Germline mutation rate predicts cancer mortality across 37 vertebrate species

Abstract

Background and objectives: Cancer develops across nearly every species. However, cancer occurs at unexpected and widely different rates throughout the animal kingdom. The reason for this variation in cancer susceptibility remains an area of intense investigation. Cancer evolves in part through the accumulation of mutations, and therefore, we hypothesized that germline mutation rates would be associated with cancer prevalence and mortality across species.

Methodology: We collected previously published data on germline mutation rate and cancer mortality data for 37 vertebrate species.

Results: Germline mutation rate was positively correlated with cancer mortality (P-value = 0.0008; R2 = 0.13). Controlling for species' average parental age, maximum longevity, adult body mass or domestication did not improve the model fit (the change (Δ) in Akaike Information Criterion (AIC) was less than 2). However, this model fit was better than a model controlling for species trophic level (ΔAIC > 2).

Conclusions and implications: The increased death rate from cancer in animals with increased germline mutation rates may suggest underlying hereditary cancer predisposition syndromes similar to those diagnosed in human patients. Species with higher germline mutation rates may benefit from close monitoring for tumors due to increased genetic risk for cancer development. Early diagnoses of cancer in these species may increase their chances of overall survival, especially for threatened and endangered species.

Keywords: animals; germ cell mutations; malignancy; tumor.

Figure 1.

The average yearly germline mutation rate is positively correlated with the percentage of animals (among the total number of individuals per species examined at necropsy) that died from cancer. Each dot represents a species, and the size of the dot indicates the number of necropsies available for that species. Error bars show 95% confidence intervals. The regression line is phylogenetically controlled using the PGLS analysis. Species’ images are from PhyloPic (https://www.phylopic.org/).

Figure 2.

The OU evolutionary model (A), rather than the BM (B) or EB model (C), best fits the species germline mutation rate data. Each point represents the reported log-likelihood at each generation of the MCMC fitting algorithm. The OU model indicates that the germline mutation rate is evolving towards an optimal value, and species evolved independently. The MCMC algorithm within the fitContinuousMCMC function generates samples of the parameter estimates from the posterior distribution. The likelihood is estimated at each generation of the algorithm to determine how well the parameter estimates fit the phylogenetic tree and the germline mutation rate data.