Gene expression in soft-shell clam (Mya arenaria) transmissible cancer reveals survival mechanisms during host infection and seawater transfer

Abstract

Transmissible cancers are unique instances in which cancer cells escape their original host and spread through a population as a clonal lineage, documented in Tasmanian devils, dogs, and ten bivalve species. For a cancer to repeatedly transmit to new hosts, these lineages must evade strong barriers to transmission, notably the metastasis-like physical transfer to a new host body and rejection by that host's immune system. We quantified gene expression in a transmissible cancer lineage that has spread through the soft-shell clam (Mya arenaria) population to investigate potential drivers of its success as a transmissible cancer lineage, observing extensive differential expression of genes and gene pathways. We observed upregulation of genes involved with genotoxic stress response, ribosome biogenesis and RNA processing, and downregulation of genes involved in tumor suppression, cell adhesion, and immune response. We also observe evidence that widespread genome instability affects the cancer transcriptome via gene fusions, copy number variation, and transposable element insertions. Finally, we incubated cancer cells in seawater, the presumed host-to-host transmission vector, and observed conserved responses to halt metabolism, avoid apoptosis and survive the low-nutrient environment. Interestingly, many of these responses are also present in healthy clam cells, suggesting that bivalve hemocytes may have inherent seawater survival responses that may partially explain why transmissible cancers are so common in bivalves. Overall, this study reveals multiple mechanisms this lineage may have evolved to successfully spread through the soft-shell clam population as a contagious cancer, utilizing pathways known to be conserved in human cancers as well as pathways unique to long-lived transmissible cancers.

Fig 1. Top differentially expressed genes and pathways in MarBTN.

Volcano plots of differentially expressed genes (A) and gene sets (B) from the comparison of MarBTN isolates (n=5) versus healthy hemocytes (n=8). Genes of note are labeled with annotations and abbreviated descriptions. Line marks false discovery rate adjusted significance threshold (p < 0.05), with genes below threshold colored in grey. Note grey points are largely not visible in A due to axis scale. “(+)” = positive regulation.

Fig 2. Fusions, copy number and transposable element insertions influence gene expression.

(A) Number of fusion transcripts per sample (dots), with mean and standard deviation for MarBTN (n=5) and healthy hemocyte (n=8) sample groupings. Black bar represents fusions found in all MarBTN samples (196). Statistical test is two-tailed t-test with unequal variance. (B) Log fold change in expression of MarBTN versus healthy hemocytes, binning genes by genomic copy number for each gene. Upper statistical tests are two-sided Wilcoxon rank-sum tests comparing medians between adjacent copy numbers, lower statistical tests are one-sample two-sided Wilcoxon rank-sum tests comparing medians versus no change. Gene counts for each group are listed below box plots. (C) Log fold change in expression of MarBTN versus healthy hemocytes for genes with a Steamer insertion within the gene itself or in the 2kB region upstream of the gene. Each set shows genes with Steamer insertions present in these samples (grey box) and, as a control, genes with no Steamer insertion present in these samples but at which Steamer insertion sites have been observed in a different sub-lineage of MarBTN (white box). Statistical tests and counts are the same as listed in (B). ns: not significant, *: p<0.05, **: p<0.005, ***: p<0.0005, ****: p<0.00005.

Fig 3. Transcriptomic response to seawater exposure.

(A) Proposed life cycle for MarBTN infections and (B) experimental design to investigate gene expression during seawater transmission. (C) Principal component analysis results from gene expression across all genes for ASW-treated and untreated MarBTN and hemocytes. ASW = artificial sea water.

Fig 4. Top differentially expressed genes and pathways in MarBTN after exposure to seawater.

Volcano plots of fold change and significance of differentially expressed genes (A/C) and gene sets (B/D) in ASW-treated MarBTN (n=3) versus untreated MarBTN (n=3). Genes of note are labeled with annotations and abbreviated descriptions. Volcano plots are filtered for genes/sets that are (A/B) or are not (C/D) differentially expressed in ASW-treated hemocytes (n=3) versus untreated hemocytes (n=3). For comparison, the reciprocal comparison is included in small grey points on each plot. Line marks false discovery rate adjusted significance threshold (p < 0.05), with genes below threshold colored in grey.