Research Article: The devil is in the details: Genomics of transmissible cancers in Tasmanian devils

Date Published: August 2, 2018

Publisher: Public Library of Science

Author(s): Andrew Storfer, Paul A. Hohenlohe, Mark J. Margres, Austin Patton, Alexandra K. Fraik, Matthew Lawrance, Lauren E. Ricci, Amanda R. Stahlke, Hamish I. McCallum, Menna E. Jones, Katherine R. Spindler.


Partial Text

Based on a transcriptomic analysis of DFT1, the progenitor tumor likely originated from a mutated Schwann cell (a type of peripheral nerve cell) in a female Tasmanian devil [8]. DFT2 is also likely to be of neuroectodermal origin, but DFT2 does not express periaxin (PRX), a Schwann cell marker present in DFT1 [10]. Both DFT1 and DFT2 likely evolved from Tasmanian devils located in eastern Tasmania, with their genetic assignments consistent with their geographic origins (in the NE and SE, respectively) [10]. The gross morphology and histology of DFT2 are different from DFT1 [9,10]. For example, DFT1 is generally composed of pleomorphic round cells in bundles, whereas DFT2 is typically characterized by pleomorphic sheets of cells [9]. Moreover, DFT2 karyotypes have a Y chromosome, indicating that this tumor arose from a male devil and thus independently from DFT1 [9]. Although DFT1 and DFT2 originated in the last 21 years, no evidence has been found for viral origin, and results are inconsistent with tumor evolution coming from exposure to anthropogenic stressors, such as increased UV light [10].

For an allograft to avoid rejection from a new host, it must circumvent recognition by major histocompatibility complex (MHC) genes [24]. MHC Class I is generally responsible for tumor recognition via identification of cell surface proteins expressed as “nonself” on cancer cells [24]; transmissible cancer cells are indeed nonself, having originated in a different individual [3,6]. Ubiquitous susceptibility of Tasmanian devils to DFTD has been hypothesized to result from low devil genetic variability overall [25], likely due to at least two historical genetic bottlenecks [25,26]. Compared to other mammals, devils have particularly low genetic variability in the MHC Class I peptide-binding region implicated in tumor recognition [27]. However, MHC diversity is not linked to variation in disease susceptibility among individuals [28], and devils reject allografts in challenge experiments [29]. Instead, DFT1 appears to down-regulate its own MHC expression, as well as MHC expression in the devil [30]. Epigenetic down-regulation of MHC expression is common in human cancers [31], as well as being a salient feature in CTVT [12,18]. In addition to MHC evasion, there are at least several other mechanisms that underlie widespread transmissibility of DFTD yet to be discovered. Moreover, there is documented variation in tumor susceptibility among devils, including rare documented cases of tumor regression and immune response, which are discussed below.

Despite widespread declines of Tasmanian devils and predictions of localized devil extinctions, continued devil survival may result from evolution of DFTD resistance, which is supported by multiple lines of genetic evidence. First, a genome scan showed large and concordant allele frequency changes and increases in linkage disequilibrium across three populations pre- and post-disease [32]. Strong support for rapid evolution (in as few as four generations) was discovered in two small genomic regions containing seven candidate genes mapped to the devil reference genome; five of these genes were associated with immune- and cancer-related functions, including cell adhesion and p53 pathways [32]. Second, a genome-wide association study showed strong evidence that a few large-effect single nucleotide polymorphisms (SNPs or single base pair changes in DNA) explain a significant proportion of observed phenotypic variation in survival following infection in females [33]. Genes of particular interest in close proximity with these SNPs also include cell adhesion, tumor suppression, and p53 pathway genes [33]. Taken together, these two studies suggest evolution resulting from a soft selective sweep, whereby selection acted on standing genetic variation in a few, large-effect loci, as opposed to on new mutations [33]. A third study showed that DFTD was capable of swamping local adaptation to weaker abiotic forces, such as altitude [34]. That is, selection by the biotic factor of disease tended to overwhelm selection by abiotic factors in the predisease environment [34].

Cytogenetic analyses currently recognize four karyotypes of DFT1 [21] and show that genomic rearrangements are limited to particular cancer regions, suggesting at least some genomic stability [19,20]. When compared to the Tasmanian devil reference genome, two strains of DFT1 collected from SE and north central Tasmania, respectively, accumulated between 15,000 and 17,000 single-nucleotide substitutions between them [18]. Assuming they share a common ancestor that emerged approximately 20 years ago, this mutation rate is higher than most human cancers (approximately 5,000) but lower than lung cancer or melanomas [18]. Evidence of within-host tumor variation is limited; of 20 devils with multiple tumors, only six individuals had tumors that were genetically distinguishable. A comparison of the mitochondrial genomes of 104 DFT1 tumors from 69 devils across Tasmania showed limited among-host genetic variability as well, with 21 somatic variants detected [18].

Transmissible cancers are indeed a frightening phenomenon, and the recent appearance of a malignancy of tapeworm origin in an HIV-infected individual [37] shows that tumors derived from nonhost DNA can emerge spontaneously in immune-suppressed humans. As a species, the Tasmanian devil has perhaps suffered most extensively, with massive population declines resulting from the emergence and spread DFTD. Fortunately, a combination of genomic and immunological studies provide compelling evidence of devil evolutionary responses that appear to be related to DFTD resistance or tolerance. Further, an in vitro drug screen showed promise for possible oral treatment therapies; both DFT1 and DFT2 are highly sensitive to several clinical compounds, and DFTs apparently show low tolerance to DNA damage [10]. The development of numerous genomic data sets to study human cancers, DFTD, CTVT, and soon likely the bivalve neoplasias provides extensive resources for the study of cancer transmissibility. Further research that uses comparative genomics and transcriptomics, such as the recent comparison of DFT1 and DFT2 [10], will likely be fruitful for understanding the origin and evolution of cancer transmissibility in general.