line have achieved an evolutionary equilibrium where the cancer cells don’t kill the dogs anymore.” This is the trend that has been documented in some clam populations, which have rebounded after being ravaged by a transmissible leukemia. However, Metzger notes, “this could be due to the cancer evolving to be less pathogenic, but it could also be a case of the clams evolving resistance to the disease.”

Andrew Storfer, an evolutionary geneticist at Washington State University, argues that, with regard to the Tasmanian devil, which normally lives five to seven years in the wild, the latter hypothesis is more likely. Because devils tend to survive for several months with the cancer before they die, they have plenty of time to transmit the disease, he says. “Given the slow-burning nature of the disease, there doesn’t seem to be a strong selection to evolve to lower virulence.” Ujvari and her colleagues are currently working to develop mathematical models to predict the evolutionary trajectory of DFT1 and DFT2 to better understand the disease in the context of host-parasite coevolution.

Like other scientists, Ujvari wonders whether transmissible cancers may be more common than realized. Perhaps they drove their hosts to extinction, she suggests, or they have reached such a stable equilibrium they are no longer detectable. To get to the root of this question, some researchers have begun to actively search for more such transmissible cancers in nature. If more examples can be identified, it could point to common factors that allow the oncogenic cell lines to jump from one individual to another.

To more directly investigate how contagious cancers arise, Fassati plans to recreate a transmissible tumor in mice. He hopes that learning how contagious cancers outsmart their hosts’ immune systems will lead to insights into how nontransmissible cancers do the same, and perhaps pave the way for more-effective immunotherapies that coax human immune cells into recognizing and attacking hard-to-treat tumor cells.

“I first got into [studying transmissible cancers] because I wanted to help the devil,” says Murchison, who is originally from Tasmania. “Now, I still want to help the devils, but it’s also just a fascinating area of research where we can learn so much about cancer in general.”

Katarina Zimmer is a freelance science writer living in New York City.

References

  1. R.J. Pye et al., “A second transmissible cancer in Tasmanian devils,” PNAS, 113:374–79, 2016.
  2. M.J. Metzger et al., “Horizontal transmission of clonal cancer cells causes leukemia in soft-shell clams,” Cell, 161:255–63, 2015.
  3. M.J. Metzger et al., “Widespread transmission of independent cancer lineages within multiple bivalve species,” Nature, 534:705–09, 2016.
  4. M.N. Leathlobhair et al., “The evolutionary history of dogs in the Americas,” Science, 361:81–85, 2018.
  5. C. Murgia et al., “Clonal origin and evolution of a transmissible cancer,” Cell, 126:477–87, 2006.
  6. M.R. Stammnitz et al., “The origins and vulnerabilities of two transmissible cancers in Tasmanian devils,” Cancer Cell, 33:607–19, 2018.
  7. D.C. Brindley, W.G. Banfield, “A contagious tumor of the hamster,” J Natl Cancer Inst, 26:949–57, 1961.
  8. B. Ujvari et al., “The evolutionary ecology of transmissible cancers,” Infect Genet Evol, 39:293–303, 2016.
  9. A. Kreiss et al., “Allorecognition in the Tasmanian devil (Sarcophilus harrisii), an endangered marsupial species with limited genetic diversity,” PLOS ONE, 6:e22402, 2011.
  10. H.V. Siddle et al., “Reversible epigenetic down-regulation of MHC molecules by devil facial tumor disease illustrates immune escape by a contagious cancer,” PNAS, 110:5103–08, 2013.
  11. L. Kosack et al., “The ERBB-Stat3 axis drives Tasmanian devil facial tumor disease,” Cancer Cell, 35:P125–39, 2019.
  12. G.K. Brown et al., “Natural killer cell mediated cytotoxic responses in the Tasmanian devil,” PLOS ONE, 6:e24475, 2011.
  13. A. Caldwell et al., “The newly-arisen devil facial tumor disease (DFT2) reveals a mechanism for the emergence of a contagious cancer,” eLife, 7:35314, 2018.
  14. C. Tovar et al., “Regression of devil facial tumor disease following immunotherapy in immunized Tasmanian devils,” Sci Rep, 7:43827, 2017.
  15. M.J. Margres et al., “The genomic basis of tumor regression in Tasmanian devils (Sarcophilus harrisii),” Genome Biol Evol, 10:3012–25, 2018.
  16. B. Ujvari et al., “Evolution of a contagious cancer: epigenetic variation in the devil facial tumor disease,” Proc Biol Sci, 280:20121720, 2013.